Question

some (most) of opencv errors are discovered in run time – using CV_64 instead of CV_32, accessing out of range, etc. this makes debugging not an easy task, at least when you can’t use debug mode from different reasons..
When an error occurs, opencv stops on assert (CV::ASSERT macro) and tell exactly where in the opencv source it happened.
is there’s anyway to tell where that error occurred in the calling code (my code)?
Assuming of course we are in release mode (no break points – only console output)
Thanks

Answer 1

Yes, it’s usually easy to tell when OpenCV fails: all you need to do is check the return of each function call, and print an error message when a fail occurs.