This is a plug-in to ViTables to view data sets as images. It uses PyQtGraph to visualize the data sets in the workspace. Valid viewable data sets include 2D arrays (H,W) as monochrome images, 3D arrays as RGB(A) (H,W,3) with or without the alpha channel, 3D arrays as N monochrome images (N,H,W), and 4D arrays as N RGB(A) images (N,H,W,3). A long term goal is to answer the problems addressed in Issue #11.
This plug-in is designed to work with the development branch of ViTables
(2.2a1), and, therefore, uses setuptools to declare the entry points
into vitables.plugins.
First, ensure that the correct version of ViTables and PyQtGraph are installed. The minimum version of ViTables necessary is (2.2a1) which is currently the development branch. The lowest know version of PyQtGraph that has been tested is 0.9.10. Once those dependencies have been installed, run:
$ python setup.py install
After running the install, launch ViTables, open the Preferences menu, select the Plugins tab, and enable Image Viewer. Once you restart ViTables, you can right click on a data set and view it as an image.
This project uses Sphinx to generate the typeset documentation. To build the HTML documentation along with the API, simply run:
$ sphinx-apidoc -o . vtimshow -M $ make html
If you are on Windows, replace the last command with make.bat.
The default ordering of the arrays in the file is assumed to be (Depth, Width, Height, RGB(A)). To change this order, go to Preferences -> image_viewer and select the proper order. To have this persist, simply hit save before exiting the preferences menu.
Two different methods are provided to do math operations on the data sets. The first performs operations within a single data set. The second performs operations on different but compatible data sets.
A panel to perform math on selected frames is under the menu button at the lower right of the image area. Toggling this on will disable the original frame selector used by the image area and the normalization panel. Instead, three frame selectors are placed in the time series plot. These correspond to R, G, and B channels. In the panel, the user may select to display these frames as a RGB image, subtracted monochrome image, or divided monochrome image.
The cube math window allows for comparing data from different data sets. After selecting up to three data sets, you can right click and select “Compare Datasets” to bring up the window. The red, green, and blue channels each have a panel at the bottom where you can select the appropriate data set for each color. You can then use the slider or the spin box to select the particular frame for processing. The available operations are listed in the panel to the right of the color assignment panels.
The left hand drop down menu is automatically populated with the name of all of the leaf data sets that have been exposed in the database tree viewer. In order for a particular dataset to appear in this list, you must have expanded the tree viewer to expose the desired leaf node.
In addition to performing mathematical operations on the monochrome
frames from a data cube, the cube math panel provides a way to apply
filters to a full data cube. On the far right of each color selector
row is a drop down menu to select a filter. The selected filter is then
applied to the data cube to reduce it to a monochrome image. Included
with this package are filters to apply a scaled human eye response for
the red, green, and blue colors. These filters assume the data cube
maps on the full visible band and the response is scaled from 0 to 1.
Additional filters can be added by defining an entry point in your
setup.py to the vtimshow.filters group. The only requirements
are that the name of the filter must be unique, the target object must
have a name attribute that is used in the drop down menu, and the
target object must have a method compute(array) that accepts a
(N,H,W) NumPy array and reduces it to a (H,W) NumPy array.
As noted in Issue #33, the current development branch of ViTables requires setuptools-git to work. If you need to install ViTables from an archived download, please refer to this patch to aid in installation.
The branching scheme is that master is the development branch and stable is only for bug fixes and tagging releases.