Skip to content

Latest commit

 

History

History
1855 lines (1434 loc) · 113 KB

CHANGES.md

File metadata and controls

1855 lines (1434 loc) · 113 KB

ChangeList

= 1.7.0 (23.07.2013) =

libpcl_common:

  • Added pcl::Intensity and pcl::Intensity8u point types
  • Added pcl::RangeImageSpherical sub-class that is more suitable than pcl::RangeImage for some kinds of 360° range images (as discussed in [PCL-users] Range Image Projection)
  • Added DefaultPointRepresentationpcl::Narf36 to allow pcl::Narf36 to be used with pcl::search (#915)

PCL Apps:

  • Added cloud_composer app
  • Added PCLModeler, with a tree view scene explorer and multiple render windows support
  • Added client app for the point cloud streaming server
  • Added new server app for point cloud streaming to mobile devices (pcl_openni_mobile_server)
  • Added a new demo for the connected component segmentation. Includes a QT gui that allows various features to be toggled on/off.
  • Added SHOT estimator wrapper using OMP
  • Added openni_organized_multi_plane_segmentation to demonstrate the OrganizedMultiPlaneSegmentation class.

libpcl_recognition:

  • Added a new tutorial for "libpcl_recognition" for Correspondence Grouping by Tommaso Cavallari (#666)
  • Added support for .LMT file loading (which are TARed files for PCD masks and SQMMT linemod templates)
  • Changes in the computation of the modality to improve performance
  • Fixed a compilation error on windows; for some reason 'NULL' needs to be explicitly casted to the pointer type
  • Added a model library class used for maintaining the object models to be recognized.
  • Changed the interface to make it less confusing to use.
  • Added a couple useful overloads for "Houg###Grouping" and "GeometricConsistencyGrouping"
  • Added CRHAlignment class.
  • Added Papazov HV method.
  • Fixed a bug in Poisson surface reconstruction that was causing the unit test to fail
  • Added option for automatic selection of number of features in extractFeature
  • Added a new greedy hypotheses verification method.
  • Added semi scale invariant linemod template detection
  • Fixed RF search radius in "Houg###Grouping"
  • Fixed some bugs in detection refinement along viewing direction
  • Fixed bug in LineRGBD::computeTransformedTemplatePoints (template point cloud's width and height fields were not set)
  • Converted uses of PointXYZRGB to PointXYZRGBA; converted std::cerr messages to PCL_DEBUG; minor reformatting to keep lines under 120 characters
  • Fixed some bugs related to bounding box computation and computation of transformed template point clouds
  • Added functionality to specify an object ID when loading templates in LineRGBD;
  • Added "GeometricConsistencyGrouping" clustering class
  • Added high level interface for RGBD version of linemod (not all parts are implemented yet)
  • Added SSE optimizations to improve detection speed
  • Added Ransac Correspondence Rejection into Houg###Grouping
  • Changed method for selecting features in color gradient modality
  • Added "CorrespondenceGrouping" abstract base class for correspondence grouping in pcl_recognitio
  • Added cosine approximation in score computation
  • Added structure for hypotheses verification methods. Base abstract class

libpcl_keypoints

  • Added implementation of the Intrinsic Shape Signature keypoint detector
  • fixed crash caused by setNormals() in HarrisKeypoint3D (related to #745)

libpcl_outofcore:

  • Added support for PointCloud + gen LOD
  • Added PointCloud2 support for outofcore queries via new "queryBBIncludes" method
  • Added "copyPointCloud" support for PointCloud2 without indices
  • Added feature: outofcore binary compressed pcd files to store point data on disk
  • Bug fix: outofcore write buffer constant limitation fixed so outofcore_process will work with large 20M+ point TRCS data sets
  • Constants for write buffer changed to 2e12 to support insertion of very large point clouds until new serialization is implemented
  • Added getVoxelSideLength to octree_base for displaying of nodes in visualizer

libpcl_search:

  • FlannSearch: fixed wrong typedef (::Ptr would break if FlannDistance!=flann::L2) and compiler error
  • Added new option in FlannSearch: FLANN KMeans Tree can now be uses as the search algorithm

libpcl_visualization:

  • Added specific methods to handle monochrome images represented by a PointCloud or a PointCloud
  • Add area selection option to PCLVisualizer so user can get a whole area indexes
  • Fix the ImageViewer shapes overlay using vtkContextItem so they now appear with transparent background

libpcl_tools:

  • Added a PNG to PCD converter

libpcl_io:

  • Added support for the Velodyne High Definition Laser (HDL)
  • Add support for foo/bar vertex property in PLY file reading

= 1.6.0 (2012-07-15) = :: "About time" =

The most notable overall changes are:

PCL Tools:

  • Added a tool for interfacing the marching cubes algorithms
  • Added a PLY To PCD conversion tool
  • Added a command line tool for transforming datasets based on their viewpoint
  • Implemented feature #605: New octree visualizer example (Contributed by Raphael Favier. Thanks!)
  • Updated "openni_save_image" to save both RGB and depth images as TIFF
  • Added two command line tools for converting PCD data into PLY and VTK formats
  • Fix for #504: Make ShapeContext and SpinImage uniform (thanks David!)

libpcl_common:

  • Fixed bug #658: Compiler error: undefined reference for getEulerAngles in pcl/common (thanks Gioia!)
  • Moved towards a unified "eigen.h" header that only includes the Eigen headers that we need.
  • Added boost 1.48 and 1.49
  • Added a default "PointRepresentation" type for "ShapeContext" (thanks Elizabeth!)
  • Added a new "PlanarPolygon" class for representing 2D planar polygon regions
  • Added "SetIfFieldExists" functor helper to copy data from a variable into a point field, if it exists
  • Added a helper functor ("CopyIfFieldExists") for copying out specific data out of a PointT field. See "test_common.cpp" for usage examples
  • Added point value initialization by default in constructors for "PointXYZI", "Normal", "PointXYZHSV", "PointXYZRGBL", and "PointXYZRGB"
  • Updating transforms.hpp to ensure that point fields are copied when applying affine transform to a specific set of indices.
  • Added error messages on failure of aux functions for PointCloud2, "pcl::concatenatePointCloud" and "pcl::concatenateFields"
  • Fixed win32 compilation error on test_plane_intersection
  • Implemented plane intersection feature (feature #644) with a related unit test
  • Added kissfft library
  • Bugfix of eigen22 version for smallest eigenvalue/vector
  • Add specialization for pcl::RGB point type
  • Intensity field accessor moved from keypoints/sift to common to be shared by others
  • Fixed some valid usages of point traits (e.g. static_cast::type>(...) ) on GCC 4.4.3 by explicitly instantiating some assert template long before it should actually be needed.
  • Added ESF Histogram 640 to point types

libpcl_filter:

  • Added command line option "-keep" to preserve the organized data structure after a "Passthrough" filter
  • Implemented feature #663: flexible comparison for conditional_removal filter (Thanks Julian!)
  • Fix for bug #672: ijk grid coordinates in VoxelGrid computed differently in different functions (thanks Nicholas!)
  • Fix for #572: Example "Euclidean Cluster Extraction" crashes / bug in VoxelGrid filter
  • Improved performance of 3x getFieldIndex and removed dependency on pcl_io
  • Fixed and implemented feature #518: VoxelGrid<> performance improvement (roughly 50+ times) (thanks Radoslaw!)
  • Fixed the visual studio compilation error described in feature #614 (thanks Remus!)
  • Fixed bug #674 (Elements in leaf_layout_ in VoxelGrid are not reset between calls to VoxelGrid::filter), Thanks Nicholas for the patch!
  • Work on issue 614: updated the PassThrough filter to derive from FilterIndices instead of Filter
  • Fix: use a makInfinite function to annihilate points in zero_padding cases
  • Add new class to handle convolution in 3D space for radial basis kernel
  • Modifying VoxelGridCovariance to allow for control over minimum number of points and eigen value inflation for singularity prevention.

libpcl_visualization:

  • Overloaded "addRectangle" in "ImageViewer" with several useful methods that take as input 3D min-max points and organized datasets, as well as image masks to create 2D rectangles on screen
  • Renamed "addBox" to "addFilledRectangle" in "ImageViewer"
  • Added "addPlanarPolygon" methods for "ImageViewer" to display planar polygonal contours in the image
  • Added new "addMask" methods to "ImageViewer" for displaying 2D image masks from a given "PointCloud"
  • "PCLVisualizer" now has a new "addCube" method for passing {x, y, z}_{min, max} directly to the screen (thanks Jeremie!)
  • Patch for two cases of mismatched new/free cases for PCL<->VTK data exchange (thanks Peter!)
  • Fixed a problem with "removeLayer" which was causing shapes to flicker on screen on sequences like: "removeLayer; addShape; spinOnce"
  • Added "addLine" to "ImageViewer" for displaying 2D lines
  • Added "PCLVisualizer::close" to close the interactor window and "PCLVisualizer::set{Size, Position}" to set the window size and position on screen added image viewer to the docs
  • Added better support for 2D image visualization (via "pcl::visualization::ImageViewer"): add/remove layers with different transparency, add 2D shapes (rectangles, circles, points, etc)
  • Added "wasStopped" to "ImageViewer" to check if the window has been closed
  • Fixed an issue in pcd_viewer where the PointPicking callback wasn't functioning properly
  • Added "setPosition" to "pcl::visualization::ImageViewer" for allowing the image viewer to be moved somewhere else on screen
  • Added two additional "addPointCloud" helpers for directly displaying "sensor_msgs::PointCloud2" data
  • Added "addPointCloud" method for "sensor_msgs::PointCloud2" data (useful to bypass the conversion to XYZ for "pcd_viewer")
  • Added the capability to remove a cloud when "removeShape" is called, to preserve API backward compatibility (a "PolygonMesh" is not treated as a "CloudActor" inside "PCLVisualizer")
  • Fixed a bug where the scalars were not updated properly on "updatePointCloud" thus causing VTK warnings on the console
  • Fixing issue #105 (PCLVisualizer::spinOnce don't work on Win32 system). Thanks Thibault and Alessio.
  • Added "PointXYZRGBA" callbacks for "OpenNIGrabber" and "PCLVisualizer".
  • Fixed two bugs where a segfault would occur in "addPolygonMesh" when the input cloud would be empty, as well as #563 : PCLVisualizer::addPolygonMesh crash (thanks Mourad!)
  • Fixed a bug where changing the point color using "setPointCloudRenderingProperties" would not update the actor's colors on screen
  • Fix for #532: "PCLVizualizer::addCoordinateSystem()" switch from "Eigen::Matrix4f" to "Eigen::Affine3f" (thanks Aurel!)
  • Fix for #524: ctrl problem with pcd_viewer (thanks Mourad!)
  • Adding opt in flag -use_vbos to pcl_visuzlier. It's still quite buggy, but shouldn't affect the visualizer unless this flag is passed.
  • Added vtkVertexBufferObject/Mapper and moved pcl_image_canvas_source2d
  • Added case handling where an actor might not have a valid viewpoint_transformation matrix resulting in a seg fault if pressing ALT+R
  • Fixed bug in displaying addCoordinateSystem
  • Added method to visualize intensity gradients
  • Added zoom functionality for ALT + Scroll-Wheel
  • Merged openni_viewer_simple with openni_viewer with all the available options except XYZI

libpcl_octree:

  • Fixed bug #693 - bounding box adaption affected change detection results
  • Applied patch by Robert Huitl, Issue #651, define maxVoxelCount in octree raycasting
  • Fixed OSX compiler warnings
  • Fixed bug #683 - octree depth changes during stream compression
  • Added new octree key class
  • Added bounding box checks in method isVoxelOccupiedAtPoint (octree pointcloud class)
  • Removed maxKeys limit in octree key generation method
  • Added range checks for integer keys in octree classes, extended octree key class
  • Fixed bug #620 (octree search fails if point cloud with indices is given) in octree pointcloud class

libpcl_io:

  • Added "DepthImage" signals/callbacks for "PCDGrabber"
  • Support for loading TAR-ed LMT files
  • Added support for TAR-PCD files for "PCDGrabber". Simply use "tar cvf file.tar *.pcd" and use "PCDGrabber" on it afterwards
  • Fixed a bug in the "PointCloud" feature estimation and I/O regarding the fields "count" property
  • Added a "saveVTKFile" method helper for saving "sensor_msgs::PointCloud2" data
  • Added support for reading PCD ascii and binary files (binary_compressed not implemented yet!) for pcl::PointCloudEigen::MatrixXf datatypes.
  • Implemented and tested a fix for feature request #558: Implement IO for PointCloud
  • Adding missing openni_device files from install
  • Fixed a bug in ply reader
  • Fixed bug #683 - dropping empty point clouds during stream compression
  • Add functions to convert between PCL and VTK data structures.
  • Handle red_diffuse, blue_diffuse and green_diffuse vertex properties
  • Fix bug where normals were written before RGB
  • Bugfix for #627 - Wrong include path for ply header

libpcl_features:

  • Fixed bug #676 in PCL Feature Search (thanks Adam!)
  • Fixes compilation problem on Windows due to size_t variable in omp loop.
  • Implemented feature request #661: border management in integral image normal estimation
  • Fixed PFHRGBEstimation bug, thanks Luis
  • Bug fix in SHOT feature, thanks wack
  • Fixed a bug which caused some normals to point in the wrong direction
  • Added Camera Roll histogram
  • Fix bug in index used for normal selection
  • Added esf feature
  • Added setViewPoint functionality and useSensorOriginAsViewPoint method to reset the viewpoint
  • Fixed bug #576 - CVFH respect setIndices now
  • Fixed issue #550: Uninformative error in VFHEstimation (thanks David!)
  • Fix #527: Bug in IntegralImageNormalEstimation (thanks Christoph!)
  • Fixed #544: overflow in implicit constant conversion warnings (thanks David!)
  • Modified SHOT omp so that the default computation of the reference frames
  • SHOT: Fixed invalid access when keypoints have increased from the previous call.
  • Setting sensor origin and orientation correctly for normal point cloud
  • Fixed a bug in estimating the orientation of the normal (view point wasn't initialized in the constructor)
  • Bug fix: if the cloud is not dense search for neighbours could raise an excpetion
  • Bug fix: SHOT accepts only radius search

libpcl_segmentation:

