Subblock Merge Signaling in Inter Prediction to Cut Bit Overhead
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images leads to increased data volume, resulting in higher transmission and storage costs, and existing inter prediction modes contribute to compression bit overhead due to diverse data representation.
Innovation Solution
An image encoding/decoding method that efficiently encodes and decodes various inter prediction modes by determining the decoding/encoding order based on subblock merge mode indicators, regular merge mode indicators, and combined intra merge modes, and adjusts encoding/decoding based on block size and slice type.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If diverse inter prediction modes are used to improve prediction performance, then prediction accuracy is improved, but compression bit overhead increases
Solution Approach 1:
The current block is divided into multiple subblocks, and different merge modes can be applied to different subblocks independently. This segmentation allows the encoder to use diverse prediction modes only where necessary rather than applying them uniformly across the entire block, thus improving prediction accuracy in complex regions while limiting bit overhead in simpler regions.
Solution Approach 2:
The patent applies subblock merge mode selectively based on block characteristics. Instead of applying diverse inter prediction modes to all blocks uniformly, the encoder determines whether to use subblock merge mode based on specific conditions, applying partial action only where it provides significant prediction improvement while avoiding unnecessary bit overhead in cases where simpler modes suffice.
2Adaptability or versatility
If multiple merge mode indicators are decoded for each block, then prediction mode flexibility is improved, but decoding complexity increases
Solution Approach 1:
The decodeing process is segmented into hierarchical levels: first a subblock merge mode indicator is decoded, and only if needed are the regular merge mode indicator and subblock merge mode information decoded separately. This segmentation of the decoding process reduces average decoding complexity while preserving the flexibility to use multiple merge modes when required.
Solution Approach 2:
The subblock merge mode indicator is decoded in advance before determining whether to decode the more detailed regular merge mode indicator and subblock merge mode information. This preliminary action allows the decoder to prepare for potential complex mode decoding while avoiding unnecessary decoding steps when simpler modes are sufficient, thus reducing overall decoding complexity while maintaining flexibility.
Data Source
AI summary
Disclosed herein is an image encoding/decoding method and apparatus. An image decoding method includes decoding a subblock merge mode indicator of a current block, decoding a regular merge mode indicator of the current block, if the subblock merge mode indicator does not indicate a subblock merge mode, and decoding subblock merge mode information of the current block, if the subblock merge mode indicator indicates a subblock merge mode.