HEVC 3DVC Depth Coding Speed via Conditional Mode Evaluation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The complexity of Depth Modeling Modes (DMM) in HEVC-based 3DVC leads to excessive invalid prediction modes, resulting in prolonged coding times due to the evaluation of many modes during depth image coding.

Innovation Solution

A method is introduced where, before coding a parent prediction unit (PU) of a depth image, a judgment module determines if the PU meets specific conditions regarding block size, pixel value differences, and uniform pixel distributions. If met, only traditional intra modes of HEVC are evaluated; otherwise, both intra modes and DMM predictions are evaluated, reducing the number of modes to assess.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If DMM prediction modes are evaluated for all parent PUs, then prediction accuracy for boundary regions is improved, but coding time increases excessively

Engineering Contradiction:
Improveprediction accuracyVSAvoidcoding time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent changes the parameter of prediction mode selection by introducing conditions (block size threshold, pixel value difference threshold, uniform pixel distribution) that determine whether to use DMM modes or traditional intra modes, thereby adapting the prediction approach to specific content characteristics and avoiding unnecessary complex evaluations

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent applies different prediction strategies to different regions: traditional intra modes are used for regions meeting specific conditions (likely smooth or uniform regions), while DMM modes are used for boundary regions that require more sophisticated prediction, optimizing the balance between accuracy and complexity for each local area

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If multiple prediction modes (both intra modes and DMM modes) are evaluated, then prediction quality is improved, but device complexity increases

Engineering Contradiction:
Improveprediction qualityVSAvoidcoding complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent introduces multiple parameters (block size, pixel value difference, pixel distribution uniformity) that control the selection between different prediction mode sets, reducing the effective complexity by limiting the search space based on content characteristics

Inventive Principle:
Principle #35Parameter changes

3Manufacturing precision

If all prediction modes are evaluated for every parent PU, then coding quality is maintained, but productivity decreases

Engineering Contradiction:
Improvecoding qualityVSAvoidcoding speed
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The patent uses parameters derived from the depth image content (block size, pixel value range, pixel distribution) to dynamically adjust the prediction mode evaluation strategy, skipping complex DMM evaluations when traditional intra modes are sufficient, thus improving coding speed without significant quality loss

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent applies partial evaluation by selectively evaluating only necessary prediction modes based on content characteristics, rather than exhaustively evaluating all modes for every block, achieving sufficient coding quality with reduced computational effort

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP2890124B1Coding method and device applied to HEVC-based 3dvc
Publication Date: 2017.12.13 ZTE CORP
  • EP2890124B1 patent drawingFigure 1
  • EP2890124B1 patent drawingFigure 2~3

AI summary

A coding method applied to HEVC-based 3DVC is provided in the example of the present invention, which includes: in the HEVC-based 3DVC, before a parent prediction unit (PU) corresponding to a coding unit (CU) of a depth image is predicted, judging whether the parent PU meets a preset condition, if the parent PU meets the preset condition, evaluating only traditional intra modes of high efficiency video coding (HEVC) when the parent PU is coded,, and if the parent PU does not meet the preset condition, evaluating the traditional intra modes of the HEVC and depth modeling modes (DMM) predictions when the parent PU is coded. A coding device applied to HEVC-based 3DVC is also provided in the example of the present invention. In the example of the present invention, by reducing the evaluated prediction modes, a coding speed is increased.