3D shape measurement in occluded environments based on stereo PMP algorithm

Abstract

Nowadays, a number of 3D measurement methods have been developed such as stereo vision, laser structured light and PMP (Phase Measuring Profilometry) method. However, they have its own limitations : 2π ambiguity, correspondence problem, long estimation time. To solve these problems, in our previous researches [9,13], we introduced a novel sensing method adopting stereo vision and PMP technique (stereo PMP algorithm). One other difficult problem is occlusion problem needed to tackle by the stereo PMP algorithm which uses the principle of stereo vision and two cameras. The occlusion problem cannot be solved by using the principle of typical stereo vision, because there is no correspondence point in occlusion area. In our previous research based on stereo PMP algorithm, however, phase information related to the projector's position is additionally used which gives more additional information. By using this additional information, we can solve the occlusion problem effectively. In order to detect occlusion area, we adopt the principle of Dynamic Programming, while to measure the depth the principle of typical PMP algorithm and the geometrical relationship of detected depth area. To verify the efficiency of the proposed method, a series of experimental tests were performed.