  • Fixed a few issues in the new segmentation classes where some comparators didn't have the appropriate "Ptr" and "ConstPtr" well defined
  • Fixed a bug where points passed onto the search method were not checked for NaN/Inf in "pcl::SegmentDifferences"
  • Added a missing "setDistanceFromOrigin" to "SACSegmentationFromNormals" for "SACMODEL_NORMAL_PARALLEL_PLANE" (thanks A. Barral)
  • Fix for #134 (Prism Extraction on Table top (flipping normal fails))
  • Fixed a segmentation fault in "SACSegmentationFromNormals" caused by calling "segment" without passing the input XYZ or normal data
  • Fixed a bug in organized connected component segmentation. Previously would crash if the first pixel had a valid depth (doesn't occur on kinect data).
  • Bugfix of issue # 675: Euclidean cluster extraction access violation. Thanks to Icy for helping find the solution
  • Fixed bug #673 in EuclideanClusterExtraction
  • Added new code for min cut segmentation
  • Fixed bug #681: member variable was not set in constructor (thanks Bhaskara!)
  • Added a threshold to MultiPlaneSegmentation for curvature, allowing us to discard smooth (but non-planar) regions.
  • Added optional projection for multi_plane_segmentation.
  • Added some new comparators for use with OrganizedConnectedComponents including RGB and edge aware
  • Added additional fucnction call for segmentAndRefine.
  • Added a comparator for doing euclidean clustering on organized point clouds.
  • Added fast SeededHueSegmentation implementations
  • Fixed segfault in multi plane refinement and functorized the comparison.
  • Improved MultiPlaneSegmentation to allow refinement of regions, and support different comparators.
  • New classes AutomatedSegmentation, AutomatedTreeSegmentation
  • Added OrganizedConnectedComponentSegmentation, which is a general class for use with organized point clouds and can take user specified comparators.
  • Added OrganizedMultiPlaneSegmentation which returns all planes in an organized cloud, along with the PlaneCoefficientComparator needed to do this.

libpcl_surface:

  • Fixed a grave bug where "setIndices" was not used in "ConvexHull"
  • Added fix for #562: pcl::ConcaveHull crashes on an empty cloud (thanks Mourad!)
  • Added "PCLSurfaceBase" base class for "MeshConstruction" and "SurfaceReconstruction", to unify the API around "setSearchMethod(&Search)" and "reconstruct(&PolygonMesh)"
  • Added GPU accelerated convex hull. 7x for Ramesses dataset (700k points)
  • Added a setDimension function to concave hull, so users can specify desired dimensionality of the resulting hull. If no dimension is specified, one will be automatically determined.
  • Fixed bug #692 - MovingLeastSquares indices issues
  • Added curve fitting and trimming of surfaces to examples/surface/example_nurbs_fitting_surface.cpp
  • Added iterative fitting routines for curve fitting surface::on_nurbs::Triangulation - added convertion functions for nurbs curve to line-polygon - added convertion functions for nurbs surface and curve to PolyMesh
  • Added flag to enable/disable usage of UmfPack for fast solving of sparse systems of equations - added triangulation functions to convert ON_NurbsSurface to pcl::PolygonMesh
  • Added bug fix in ConcaveHull, thanks to summer.icecream
  • Added marching cubes using RBF and Hoppe SDF
  • Pushed new functions that perform texture-mapping on meshes.
  • Fix: issue #646 (vtk_smoothing not copied)
  • Added new functionalities to TextureMapping: Find occlusions based on raytracing in octrees, UV mapping based on Texture resolution and camera focal lenght.
  • Relaxing dimensionality check threshold on concave_hull, so that 3D hulls should no longer be calculated on 2D input.
  • Added poisson filter

libpcl_keypoints:

  • Added combined Harris keypoint detector to detect corners in image, in 3D and combined ones
  • Fix bug #691 (Compiler error: undefined reference for setMinimalDistance in pcl/keypoints/harris_keypoint2D)
  • Fixed the #637 pitfall in harris3D and harris6D
  • Added implementation of 2D corner detectors

libpcl_geometry:

  • Added a new PCL library for computational geometry
  • Added bugfix to remove self intersecting polygons
  • Fixed some refinement corner cases for polygon approximation
  • Added line iterator class for iterating over e.g. an image/organized cloud in the pixel space.

libpcl_search:

  • Added NaN checks in "pcl::search::FlannSearch"
  • Skip infinite neighbor candidates in pcl::search::OrganizedNeighbor::radiusSearch
  • Added point projection method for organized search

libpcl_tracking:

  • Fixed issue #514: SpinImages segfault when using setKSearch (thanks David!)
  • gcc-4.7 compatibility issues in pcl::tracking fixed. Thanks to Rich Mattes

libpcl_sample_consensus:

  • Implemented feature #589: Add sac_model_cone and sac_model_normal_sphere (contributed by Stefan Schrandt. Thanks!)
  • Fix for #616: pcl::ProgressiveSampleConsensus::getRandomSamples unimplemented (thanks Mourad!)
  • Fix for #498: Bug in setOptimizeCoefficients for pcl::SACMODEL_SPHERE.
  • Fixed bug #648: code was checking for plane perpendicular to given axis instead of parallel
  • Applied patch for feature request #649: Locally constrained model generation for Sample Consensus

libpcl_registration:

  • Fixed issue #619: Cannot call pcl::registration::TransformationValidationEuclidean::validateTransformation (thanks Raphael!)
  • Added code for #592: correspondence estimation based on normal shooting (coded by Aravindhan. Thanks!)
  • Fixed #533: WarpPointRigid missing EIGEN_MAKE_ALIGNED_OPERATOR_NEW (thanks Radoslaw!)
  • Correct overlapping grid distribution in 2D NDT implementation

= 1.5.0 (2012-02-22) = :: "Better late than never" =

The most notable overall changes are:

  • new framework for PCL examples in the example/ folder
  • added the ICCV 2011 tutorial code to the list of bundled applications
  • a brand new PCL tracking library (libpcl_tracking)
  • new, well-tested implementations for OrganizedNeighbor and BruteForceSearch in libpcl_search
  • considerable speedups for: eigendecomposition, covariance matrix estimation, and normal estimation (now running at 30Hz on VGA data)
  • a lot of bug fixes and code refactorizations and optimizations (thanks to everyone that contributed!)
  • compilation now defaults to Eigen column major again, but all the code in PCL should compile and work against both row major and column major matrices.
  • libpcl_range_image is now incorporated into pcl_common, no extra linking required.

libpcl_common

  • added two new methods to compute centroid and covariance at once. Both are faster than calculating first the centroid and afterwards the covariance matrix. The float versions of the new methods are twice as fast but less accurate, whereas the double versions are more accurate and about 35% faster.
  • fixed a few issues in our ScopeTime class where the creation of the string as well as the cerr were contributing negatively to the time measurements
  • added empty() to PointCloudEigen::MatrixXf to return whether the matrix is empty or not
  • added size to PointCloudEigen::MatrixXf and fixed the constructor warnings exposed by -pedantic
  • added two new Point Types: Axis and ReferenceFrame as they are used in the Reference Frame calculations
  • added a helper functor (CopyIfFieldExists) for copying out specific data out of a PointT field. See test_common.cpp for usage examples
  • implemented and tested a fix for feature request #558: Implement IO for PointCloud

libpcl_filters

  • refactored RandomSample to return sorted indices. In addition it now runs nearly twice as fast at it did previously.
  • fixed and implemented feature #518: VoxelGrid<> performance improvement (roughly 50+ times)

libpcl_kdtree

  • removed unneeded mutices, added asserts instead of constant checks (the user is responsible for passing in valid data!) and sped up a few other things in KDTreeFLANN

libpcl_search

  • adapted use of flann::Matrix to new semantics of the 'stride' parameter, fixing incorrect results and a crash in FlannSearch

  • changes that affect pcl_features, pcl_search, pcl_octree, pcl_registration, pcl_surface:

    • we used to have 3 "overloads" for nearestKSearch and 3 for radiusSearch that every class inheriting from pclk::KdTree or pcl::search::Search needed to implement. Because most of the children always had the same implementation (copy&paste) we moved it up in the base class, made the methods virtual, and only children that need to change that behavior will overwrite it.
    • the max_nn parameter from radiusSearch has been changed from int with a default value of -1 to an unsigned int with a default value of 0. Each radiusSearch implementation should check for 0 or >= max_points_in_cloud and branch accordingly (if needed).
    • const correctness is now implemented everywhere in all nearestKSearch and radiusSearch operations
  • added an implementation of KdTreeFLANN for PointCloudEigen::MatrixXf data structures

  • FlannSearch: don't include flann/matrix.h, move all FLANN includes to impl/flann_search.hpp. Also adapt max_nn parameter in radius search to change r3962 (pass 0 to search for all neighbors instead of -1)

  • completely new and shiny BruteForceSearch and OrganizedNeighbor classes

libpcl_octree

  • fixed issue #513 (bug in octree iterator)
  • fixed octree bounding box adaption
  • fixed range test in getPointByIndex method (octree)
  • fixed bug #604 & removed compiler warnings in Visual Studio 10
  • fixed numerical problem in octree pointcloud class when points lie very close to the bounding box bounds
  • improved performance with std::deque in OctreeBreadthFirstIterator class
  • added breadth-first octree iterator and depth-first octree iterator classes, restructured octree iterator classes, moved "getVoxelBounds" method from OctreeIterator class to OctreePointCloud class

libpcl_sample_consensus

  • added explicit checks + errors for NaN covariance matrices
  • fixed a crash on 32bit platforms (thanks bharath)
  • fix for #498: Bug in setOptimizeCoefficients for pcl::SACMODEL_SPHERE. The issue actually exposed a grave bug introduced with r2755, when we switched from CMinPack to Eigen. The fix touches all segmentation models in pcl_sample_consensus using Levenberg-Marquardt, and TransformationEstimationLM in libpcl_registration. Also added unit test to catch this in the future.

libpcl_io

  • fix row-major/column-major issues in I/O (#578 related)
  • fixed #593 by removing unimplemented method declarations
  • added a saveVTKFile method helper for saving sensor_msgs::PointCloud2 data
  • added support for reading PCD ascii and binary files (binary_compressed not implemented yet!) for pcl::PointCloudEigen::MatrixXf datatypes.
  • implemented and tested a fix for feature request #558: Implement IO for PointCloud
  • added PointXYZRGBA callbacks for OpenNIGrabber and PCLVisualizer
  • fixed issue #565
  • fixed a few bugs in PCDReader regarding uint8/int8 data values and added a comprehensive unit test for all possible combinations of data types that we support in PCL
  • added isValueFinite for checking whether a value T is finite or not, and rewrote the PCDReader bits of code that perform the is_dense checks
  • integrated libply written by Ares Lagae as an independent component in io
  • rely on ply library for PLY parsing, remove the old parser and use the freshly integrated libply, change the namespace ::ply to ::pcl::io::ply to avoid any potential issue

libpcl_surface

  • Updated convex hull. It should now be somewhat faster, and allows users to optionally specify if the input is 2D or 3D, so the input dimension need not be calculated every time.
  • fix for #562: pcl::ConcaveHull crashes on an empty cloud (thanks Mourad!)
  • added option to ensure that triangle vertices are ordered in the positive direction around the normals in GreedyProjectionTriangulation
  • fixed issue #489 in GreedyProjectionTriangulation
  • fixed triangle vertex ordering issues in OrganizedFastMesh
  • MarchingCubes now adheres to the new pcl_surface base class architecture
  • added PCLSurfaceBase base class for MeshConstruction and SurfaceReconstruction, to unify the API around setSearchMethod(&Search) and reconstruct(&PolygonMesh)
  • added marching_cubes_reconstruction tool

libpcl_features

  • fixed issue #550: Uninformative error in VFHEstimation (thanks David!)
  • fix for #504: Make ShapeContext and SpinImage uniform (thanks David!). Note: this is an API breaking change, as setInputCloudWithNormals, which was deviating from the FeatureEstimation API is now replaced with patched setInputCloud and setInputNormals, thus no longer causing an error. However, we are confident with this change as the API was broken, so this is a bug fix.
  • fixed a bug in the PointCloud<MatrixXf> feature estimation and I/O regarding the fields count property
  • fixed an issue where the header stamp should only be copied from a PointCloud into a PointCloud in PCL standalone, and not in PCL in ROS
  • optimized RSD computation
  • speeded up normal estimation in IntegralImageNormalEstimation
  • renamed computeFeature (MatrixXf&) into computeFeatureEigen (MatrixXf&) and compute (MatrixXf&) into computeEigen (MatrixXf&) to keep the compiler happy, avoid multiple inheritance, and use clearer semantics between the two different output estimation methods
  • fixed an issue where if radius would be defined for VFHEstimation, Feature::initCompute would error out
  • fixed issue #539 (fpfh_estimation tutorial)
  • fixed issue #544 (overflow in implicit constant conversion warnings)
  • fixed issue #527 (bug in IntegralImageNormalEstimation)
  • fixed issue #514 (SpinImages segfault when using setKSearch)

