Disable model updating ?

[chacha21]

I would like to test the BGS library for post-processing some short very static sequences.

When I do that with OpenCV, I usually : -make a first pass with an auto-learning rate on the sequence (or a set of images matching the learning "time window" of the algorithm) -then I can pick images randomly in my sequence, and "apply()" the algorithmn with a special value for the learning rate (0) to prevent the bg model from being updated any more. (https://docs.opencv.org/master/d7/df6/classcv_1_1BackgroundSubtractor.html#aa735e76f7069b3fa9c3f32395f9ccd21) This way, I have reproducible results that do not depend on the order I look at the images.

There seem to be no such "learningRate" abstraction for the BGS library, and I cannot "disable" the BG model updating.

Did I miss something ?

[andrewssobral]

Hi @chacha21 , sorry for my delayed feedback to you. Yes, currently the bgslibrary doesn't have this feature, I will work on it. Thanks, Andrews

[andrewssobral]

@chacha21 , for info, I just released a new version of the BGSLibrary, now v3.1.0. I tested it on Windows, Mac and Linux, both compiling the C++ library on OpenCV 3.4.7, 3.4.16 and 4.6.0. I also tested its respective Python wrapper and GT user interface. Please feel free to get back to me if you have any issues, Andrews