Model-Based Joint Bit Allocation Between Geometry and Color for Video-Based 3D Point Cloud Compression

In video-based 3D point cloud compression, the quality of reconstructed depends on both geometry, and color distortions. Finding an optimal allocation total bitrate between geometry coder, coder is a challenging task due to large number possible solutions. To solve this bit problem, we first propose analytical distortion, rate models for information. Using these models, formulate joint problem as constrained convex optimization it with interior method. Experimental results show that rate-distortion performance proposed solution close obtained exhaustive search but at only 0.66 $\%$ its time complexity.