Triangle Mesh Encoding With Adaptive Angles to Reduce Code Size
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for encoding three-dimensional data face challenges in efficiently processing vertex information and angle information, leading to increased code amounts.
Innovation Solution
An encoding device that encodes first, second, and third vertex information, along with angle information to identify a fourth vertex, using a bitstream, where the angle information is within a specific range and encoded adaptively based on prediction errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If vertex information and angle information are encoded separately using traditional methods, then the encoding process is simple, but the code amount increases
Solution Approach 1:
The patent combines the encoding of vertex information and angle information into a unified encoding process. The angle information is encoded adaptively based on prediction errors from previously decoded triangles, and both vertex and angle data are integrated into a single bitstream structure, reducing overall code amount while maintaining encoding efficiency
Solution Approach 2:
The patent changes the encoding parameters by introducing angle information within a specific range (0 to π) and using adaptive encoding based on prediction errors. This parameter transformation allows for more compact representation of geometric data, reducing the code amount required for three-dimensional mesh encoding
2Measurement precision
If angle information is encoded with high precision, then the position accuracy of the fourth vertex improves, but the code amount increases
Solution Approach 1:
The patent transforms the angle information into a normalized range (0 to π) and applies adaptive encoding based on prediction errors. This parameter transformation maintains position accuracy while significantly reducing the code amount by encoding only the necessary precision required for accurate vertex position reconstruction
Solution Approach 2:
The patent implements a feedback mechanism where angle information is encoded adaptively based on prediction errors from previously decoded triangles. The encoder uses the decoded vertex positions to predict subsequent vertex positions and encodes only the differences, maintaining high position accuracy while minimizing code amount through error-based adaptive encoding
Data Source
AI summary
An encoding device includes memory and a circuit accessible to the memory. In operation, the circuit: encodes first vertex information, second vertex information, and third vertex information into a bitstream, the first vertex information indicating a position of a first vertex of a first triangle, the second vertex information indicating a position of a second vertex of the first triangle, and the third vertex information indicating a position of a third vertex of the first triangle; and encodes angle information into the bitstream, the angle information indicating one or more angles used for identifying a position of a fourth vertex of a second triangle by reference to the first triangle, the second triangle being a triangle that shares a common side with the first triangle and is on the same place as the first triangle.