Efficient disparity map generation for moving multi-camera system using TOF depth sensor

摘要

In this paper, we propose an efficient disparity map generation method for multi-view video sequences captured by a forward moving multi-camera system. Forward camera moving increases the minimum and maximum disparity value of each frame with time. Thus, computational complexity can be increased and the quality of generated disparity can be degraded since the stereo matching operation has to deal with a wide range of the disparity candidates. In order to solve this problem, we employ a time-of-flight (TOF) depth sensor as a guide to find the minimum and maximum disparity values for each frame. Then, the stereo matching process for video sequences captured by the moving multi-camera system becomes simple without quality degradation.