Hybrid projection-based point cloud texture coding

A method, comprising:

• receiving, by circuitry in an apparatus, a video that includes at least one three-dimensional (3D) object in a 3D physical space;
• generating, by the circuitry, a 3D geometrical representation of a point cloud corresponding to the at least one 3D object based on the received video, wherein the 3D geometrical representation of the point cloud comprises a first set of points associated with geometrical information and texture information corresponding to the at least one 3D object;
• generating, by the circuitry, a first plurality of two-dimensional (2D) projections from the 3D geometrical representation of the point cloud, wherein each 2D projection of the first plurality of 2D projections represents a projection of texture from a first viewing angle of a plurality of different viewing angles of the 3D geometrical representation of the point cloud of the at least one 3D object;
• detecting, by the circuitry, a second set of points that is occluded in the first set of points of the 3D geometrical representation of the point cloud, wherein the detected second set of points corresponds to the first plurality of 2D projections;
• distinctly encoding, by the circuitry, the first plurality of 2D projections and the detected second set of points;
• and
• removing, by the circuitry, a third set of points from the first set of points of the 3D geometrical representation of the point cloud, wherein the third set of points is different from the detected second set of points.