Intra Prediction Mode Fusion for Video Decoding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in achieving improved coding efficiency through the development of new intra prediction modes.
Innovation Solution
A method and device for decoding video data that utilizes a combination of decoder-side intra mode derivation (DIMD) and template-based intra mode derivation (TIMD) modes to generate a third intra prediction, which is then used to reconstruct block units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple intra prediction modes (DIMD and TIMD) are combined to improve coding efficiency, then video decoding performance is improved, but device complexity increases
Solution Approach 1:
The patent combines DIMD (decoder-side intra mode derivation) and TIMD (template-based intra mode derivation) modes into a unified intra prediction framework. The decoder integrates both mode derivation approaches, using template matching techniques from TIMD alongside the gradient-based angular mode selection from DIMD, thereby improving coding efficiency through fused prediction capabilities while managing computational complexity through shared data structures and processing pipelines.
2Measurement precision
If template-based intra mode derivation (TIMD) is used to determine intra prediction modes, then prediction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent implements a threshold-based mechanism where template costs are compared against a predefined threshold to determine whether full template-based mode derivation is performed. When the template cost ratio exceeds the threshold, the system selectively applies only the most critical template matching operations or falls back to simpler mode selection, thereby maintaining prediction accuracy for difficult cases while reducing computational complexity for easier cases.
3Adaptability or versatility
If decoder-side intra mode derivation (DIMD) is used to derive intra prediction modes, then adaptability is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary gradient calculations and angular mode estimations during the early stages of block processing, before full template matching is executed. By pre-computing gradient histograms and identifying candidate angular modes in advance, the system reduces the search space for subsequent template-based refinement, thereby maintaining adaptability through comprehensive mode exploration while reducing overall processing time through staged computation.
Data Source
AI summary
A method of decoding video data is provided. The method includes: receiving the video data; determining, from an image frame of the video data, a block unit; determining a first intra prediction of the block unit based on a decoder-side intra mode derivation (DIMD) mode; determining a second intra prediction of the block unit based on an intra prediction mode other than the DIMD mode; generating a third intra prediction based on the first intra prediction and the second intra prediction; and reconstructing the block unit based on the third intra prediction. In addition, an electronic device using the method is also provided.