Color vector similarity-based stereo matching for single-shot 3D measurement

We propose a simple yet robust stereo matching scheme that uses the inner product of color vectors to measure the similarity between a pair of color images. When combined with color structured-light projection, our technique has the advantages of stereo matching stability, robustness to differences in light brightness, and faster computation time. We demonstrate these strengths through experiments that show that our approach is applicable to single-shot three-dimensional (3D) measurement of moving objects.