Motion detection based on temporal-to-spatial conversion of depth maps for multi-view video

Motion detection and extraction for multi-view video (MVV) is one of the key technologies for MVV applications, and it has become a hot in recent years. In this paper, a new method that will use semantic information and movement information based on depth maps for motion detection is pro-posed. First, a sequence of consecutive depth images is converted into a sequence of consecutive temporal-to-spatial slices in the horizontal and vertical directions. In each one temporal-to-spatial slice, semantic information and motion information are both included, the background region forms a vertical line pattern, and a moving object creates an irregular, non-vertical structure. Second, to binarize the temporal-to-spatial slices are binarized by a dynamic threshold. Then we reconstruct the converted slices in the horizontal and vertical directions into temporal image masks, and reserve the common part of two corresponding masks. At last, by post-processing the reconstructed images, moving object masks can be obtained. Experiment results show that the proposed method exhibits a good performance for motion detection.