Authors: Alexander Vakhitov, based on the ORB-SLAM2 system by Raul Mur-Artal, Juan D. Tardos, J. M. M. Montiel and Dorian Galvez-Lopez (DBoW2)
03.06.2019 Release of the LLD-SLAM source code
LLD-SLAM is a real-time point+line Stereo SLAM library based on a well known ORB-SLAM2 system. It computes the camera trajectory and a sparse 3D reconstruction consisting of points and lines.
[Point+Line LLD SLAM] A. Vakhitov and V. Lempitsky. Learnable Line Segment Descriptor for Visual SLAM. IEEE Access, 2019.
LLD-SLAM is released under a GPLv3 license. For a list of all code/library dependencies (and associated licenses), please see Dependencies.md.
If you use LLD-SLAM in an academic work, please cite:
@article{murTRO2015,
title={Learnable Line Segment Descriptor for Visual SLAM},
author={Vakhitov, Alexander and Lempitsky, Victor},
journal={IEEE Access},
year={2019}
}
We have tested the library in Ubuntu 12.04, 14.04 and 16.04, but it should be easy to compile in other platforms. A powerful computer (e.g. i7) will ensure real-time performance and provide more stable and accurate results.
We use the new thread and chrono functionalities of C++11.
We use LBDMOD for line detection and description. Please download, make, install and insert path to the CMakelists.txt.
We use Pangolin for visualization and user interface. Dowload and install instructions can be found at: https://github.com/stevenlovegrove/Pangolin.
We use OpenCV to manipulate images and features. Dowload and install instructions can be found at: http://opencv.org. Required at leat 2.4.3. Tested with OpenCV 2.4.11 and OpenCV 3.2.
Required by g2o (see below). Download and install instructions can be found at: http://eigen.tuxfamily.org. Required at least 3.1.0.
We use modified versions of the DBoW2 library to perform place recognition and g2o library to perform non-linear optimizations. Both modified libraries (which are BSD) are included in the Thirdparty folder.
We provide some examples to process the live input of a monocular, stereo or RGB-D camera using ROS. Building these examples is optional. In case you want to use ROS, a version Hydro or newer is needed.
Clone the repository:
git clone git@github.com:alexandervakhitov/lld-slam.git LLD_SLAM
We provide a script build.sh to build the Thirdparty libraries and LLD-SLAM. Please make sure you have installed all required dependencies (see section 2).
Please modify the CMakelists.txt: insert a correct path to the LBDMOD library instead of . Then execute:
cd LLD_SLAM
chmod +x build.sh
./build.sh
This will create libLLD_SLAM.so at lib folder and the executables stereo_kitti and stereo_euroc in Examples folder.
-
Download the KITTI dataset (grayscale images) from http://www.cvlibs.net/datasets/kitti/eval_odometry.php
-
Download a sequence (ASL format) from http://projects.asl.ethz.ch/datasets/doku.php?id=kmavvisualinertialdatasets
-
Download the precomputed line detections and LBD descriptors from https://yadi.sk/d/D5QEuced7y5I1w. Modify the
KITTIX_Y.yaml, 'EuRoC_Y.yaml' files to include the paths to the downloaded and unpacked dataset. -
Execute one of the following commands depending on the dataset you are going to use. Change
KITTIX_Y.yamlto KITTI00-02_Y.yaml, KITTI03_Y.yaml or KITTI04-12_Y.yaml for Y = {LBD,Empty} for sequence 0 to 2, 3, and 4 to 12 respectively. ChangePATH_TO_DATASET_FOLDERto the uncompressed dataset folder. ChangeSEQUENCE_NUMBERto 00, 01, 02,.., 11. Substitute $SSS with SEQUENCE_NUMBER in the *.yaml file.
./Examples/Stereo/stereo_kitti Vocabulary/ORBvoc.txt Examples/Stereo/KITTIX.yaml PATH_TO_DATASET_FOLDER/dataset/sequences/SEQUENCE_NUMBER
./Examples/Stereo/stereo_euroc Vocabulary/ORBvoc.txt Examples/Stereo/EuRoC_Y.yaml PATH_TO_SEQUENCE/mav0/cam0/data PATH_TO_SEQUENCE/mav0/cam1/data Examples/Stereo/EuRoC_TimeStamps/SEQUENCE.txt
./Examples/Stereo/stereo_euroc Vocabulary/ORBvoc.txt Examples/Stereo/EuRoC_Y.yaml PATH_TO_SEQUENCE/cam0/data PATH_TO_SEQUENCE/cam1/data Examples/Stereo/EuRoC_TimeStamps/SEQUENCE.txt
| No. in combined | No. in KITTI | No. in EuRoC |
|---|---|---|
| 0 | 0 | |
| 1 | 1 | |
| 2 | 2 | |
| 3 | 3 | |
| 4 | 4 | |
| 5 | 0 | |
| 6 | 1 | |
| 7 | 3 | |
| 8 | 5 | |
| 9 | 7 | |
| 10 | 7 | |
| 11 | 9 | |
| 12 | 6 | |
| 13 | 5 | |
| 14 | 8 | |
| 15 | 9 | |
| 16 | 10 | |
| 17 | 2 | |
| 18 | 4 | |
| 19 | 6 | |
| 20 | 8 | |
| 21 | 10 |
You will need to create a settings file with the calibration of your camera. See the settings file provided for the KITTI and EuRoC datasets for stereo cameras. We use the calibration model of OpenCV. See the examples to learn how to create a program that makes use of the LLD-SLAM library and how to pass images to the SLAM system. Stereo input must be synchronized and rectified. Use the LBD descriptor and detector available in the LBDMOD library, as we show in the lld-public.
You can change between the SLAM and Localization mode using the GUI of the map viewer.
This is the default mode. The system runs in parallal three threads: Tracking, Local Mapping and Loop Closing. The system localizes the camera, builds new map and tries to close loops.
This mode can be used when you have a good map of your working area. In this mode the Local Mapping and Loop Closing are deactivated. The system localizes the camera in the map (which is no longer updated), using relocalization if needed.