Image Processing Apparatus Motion Vector Accuracy Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing technologies, such as AVC and HEVC, face limitations in coding efficiency due to excessive memory bandwidth requirements, which restrict prediction modes and block sizes, and do not adequately consider the accuracy of motion vectors, leading to unnecessary decreases in coding efficiency.
Innovation Solution
An image processing apparatus and method that control inter prediction modes based on the accuracy of motion vectors, including integral and fractional accuracy, to optimize motion compensation and reduce memory bandwidth demands, while ensuring decoding feasibility by determining the possibility of decoding based on data reading amounts and resource constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If motion compensation with fractional accuracy is used to improve prediction precision, then manufacturing precision is improved, but loss of substance increases due to excessive memory bandwidth requirements
Solution Approach 1:
The patent dynamically adjusts the motion vector accuracy parameter (integer vs. fractional) based on the prediction mode and block size. By changing this parameter adaptively, the system achieves high prediction precision when needed while reducing memory bandwidth consumption when high precision is not necessary, thus resolving the contradiction between manufacturing precision and loss of substance
Solution Approach 2:
The patent introduces dynamic control of motion compensation parameters based on prediction mode and block size conditions. The system transitions between integer and fractional accuracy modes dynamically, allowing optimal balance between prediction precision and memory bandwidth usage different for each prediction scenario
2Loss of substance
If prediction modes and block sizes are restricted to reduce memory bandwidth, then loss of substance decreases, but productivity decreases due to limited coding options
Solution Approach 1:
The patent changes the motion vector accuracy parameter based on prediction mode and block size, enabling a broader range of prediction options without excessively increasing memory bandwidth. This allows the system to maintain high coding efficiency while controlling memory bandwidth usage through intelligent parameter selection
Solution Approach 2:
The patent segments the prediction space into different categories (prediction modes and block sizes) and applies different motion vector accuracy requirements to each segment. This segmentation allows restrictive memory bandwidth usage overall while permitting high-efficiency coding options where appropriate
3Loss of substance
If integer accuracy motion vectors are used to reduce memory bandwidth, then loss of substance decreases, but manufacturing precision deteriorates
Solution Approach 1:
The patent dynamically changes the motion vector accuracy parameter from integer to fractional based on prediction mode and block size conditions. This ensures memory bandwidth is reduced only when integer accuracy suffices, while fractional accuracy is applied when higher precision is needed for optimal prediction
Data Source
AI summary
An image processing apparatus includes a controller which controls candidates for an inter prediction mode in which motion compensation is performed based on accuracy of a motion vector, a motion search unit which performs a motion search only in the inter prediction modes allowed by the controller to be the candidates to generate a motion vector, and a motion compensation unit which performs the motion compensation in the inter prediction mode selected from the candidates in each of which the motion vector is generated by the motion search unit to generate a predicted image. The present disclosure is applicable to the image processing apparatus.