Multiview-Video-Based Compression of Plenoptic Point Clouds

The plenoptic point cloud that has associated colors from various directions, is a more complete representation of a 3D object than the general point cloud that has only one color. It is more realistic but also brings a larger volume of data that needs to be compressed. The state-of-the-art method to compress the plenoptic point cloud is the multiple attributes extension of the region-based adaptive hierarchical transform (RAHT). In addition to RAHT, the video-based point cloud compression (V-PCC) is also an efficient method to compress the point cloud. However, there are not any works using a video-based solution to compress the plenoptic point cloud yet. Therefore, in this paper, we provide a Multiview-video-based framework utilizing the high efficiency of the multiview video coding standard to compress the plenoptic point cloud efficiently. Under the proposed multiview-video-based framework, a plenoptic point cloud is projected to its bounding box to generate multiple attribute videos since it has multiple colors from various directions. Then the multiple attribute videos are compressed efficiently using Multiview High Efficiency Video Coding (MV-HEVC). To the best of our knowledge, this is the first work to compress the plenoptic point cloud using a video-based solution. To further improve the performance of the proposed framework, we propose two methods to reduce the bit cost of unoccupied pixels that are useless for the reconstructed quality of the plenoptic point cloud. First, we propose a blockbased group padding scheme to unify the unoccupied pixels across the attribute direction to minimize the bit cost of the unoccupied pixels. Second, we propose ignoring the distortion of the unoccupied pixels during the rate distortion optimization in MV-HEVC. The proposed algorithms are implemented in the V-PCC and the MV-HEVC reference software. The experimental results show that the proposed algorithms can lead to significant bitrate savings compared with the state-of-the-art method.