libpcl_registration

  • added TransformationValidation base class and TransformationValidationEuclidean as a measure to validate whether a final transformation obtained via TransformationEstimation when registering a pair of clouds was correct or not
  • added getter/setter for maximum number of inner iterations to control number of optimizations in GeneralizedIterativeClosestPoint
  • fixed issue #507
  • fixed issue #533 (WarpPointRigid missing EIGEN_MAKE_ALIGNED_OPERATOR_NEW)
  • fixed issue #464 (already defined function pointers)
  • fixed a grave bug on 32bit architectures where MapAligned was used incorrectly

libpcl_segmentation

  • fixed a segmentation fault in SACSegmentationFromNormals caused by calling segment without passing the input XYZ or normal data
  • added a missing setDistanceFromOrigin to SACSegmentationFromNormals for SACMODEL_NORMAL_PARALLEL_PLANE (thanks A. Barral)

libpcl_visualization

  • added PointXYZRGBA callbacks for OpenNIGrabber and PCLVisualizer
  • fixed two bugs where a segfault would occur in addPolygonMesh when the input cloud would be empty, as well as #563 : PCLVisualizer::addPolygonMesh crash (thanks Mourad!)
  • added the capability to remove a cloud when removeShape is called, to preserve API backward compatibility (a PolygonMesh is not treated as a CloudActor inside PCLVisualizer)
  • fix #472: at startup the camera view pointis set to last added point cloud, but with the key combination ALT+r on can iterate through the view points of the visualized point clouds.
  • added addPointCloud method for sensor_msgs::PointCloud2 data (useful to bypass the conversion to XYZ for pcd_viewer)
  • added PCL_VISUALIZER_IMMEDIATE_RENDERING for processing large datasets
  • refactorized pcd_viewer to consume less memory by skipping the intermediate conversion to PointCloud<PointXYZ>, clearing the PointCloud2 data once converted to VTK, and enabling intermediate mode
  • added setPosition to pcl::visualization::ImageViewer for allowing the image viewer to be moved somewhere else on screen
  • changed the openni_image viewer tool to display both RGB and depth images
  • fixed a nasty bug where the scalars range were not correctly updated in the mapper on updatePointCloud thus causing weird VTK warnings at the console
  • added setupInteractor to allow QVTK customizations
  • pcd_grabber_viewer now shows the recorded images in PointXYZRGB clouds, and can use fps=0 and trigger each new frame manually by pressing SPACE
  • fixed an BGR/RGB issue in image_viewer
  • added setKeyboardModifier to change the default keyboard modifier from Alt to Shift or Ctrl so that in QVTK or other widgets, we can still use shortcuts like Mod+r, Mod+f, etc
  • PCLVisualizer is now working with VTK compiled for cocoa, no X11 necessary on OS X
  • build bundle executable for pcd_viewer if VTK is compiled with Cocoa. With that the keyboard events are handled by the application
  • modified openni_passthrough to use QVTK instead of VTK
  • added PCLVisualizer zoom in/out via /- [ alt]
  • implemented feature #516 (pcd_viewer multiview with .vtk files)
  • added showMonoImage for displaying a monochrome unsigned char 2D image in PCLVisualizer
  • added setFullScreen and setWindowBorders methods to allow an user to change the full screen/border properties of the PCLVisualizer GUI
  • fix for issue #519 (set the minimum number of cloud points to 1 in vtkLODActor)
  • fix for issue #524 (ctrl problem with pcd_viewer)
  • fix for issue #532 (PCLVizualizer::addCoordinateSystem() switch from Eigen::Matrix4f to Eigen::Affine3f)
  • bug fix in pcd_grabber_viewer: now sorting the read .pcd file by name before showing them (was in random order before)
  • fixed issue #525 (pcd_viewer did not handle nan values correctly in fields other than x, y, z)
  • fixed a bug where changing the point color using setPointCloudRenderingProperties would not update the actor's colors on screen
  • workaround for displaying FPS rate on PCLVisualizer - disabled backface culling because of known VTK issue

= 1.4.0 (2011-12-31) = :: "Happy New Year" =

The most notable overall changes are:

libpcl_common

  • added float union to the pointXYZRGBL type, in order to aid PCD_viewer compatibility
  • bugfix: pcl::computeCovarianceMatrixNormalized did not normalize for each implementation
  • fixed bug #421: at(u,v) should return a const reference
  • added InitFailedException and UnorganizedPointCloudException. Created PCL_EXCEPTION helper macro
  • added the GFPFH global descriptor point type
  • added gaussian kernel class and utilities
  • added isTrivial to pcl::PointRepresentation which allows avoiding copy operations in special cases, thus speeding up operations like rep->isValid (p)
  • added eigen33 for smallest eigenvalue/vector and only eigenvalues version
  • added PointCloudColorHandlerHSVField implementation
  • deleted wrong definition of density in pcl::transformPointCloud
  • added copy constructor for PCLBase
  • fixed PointCloud's in range inserter: width and height were not updated
  • added PCL_MAJOR_VERSION and PCL_MINOR_VERSION #defines per Ryan's request
  • fixed the getMatrixXf advanced user API call to return the correct values now that we are forcing Eigen Matrices to be row major in 1.x
  • moved the content of win32_macros.h into pcl_macros.h
  • added an isFinite method to check for x/y/z = NaN in point_types.hpp
  • moved internal headers from pcl/ros/point_traits.h into pcl/point_traits.h and pcl/ros/for_each_type.h to pcl/for_each_type.h (these headers are for internal use only and should have never been imported by user code - however pcl_ros in perception_pcl_unstable will need to be modified when following trunk after this revision, as that is the only external piece of code that should import these headers)
  • fixed one of the constructors for pcl::PointCloud<T> (cloud, indices) that was incomplete
  • no longer checking "frame_id" parameters in header for operator += in pcl::PointCloud<T>
  • added operator + for pcl::PointCloud<T>
  • improved doxygen documentation all around libpcl_common
  • added new specialization for pcl::PointCloud<Eigen::MatrixXf> - for advanced users only! (expect the API to change in trunk if needed)
  • moved NdCopyPointEigenFunctor and NdCopyEigenPointFunctor from voxel_grid.h to point_cloud.h
  • added CloudProperties class for optional properties for pcl::PointCloud<Eigen::MatrixXf>, and moved sensor_origin_ and sensor_orientation_ there (as properties.sensor_origin and properties.sensor_orientation), and removed header, while keeping header.stamp as properties.acquisition_time
  • fixed the copy constructor (vector<size_t> -> vector) and added an implementation for the + operator
  • added a general convolution class. Convolution handles point cloud convolution in rows, columns and horizontal directions. Convolution allows for 3 policies: zero padding (default), borders mirroring and borders duplicating through PointCloudSpring class
  • refactorized PointCorrespondence and Correspondence into the same thing, per issue #458
  • added ChannelProperties in preparation for 2.0
  • fixed bug #445: Overloaded the setIndices method in PCLBase so that it can accept boost::shared_ptr<const std::vector> >
  • disabled unit test for getMatrixXfMap in DEBUG mode, as it was failing due to eigen's strictness on aligned/unaligned data, and added const+non consts versions

libpcl_filters

  • fixed bug in random_sample.cpp; The PointCloud2 filter wasn't working causing the unit test to fail
  • crop_box filter now supports PointT and PointCloud2 point clouds
  • added a normal space sampling filter
  • fixed bug #433: pcl::CropBox doesn't update width and height member of output point cloud
  • fixed bug #423 CropBox + VoxelGrid filters, order changes behaviour (using openni grabber)
  • add CropHull filter for filtering points based on a 2D or 3D convex or concave hull. The ray-polygon intersection test is used, which relys on closed polygons/surfaces
  • added clipper3D interface and a draft plane_clipper3D implementation
  • removed spurious old ColorFilter class (obsolete, and it was never implemented - the skeleton was just lurking around)
  • better Doxygen documentation to PassThrough, VoxelGrid and Filter
  • moved the set/getFilterLimits, set/getFilterFieldName, and get/setFilterLimitsNegative from Filter into VoxelGrid and PassThrough, as they were the only filters using it. The rest were not, thus leading to user confusion and a bloated API.

libpcl_kdtree

  • fixed bug #442: reversed template parameters preventing use of nearestKSearchT and radiusSearchT in pcl::search::Search and pcl::kdtree:KdTree

libpcl_search

  • fixed a bug in OrganizedNeighbor where one of the radiusSearch signatures was working only if the cloud is not organized (#420 - thanks Hanno!)
  • updated pcl::search:OrganizedNeighbor to be back functional, radiusSearch is working back, 4-6x faster then KdTree, Knearest still need to implement this
  • changed pcl::search::FlannSearch: fixed style of class and tests, moved impl to hpp
  • cleaning up OrganizedNearestNeighbor as well as implementing one of the missing radiusSearch methods
  • improved documentation for libpcl_search and silenced all Doxygen warnings
  • removed auto (it needs to be completely rewritten and was left in a very bad state)
  • made sure all the output of the search method is consistently returning 0 in case the nearest neighbor search failed and not -1
  • updated kdtree wrapper code for the latest FLANN release

libpcl_octree

  • added method getVoxelBounds to octree iterator & removed result vector reserves
  • improved documentation for libpcl_octree and silenced all Doxygen warnings

libpcl_sample_consensus

  • fixed an omission of the sample size declaration for SACMODEL_PARALLEL_LINES (thanks Benergy)
  • replaced rand () with boost random number generators and fixed all failing unit tests
  • workaround to get rid of infinite loops if no valid model could be found

libpcl_io

  • refactorization and consistent code indentation + make sure the timestamp is not set if we use PCL in ROS
  • fixed an older issue where the is_dense flag was not set appropriately when reading data from a binary file (thanks Updog!)
  • fixed an issue in the PLYWriter class where the element camera was not correctly set thus leading to crashes in Meshlab (thanks Bruno!)
  • the VTK library loads vertex colors in PLY files as RGB. Added that to the polymesh loader.
  • fixed 2 bugs in writing PolygonMesh and unpacking RGB
  • fixed a bug in the .OBJ exporter to strip the path from the material library filename (thanks Robert!)
  • added support for exporting vertex normals to the .OBJ (pcl::io::saveOBJFile) file exporter (thanks Robert!)
  • fixed a 32bit/64bit issue in pointcloud compression
  • added method to write obj files from PolygonMesh
  • added serial number support for Windows using mahisorns patch. Thanks to mahisorn
  • overloaded callbacks for OpenNIGrabber to output PointCloud<Eigen::MatrixXf> datasets
  • added the possibility to write binary compressed Eigen data to disk using two new methods: generateHeaderEigen and writeBinaryCompressedEigen. Performance improvements to 30Hz I/O
  • fix for #463 (Missing Symbol rgb_focal_length_SXGA_)
  • fix: rgb values need to be packed before saving them in PointCloud2 for PLYWriter
  • added example code for accessing synchronized image x depth data
  • added support for the Stanford range_grid element and obj_info for PLY files. If you chosse to use range_grid instead of camera then only valid vertices will be written down to the PLY file.

lipcl_keypoints

  • added refine method for Harris3D corner detector
  • rewrote big parts of the NARF keypoint extraction. Hopfully fixing some stability issues. Unfortunately still pretty slow for high resolution point clouds.
  • fixed bug #461 (SIFT Keypoint result cloud fields not complete); cleaned up the line-wrapping in the error/warning messages

libpcl_surface

  • fixed a bug in MarchingCubes's getIndexIn1D which led to the result variable to overflow when the data_size was larger than something around 2^10 (thanks Robert!)
  • reviewed and slightly modified mls implementation. Added MovingLeastSquaresOMP OpenMP implementation
  • new architecture for the mesh processing algorithms using VTK: MeshProcessing
  • overloaded reconstruction and performReconstruction in SurfaceReconstruction to output a PointCloud<T> and vector<Vertices> as well
  • added a new class called MeshConstruction per API design decision to split the surface reconstruction methods into topology preserving (EarClipping, OrganizedFastMesh, GreedyProjectionTriangulation) and the rest. The new class implements a reconstruction/performReconstructionforPolygonMesh(for backwards compatibility purposes) and a faster set of the same methods forvector`
  • refactorized GreedyProjectionTriangulation by making it inherit from the new MeshConstruction base class, and removed a lot of old buggy code
  • overloaded performReconstruction (PointCloud<T> &, vector<Vertices> &) per the new SurfaceReconstruction API
  • fixed a bug in OrganizedFastMesh where the x/y/z indices were assumed to be 0/1/2, and made it part of the MeshConstruction API
  • optimizations for OrganizedFastMesh. Now in 30Hz+ flavor.
  • fixed a segfault from last night's EarClipping refactorization and improved the efficiency of the algorithm considerably
  • updated ConvexHull and ConcaveHull to inherit from the new MeshConstruction class
  • renamed mesh_processing.h to processing.h and performReconstruction/reconstruct to performProcessing/process for the new MeshProcessing API

libpcl_features

  • fixed bug #439: 3DSC unit test fails on MacOS
  • fixed IntegralImage2Dim : setting first line to zero was buggy producing undefined output data.
  • fixing an issue in PrincipalCurvaturesEstimation where the pc1 and pc2 magnitudes were not normalized with respect to the neighborhood size, thus making comparisons of different neighborhoods impossible (thanks Steffen!)
  • fixed ShapeContext3DEstimation computation and unit tests by switching from stdlib.h's random () to Boost.
  • fixed a bug in IntensityGradient: need to demean also intensity values, otherwise its assumed that the hyperplane goes always through origin which is not true
  • overloaded compute in pcl::Feature to save the output in an PointCloud<Eigen::MatrixXf>. Added partial specialization on Eigen::MatrixXf to all features and implemented computeFeature (PointCloud<MatrixXf> &output)
  • major doxygenization work
  • added optimization checks for is_dense for some features, implemented NaN output for most (per http://dev.pointclouds.org/issues/457)
  • added new unit tests for the Eigen::MatrixXf output
  • fixing the MacOS GCC 4.2.1 compiler segfault by removing the OpenMP #pragma (related to http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35364 ?).
  • fixed bug #468: refactored to make floor explicit.
  • added a setUseInternalCache method for PFHEstimation to use the internal cache on demand only. Other fixes to PFH/FPFH/PPF signatures (histogram sums were not adding up to 100.0 in some cases)
  • major improvements for integral images + added template specialization for one dimensional integral images

libpcl_registration

  • fixed some bugs in ELCH, thanks Frits for pointing them out

  • Big ELCH cleanup.

