OpenPTV library liboptv is 0.3.1 with a 3D tracker

[Alex Liberzon]

and PyPTV is 0.3.7 with some bugs fixed on naming. 

You can now use JPEG, TIFF, PNG, BMP, etc. and you can mix those

You can now use complex path locations and names without any need to rename the images

You can now work with some AI/ML plugins, e.g. rembg - shut out to the new stuff

The new stuff is in Python, but can be much faster with numba

pyptv_batch_parallel allows your to run 10000 images segmentation in 1/3 of time cause the images processed on concurrent multiprocessing

Looking for GPU contributors to make stuff even faster. 

Looking for good C programmers to make it all transparent to Python and include Tracking with Display (this part is sometimes very useful for visual adjustments)