Hello all
I'm using the funcion cv::warpPerspective to correct an image which was captured by webcam.
The fact is: If I use NO perspective correction, my capture loop runs at 23 frames per second (FPS) (almost the camera limit), however, if I use perspective correction, my FPS rate drops to almost 5 fps.
I'm using a very limited machine (a netbook) but I would like to know if there is a more computatinal efficient way ( a function that takes less time to evaluate) to correct perspective images captured from webcams.
Regards
Edson Valle