    • elch.h (API changes):
    • Use Boost bundled properties for LoopGraph, every vertex stores a pointer to the corresponding point cloud in graph[vertex].cloud.
    • loop_graph_ is now stored as a shared_ptr.
    • loop_start_ and loop_end_ are saved as ints for now (will change in the future).
    • loop_transform_ is not stored as a pointer (may change in the future).
    • elch.hpp:
    • Remove call to PCLBase::initCompute() as ELCH doesn't use a single input cloud or indices.
    • Change loopOptimizerAlgorithm to use start and end of the loop as stored in the object.
    • Adapt to API changes in elch.h.
  • added 2D implementation of Normal Distributions Transform for registration (Biber, Strasser; 2003), including example tool program.

  • reworked internal Registration API: Remove private computeTransformation without guess, make computeTransformation with guess abstract, Adapt all classes implementing Registration accordingly.

  • implemented SampleConsensusInitialAlignment with initial guess (thanks Dennis Guse)

  • cleaned up the CorrespondenceRejector API, per #375

  • apply Mourad's patch for #454; add more doc comments

libpcl_visualization

  • fixed an issue in pcl::visualization::PointPickingCallback where iren->GetMousePosition () didn't seem to work on some VTK versions/systems, so we replaced it instead with iren->GetEventPosition () (#444 - thanks Adam!)
  • fixed a bug in PointCloudColorHandlerRGBField<sensor_msgs::PointCloud2> where one of the offsets for Z was incorrectly calculated thus leading to erroneous data (#404 - thanks Lucas!)
  • set the lighting off on setShapeRenderingProperties when the user tries to control the color of the object precisely (maybe add a different lighting on/off property later?)
  • revert pcl::visualization::PCLVisualizer::setShapeRenderingProperties vtk5.2, default in ubuntu LTS, doesn't support the SetLighting method on vtkActor
  • fixed an issue in PointCloudColorHandlerRGBField where using an "rgba" field was causing an error in pcd_viewer (#428 - thanks Maurice!)
  • added PointCloudColorHandlerHSVField
  • fixed the PointPickingCallback behavior on Windows 7: iren->GetShiftKey () returns 4 instead of 1 (thanks bepe)
  • added updateFeatureHistogram functionality to PCLHistogramVisualizer, per #456 (thanks Asil!)
  • added patch from Adam Stambler (Bug #396)
  • added removePolygonMesh to make the API more user friendly (addPolygonMesh was already existing, but the polygon meshes were removed using removeShape until now)
  • made the polygon meshes actually implement CloudActor structures internally, so that we can use different colors per vertex
  • added updatePolygonMesh and improved the efficiency of addPolygonMesh for fast online rendering (i.e., produced via OrganizedFastMesh)

= 1.3.1 (2011-11-30) =

  • fixed bug #428: in PointCloudColorHandlerRGBField where using an "rgba" field was causing an error in pcd_viewer (thanks Maurice!)
  • fixed bug #404: in PointCloudColorHandlerRGBField<sensor_msgs::PointCloud2> where one of the offsets for Z was incorrectly calculated thus leading to erroneous data (thanks Lucas!) fixed bug #420: in OrganizedNeighbor where one of the radiusSearch signatures was working only if the cloud is not organized (thanks Hanno!)
  • fixed bug #421: toROSMsg referenced the address of a temporary return
  • fixed bug #422: SHOT estimation did not work with different surface than keypoints
  • fixed bug #430: an infinite loop in SAC methods if no valid model could be found
  • fixed bug #433: CropBox did not update width and height member of output point cloud
  • added missing using declarations for shot
  • added a missing include file to transformation_estimation_point_to_plane.h;
  • added transformation_estimation_point_to_plane_lls.h/hpp to the CMakeLists file
  • fixed an issue where the is_dense flag was not set appropriately when reading data from a binary file (thanks Updog!)
  • fixed a bug in the PLYWriter class where the element camera was not correctly set thus leading to crashes in Meshlab (thanks Bruno!)
  • fixed a bug in PrincipalCurvaturesEstimation where the pc1 and pc2 magnitudes were not normalized with respect to the neighborhood size, thus making comparisons of different neighborhoods impossible (thanks Steffen!)
  • set the lighting off on setShapeRenderingProperties when the user tries to control the color of the object precisely fixed bug in PointCloudColorHandlerRGBField<sensor_msgs::PointCloud2> where y_point_offset and z_point_offset were not incremented
  • fixed a bug in tools/compute_cloud_error
  • fixed a crash in test_ii_normals.cpp
  • removed unnecessary header from point_correspondence.h
  • fixed unnormalized covariance matrix in computeCovarianceMatrixNormalized
  • added missing include in std_msgs/Header.h
  • added error checking for radius search in extractEuclideanClusters
  • fixed a bug in PassThrough where NaN values were not correctly replaced by the user-specified value
  • fixed a bug in ply_io when writing PolygonMesh
  • fixed a bug in ply_io when unpacking rgb values

= 1.3 (2011-10-31) = :: "Gaudi" =

From 1.3 we are constructing the Changelist for each library separately, as they can in theory be mixed and matched with older versions (though we officially do not support this yet). The most notable overall changes are:

  • removed wxWidgets as a dependency from PCL and implemented the Image visualization classes using VTK
  • removed cminpack as a dependency from PCL and implemented the LM optimization functionality using Eigen
  • added a new library called PCL Search (libpcl_search) that has a more uniform API for nearest neighbor searches and stripped the unneeded functionality from libpcl_kdtree and libpcl_octree. Search now wraps KdTree and Octree for NN calls. This is a MINOR API BREAKING CHANGE. To change your code switch from:
pcl::KdTree -> pcl::Search (if the purpose is to use one of its children for search)
pcl::KdTreeFLANN -> pcl::search::KdTree
pcl::OrganizedDataIndex -> pcl::search::OrganizedNeighbor
  • improved MacOS support
  • improved documentation and unit tests
  • added lots of application examples and demos. Note: not all have a complete functionality... we will try to clean this up in 1.4 and future releases.

Build system changes

  • define PCL_ROOT environment variable using the NSIS installer
  • removing test+python from the dependency graph
  • fixed bug #374
  • remove cminpack from PCLConfig.cmake cmake and 3rdparty (switched to Eigen's Levenberg-Marquardt implementation)
  • update NSIS.template.in : use CPACK_PACKAGE_INSTALL_REGISTRY_KEY as the registry key
  • added the capability to build an all-in-one installer
  • fixed the build system to work 100% on Android and MacOS

libpcl_common

  • add overriding for operator[] to shorten code
  • add a setIndices method that computes indices of points in a region of interest
  • add demeanPointCloud with indices as PointIndices as well
  • added eigen allocator to correspondence vectors (pcl::Correspondences) and adapted all registration modules => be sure to use pcl::Correspondences instead of std::vector<pcl::Correspondence>
  • fixed a few doxygen errors
  • added simple stop watch for non-scoped function time measurements and made ScodeTime derived from StopWatch
  • fixed a bug in getRejectedQueryIndices, wrong output when order of correspondences have been changed
  • moved getRejectedQueryIndices to pcl/common/correspondence.h
  • added more doxygen documentation to the registration components
  • marked all getRemainingCorrespondences-functions as DEPRECATED, we sould replace them with purely stateless version outside the class body
  • fixed a const missing in PolynomialCalculationsT (#388 - thanks Julian!)
  • add PCL_DEPRECATED macro, closes #354.
  • added PointXYZHSV type and the conversions for it
  • added check for endianness for the Android platform

libpcl_search

  • BIG changes introduced - migration from KdTree to pcl::Search: moved OrganizedDataIndex and OrganizedNeighbor classes to libpcl_search
  • added new templated methods for nearestKSearch and radiusSearch for situations when PointT is different than the one the KdTree object was created with (e.g., KdTree vs nearestKSearch (PointT2 &p...)
  • added two new methods for getApproximateIndices where given a reference cloud of point type T1 we're trying to find the corresponding indices in a different cloud of point type T2
  • refactorized a lot of code in search and octree to make it look more consistent with the rest of the API
  • fixed a bug in octree_search which was semantically doing something bad: for each radiusSearch/nearestKSearch/approxNearestSearch call with a PointCloudConstPtr, the octree was getting recreated. Changed the API to be consistent with the rest of PCL (including pcl_search and pcl_kdtree) where we pass in a PointCloud instead of a PointCloudConstPtr which simply calls searchMethod (cloud.points[i], ...)
  • minor code optimizations
  • renamed organized_neighbor.h header in pcl_search (unreleased, therefore API changes OK!) to organized.h
  • disabled the auto-search tuning as it wasn't really working. must clean this up further
  • remove all approxNearestSearch methods from the base pcl::Search class as they did not belong there

libpcl_io

  • use stringstream instead of atof because of locale issues in IO (note that we can't use the ifstream directly because we have to check for the nan value)
  • added locale independent PCD ASCII i/o
  • implemented setMapSynchronization for pcl::PCDWriter. When set, msync() will be called before munmap() in the pcl::PCDWriter::write* calls, which guarantees data reliability. Though I/O performance is 300% better when unset, data corruption might occur on NFS systems, as indicated by http://www.pcl-developers.org/PCD-IO-consistency-on-NFS-msync-needed-td4885942.html.
  • added new writeBinaryCompressed functionality for general purpose sensor_msgs::PointCloud2 data (which is still our generic data container in PCL 1.x)
  • added additional unit tests for binary_compressed
  • fixed a grave bug in PCDReader (most likely introduced a few releases ago due to code refactorization) where the data was incorrectly copied if a PCD ASCII file had a field with multiple count elements (field.count) as first. Binary files are not affected by this bug. Added an unit test to catch this in the future.
  • added functionality for openni_grab_frame (added optional command line options, optional output filename, chose output format)
  • changed to new location of samplesconfig.xml for OpenNI
  • added signal and slot blocking into grabber. Using blocking to skip first frame in openni_grabber, since it is corrupted
  • added PLY format file support in binary and ascii mode (requires boost::iostreams library)

libpcl_keypoints

  • added 3D versions of Harris/Noble/Lowe/Tomasi and Curvature-based keypoint detection... scale space still missing
  • work on making SIFTKeypoint more flexible by saving scale only when the output point type contains "scale" (the catch is that all point types must be correctly declared via our macros - see the modifications in test_keypoints.cpp). This allows us to use a SIFTKeypoint<PointXYZRGB, PointXYZRGB> and thus removes the constraint on using copyPointCloud afterwards.
  • fixed an issue in SIFTKeypoint where width/height were not correctly set

libpcl_features

  • specialize std::vector for Eigen::Matrix4f (alignment issue with msvc 32bit) in SHOTEstimation
  • added a faster (eigen-based) integral image calculation => sped up normal estimation to 15Hz
  • added Unique Shape Context (USC) feature descriptor
  • added Shape Context 3D feature descriptor
  • fixed a bug in the normalization factor of VFH for the distance component (only affecting if set distance component is true)
  • fixed a few bugs regarding Windows build introduced in earlier commits
  • BIG changes introduced - migration from KdTree to pcl::Search
  • merged libpcl_range_image_border_extractor into libpcl_features. There's absolutely no reason why we should have 2 libraries generated from the features module.

