In-Between Point Cloud Texture Patches for Lower Bit-Rate Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud compression methods struggle to efficiently distribute dynamic point clouds to end-users while maintaining acceptable quality and minimizing bit-rate consumption.
Innovation Solution
A method involving signaling texture patches representing texture values of in-between 3D samples, projecting them onto a projection plane, and packing them into a texture image, along with orthogonal projection and mapping to derive texture values, which are then encoded into a bitstream for transmission and retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dynamic point clouds are distributed to end-users, then quality of experience is maintained, but bit-rate consumption increases
Solution Approach 1:
The patent segments the point cloud data by separating in-between 3D samples from nearer and farther samples, and divides texture values into distinct texture patches. This segmentation allows selective transmission of only the in-between sample texture patches when needed, reducing overall bit-rate consumption while maintaining quality of experience for critical depth regions.
Solution Approach 2:
The patent introduces texture patches as intermediary structures that bridge the gap between nearer and farther samples. These texture patches contain only the essential texture information for in-between samples, acting as a mediator that enables quality reconstruction without transmitting the full point cloud data, thus reducing bit-rate consumption.
2Manufacturing precision
If in-between 3D samples are processed, then texture quality is improved, but processing complexity increases
Solution Approach 1:
The patent segments the processing task by handling in-between samples separately from nearer and farther samples. The in-between samples are processed only when needed, and their texture values are organized into distinct texture patches. This segmentation reduces processing complexity by avoiding the need to process all samples uniformly, while still achieving improved texture quality for critical regions.
Solution Approach 2:
The patent applies partial action by processing and transmitting texture information for in-between samples only when necessary, rather than processing all samples uniformly. This selective processing approach improves texture quality for critical depth regions while reducing overall processing complexity and computational load.
3Loss of energy
If texture patches are signaled and extracted, then bit-rate efficiency is improved, but decoding complexity increases
Solution Approach 1:
The patent segments the bitstream into distinct texture patches corresponding to different depth regions. The syntax elements clearly delimit these patches, allowing the decoder to extract only the relevant in-between sample texture patches when needed. This segmentation improves bit-rate efficiency by transmitting only essential information, while the structured format actually simplifies decoding compared to processing complete point clouds.
Solution Approach 2:
The patent uses syntax elements that signal and describe texture patches as copies of the essential texture information. These syntax elements contain compressed representations of the in-between sample texture values, allowing efficient extraction and reconstruction without requiring complex decoding of the entire point cloud, thus improving bit-rate efficiency while managing decoding complexity.
Data Source
AI summary
At least one embodiment relates to signaling at least one texture patch representing a texture value of at least one in-between 3D sample, a texture patch being a set of 2D samples representing texture values of orthogonally projected 3D samples of a point cloud along projection lines onto a projection plane, and said at least one in-between 3D sample being a 3D sample of the point cloud having a depth value greater than a nearer 3D sample of the point cloud and lower than a farther 3D sample of the point cloud, said at least one in-between 3D sample and said nearer and farther 3D samples being projected along the same projection line.