Switch branches/tags
Find file History
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
..
Failed to load latest commit information.
build
camera
disparity
microlens
plenopticIO
rendering
samples
.DS_Store
README.md
plenopticGUI.py
setup.py

README.md

PythonLibrary v1

This folder contains the python library to work with Plenoptic Images.

It contains the following folders:

  • plenopticIO: Input and Output management for such images
  • microlens: The object structure used for each micro-lens is built here
  • disparity: It contains the code for estimating disparity map from a single image
  • rendering: It takes care of the rendering process, showing the image and calculating the all-in-focus images
  • samples: Some samples to show how the code works (check the sample folder for a more detailed explanation)

And the following files:

  • setup.py: used for the cython compilation of the sgm file
  • plenopticGUI.py: a basic graphical user interface to load images and estimate disparity. It is still a beta version for testing, it probably has some bugs

Dependencies and requirements

In order to work with the python library, you need:

  • version of python 3 or higher (using python3.6)
  • a C++ compiler and Cython (version 3 or higher is preferable)
  • matplot libraries (python3-matplotlib)
  • tk for the GUI (python3-tk)
  • numpy package (python3-numpy)
  • scipy package (python3-scipy)
  • header files (python3-dev)
  • opencv for python (python3-opencv)

Cython

Once you have installed dependencies, you have to run the setup file setup.py that will cythonize the Cython files. Use the command

python setup.py build_ext --inplace

It should create one or two files (.so) in the python folder, and they have to be moved to the disparity folder (still have to figure out a way to do it automatic, sorry)

More detail about the cython compilation here

Depending on the OS (it was developed on Linux and tested on Linux and Mac), there may be some small issues due to compatibility of some libraries. If you encounter any problem, please send some feedback, it would be useful for correcting.

Known issues:

fatal error: 'numpy/arrayobject.h' file not found - Mac OS X 10.10.5 Yosemite, python3.6

workaround found it here worked in this case, other possible solutions here and here

PYTHONPATH

Moreover, the structure of the python code requires that you set the environmental variable of your computer PYTHONPATH=/thepathwhereyoudownloadedthefolder/PlenopticToolbox/python

This can be done temporarily in two ways (temporarily means this has to be done everytime the python scripts have to be launched):

  • in python, at the beginning of the code, using sys
import sys
sys.path.append('/thepathwhereyoudownloadedthefolder/PlenopticToolbox/python')
  • in the terminal window, by using the command
export PYTHONPATH=/thepathwhereyoudownloadedthefolder/PlenopticToolbox/python

There is another way to set the environmental variable at login, by editing the right configuration file, either ~/.bash_profile or ~/.bashrc (here there is a more detailed explanation about which file to edit)

In the file it should be added the line

export PYTHONPATH=/thepathwhereyoudownloadedthefolder/PlenopticToolbox/python

TEST

To prove that the environmental variable is set, a simple test can be done using the command printenv from terminal, that will print the content of all environmental variables.

RUNNING A SCRIPT

In order to correctly run a script, you need to have downloaded at least one image (not from github as they don't fit here for reason of space, but from the google drive folder or the figshare dataset) and its corresponding configuration file (it ends in .xml)

Once you have both, the script can be run using the corresponding command. For ease of use, the script assumes that the name of the configuration file and the image are the same. The easiest way to do this is just renaming the configuration file (it is the same for many pictures) as the image. For example, if you want to use Dragon_Processed.png image, change .xml file to Dragon_Processed.xml. You are also free to modify the code (in the folder plenopticIo the file imgIO.py is taking care of this, at the lines 149 and 163 there are two methods that are using the filename, if you edit there putting the name you wish, this can be easily changed).

For more example about scripts, go to the samples page.

For further information, write to lpa@informatik.uni-kiel.de