libpcl_filters

  • added FilterIndices functionality #315
  • added a RandomSample filter which makes use of indices #323
  • added a new (very fast) class for data decimation: ApproximateVoxelGrid
  • fix for #369 (StatisticalOutlierRemoval crash when input dataset is empty)
  • implemented feature request #346

libpcl_octree

  • added function genVoxelBounds to octree pointcloud class
  • added octree neighbor search class
  • added octree-ray-tracing patch to octree_search class
  • buxfix in octree ray traversal function getIntersectedVoxelCentersRecursive
  • added unit test for getIntersectedVoxelCentersRecursive
  • added method getIntersectedVoxelIndices for getting indices of intersected voxels and updated unit test
  • refactorized a lot of code in search and octree to make it look more consistent with the rest of the API
  • fixed a bug in octree_search which was semantically doing something bad: for each radiusSearch/nearestKSearch/approxNearestSearch call with a PointCloudConstPtr, the octree was getting recreated. Changed the API to be consistent with the rest of PCL (including pcl_search and pcl_kdtree) where we pass in a PointCloud instead of a PointCloudConstPtr which simply calls searchMethod (cloud.points[i], ...)
  • minor code optimizations
  • renamed organized_neighbor.h header in pcl_search (unreleased, therefore API changes OK!) to organized.h
  • disabled the auto-search tuning as it wasn't really working. must clean this up further
  • remove all approxNearestSearch methods from the base pcl::Search class as they did not belong there

libpcl_registration

  • fixed a minor bug in CorrespondenceRejectorSampleConsensus: getRemainingCorrespondences tried to nput_ although it should only use the given input correspondences
  • added missing implementation for TransformationEstimationLM on correspondence vectors
  • added eigen allocator to correspondence vectors (pcl::Correspondences) and adapted all registration modules --> be sure to use pcl::Correspondences instead of std::vector<pcl::Correspondence>
  • fixing the API: a few left inconsistencies between vector<Correspondence> and Correspondences. The latter is to be preferred as it contains the Eigen aligned allocator.
  • added new ELCH loop correction API (New pcl::registration::ELCH class (WIP), add Registration::Ptr typedef)
  • added unit tests for the (new) registration API and all registration components
  • Further cleaning up registration code and writing documentation.
    • fixed bug in getRejectedQueryIndices, wrong output when order of correspondences have been changed
    • moved getRejectedQueryIndices to pcl/common/correspondence.h
    • added more doxygen documentation to the registration components
    • marked all "getRemainingCorrespondences"-functions as DEPRECATED, we sould replace them with purely stateless version outside the class body
  • Update: remove ciminpack dependency and rely on eigen for LM
  • Fixed a bug in ICP-NL by modifying WarpPointRigid to preserve the value of the 4th coordinate when warping; Re-enabled missing unit tests for ICP and ICP-NL
  • Added point-to-plane ICP
  • added nr_iterations_ and max_iterations_ to the initializer list (were mising)
  • Fixed bugs in WarpPointRigid3D and TransformationEstimationLM
  • fixed a problem where if no correspondences would be found via nearestKSearch, the RANSAC-based rejection scheme would fail (thanks to James for reporting this)
  • changed the default LM implementation to use L2 instead of L2SQR
  • Added a new TransformationEstimationPointToPlaneLLS class that uses a Linear Least-Squares approximation to minimize the point-to-plane distance between two point clouds
  • Added the ability to specify the error function to be minimized in SAC-IA (see Feature #362)

libpcl_sample_consensus

  • reimplemented the Levenberg Marquardt code that was using cminpack with Eigen, thus dropping the cminpack dependency for PCL

libpcl_surface

libpcl_segmentation

  • added a labeled cluster euclidean segmentation method

libpcl_visualization

  • fixed an issue where saveScreenshot was not correctly initialized with the proper renderer, thus not saving the data
  • supporting fontsize in addText (per feature #365)
  • fixing the interactor style by ignoring all combinations in OnChar that we cover in OnKeyDown
  • added removeAllShapes and removeAllPointClouds per #353
  • renamed deleteText3D to removeText3D to consolidate the API
  • fixing the API: a few left inconsistencies between vector<Correspondence> and Correspondences. The latter is to be preferred as it contains the Eigen aligned allocator.
  • added patch from Lucas Walter to fix pcl::visualization::PCLVisualizer::removeAllPointClouds and removeAllShapes
  • fixed a few doxygen errors
  • cleaned up the PCLHistogramVisualizer API by moving to a vtk interactor instead of ours, and removed unnecessary calls such as saveScreenShot (never used with the histogram visualizer, and most likely buggy), wasStopped and resetStoppedFlag (never used).
  • removed PCLVisualizerInteractor and switched back to a better default vktWindowRenderInteractor for VTK >= 5.6. Adjusted and fixed the internal API. No public API changes were made (except the removal of the PCLVisualizerInteractor class which was never meant to be used externally anyway).
  • cleaned up and implemented the ImageViewer class properly and demonstrated it in a single thread via openni_viewer. The other tool (openni_viewer_simple) will not work due to CloudViewer starting its own thread and thus clashing on the same OpenGL context.
  • fixed the correct order in initializer list for ImageViewer
  • removed wxWidgets completely from the codebase
  • added implementation for markPoint to ImageViewer to mimic older wxWidgets-based code
  • fixed an issue in PCLVisualizerInteractorStyle where mouse events were not properly mapped in Qt (#389 - thanks Adam!)
  • added an extra field of view to the camera paramers (#392 - thanks Adam!)
  • small bugfix when the radius_sphere was different than 1 for renderViewTesselatedSphere

= 1.2 (2011-09-30) = :: "Enough small talk" (LM) =

Additions, improvements, and optimizations

  • Eliminated the transform.h header file < API breaking change > (r2517)

    • the following functions have moved to transforms.h:

      • transformXYZ (renamed to transformPoint)
      • getTransformedPointCloud (replaced with transformPointCloud)
    • the following methods have been replaced with built-in Eigen methods:

      • operator<< (use .matrix () )
      • getRotationOnly ( .rotation () )
      • getTranslation ( .translation () )
      • getInverse ( .inverse () )
    • the following functions have move to eigen.h:

      • getTransFromUnitVectorsZY
      • getTransFromUnitVectorsZY
      • getTransFromUnitVectorsXY
      • getTransFromUnitVectorsXY
      • getTransformationFromTwoUnitVectors
      • getTransformationFromTwoUnitVectors
      • getTransformationFromTwoUnitVectorsAndOrigin
      • getEulerAngles
      • getTranslationAndEulerAngles
      • getTransformation
      • getTransformation
      • saveBinary
      • loadBinary
  • Made major changes in pcl::registration (r2503)

    • all registration code now uses TransformEstimation objects (TransformEstimationSVD and TransformEstimationLM in particular) rather than the older estimateRigidTransformationSVD code. Each class inheriting from pcl::Registration can pass in a different estimator via setTransformationEstimation
    • simplified TransformEstimationSVD code
    • implemented TransformEstimationLM by moving away code from IterativeClosestPointNonLinear (which now uses the transformation object)
  • replaced the io/io.h header file with common/io.h (for backwards compatibility, io/io.h will remain, but its use is deprecated)

  • added unit test for lineWithLineIntersection (r2514)

  • improved the VTK installation from source documentation for MacOS (r2589)

  • updated the tutorials regarding usage of FindPCL.cmake vs. PCLConfig.cmake (r2567)

  • added a new PointCloud constructor for copying a subset of points (r2562)

  • made wxwidgets an optional dependency for visualization (r2559)

  • added feature #334 (Enabling a library should enable all its library dependencies in CMake) implementation, (r2551)

  • added an internal estimateRigidTransformationSVD method to SampleConsensusModelRegistration (r2502)

  • added a PCL_VISUALIZER_REPRESENTATION property for setShapeRenderingProperties with three possible values:

    • PCL_VISUALIZER_REPRESENTATION_POINTS for representing data as points on screen
    • PCL_VISUALIZER_REPRESENTATION_WIREFRAME for representing data as a surface wireframe on screen
    • PCL_VISUALIZER_REPRESENTATION_SURFACE for representing data as a filled surface on screen (r2500)
  • optimized performance of BoundaryEstimation (approximately 25% faster) (r2497)

  • added reference citation to estimateRigidTransformationSVD (r2492)

  • added a new keypoint for uniformly sampling data over a 3D grid called UniformSampling (r2413)

  • added a destructor to VoxelGrid<sensor_msgs::PointCloud2> (r2412)

  • optimized the performance of SampleConsensusModelLine (r2404)

  • changed the behavior of toc_print and toc from pcl::console:TicToc to return milliseconds (r2402)

  • added a new model, SAC_MODEL_STICK, for 3D segmentation (r2400)

  • added 2x point picking to PCLVisualizer; use Alt + Mouse Left click to select a pair of points and draw distances between them (r2388)

  • added two new functions (pcl::getMaxSegment) for determining a maximum segment in a given set of points (r2386)

  • moved filters/test/test_filters to test/test_filters (r2365)

  • renamed the binary executable in the compression.rst tutorial (r2345)

  • Updated library dependencies

    • removed the libfeatures dependency for `libpcl_surface (r2337)
    • removed the libpcl_io dependency from libpcl_surface (r2354)
    • removed libpcl_io dependency from libpcl_keypoints (r2364)
    • removed libpcl_io dependency for libpcl_filters (r2365)
    • removed libpcl_io dependency for libpcl_registration (r2372)
  • added a new function, pcl::concatenateFields for concatenating the fields of two sensor_msgs::PointCloud2 datasets (1933)

  • added a new countWithinDistance method to SampleConsensusModel (r2326), and optimized RANSAC and RRANSAC by replacing calls to selectWithinDistance with countWithinDistance (r2327)

  • added feature #330: new PCLVisualizer::addCube and pcl::visualization::createCube methods (r2322)

  • added correspondence.h to the includes in common/CMakeLists.txt (r2260)

  • implemented defaults for the search method, per http://www.pcl-developers.org/defaults-for-search-method-td4725514.html (r2219,r2220,r2222,r2225,r2226,r2228)

  • exposed pcl::PCLVisualizer::saveScreenshot (const std::string &filename) (r2095)

  • enabled pcd_io.hpp (r2085)

  • added a new faster binary writer + cloud/indices (r2080)

  • added pcl::RGB structure and switched bitshifting RGB code to use the pcl::RGB structure's members (r2077,r2078)

  • added a new IOException class for read/write exceptions (r2068)

  • added missing set/getAngleThreshold for BoundaryEstimation. Made class parameter protected. (r2067)

  • added functions in pcl::console for parsing CSV arguments of arbitrary length (r2052)

  • added new functionality to OrganizedFastMesh (r1996); Now support for modes:

    • fixed triangle meshing (left and right) that create quads and always cut them in a fixed direction,
    • adaptive meshing that cuts where possible and prefers larger differences in 'z' direction, as well as
    • quad meshing
  • improved OrganizedFastMesh's removal of unused points (r1996)

  • CMake changes (r2592)

    • changed the install dir of PCLConfig.cmake in Windows
    • renamed CMAKE_INSTALL_DIR into PCLCONFIG_INSTALL_DIR
    • NSIS installer will add a key in Windows Registry at HKEY_LOCAL_MACHINE\Software\Kitware\CMake\Packages\PCL to help CMake find PCL (CMake >= 2.8.5) (NSIS.template.in)
    • reordered CMAKE_MODULE_PATH entries, for CMake to pick up our NSIS.template.in
    • fixed CPACK_PACKAGE_INSTALL_REGISTRY_KEY value

Bug fixes

  • fixed bugs in PointCloud

    • in swap(), the point data was swapped, but the width and height fields were not (r2562)
    • in push_back(), adding points did not update the width/height of the point cloud (r2596)
    • in insert()/erase(), inserting or erasing points did not update the width/height of the point cloud (r2390)
  • fixed bugs in SampleConsensusModelRegistration

    • if target_ wasn't given, would crash with a boost shared uninitialized error (r2501)
    • homogeneous coordinates were wrong (r2502)
  • fixed a bug in BoundaryEstimation in case "angles" is empty (r2411)

  • fixed a bug in OrganizedFastMesh's adaptive cut triangulation; added a new unit test (r2138)

  • fixed a bug in the openni_image viewer tool (r2511)

  • fixed problems with Windows/MacOS ALT bindings in PCLVisualizer (r2558)

  • fixed issues

    • #139 (FPFHEstimation for non-trivial indices) (r2528)
    • #303 (make pcd_viewer reset the camera viewpoint when no camera given) (r1915)
    • #338 (cylinder segmentation tutorial was referring to a different file) (r2396)
    • #339 (the direction of the normal was wrongly estimated when setSearchSurface was given) (r2395)
    • #340 (keep_organized_ bug in ConditionalRemoval) (r2433)
    • #342 (Allow QT to be used with PCL Visualizer) (r2489)
    • #343 (empty() member function of a point cloud is not const) (r2440)
    • #350 (Misspelling in GreedyProjectionTriangulation::setMaximumSurfaceAgle()) (r2556)
  • added missing include in correspondence_rejection.h (r2393)

  • corrected the headers included by sample_consensus/sac_model.h (r2550)

  • removed duplicate content of NSIS.template.in (r2601)

  • fixed various casting related compiler warnings (r2532)

  • corrected typos in

    • Windows tutorial (CMAKE_INSTALL_DIR => CMAKE_INSTALL_PREFIX) (r2595)
    • registration/icp.h documentation (reg => icp) (r2515)
    • several apps/tools console messages (wotks => works) (r2491)
    • how_features_work.rst tutorial (Muechen => Muenchen) (r2484)

= 1.1.1 (2011-08-31) = :: "Low Entropy" =

  • issues fixed: #224, #277, #288, #290, #291, #292, #293, #294, #295, #296, #297, #299, #302, #318, #319, #324, #325, #329

  • Additional fixes:

    • fixed a segfault in PCLVisualizer::addPointCloudNormals
    • fixed PCLVisualizer hanging on 'q' press
    • fixed a bug in MLS
    • fixed a bug in test_io
    • fixed a bug in PointCloudColorHandlerGenericField
    • fixed a bug when writing chars to ASCII .PCD files
    • fixed several errors "writing new classes" tutorial
    • added missing parameter setter in the "concave hull" tutorial

= 1.1.0 (2011-07-18) = :: "Deathly Hallows" =

  • new 3D features:

    • SHOT (Signature of Histograms of Orientations)
    • PPF (Point-Pair Features)
    • StatisticalMultiscaleInterestRegionExtraction
    • MultiscaleFeaturePersistence
  • improved documentation:

    • sample consensus model coefficients better explained
    • new tutorials (RadiusOutlierRemoval, ConditionalRemovalFilter, ConcatenateClouds, IterativeClosestPoint, KdTreeSearch, NARF Descriptor visualization, NARF keypoint extraction, ConcaveHull, PCLVisualizer demo)
  • new surface triangulation methods:

    • MarchingCubes, MarchingCubesGreedy
    • OrganizedFastMesh
    • SurfelSmoothing
    • SimplificationRemoveUnusedVertices
  • new registration methods:

    • PyramindFeatureMatching
    • CorrespondenceRejectorSampleConsensus
  • new general purpose classes/methods:

    • Synchronizer
    • distance norms
    • TextureMesh
    • PointCloud.{isOrganized, getMatrixXfMap)
    • SACSegmentation now works with PROSAC too
    • PCDViewer now reads VTK files
    • new Mouse and Keyboard events for PCLVisualizer
    • PCLVisualizer.{addText3D, addCoordinateSystem (Eigen::Matrix4f), deleteText3D, updatePointCloud, renderViewTesselatedSphere, resetCameraViewpoint, getCameras, getViewerPose}
    • ONIGrabber, DeviceONI
    • ImageRGB24, IRImage
    • generic FileReader + FileWriter
  • optimizations:

