Real time full frame dense stereo vision using indexed histogram
Jitesh Butala, V. Udpikar · 2009
In this paper, we propose a novel algorithm for real time dense disparity calculations. The algorithm is based on the concept of indexed histogram for matching. The matching algorithm works on a single pair of scanlines while a postprocessing block cleans up the disparity map. The algorithm presented takes constant time irrespective of the maximum disparity given to the algorithm. The implementation on a Blackfin DSP works in real time on a full PAL frame rate (720 × 576 active pixels @ 25 fps). (4 pages)