MMVD Candidate List Reordering for Efficient Image Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images leads to a surge in transmission and storage costs due to the increase in bit rate, necessitating a high-efficiency image compression technology.
Innovation Solution
An image encoding/decoding method that constructs a merge candidate list using a merge mode with Motion Vector Difference (MMVD), reorders the MMVD candidate list, and derives motion information based on the reordered list to improve encoding/decoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution and high-quality images are transmitted, then image quality is improved, but transmission costs and storage costs increase due to increased bit rate
Solution Approach 1:
The patent changes the parameter of prediction candidate ordering in MMVD mode. By reordering the MMVD candidate list based on template matching error (SAD) values, the encoder can more efficiently select the best prediction candidate, improving compression efficiency and reducing bit rate for the same image quality
Solution Approach 2:
The patent replaces the conventional fixed-order candidate selection mechanism with a template matching-based reordering mechanism. Instead of using a predetermined candidate list order, the system calculates SAD values between templates and reference blocks to dynamically reorder candidates, achieving better compression without increasing computational complexity significantly
2Measurement precision
If the number of prediction candidates is increased, then encoding accuracy is improved, but the amount of bits required to represent motion information increases
Solution Approach 1:
The patent changes the ordering parameter of the MMVD candidate list based on template matching performance. By sorting candidates according to their SAD values rather than using a fixed order, the system can represent motion information more efficiently with fewer bits while maintaining or improving encoding accuracy
Solution Approach 2:
The patent performs preliminary template matching and SAD calculation to pre-order the candidate list before actual prediction. This preliminary action allows the decoder to efficiently select the best candidate with minimal bit overhead, as the reordering information can be compactly encoded based on the pre-calculated SAD values
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image encoding/decoding method, a bitstream transmission method, and a computer-readable recording medium for storing a bitstream are provided. The image decoding method according to the present disclosure is an image decoding method performed by an image decoding apparatus, and may comprise the steps of: constructing a merge candidate list on the basis that a prediction mode of a current block is a merge mode; constructing a merge mode with motion vector difference (MMVD) candidate list on the basis of a motion vector difference (MVD) available in an MMVD mode and a first merge candidate included in the merge candidate list; reordering the MMVD candidate list; and deriving motion information of the current block on the basis of the reordered MMVD candidate list.