Stereo matching method for raster binocular stereo vision sensor

Stereo matching is one of the difficulties of raster binocular stereo vision sensor. For this problem, a stereo matching method was presented based on epipolar constraints and searching minimum distance of space points. The method treated raster stereo vision system as two raster structur light sensors. After calibration, the 3D coordinate for centre point of light stripe to the structure light model was measured by using this method. If two points were matching points, their distance became zero. Leading in epipolar constraint, a feature point could be found in the left picture. Then a epipolar line could be found correspondingly in the right picture. The matching point could be found in the intersection points of epipolar line and structure line stripes. This method achieved points matching in 3D space. It has small account of calculation, and can obtain unique matching point. The simulation result shows the effectiveness of the method.