Stereo Matching Algorithm Based on HSV Color Space and Improved Census Transform

Aiming at the problem that stereo matching accuracy is easily affected by noise and amplitude distortion, a algorithm based on HSV color space improved census transform proposed. In cost calculation stage, image first converted from RGB to space; moreover, hue channel used as primitive establish absolute difference (HAD) function, which reduces amount of enhances robustness matching. Then, solve traditional overrelying central pixel improve resistance algorithm, an method neighborhood weighting also Finally, HAD are nonlinearly fused initial cost. aggregation outlier elimination confidence interval By calculating window, this paper eliminates value not in subsequently filters well aggregates remaining costs further reduce interference accuracy. Experiments show proposed can only effectively suppress influence noise, but achieve more robust effect scenes with changing exposure lighting conditions.