    • faster pipelinening by not recreating a fake set of indices everytime
    • rendering optimizations for PCLVisualizer
    • copyPointCloud is now faster and can copy the intersections of the fields in both input datasets
    • VoxelGrid is now ~20Hz for Kinect data
  • new applications:

    • VFH NN classification
    • 3D concave hulls
    • 3D convex hulls
    • ICP registration
    • Planar segmentation
    • Stream compression
    • Range image viewer
    • Voxel Grid
    • IntegralImage normal estimation
  • issues fixed: #75, #106, #118, #126, #132, #139, #156, #182, #184, #189, #197, #198, #199, #201, #202, #203, #210, #213, #211, #217, #223, #225, #228, #230, #231, #233, #234, #240, #247, #250, #251, #254, #255, #257, #258, #259, #261, #262, #264, #265, #266, #267, #268, #269, #273, #276, #278, #279, #281, #282, #283

= 1.0.1 (2011-06-29) = :: "First Steps" =

please note that version 1.0.0 had a flaw when creating ASCII pcd files. This version includes the tool pcd_convert_NaN_nan to fix this

  • added VTK file visualization to pcd_viewer
  • hiding the cminpack/FLANN headers, thus reducing compile time for user code
  • fixed IntegralImageNormalEstimation
  • tutorial updates and fixes + new tutorials. Changed tutorial structure to split CPP files from RST text.
  • better doxygen documentation for many functions
  • fixed a bug in ConditionalRemovalFilter where the keep_organized condition was reversed
  • removed BorderDescription and Histogram<2> from the list of explicit template instantiations
  • added PointXY point registration macros
  • added ExtractIndicesSelf unit test
  • fixed a lot of alignment issues on 32bit architectures
  • PCD ascii files now have each individual line trimmed for trailing spaces
  • internal changes for PCDReader/PCDWriter, where NAN data is represented as "nan"
  • sped up compilation with MSVC by adding /MP for multiprocessor builds
  • added a voxel grid command line tool filter
  • issues fixed: #242, #207, #237, #215, #236, #226, #148, #214, #218, #216, #196, #219, #207, #194, #192, #183, #178, #154, #174, #145, #155, #122, #220
  • added support for PathScale pathcc compiler
  • added support for Intel icc C++ compiler
  • added support for GCC 4.6 C++ compiler
  • added preliminary support for Clang C++ compiler
  • FindPCL.cmake and PCLConfig.cmake completed

= 1.0.0 (2011-05-11) = :: "A new beginning" =

  • completely standalone build system using CMake (not dependent on ROS anymore)
  • tested on Win32/Win64, Linux (32/64) and MacOS (installers provided for all 3rd party dependencies and PCL)
  • separated the entire codebase into multiple libraries that depend on each other. separate CMake declarations for each individual library.
  • provide a FindPCL.cmake to ease integration
  • many new unit tests + tutorials + improved Doxygen documentation (check http://www.pointclouds.org/documentation and http://docs.pointclouds.org)
  • new liboctree interface for nearest neighbor/radius search with Octrees, change detection, and Point Cloud compression!
  • added concave/convex hulls using QHull (dropped our old ConvexHull2D class)
  • pcl::Registration improvements and bugfixes, sped up inner loops, added alignment methods with initial guess
  • pcl::visualization bugfixes: multiple interactors are now possible, better threading support, sped up visualization (2-3x), addPolygonMesh helper methods
  • new grabber interface for OpenNI-compatible camera interface: all 3 OpenNI cameras are supported (Kinect, Asus XTionPRO, PSDK)
  • new grabber interface for streaming PCD files
  • pcl::io bugfixes for PCD files using binary mode
  • ROS_ macros are no longer valid, use PCL_ instead for printing to screen
  • new PCA implementation, up to 4-5 times faster eigen decomposition code
  • new CVFH feature, with higher performance than VFH for object recognition

Pre 1.0 (copied from http://pcl.ros.org/ChangeList) - for historical purposes.

The version numbers below belong to the perception_pcl stack in ROS, which coordinated both the release of PCL, as well as the release of the ROS PCL wrappers. From 1.0 onwards PCL got separated from ROS, and the release process of perception_pcl and PCL diverged.

  • [[flann]]

  • Updated to version 1.6.8

  • [[pcl]]

  • bug fix in pcl::PassThrough filter, where the output data size was not set correctly when keep_organized was set to false

  • bug fix in pcl::GreedyProjectionTriangulation, where point distances were incorrect, thus poroducing too many holes points during reconstruction ()

  • added more elaborate unit tests for surface reconstruction

  • Added possibility to restrict no of returned interest points to pcl::NarfKeypoint

  • Added parameter to control no of openmp threads to pcl::NarfKeypoint

  • [[pcl_ros]]

  • added the keep_organized option to pcl_ros::Passthrough (#4627)

= 0.10.0 (2010-02-25) =

  • Re-versioned 0.9.9 to follow patch-versioning semantics.

= 0.9.9 (2010-02-22) :: "Almost there" edition =

  • [[pcl]]

  • removed ConvexHull2D (API breaking change!) You need to change your code. From:

    pcl::ConvexHull2D<...> ...;
    

    to

    pcl::ConvexHull<...> ...;
    
  • added a new general purpose 2D/3D ConvexHull class based on QHull

  • added a new general purpose 2D/3D ConcaveHull class based on QHull

  • added helper transformPointCloud method for cloud+indices as input

  • fixed: segfaults when ICP finds no correspondences (#4618)

  • improved the PCD I/O capabilities (for binary PCD) to deal with SSE padding

  • Added possibility to create a RangeImagePlanar from a point cloud

  • Corrected is_dense to false in all range images.

  • Reimplemented big parts of the NARF keypoint extraction - should be more reliable now (and unfortunately slower) - uses polynomials to search maxima now.

  • Added helper classes for polynomial approximations to common

  • Added a new RANSAC-like algorithm: PROSAC (much faster when there is a confidence for the matches)

  • Fixed normalization factor in the VFH's scale component.

  • Made MLS more flexible: output has all fields the input has, only with XYZ smoothed, and normals are provided separately (and optionally)

  • Added multi-scale calculation to NARF keypoint to make it faster again, fixed a bug in BorderExtractor and fixed some issues in RangeImagePlanar.

  • Added functions in common to compute max distance from a point to a pointcloud. Fixed distance component of VFH, normalization is now also invariant to rotation about the roll axis.

  • Added pcl::PointSurfel to known point types.

  • eigen-decomposition for symmetric positive-semi-definite 3x3 matrices: 1) bug fix so eigenvects are orthogonal, 2) is more robust for degenerated cases

  • [[pcl_ros]]

    • MovingLeastSquares nodelet improvements
    • Changed serialization of point clouds to ship the data as-is over the wire, padding included (#4754). Implemented subscriber-side optimizations to minimize memcpys. Will do one memcpy for the whole point cloud if the data layout is an exact match.

= 0.9.0 (2011-02-08) :: "Darn the torpedoes" Edition =

  • [[pcl]]

  • optimizations for dense dataset (no checking for isfinite inside the loop unless is_dense is false)

  • improved eigen decomposition pcl::eigen33 by better handling degenerate cases. Normal estimation should work better now.

  • more 32bit alignment fixes

  • improved Doyxgen documentation

  • added POINT_CLOUD_REGISTER_POINT_STRUCT for Narf36

  • fixed a minor bug in poses_from_matches where a distance was not computed correctly

  • disabled TBB classes until 1.0 comes out

  • fixed a few bugs in ICP_NL registration (thanks Mike!)

  • enforced const-ness for pcl::ArrayXfMap and pcl::VectorXfMap, by creating pcl::ArrayXfMapConst and pcl::VectorXfMapConst for maps over const types

  • improved Windows support

  • fixed a bug in StatisticalOutlierRemoval where the output data array was set incorrectly.

  • fixed a boost::split bug in pcd_io, for cases when point data starts with spaces or tabs

  • finalized the surface reconstruction (pcl::SurfaceReconstruction) API

  • added a new method for surface reconstruction using grid projections (pcl::GridProjection)

  • added a new generalized field value filter (pcl::ConditionalRemoval)

  • cleaned up normal estimation through integral images

  • PCL rift.hpp and point_types.hpp fixed for Windows/VS2010

  • fixed all is_dense occurances

  • unmanged all Eigen3:: calls to Eigen::

  • changed all !isnan checks to isfinite in order to catch INF/-INF values

  • added vtk_io tools for saving VTK data from PolygonMesh structures

  • fixed SACSegmentation::SACMODEL_CIRCLE2D: to accept setRadius limits

  • use SACSegmentation when the model doesn't fit SACSegmentationFromNormals (assert fix)

  • [[pcl_ros]]

  • added a mutex for configuration parameters and filter computation, to prevent changing parameters while the algorithm is running. Alternative: copy parameters before the compute loop.

  • pcd_to_pointcloud has been updated to take a private parameter "frame_id" which is used in publishing the cloud

  • unmangled all Eigen3:: calls to Eigen::

  • Document cloud to image conversion.

  • http://www.ros.org/wiki/pcl_ros/Tutorials/CloudToImage

  • Removed the tool from pcl , and moved to pcl_ros

= 0.8.0 (2011-01-27) :: Operation CloudMan =

  • [[flann]]

  • Updated to version 1.6.7

  • [[pcl]]

  • improved doxygen documentation overall

  • added support for indices for SampleConsensus models (previously broken)

  • fixed a grave bug in ProjectInliers, where row_step was not correctly set thus leading to (de)serialization issues in the new ros::Subscriber<PointCloud > scheme

  • RangeImagePlanar can now also be created from a depth image.

  • Fixed a bug in RangeImagePlanar.

  • Fixed possible segmentation fault in RangeImageBorderExtractor.

  • RangeImage now has is_dense=false, since there typically are NANs in there.

  • Added Correspondence as a structure representing correspondences/matches (similar to OpenCV's DMatch) containing query and match indices as well as the distance between them respective points.

  • Added CorrespondenceEstimation for determining closest point correspondence, feature correspondences and reciprocal point correspondences.

  • Added CorrespondenceRejection and derivations for rejecting correspondences, e.g., based on distance, removing 1-to-n correspondences, RANSAC-based outlier removal (+transformation estimation).

  • Further splitted up registration.h and added transformation estimation classes, e.g., for estimating rigid transformation using SVD.

  • Added sensor_msgs::Image image; pcl::toROSMsg (cloud, image);See tools/convert_pcd_image.cpp for a sample.

  • Added a new point type, PointWithScale, to store the output of SIFTKeypoint detection.

  • Fixed a minor bug in the error-checking in SIFTKeypoint::setScales(...)

  • Fixed small bug in MovingLeastSquares

  • Fixed small bug in GreedyProjectionTriangulation

  • Added unit tests for RSDEstimation, MovingLeastSquares and GreedyProjectionTriangulation

  • Fixed and improved multiple point to line distance calculations, and added it to distance.h and unit testing

  • [[pcl_ros]]

  • Fixed a grave bug introduced in 0.7 where the NullFilter approach doesn't work

  • Implemented the exact time synchronizer path for PointCloudConcatenateDataSynchronizer

  • replaced subscribers/publishers in surface, segmentation, and feature nodelets from sensor_msgs::PointCloud2 to pcl::PointCloud, thus reducing complexity and getting rid of all the intermediate from/toROSMsg (de)serializations

  • moved away from storing state in the nodelets. The right way to do it is: callback -> PointCloud -> process -> publish, without storing the PointCloud as a member into the base class

  • introduced a mandatory emptyPublish which nodelets should use to send an empty result with the header frame_id and stamp equal with the input's.

= 0.7.1 (2011-01-09) =

  • [[flann]]

  • Makefile patch for LateX PDF build

  • [[pcl]]

  • fixed roslib/Header warnings by enabling std_msgs/Header if ROS_VERSION > 1.3.0

  • working on PCL port to Android - mainly CMake scripts, and a few #if ANDROID statements.

  • _pcl::VoxelGrid _now saves and gives access to the created voxel grid, with lots of helper functions

  • [[pcl_ros]]

  • dropped pcl_ros::Subscriber due to the changes introduced in 0.7

= 0.7 (2010-12-20) :: Special "FLANN is awesome" Edition =

  • [[eigen3]]

  • removing eigen3 as it's already present in geometry (unstable) and geometry_experimental (cturtle)

  • [[pcl]]

  • added general purpose getMinMax() method for n-dimensional histograms (r34915).

  • more 32bit architecture alignment fixes

  • fixing \r string splitting issues on Windows systems for PCDReader (r34830).

  • Added new range image class RangeImagePlanar. E.g., for direct usage with kinect disparity images.

  • added the option to keep the data organized after filtering with pcl::PassThrough and replace values in place rather than removing them (r34776)

  • removed boost::fusion macros

  • moving the Feature estimation architecture to the new PCL structure

  • moved sorted_ as a parameter in the KdTree base class

  • added PCL_INSTANTIATE macros that easily instantiate a specific method/class given a set of point types

  • Better checks for is_dense plus bugfixes.

  • Added RSD (Radius Signature Descriptor) feature.

  • Updated the entire PCL tree to use FLANN as a default KdTree

  • Optimized the KdTreeFLANN::radiusSearch so it doesn't do any memory allocation if the indices and distances vectors passed in are pre-allocated to the point cloud size.

