Computational stereo vision using color

An investigation of the efficiency of color or chromatic information as an aid in solving the stereo correspondence problem is described. Two very simple stereo algorithms differing only in the use of chromatic information are applied to images containing various chromatic characteristics. The intensity-matching algorithm matches the zero crossings in a Laplacian-of-Gaussian filtered image, while the color-matching algorithm includes chromatic gradients that characterize the intensity zero crossings. The algorithms have been defined with a very simple matching structure in an effort to isolate the effects of including chromatic information in early stereo processing. The results indicate that the use of chromatic information can significantly reduce the ambiguity between potential matches while increasing the accuracy of the resulting matches.<>