Fast stereo matching using multilevel enhancement

Presents a new approach for fast and accurate stereo correspondence. We address some of the efficient and robust implementation aspects of stereo matching algorithms based on thresholding the correlation values of previous pyramid structures that reduce the search area for finding the best matching point. The disparity for each scan line is chosen by selecting the position that gives the maximum correlation coefficient from the pyramid structure. However, sometimes this maximum value may be overridden by a spurious maximum. The ambiguities mainly come from image noise, lack of intensity variation or geometric distortion. For disambiguation, the adaptive window theory is applied to avoid mismatches in featureless areas. Adjusting the threshold parameter at each level of the pyramid structure reduces the computation time and increases the accuracy for the next pyramid level. Both synthetic and real image sets have been tested, and good results have been obtained.