  • Changed KdTree::radiusSearch method signature to return int (number of neighbors found) intead of bool

  • added new pcl::View class that holds a PointCloud, an Image, and a CameraInfo message (r34575)

  • Moving the Surface reconstruction framework to the new structure (r34547)

  • Moving the Segmentation framework to the new structure (r34546)

  • Creating a libpcl_sample_consensus and moving the SAC methods and estimators to the new structure (r34545)

  • moving Filter to the new structure (r34544)

  • Creating a libpcl_kdtree and moving the KdTree to the new structure (r34543)

  • added support for PCD v0.7 files (+VIEWPOINT), minor API changes and consolidations

  • adding a sensor origin + orientation for pcl::PointCloud (r34519)

  • Improved Boost CMake macros

  • [[pcl_ros]]

  • greatly simplified the PointCloudConcatenateDataSynchronizer nodelet (r34900).

  • pcl::PointCloud works natively with ros::Publisher and ros::Subscriber. Separate pcl_ros::Publisher and pcl_ros::Subscriber are no longer necessary.

  • [[flann]]

  • Updated to FLANN 1.6.6

  • Updated to FLANN 1.6.5.

  • Updated to FLANN 1.6.4

  • Updated to FLANN 1.6.3

  • [[pcl_tf]]

  • Moved functionality to pcl_ros. RIP.

  • [[ann]]

  • Removed ANN as a dependency. RIP.

  • [[cminpack]]

  • updated cminpack to 1.1.1

  • fedora rosdep fixes

  • Patch for building and installing cminpack correctly (thanks to Nicholas)

= 0.6 (2010-12-01) :: Special "Happy Birthday Bastian!" Edition =

  • [[pcl]]

  • sort clusters in ExtractEuclideanCluster in descending order before returning (r34402)

  • VoxelGrid patch for handling RGB colors (thanks to Yohei)

  • lots of work to make PCL compile in Windows (thanks to Stefan)

  • bumping up the ASCII precision from 5 to 7 on PCD I/O operations

  • moved PCL to the new stack: perception_pcl

  • header include cleanup for faster compiles

  • switched StatisticalOutlierRemoval to FLANN

  • Updated kdtree_flann to use the C++ bindings from FLANN

  • [[pcl_tutorials]]

  • Added new tutorial to visualize the NARF descriptor and look at descriptor distances.

  • fixed a problem that was preventing tutorials to be built on MacOS (r34287)

  • [[flann]]

  • Updated to latest version of FLANN (1.6.2)

  • Updated to latest version of FLANN (1.6.1)

  • Updated to latest version of FLANN (1.6)

  • [[cminpack]]

  • Updated cminpack to version 1.1.0

  • Updated cminpack to version 1.0.90 (r34303)

  • Fixed a portability issue where the library was copied as .so, while MacOS systems need .dylib, thanks to Cyril per #4596 (r34286)

  • [[pcl_ros]]

  • Fixed a portability issue where uint was used without being defined on MacOS systems, thanks to Cyril per #4596 (r34285)

  • Patch for ConvexHull2D to publish PolygonStamped (thanks to Ryohei)

= 0.5 (2010-11-25) :: Special Thanksgiving Edition =

  • [[pcl]]

  • got rid of ROS_ASSERTS in favor of ROS_ERROR + return (true/false)

  • improvements on CMake build scripts and custom-made ROS header includes to make PCL work outside of ROS

  • pow() fixes for macos (#4568)

  • fixing a bug where if inliers are always 0, we get stuck in an (almost) infinite loop in Sample Consensus methods (r34191)

  • SegmentDifferences is now optimized for null targets (simply copy the input)

  • Changed operator(x,y) in PointCloud to return references, both const and non-const.

  • set the default behavior of ExtractIndices to return the complete point cloud if the input indices are empty and negative is set to true

  • resize the clusters to 0 on failed initCompute for EuclideanClusterExtraction

  • fixed all filtering methods to return empty datasets (points/data.size () = 0, width = height = 0) on failures and make sure that header (+fields for PointCloud2) are always copied correctly in all cases (r34099).

  • make sure header and fields are copied even if the indices are invalid for ProjectInliers

  • fixed error message typo in conversions.h

  • fixing the meaning of is_dense per #4446

  • [[pcl_ros]]

  • added min_inliers as the minimum number of inliers a model must have in order to be considered valid for SACSegmentation

  • added a SegmentDifferences nodelet

  • fixed a grave bug in Filter where the output TF frame was not always set correctly (r34159)

  • added eps_angle as a dynamic reconfigure parameter to SACSegmentation

  • added error output on invalid input for all nodelets

  • Allow pcl_ros publishers to take shared pointers - fix #4574 (r34025)

  • change the default behavior of isValid from data.empty || width * height = 0, to width * height * step = data.size for PCLNodelet

  • make sure data is published even if empty

  • making sure publishing with a rate of 0 works for pcd_to_pointcloud

  • added TF transform capabilities to bag_to_pcd

  • Enforce that the TF frame and the timestamp are copied before publish

  • [[pcl_tf]]

  • got rid of unneeded transforms (if target_frame = input.header.frame_id, simply return the input)

  • [[pcl_visualization]]

  • removed -DNDEBUG which was causing some weird random bugs (-cam "..." was one of them) (r33984)

  • fixing an issue with VTK exports which lead to 3rd party user packages needing to know and explicitly import VTK (r33980)

= 0.4.2 (2010-11-01) =

  • [[pcl_ros]]

  • removed the vtk metapackage (replaced with findVTK macros)

= 0.4.1 (2010-11-01) =

  • [[pcl_ros]]

  • adding EIGEN_MAKE_ALIGNED_OPERATOR_NEW to all nodelets to make sure that we are aligned on 32bit architectures

= 0.4.0 (2010-10-30) =

  • [[pcl]]

  • changes needed for the eigen3 upgrade (Eigen3::Transform3f -> Eigen3::Affine3f)

  • Added SIFTKeypoint keypoint detection algorithm for use on xyz+intensity point clouds.

  • Fixed a bug in Registration::FeatureContainer::isValid () where the source and target feature clouds were required (incorrectly) to be the same size

  • added NARF features (descriptor) to features, updated NARF keypoints, made RangeImageBorderExtractor derived from Feature.

  • added set/get AllFields/AllData helper methods to ProjectInliers (r33284)

  • added helper bool getFilterLimitsNegative () to Filter (r33283)

  • added helper set/get LocatorType methods to EuclideanClusterExtraction (r33282)

  • added missing dependency on roscpp

  • added a check for insufficient number of points when fitting SAC models to avoid infinite loops (r33195)

  • added makeShared () to pcl::PointCloud, and fixed the Ptr/ConstPtr typedefs (r33147)

  • fixed a bug in SamplesConsensusModelRegistration where the sample selection could get stuck in an infinite loop

  • added line to line intersection routines (r33052)

  • corrected transformPointCloudWithNormals in transforms.hpp to properly handle normals

  • [[pcl_ros]]

  • switching rosrecord API to rosbag (r33602)

  • Complete refactorization of PCL_ROS nodelets. Got rid of multiple inheritance in favor of delegation. Using template specializations now the code compiles faster and GCC processes occupy less RAM during compilation.

  • new PointCloudConcatenateDataSynchronizer nodelet: concatenates PointCloud2 messages from different topics (r33241)

  • added a max_clusters int_t nodelet option to EuclideanClusterExtraction: The maximum number of clusters to extract (r33258)

  • added a filter_limit_negative bool_t nodelet option to Filter. Set to true if we want to return the data outside [filter_limit_min; filter_limit_max] (r33257).

  • fixed bug in PassThrough nodelet that prevented dynamic reconfigure from setting the FilterFieldName properly.

  • added implementation of MovingLeastSquares nodelet

  • [[eigen3]]

  • upgraded to eigen3-beta2

  • [[cminpack]]

  • Updated cminpack to 1.0.4 (r33419)

  • [[pcl_tutorials]]

  • added/updated tutorials for range image creation, range image visualization, range image border extraction, NARF keypoint extraction and NARF descriptor calculation.

  • [[point_cloud_converter]]

  • added missing dependency on roscpp (r33045)

= 0.3.3 (2010-10-04) =

  • [[pcl]]

  • fixed a bug in ProjectInliers for 32bit architectures (r33035)

  • Added unit tests for computeCovarianceMatrixNormalized (r32985)

  • making PointCloud2/PointCloud typedefs consistent (r32989)

  • field W is no longer needed in pcl::PointNormal, as it was only used for SSE optimization. Same for pcl::PointXYZW, which means we can remove them (r32960)

  • [[point_cloud_converter]]

  • Changed INFO to DEBUG on print statements

  • [[pcl_tf]]

  • added code to check for TF exceptions

  • [[pcl_ros]]

  • changes to consistently output/publish empty datasets + made sure all inputs are checked for validity in each nodelet before processing (r32993)

  • added dynamic_reconfigure TF input_frame/output_frame parameters to pcl_ros::Filter (r32990)

= 0.3.2 (2010-09-28) =

  • [[pcl]]

  • fixed a bug in ExtractPolygonalPrismData that got introduces in r31172

  • added new 3D Keypoint extraction base class and the NARF interest points.

  • added iterator, const_iterator, begin, end for pcl::PointCloud

  • fixed bug in radius search of kdtree, that occurred when indices_ is used

  • renamed sac_model_oriented_[line|plane] to *parallel_line and *perpendicular_plane to be more clear

  • added a new SAC model for finding planes that are parallel to a given axis. Useful for e.g. finding vertical planes (parallel to "up" axis)

  • added a smarter sample selection to SampleConsensusModelRegistration that uses a minimum distance between samples based on the principal directions/magnitudes of the dataset (r32873)

  • enabling SSE optimizations by default for all libraries

  • normalized covariance matrix estimation pcl::computeCovarianceMatrixNormalized (r32872)

  • Added methods to pcl::Registration for finding point correspondences between pairs of feature clouds

  • fixed a bug in pcl::SampleConsensusModelLine where the number of iterations was not correctly checked/set in getSamples

  • Added missing includes in icp_nl.h

  • [[pcl_ros]]

  • changing the order of the members to allow nodelet unloading (r32941)

  • enabling SSE optimizations for all libraries (r32909)

  • enabling parameter setting via dynamic_reconfigure for StatisticalOutlierRemoval filter (r32876)

  • set the default number of '''gcc''' build processes to 2, to avoid large memory allocation thus rendering the machine unusable during compilation

  • [[flann]]

  • upgraded to FLANN 1.5

= 0.3.1 (2010-09-20) =

  • [[pcl]]

  • fixed a grave bug that caused 32bit architectures to exhibit errors regarding eigen alignment issues (32769)

  • fixed a bug with isBoundaryPoint() and pcl::BoundaryEstimation::computeFeature() relating to the use of different surface_ and input_clouds and/or non-default indices_, and added a new way of calling isBoundaryPoint()

  • fixed the problem of mls.hpp depending on eigen 2

  • added computeCovarianceMatrixNormalized methods <<Ticket(ros-pkg 4171)>>

  • Renamed namespace of eigen3 to Eigen3 to prevent conflicts with eigen2

  • [[laser_scan_geometry]]

  • fixed #4437 where PointCloud2 projectLaser fails when laser intensity is off (32765)

  • [[pcl_tf]]

  • fixed #4414 where lookupTransform was being called without tf_listener object (32764)

= 0.3.0 (2010-09-05) =

  • [[pcl]]

  • VectorAverage uses fast spezialized PCA method now for dimension=3

  • new version of VoxelGrid filter that downsamples data faster and doesn't use as much memory (the previous version could easily cause std::bad_alloc exceptions)

  • fixed a major bug where: 1) PointCloud2 structures were being sent with extra unusued (padding) data (i.e., directly copied from pcl::PointCloud); 2) writing PCD files in binary mode resulted in corrupted files (r32303)

  • fixed a bug where the fields were not cleared on pcl::getFields (pcl::PointCloud) (r32293)

  • fixed a bug where PCDWriter was writing to the same file instead of creating new ones

  • improvements in non linear registration (reduced the number of estimates from 7 to 6, SSE optimizations, etc)

  • reduced the number of operations for covariance matrix estimation from 18N to 10N+3 => big speedup increase together with pcl::eigen33 !

  • Big patch for PCL to work with Eigen3. Fixed alignment issues for Sample Consensus models. Added a faster (experimental) eigen decomposition method from Eigen3 trunk.

  • fixed a bug where the axis_ was not properly aligned on 32bit architectures thus causing an assert

  • Added ICP functionalities to range image, add new helper functions header file_io, minor fixes.

  • [[pcl_tf]]

  • Added function to just transform a point cloud using an Eigen matrix.

  • [[pcl_visualization]]

  • fixed a bug where NaN values invalidated certain color handlers

  • Range image example uses viewpoint information from file now

  • [[eigen3]]

  • added Eigen3 library

= 0.2.9 (2010-08-27) =

  • [[pcl]]

  • added copyPointCloud (sensor_msgs::PointCloud2 &in, indices, sensor_msgs::PointCloud2 &out)

  • [[pcl_visualization]]

  • fixed a bug in PCLVisualizer where NULL scalars were causing a segfault

  • [[pcl_ros]]

  • changed the default max queue size to 3

  • added implementations for approximate time synchronizers

  • ExtractPolygonalPrismData can now rotate the hull in the point cloud frame using TF

  • [[pcl_tf]]

  • transforms.cpp add missing out.header.frame_id = target_frame;

  • [[laser_scan_geometry]]

  • Max range handling from Hokuyo should work properly now.

= 0.2.8 (2010-08-24) =

  • [[pcl]]

  • added a Huber kernel to LM optimization

  • fixed a bug where the order of setInputCloud and setPointRepresentation triggered an assert condition (31938)

  • added a Sample Consensus model for outlier rejection during ICP-like registration

  • added implementations of the intensity-domain spin image and Rotation Invariant Feature Transform, as described in "A sparse texture representation using local affine regions," by Lazebnik et al. (PAMI 2005).

  • added a templated N-dimensional histogram point type

  • Updated code to work with point wrapper registration.

