Range data acquisition using color structured lighting and stereo vision

Abstract

This paper presents a new color-lighting/stereo method for 3D range data acquisition by combining color structured lighting and stereo vision. A major advantage of using stereo vision together with color stripes lighting is that there is no need to solve the problem of finding the correspondence between the color stripes projected by the light source and the color stripes observed in the images. That is, the more difficult problem of finding the correct color stripe correspondence problem between the light source and the image is replaced by an easier image-to-image stereo correspondence — which is not only easier than the above lighting-to-image correspondence problem, but also easier than the traditional stereo correspondence because a good color pattern has been projected onto the object. Another advantage of using stereo vision is that there is no need to calibrate the position and orientation for each of the projected light stripes in 3D space. In this work, a pattern of color stripes is projected onto the objects when taking images, after which edge segments are extracted from the acquired stereo image pair, and then used for finding the correct stereo correspondence. A systematic procedure is proposed in this paper for generating good color stripe patterns. To find the correct stereo correspondence, a global search method based on intra-scanline dynamic programming is adopted. A winner-take-all scheme using edge-based inter-scanline consistency is then proposed to refine the results obtained from intra-scanline search. Experimental results have shown that the proposed method can successfully generate a dense range map with only one pair of stereo images.