User:Marpred/sandbox

V-PCC (Video-based Point Cloud Compression) is an MPEG standard for encoding dynamic point clouds. The main goal for V-PCC is to enable rapid deployment of point cloud representations on current devices through the reuse of existing hardware, even though that may imply some limitations in coding and reconstruction performance. This is achieved by first transforming the 3D geometry and attribute data of the point cloud into a set of 2D patches. These patches are then mapped to a predefined set of 2D planes through orthogonal projections, without self-occlusions and with limited distortion. A mapping between the point cloud and a regular 2D grid is then obtained by packing the projected patches. This mapping is then used to generate 2D images representing the point cloud geometry and its attributes. Since the mapping between the point cloud and the 2D grid is not bijective, an extra binary image, referred to as the occupancy map, is needed in order to distinguish between the filled (i.e., associated with a point) and the empty (i.e., not associated with any point) cells of the grid. Similar to the geometry and attribute video sequences, the occupancy map video sequence is compressed using 2D based image or video codecs. In order to be able to reconstruct the 3D point cloud from the 2D geometry, attribute, and occupancy videos, additional information specifying per patch information (e.g., 3D/2D patch location and projection plane index) is required. V-PCC introduces a new codec specifically optimized to handle the patch information sub-stream. This sub-stream occupies a relatively small amount of the overall bitstream (i.e., less than 5%). Additional information needed to synchronize and link the video and patch sub-streams is also signaled in the bitstream. All these components could be multiplexed into a single bitstream or encoded as separate tracks of an ISOBMFF container. Similar to video encoding standards, V-PCC only specifies the decoding process. The encoding process is out of scope of the standard and it is left to an encoder manufacturer to create the most appropriate encoder that generates bitstreams that would conform to the V-PCC specification.