  • [[pcl_visualization]]

  • scale the colors lookup table to the color handler limits (31942)

  • fixed a bug where the screenshot function was not using the last updated screen

  • addPointCloudPrincipalCurvatures renders the principal curvatures of a point cloud on screen + set color properties

= 0.2.7 (2010-08-19) =

  • [[pcl]]

  • SSE improvements in *PFHEstimation

  • more SSE improvements in computeCentroid, computeCovarianceMatrix, etc

  • several improvements regarding SSE optimizations in SACModelPlane

  • make all data points that deal with XYZ SSE aligned (31780)

  • work on improving the VFH signatures (changed the signature from 263 to 308, added distances)

  • [[pcl_visualization]]

  • fixed a silly X11 include bug, where Bool was defined globally therefore breaking BOOST_FOREACH

  • added font size property setter for text actors

  • new PCLVisualizer::addText method for adding texts to screen (r31822)

  • implemented window position ordering and range min/max for PCLHistogramVisualizer (r31805)

  • implemented a basic histogram visualizer for multi-dimensional count=1 features (e.g., VFH)

  • [[pcl_tutorials]]

  • added a tutorial for online PointCloud2 message viewing

= 0.2.6 (2010-08-17) =

  • [[pcl]]

  • fixed another bug where internal matrices were not always cleared in FPFH/VFH estimation

  • fixed a few bugs regarding sizeof(PointT)=16 for demeanPointCloud (r31762)

  • fixed a grave bun in FPFHOMP/VFH estimation where the matrices were not correctly initialized on consequent runs (r31757)

  • introduced PCL exceptions, and added checks for conversion and cloud.at (u, v) operations

  • [[pcl_ros]]

  • fixed a bug where SACSegmentation was accepting two parameters with the same name (distance_threshold vs model_threshold)

  • disable empty data publishing in PCDReader

  • [[pcl_visualization]]

  • added lookup table to the renderer

  • [[cminpack]]

  • enabled shared library by default

= 0.2.5 (2010-08-14) =

  • [[pcl]]

  • added a convenience method in pcl::io::loadPCDFileHeader for loading up files fast and checking their fields, data types and sizes, etc (r31740)

  • added implementation of the Sample Consensus Initial Alignment (SAC-IA) algorithm described in "Fast Point Feature Histograms (FPFH) for 3D Registration," Rusu et al., (ICRA 2009)

  • moved the estimateRigidTransformationSVD methods out of IterativeClosestPoint and made them static functions defined in registration.h.

  • [[pcl_visualization]]

  • fixed the minmax color handler visualization (r31710)

  • few addPointCloud convenience method for geometry handlers + bugfix in pcd_viewer (the last field was added twice as a color handler when more than one geometry handler was available)

  • addPointCloudNormals displays surface normals on screen (PCLVisualizer)

  • added -normals and -normals_scale options to pcd_viewer

  • [[pcl_ros]]

  • grave bug fixed in PointCloudConcatenateFieldsSynchronizer where data was not correctly copied (r31708)

  • PCDReader nodelet uses a latched topic to publish data

= 0.2.4 (2010-08-10) =

  • [[pcl]]

  • pairwise registration tutorial now has on-screen visualization (r31698)

  • new laser scan geometry for LaserScan->PointCloud2 conversion

  • added general purpose transformation routines for sensor_msgs::PointCloud2 using TF (r31678)

  • added basic shapes (line, sphere, cylinder, polygon, etc), split rendering properties for shapes and point clouds (r31665)

  • fixed a few bugs regarding fields.count

  • small bug fix where triples on the parameter server didn't match with the value in dynamic reconfigure (VoxelGrid)

  • added a convenience method for conversion from PointCloud2 (xyz) to Eigen format

  • added a method to get the radius of a circumscribed circle

  • added methods for saving/loading camera parameters in PCL Visualizer (c, j, -cam)

  • added -multiview to pcd_viewer; PCLVisualizer can now create and use different viewports (r31600)

= 0.2.3 (2010-08-02) =

  • [[pcl]]

  • tiny bug fixes for cturtle

= 0.2.2 (2010-08-02) =

  • [[pcl]]

  • work to improve the PCL Visualization package, standalone pcd_visualizer completed

  • added cloud(u,v) accessors for organized datasets.

  • added constrained implementations (maximum distance between correspondences) for registration methods

  • added PCD file format v.6 which includes field count (r31367)

  • added a PointRepresentation class for spatial search methods that allows the user to specify what dimensions are important

  • added a maximum distance threshold for point correspondences in registration

  • fixed IterativeClosestPoint registration, added a L1 kernel to icp_nl, and unit tests (31283)

  • added a new segmentation method for obtaining the differences between two point clouds (r31281)

= 0.2.1 (2010-07-25) =

  • [[pcl]]

  • test build fixes

  • made BAGReader inherit directly from nodelet::Nodelet

= 0.2.0 (2010-07-24) =

  • [[pcl]]

  • added PCL_ROS package for all PCL_ROS interactions. Moved all nodelets and all ROS specific interfaces from PCL to PCL_ROS.

  • moved vectorAverage to common

  • added class to calculate transformation from corresponding points to common

  • added index remapping for the kdtrees so that NAN points are automatically ignored.

  • Split kdtrees in h and hpps

  • Added some new functionality to range images and fixed bugs in creation process.

  • separated the PCL tutorials from the library into a separate package (pcl_tutorials)

  • added a PCD Visualizer and implemented several point cloud visualization techniques (handlers) for geometry and color

  • added getFieldsList (PCL/io) for a pcl::PointCloud

  • added VTK as a dependency for visualization in an attempt to be compatible with older Linux distributions

= 0.1.9 (2010-06-28) =

  • [[pcl]]

  • added radius search for organized_data

  • fixed some bugs in the range image class

  • structural changes in range images related classes

  • split point_types.h in point_types.h and point_types.hpp and added output operators for all points

  • Added angles.h, norms.h, time.h, transform.h and common_headers.h (which includes all the others) to provide some basic functions for convenience.

  • Added border_extraction for range images

  • Added vectorAverage to features, which calculates mean and covariance matrix incrementally without storing added points.

  • Added macros.h, which provides some basic macros for convenience.

  • added an initial PCL Visualizer implementation

  • replaced rosrecord with rosbag for all PCL tools

  • added cluster_max size to Euclidean Clustering (r30210)

  • adding cylinder direction axis constraints + sphere min/max radius constraints for model segmentation (r30195)

  • fixed a grave bug which was preventing objects containing ANN trees to be used in multiple threads (r30194)

  • fixed a bug where patches for ANN 1.1.2 were not applied correctly (r30193)

  • all nodelets now have use_indices and max_queue_size checked at startup via PCLNodelet base class

  • switched from rosrecord to the rosbag API for BAGReader

  • enable max_queue_size parameter (default to 1) for all nodelets

  • set the indices/distances to 0 if the search failed for ANN trees (r30130)

  • PCD reader fix when an invalid file name is given (r30004)

  • added range image class (r29971)

  • added axis and eps_angle as ROS parameters for segmentation (r29938)

  • setting the maximum number of clusters that are to be published via max_clusters (r29935)

  • fixed a grave bug where SACModelOrientedPlane was performing an infinite loop inside selectWithinDistance (r29934)

= 0.1.8 (2010-05-24) =

  • [[pcl]]

  • simplified API in point_cloud_converter

  • new general purpose 3D point in 2D polygon check method (r29645)

  • added a getPointsInBox() method

  • fixed a bug where some points were not considered as being part of the polygonal prism due to polygonal bound checking errors (r29597)

  • added a PointXYZI type to represent intensity data.

  • updating flann to 1.5_pre2 (r29549)

  • added MovingLeastSquares tutorial (r29539)

= 0.1.7 (2010-05-13) =

  • [[pcl]]

  • getMinMax3D methods now reside in common

  • fixed a major bug in VoxelGrid where centroids were not initialized correctly

  • fixed a bug where NdCopyEigenPointFunctor and NdCopyPointEigenFunctor were given incorrect values

  • reorganizing the library for faster compilation

  • fix for linker errors when linking together multiple translation units that include the same point type registration. Static tag name strings would be defined more than once. #4071 (r29435)

= 0.1.6 (2010-05-01) =

  • [[pcl]]

  • added parameter sets for Filter, VoxelGrid, ExtractIndices (r29229)

  • refactorized the dynamic reconfigure parametrization for nodelets

  • added tutorial/sample for cylinder segmentation via SAC (r29224)

  • introduced bad model handling in the sample consensus loop (r29218)

  • added the field name to the dynamic reconfigure parameter list for Filters (r29211)

  • improved NaN handling in kdtrees (r29204)

  • fixed a bug where PassThrough was incorrectly declared as a nodelet (r29202)

  • added tutorial/sample for the ConvexHull2D algorithm (r29198)

  • added set/get radius limits (to be extended), implemented radius limits for cylinder models, and better RANSAC model checking (r29194)

  • default input topic on pointcloud_to_pcd changed from "tilt_laser_cloud" to "input" (r29137)

  • CMakeLists split into individual files for better organization

  • fixed a couple of boost linking errors on some machines

  • transformed point_cloud_converter from a node into a library+node so it can be reused elsewhere

  • added a command line tool for concatenating PCD files (r29035)

  • added an implementation of the Viewpoint Feature Histogram (VFH) (r29032)

= 0.1.5 (2010-04-22) =

  • [[pcl]]

  • added tutorial/sample for the ExtractIndices filter (r28990)

  • added tutorial/sample for the VoxelGrid filter (r28986)

  • added tutorial/sample for the StatisticalOutlierRemoval filter (r28985)

  • added templated version of loadPCDFile (r28978) and PCDReader::read

  • added unit tests for the StatisticalOutlierRemoval filter (r28971)

  • added unit tests for the ProjectInliers filter (r28970)

  • added unit tests for the VoxelGrid filter (r28969)

  • added negative limits and fixed a bug where the centroid was not correctly initialized in VoxelGrid (r28967)

  • added unit tests for the PassThrough filter (r28966)

  • added unit tests for PCL filters: RadiusOutlierRemoval and ExtractIndices (r28961)

  • added extra unit tests for concatenate points/fields (r28959)

  • added tutorial/sample code for BAGReader (r28955)

= 0.1.4 (2010-04-21) =

  • [[pcl]]

  • added a few nodelet tests (talker, listener, pingpong)

  • added unit tests for BAGReader (r28924)

  • added an implementation for BAGReader (r28922)

  • added unit tests for PCDReader/PCDWriter

  • fixed a few bugs where certain include headers were not exposed properly

  • added an implementation of a statistical outlier removal filter (r28886)

  • added an implementation of radius outlier removal filter (r28878)

  • fixed a bug (#3378) where the output file name in several tools was incorrect (r28858)

  • added an implementation of non-linear ICP (28854)

  • fixed a bug where the min/max estimation was incorrect for sensor_msgs/PointCloud2 types in voxel_grid (r28846)

= 0.1.3 (2010-04-15) =

  • [[pcl]]

  • final transformation is now accessible after registration (r28797)

  • added filter_limit_negative to switch between inside/outside interval (r28796)

  • added =/+= operator to PointCloud (r28795)

  • added a set of tutorials for IO, Segmentation, Filtering

  • fixed a very ugly bug regarding covariance matrices having 0 elements (r28789)

  • IterativeClosestPoint tested and works (r28706)

  • added set/get maximum distance for organized data index and virtualized the ANN radiusSeach (r28705)

  • added demeaning methods to registration/transforms.h

  • updated ANN library to 0.1.2

  • fixed a major bug in ExtractPolygonalPrismData where distances were incorrectly checked

  • added plane normal flip via viewpoint to ExtractPolygonalPrismData

  • added axis and eps_angle (set/get) to SACSegmentationFromNormals

  • added methods for transforming PointCloud datasets containing normals

= 0.1.2 (2010-04-05) =

  • [[pcl]]

  • added a first draft implementation of tf_pcl

  • added missing descriptions to PCL nodelets list

  • fixed a few bugs in voxel_grid where the offset was not incremented correctly (r28559)

  • added a new passthrough filter

  • fixed a few bugs regarding Eigen/Boost's make_shared (r28551)

  • solved numerical instability issue in MLS, preferred the solution using doubles over the one using SVD (r28502). MLS works great now.

  • simplified the registration API (still in progress)

  • changed point_cloud_converter to accept 2 input and 2 output topics, thus fixing the "subscribe when nothing is being published" problem (r28540)

= 0.1.1 (2010-03-29) =

  • [[pcl]]

  • fixed a very nasty bug in VoxelGrid where leaves were not properly initialized (r28404)

  • fixed extractEuclideanClusters (was not working correctly)

  • added min/max height for polygonal prism data extraction (r28325)

  • added a bag_to_pcd tool that doesn't need ROS to convert BAG files to PCD (r28469)

  • fixed a bug where PCL classes couldn't be used unless ros::init was called first

  • classes now inherit from PCLNodelet and PCLBase

  • setInputCloud and setIndices are defined in PCLBase now, and used everywhere

  • fixed a bug where kdtree structures were incorrectly created

  • added get/set helper methods everywhere

  • PointCloud2 structures can now contain NaN (Not a Number) values, so extra checks have been added in a lot of places

  • added new distance methods to sample consensus models

  • removed ROS_DEBUG statements in the feature constructors

  • indices can now be set via pcl::PointIndices too

  • fixed a bug where the surface was not getting re-initialized in Feature (r28405)

  • added explicit checks for invalid covariance matrices (r28306)

  • fixed a bug where indices are created to mimic the input point cloud data but never reset (r28306)

  • cleaned the C++ API documentation

  • initial MLS port (untested)

= 0.1.0 (2010-03-10) =

  • [[pcl]]

  • initial release