acgtyrant's profile - activity

@LBerger This post does not answer about the license of documentations.

I want to add this documentation to devdocs.io, but the documentation's license must permit alteration, redistribution and commercial use.

I want to add this documentation to devdocs.io, but the documentation's license must permit alteration, redistribution and commercial use.

How to do hard negative mining? The false positive created by detector can not use as negative immediately while the opencv_traincascade will use a slide window to extracted all candidates from the negative image. I have tried to scale the false positive down to the window size, but I find that the opencv_traincascade does not think this is a valid negative so does not take it into training.

I'm looking forward for your help, thank you very much!

Here is my script:

def main():
    subprocess.call([
            'opencv_createsamples',
            '-info', label_pathname,
            '-vec', vec_pathname,
            '-num',  str(positive_samples_count),  # not that it is int!
            '-w', str(positive_sample_scale),  # not that it is int!
            '-h', str(positive_sample_scale),  # not that it is int!
            '-show', 'NO'])

if __name__ == "__main__":
    main()

And I execute it, it still shows the sample! I have to pressing Esc will continue the samples creation process without. So what is wrong? Thank you!

uname -a: Linux YY 3.19.0-15-generic #15-Ubuntu SMP Thu Apr 16 23:32:37 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux.

OpenCV version: 2.4.9

Here is my script:

def main():
    subprocess.call([
            'opencv_createsamples',
            '-info', label_pathname,
            '-vec', vec_pathname,
            '-num',  str(positive_samples_count),  # not that it is int!
            '-w', str(positive_sample_scale),  # not that it is int!
            '-h', str(positive_sample_scale),  # not that it is int!
            '-show', 'NO'])

if __name__ == "__main__":
    main()

And I execute it, it still shows the sample! I have to pressing Esc will continue the samples creation process without. So what is wrong? Thank you!

uname -a: Linux YY 3.19.0-15-generic #15-Ubuntu SMP Thu Apr 16 23:32:37 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux.

OpenCV version: 2.4.9

@berak You are right, but I no not want to include too many header files while I use only a little API.

I can not find any documentations about header files, so I do not know which header file should I include when I use some OpenCV API.

By the way, I find the the Rect class in core/core.hpp and the operator & function which accept two Rects objects in core/operations.hpp. But the core/operations.hpp does not include the core/core.hpp, so if I want to define two Rect objects and use the operator & function`, I should include these two header files manually? Is there the other header file can include them instead at once?

I can not find any documentations about header files, so I do not know which header file should I include when I use some OpenCV API.

By the way, I find the the Rect class in core/core.hpp and the operator & function which accept two Rects objects in core/operations.hpp. But the core/operations.hpp does not include the core/core.hpp, so if I want to define two Rect objects and use the operator & function`, I should include these two header files manually? Is there the other header file can include them instead at once?

I am writing a Python script, and I am shocked that I find that there is no such cv2.Mat and cv2.Rect classes, so that I can not construct variables.

I do not understand why does Python interface miss these key classes.