Mesh Decoding With Base-Mesh Intra Prediction for Efficient Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mesh encoding technologies fail to reduce the amount of information effectively, leading to low encoding efficiency due to the lossless encoding of all vertex coordinates and connectivity information, even when loss is allowed.
Innovation Solution
A mesh decoding device and method that performs intra prediction on subdivided vertices using a base mesh to calculate an intra prediction value and add it to an intra prediction residual, improving encoding efficiency by reducing the amount of information required.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all vertex coordinates and connectivity information are encoded losslessly, then mesh decoding accuracy is maintained, but encoding efficiency deteriorates due to inability to reduce information amount
Solution Approach 1:
The patent segments the mesh encoding process into multiple levels: base mesh encoding and subdivision residual encoding. The base mesh is encoded losslessly to maintain overall structure accuracy, while only the subdivision residuals (difference information) are encoded, allowing for information reduction and improved encoding efficiency without sacrificing final mesh accuracy.
Solution Approach 2:
The patent performs preliminary action by first encoding the base mesh completely and losslessly, then using this pre-encoded base mesh as a reference for subsequent residual encoding. This preliminary complete encoding establishes the foundation that allows later efficient residual encoding to achieve both accuracy and efficiency.
2Reliability
If lossless encoding is applied to all mesh data, then data integrity is preserved, but the amount of information cannot be reduced leading to low encoding efficiency
Solution Approach 1:
The patent applies local quality by differentiating the encoding approach for different parts of the mesh data. The base mesh is encoded losslessly to preserve overall data integrity and structure, while the subdivision residuals are encoded with loss tolerance acceptable for the fine-detail adjustment. This localized application of lossless vs. lossy encoding reduces total information quantity while maintaining critical data integrity.
Solution Approach 2:
The patent segments the mesh data into base mesh (essential structure) and subdivision residuals (fine adjustments). By segmenting, the system can apply lossless encoding only where critical (base mesh) and lossy encoding where acceptable (residuals), thereby reducing total information amount while preserving data integrity for the most important components.
3Productivity
If intra prediction is performed on subdivided vertices using base mesh, then encoding efficiency is improved through information reduction, but decoding complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-encoding the base mesh completely before processing subdivision residuals. This pre-computed base mesh serves as a ready-to-use reference during residual decoding, eliminating the need for complex real-time prediction calculations and reducing decoding complexity while maintaining improved encoding efficiency.
Solution Approach 2:
The patent uses the base mesh as a template or copy that is decoded once and then reused as reference information for decoding all subdivision residuals. This copying approach avoids repeatedly performing complex prediction calculations during residual decoding, simplifying the decoding process while maintaining the efficiency benefits of information reduction.
Data Source
AI summary
A mesh decoding device 200 includes a circuit that: outputs a base mesh, outputs an intra prediction residual, and decodes a displacement by performing intra prediction of a displacement of a subdivided vertex based on the outputted base mesh to calculate an intra prediction value, and adding the calculated intra prediction value and the outputted intra prediction residual.