Point Cloud Compression Using 2D Projection and Boundary Color Smoothing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Point clouds require significant bandwidth for transmission due to their large size, and existing methods often necessitate specialized hardware for compression, which can be costly and not widely available.
Innovation Solution
Convert point clouds from a 3D representation to a 2D representation by projecting points onto frames, using existing video codecs like HEVC, AVC, VP9, VP8, and VVC to compress the 2D frames, and implement low complexity color smoothing at patch boundaries to reduce visual artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If point clouds are compressed using specialized hardware, then compression efficiency is improved, but device complexity and cost increase
Solution Approach 1:
The patent uses 2D video frames as copies or projections of the 3D point cloud data. By projecting the 3D point cloud onto 2D planes and encoding these projections using standard 2D video codecs, the system achieves compression without requiring specialized 3D compression hardware, thus resolving the contradiction between compression efficiency and hardware complexity
Solution Approach 2:
The patent transforms the 3D point cloud data into 2D projections by collapsing the z-dimension. This dimensionality reduction allows the use of mature 2D video compression standards instead of requiring complex specialized 3D compression hardware, effectively resolving the technical contradiction
2Manufacturing precision
If point clouds are transmitted uncompressed, then data quality is improved, but bandwidth requirements increase
Solution Approach 1:
The patent segments the 3D point cloud data into multiple 2D projection views (front, back, left, right, top, bottom). Each projection is encoded separately using efficient 2D video compression, achieving significant bandwidth reduction while preserving the quality information needed for accurate 3D reconstruction
Solution Approach 2:
By projecting 3D point cloud data onto 2D planes, the patent reduces the data dimensionality from three dimensions to two, enabling the use of highly efficient 2D video compression algorithms that significantly reduce bandwidth requirements while maintaining reconstructable quality
3Manufacturing precision
If color smoothing is applied at patch boundaries, then visual quality is improved, but processing complexity increases
Solution Approach 1:
The patent applies color smoothing selectively only at patch boundaries where visual artifacts occur, rather than uniformly across the entire point cloud. This localized approach improves visual quality at the specific problem areas while minimizing additional processing complexity
Solution Approach 2:
The patent copies color information from neighboring patches and uses it to smooth artifacts at patch boundaries. By leveraging existing color data from adjacent regions rather than performing complex computations, the system improves visual quality while keeping processing complexity low
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method and decoder for point cloud decoding. The method includes receiving and decoding a bitstream into multiple frames that include patches corresponding to respective clusters of points from a 3D point cloud. The method also includes generating a grid that includes a plurality of 3D cells, wherein the 3D point cloud is within the grid. The method further includes identifying a first cell of the plurality of 3D cells that includes a query point corresponding to a pixel on a boundary of one of the patches. The method additionally includes identifying a luminance value of the first cell. The method also includes performing color smoothing on the query point based on comparison of the luminance value of the first cell to a threshold.