Dynamic Mesh Geometry Refinement With Block-Aligned LoD Packing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for mapping 3D displacement coefficients to a 2D surface in video coding fail to clearly distinguish samples at specified levels of detail, leading to inefficient memory allocation and partial reconstruction scenarios.
Innovation Solution
A block-boundary-aligned arrangement is used to unpack and pack samples belonging to multiple levels of detail (LoD) associated with geometry displacements, aligning them with the boundaries of blocks in a two-dimensional plane, allowing for efficient encoding and decoding of dynamic mesh geometry.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If maximum memory is allocated to ensure complete geometry reconstruction, then reconstruction reliability is improved, but memory efficiency deteriorates in partial reconstruction scenarios
Solution Approach 1:
The geometry displacement data is segmented into multiple levels of detail (LoD) with distinct sample sets. Each LoD can be independently processed and reconstructed, allowing the system to allocate memory only for the required LoD levels rather than allocating maximum memory for complete reconstruction of all details.
2Productivity
If samples from multiple levels of detail are packed into a two-dimensional plane, then coding efficiency is improved, but sample distinction by level deteriorates
Solution Approach 1:
The two-dimensional plane is segmented into multiple blocks, with each block containing samples from a specific level of detail. This segmentation allows efficient packing of multiple LoD samples while maintaining clear distinction between levels through block boundaries, enabling both coding efficiency and sample distinguishability.
Solution Approach 2:
Different blocks in the two-dimensional plane are assigned different local qualities by storing samples from specific LoD levels in each block. This allows the system to maintain efficient packing while ensuring that samples from different levels are clearly distinguished through their spatial localization in the block structure.
3Quantity of substance
If block-boundary-aligned arrangement is used for packing samples, then memory allocation efficiency is improved, but processing complexity increases
Solution Approach 1:
The sample data is segmented into blocks that align with memory boundaries. This segmentation enables efficient memory allocation by ensuring that each block fits within standard memory page boundaries, while the segmentation structure itself provides a clear framework for processing that reduces actual implementation complexity.
Data Source
AI summary
Computer-implemented methods and systems for processing geometry replacements are disclosed. The methods include: packing/unpacking, in a block-boundary-aligned arrangement, a plurality of samples belonging to a plurality of levels of detail (LoD) associated with geometry displacements, into/from a two-dimensional plane including a plurality of blocks.