Revision history [back]

Why are you studing opencv_haartraining application? It's obsolete version of a cascade training. In my opinion opencv_haartraining is more difficult for study than the newer opencv_traincascade. In any case the easiest way to achieve your goals is to build OpenCV from source in debug mode, open it in your favorite IDE and debug the choosen application as you usually do.

I also recommend you to try LBP features instead of HAAR. They are real-time even on mobile platforms. But they are supported by opencv_traincascade only ;)

Why are you studing opencv_haartraining application? It's obsolete version of a cascade training. In my opinion opencv_haartraining is more difficult for study than the newer opencv_traincascade. In any case the easiest way to achieve your goals is to build OpenCV from source in debug mode, open it in your favorite IDE and debug the choosen application as you usually do.do. You don't need to create your 'own objects'.

I also recommend you to try LBP features instead of HAAR. They are real-time even on mobile platforms. But they are supported by opencv_traincascade only ;)