Intra Prediction Coding Mode Selection via Syntax Flag Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards, such as ITU-T H.264 and HEVC, face limitations in efficiently enabling non-most probable modes (MPM) for intra prediction, which can lead to suboptimal coding performance and increased complexity in video encoding and decoding processes.
Innovation Solution
The proposed solution involves parsing syntax elements to determine whether a most probable mode (MPM) or a non-most probable mode (MPM) is used for intra prediction, and conditionally parsing additional syntax elements based on this determination, allowing for the selection and signaling of non-MPM modes for intra sub-partitions (ISP) and non-zero reference line (NZRL) predictions, thereby enabling more flexible and efficient intra prediction techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If non-most probable modes (MPM) are enabled for intra prediction, then coding performance is improved, but device complexity increases
Solution Approach 1:
The intra prediction mode selection is segmented into two distinct pathways: MPM mode and non-MPM mode. The patent introduces separate syntax elements (intra_luma_mpm_flag and intra_luma_mpm_idx) to control and signal the selection between these segments, allowing the decoder to efficiently determine which mode is used without exhaustive checking of all possible modes.
Solution Approach 2:
The patent performs preliminary classification by introducing the intra_luma_mpm_flag syntax element that预先 indicates whether the MPM mode or non-MPM mode will be used. This preliminary action allows the decoder to prepare the appropriate processing path in advance, avoiding the need to check multiple conditions during decoding and thereby reducing computational complexity.
2Adaptability or versatility
If multiple syntax elements are always parsed for intra prediction mode, then all modes are supported, but the number of checks and processing overhead increases
Solution Approach 1:
The syntax element structure is made dynamic through the intra_luma_mpm_flag. When this flag indicates MPM mode, the decoder parses intra_luma_mpm_idx to select from the most probable mode list. When it indicates non-MPM mode, a different syntax element structure is used. This dynamic adaptation allows the system to parse only the necessary syntax elements for the selected mode, reducing processing overhead while maintaining support for all intra prediction modes.
3Ease of operation
If intra prediction uses fixed mode selection, then decoding is simpler, but coding efficiency is reduced
Solution Approach 1:
The intra_luma_mpm_flag is parsed in advance to preliminarily determine which mode category (MPM or non-MPM) will be used. This preliminary classification enables the decoder to follow a simplified, predetermined path based on the flag value, maintaining decoding simplicity while still allowing access to the full range of prediction modes for optimal coding efficiency.
Solution Approach 2:
The intra_luma_mpm_flag acts as an intermediary that mediates between the simplified decoding process and the efficient coding performance. By using this intermediate flag to guide the selection between MPM and non-MPM pathways, the system achieves both decoding simplicity and coding efficiency without requiring complex decision logic at each decoding step.
Data Source
AI summary
This disclosure relates to video coding and more particularly to techniques for intra prediction coding. In one example, a method of decoding video data, the method comprising: parsing a first syntax element having a value indicating an intra sub partitions (ISP) intra prediction is used for coding a current video block; parsing a second syntax element indicating whether a most probable mode (MPM) or a non-most probable mode (MPM) is an intra prediction mode used for generating a prediction for the current video block according to the ISP intra prediction; based on whether the second syntax element indicates the MPM or the non-MPM is an intra prediction mode used for generating a prediction, conditionally parsing one of: a third syntax element indicating the MPM, or a fourth syntax element indicating the non-MPM; and generating a prediction for the current video block based on the first syntax element.