Video Block Intra Prediction Using DIMD and Multiple Reference Lines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently utilizing intra prediction modes and reference lines for video processing, leading to high signaling overhead and suboptimal coding efficiency.
Innovation Solution
The proposed method combines decoder-side intra prediction mode derivation (DIMD) with multiple reference lines (MRL) to derive intra prediction modes, utilizing non-adjacent reference lines for improved video block reconstruction, thereby reducing signaling overhead and enhancing coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional intra prediction modes are used, then prediction accuracy is maintained, but signaling overhead is high
Solution Approach 1:
The decoder derives intra prediction modes autonomously using decoder-side intra prediction mode derivation (DIMD) based on previously decoded blocks, eliminating the need for explicit mode signaling. The system serves itself by generating prediction modes through local signal analysis rather than receiving pre-specified modes from the encoder.
Solution Approach 2:
The method uses feedback from previously decoded blocks to derive intra prediction modes for current blocks. The decoder analyzes the structure and characteristics of already-reconstructed blocks to infer appropriate prediction modes, creating a feedback loop that adapts prediction strategies based on actual decoded content.
2Productivity
If multiple reference lines are used, then coding efficiency is improved, but device complexity increases
Solution Approach 1:
The current block is divided into multiple segments or regions, each utilizing different reference lines (adjacent and non-adjacent) for prediction. This segmentation allows selective application of multiple reference lines to different parts of the block, improving coding efficiency while managing processing complexity through localized operations.
Solution Approach 2:
The method extends reference line usage from traditional single-line references to multiple lines including non-adjacent lines, adding a dimensional aspect to reference utilization. This multi-line approach enriches the prediction space by incorporating information from multiple vertical or horizontal positions simultaneously.
3Loss of information
If decoder-side intra prediction mode derivation is applied, then signaling overhead is reduced, but processing time increases
Solution Approach 1:
Intra prediction modes are derived in advance during the decoding process using previously decoded blocks as a foundation. By preparing and analyzing prior blocks before processing current blocks, the system performs preliminary actions that enable efficient mode derivation without requiring additional signaling, balancing computational overhead with information reduction benefits.
Data Source
AI summary
Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. The method comprises: processing, during a conversion between a picture of a video and a bitstream of the video, a first video block in the picture based on a combination of decoder-side intra prediction mode derivation (DIMD) and multiple reference line (MRL), at least one intra prediction mode (IPM) being derived from the DIMD, and at least one non-adjacent reference line of reconstructed samples for the first video block being used in the combination of the DIMD and the MRL; and performing the conversion based on the first video block. Compared with the conventional solution, the proposed method can advantageously improve the coding effectiveness and coding efficiency.