Video Coding Reference Picture List Construction for Bitrate Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video coding techniques suffer from low coding efficiency due to limitations in the reference structure for video coding, particularly in handling resolution changes and temporal layer dependencies, which leads to increased bitrate and reduced performance in adaptive video transmission.
Innovation Solution
The proposed method enhances video coding efficiency by enriching the reference picture list with high-resolution pictures that have a higher similarity or temporal layer identity to the target picture, allowing for improved inter-coding and adaptive resolution changes without inserting IDR or TRAP pictures, thereby reducing bitrate and enhancing coding performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional video coding techniques are used with traditional reference structures, then device complexity is reduced, but coding efficiency deteriorates
Solution Approach 1:
The patent implements dynamic reference picture list construction by allowing pictures from the same temporal layer to be conditionally added to the reference picture list based on motion characteristics and prediction modes. This dynamic adaptation enables the system to optimize reference picture selection for each coding unit, improving coding efficiency without requiring complex static structures
Solution Approach 2:
The patent changes the parameter of temporal layer identity restriction by allowing pictures with the same temporal layer ID to serve as reference pictures under specific conditions. This parameter change enables richer reference picture availability while maintaining temporal layer management structures, resolving the contradiction between efficiency and complexity
2Productivity
If high-resolution reference pictures are used for inter-coding, then coding efficiency is improved, but bitrate increases
Solution Approach 1:
The patent applies local quality by allowing high-resolution reference pictures to be used selectively for specific prediction modes (e.g., inter-layer prediction) and coding units that benefit most from them. Not all coding units use high-resolution references, which optimizes the bitrate-efficiency tradeoff by applying high-quality references only where necessary
Solution Approach 2:
The patent uses partial action by incorporating only a subset of available high-resolution reference pictures into the reference picture list based on motion characteristics and prediction requirements. This selective inclusion improves coding efficiency for relevant blocks while avoiding the bitrate cost of transmitting and processing all high-resolution references
3Reliability
If reference pictures from the same temporal layer are used, then quality of reference pictures is improved, but device complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-establishing temporal layer identities and organizing pictures into temporal layers before the inter-coding process. This preliminary organization allows the system to efficiently query and select reference pictures from the same temporal layer without complex runtime management, improving reference picture quality while keeping temporal layer management tractable
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: obtaining a first reference picture list (RPL) for a target picture of a video; determining a set of candidate pictures from a plurality of coded pictures of the video, each of the set of candidate pictures having a same temporal layer identity (ID) as the target picture; obtaining a second RPL based on the set of candidate pictures and the first RPL; and performing a conversion between the target picture and a bitstream of the video based on the second RPL. Compared with the conventional solution, the proposed method can advantageously improve the coding efficiency.