Coding Unit Size Dependent Simplified Depth Coding for 3D Video
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The 3D-HEVC Test Model 3 has limitations in coding block size for intra_DMM, restricting 64×64 SDC-coded coding units to intra_DC or intra_planar modes, leading to redundant context allocations and increased complexity in hardware implementations.
Innovation Solution
Providing options to either omit simplified depth coding for 64×64 coding units, partition them into smaller prediction units, or limit intra-prediction modes to intra_DC or intra_planar for 64×64 coding units, allowing flexibility in coding strategies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If 64×64 SDC-coded coding units are restricted to intra_DC or intra_planar modes, then implementation complexity is reduced, but coding efficiency deteriorates due to redundant context allocations
Solution Approach 1:
The patent introduces dynamic mode selection that adapts the intra-prediction mode based on coding unit size. For 64×64 CUs, the system dynamically chooses between intra_DC, intra_planar, and intra_DMM modes based on content characteristics, rather than statically restricting to only two modes. This dynamic adaptation resolves the contradiction by maintaining simplicity for small CUs while enabling efficient coding for large CUs when needed.
Solution Approach 2:
The patent changes the parameter of available prediction modes based on the coding unit size parameter. When CU size is 64×64, the system conditionally enables additional modes (intra_DMM) based on content characteristics, effectively changing the set of available parameters (prediction modes) to optimize coding efficiency without significantly increasing overall system complexity.
2Ease of manufacture
If intra_DMM is restricted for 64×64 coding units, then hardware implementation cost is reduced, but coding flexibility deteriorates
Solution Approach 1:
The patent segments the coding unit processing based on size. For smaller CUs (non-64×64), full flexibility including intra_DMM is available. For 64×64 CUs, the system segments the mode selection to conditionally allow intra_DMM only when beneficial, thereby reducing hardware complexity for the majority case while preserving flexibility when needed.
Solution Approach 2:
The patent creates a universal coding framework that handles multiple CU sizes and mode combinations through a unified structure. The same hardware infrastructure supports both restricted mode for 64×64 CUs and full mode availability for other sizes, achieving multi-functionality without requiring separate dedicated hardware paths for each case.
3Loss of information
If simplified depth coding is omitted for 64×64 coding units, then context allocation redundancy is reduced, but processing efficiency deteriorates
Solution Approach 1:
The patent performs preliminary assessment of coding unit characteristics before selecting the prediction mode. For 64×64 CUs, the system preliminarily evaluates whether intra_DMM would benefit the specific content, and only then allocates the additional context. This preliminary action prevents redundant context allocation for cases where it wouldn't help, while preserving processing efficiency for cases where it would be beneficial.
Data Source
AI summary
Systems, articles, and methods for coding unit size dependent simplified depth coding for 3D video coding.