Video Coding with Decoder-Side Motion Compensation and Sub-Block SAD Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies face challenges in efficiently compressing high-resolution images and 3D video content, particularly in adapting motion modification for improved encoding/decoding efficiency.
Innovation Solution
The method determines motion modification in units of sub-blocks, calculates SAD candidates within a defined search region, and derives delta motion information using pre-generated motion information to adaptively modify motion for improved encoding/decoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If motion modification is performed using traditional methods, then encoding complexity is reduced, but compression efficiency deteriorates for high-resolution and 3D video content
Solution Approach 1:
The current block is divided into multiple sub-blocks, and motion modification is performed independently for each sub-block. This segmentation allows the decoder to adaptively refine motion information in specific regions without increasing overall encoding complexity, as the encoder transmits only the necessary modification parameters rather than complete motion data for each sub-block.
Solution Approach 2:
The decoder performs motion modification autonomously using received motion information and pre-stored reference block data. The decoder calculates SAD candidates, determines optimal motion vectors, and applies modifications without requiring additional encoding steps, thereby improving compression efficiency while maintaining encoding simplicity.
2Measurement precision
If motion modification is performed for all blocks, then decoding accuracy is improved, but processing time increases
Solution Approach 1:
Motion modification is applied selectively based on block characteristics such as size, prediction mode, and motion complexity. The decoder determines whether to perform motion modification for each block or sub-block individually, applying the technique only where it provides significant accuracy improvement, thereby avoiding unnecessary processing time for blocks where simple motion compensation suffices.
Solution Approach 2:
The decoder uses parameters such as block size, prediction mode, and motion vector differences to dynamically adjust the extent of motion modification performed. By changing processing parameters based on content characteristics, the system achieves high decoding accuracy for complex regions while maintaining fast processing for simpler regions.
3Measurement precision
If search region for motion modification is expanded, then motion accuracy is improved, but computational complexity increases
Solution Approach 1:
The encoder transmits pre-calculated motion information and reference block data to the decoder in advance. This preliminary action allows the decoder to perform motion modification with an expanded search region without requiring real-time computation of all possible motion vectors, as the reference data is already available from previous encoding steps.
Solution Approach 2:
The decoder performs motion modification with a search region that may be larger than strictly necessary, using SAD candidate calculations to evaluate multiple potential motion vectors. This partial excessive action ensures high motion accuracy by considering more candidates than the minimum required, while the actual computational burden is reduced by leveraging pre-transmitted reference data and selective processing based on block characteristics.
Data Source
AI summary
A video coding method and device, according to the present invention, determine whether motion compensation is performed by sub-block unit, determine a search area for motion compensation of a current block, calculate a plurality of SAD candidates with respect to the search area, derive delta motion information of the current block on the basis of the plurality of SAD candidates, and can compensate for motion information of the current block by using pre-generated motion information and the delta motion information of the current block.