Adjacent-Block Weighted Prediction for Lower-Bitrate Image Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding techniques, such as VVC, do not sufficiently reduce code amount when using prediction images synthesized by averaging inter and intra prediction images.
Innovation Solution
An image encoding technique that generates a prediction image by synthesizing an in-screen block copy prediction and an intra prediction, with a weighting process determining a weighting parameter based on adjacent block information, and encodes the difference between the generated prediction image and the image block.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a prediction image is generated by simply averaging inter prediction image and intra prediction image, then the synthesis prediction can be obtained, but the code amount reduction is insufficient
Solution Approach 1:
The patent changes the parameter of synthesis by introducing weighted averaging instead of simple averaging. The weighting parameter α allows dynamic adjustment of the contribution from intra prediction and inter prediction, enabling optimization of prediction accuracy and code amount reduction based on block characteristics
Solution Approach 2:
The patent makes the synthesis process dynamic by determining the weighting parameter α based on prediction mode information of adjacent blocks. This allows the system to adaptively adjust the synthesis strategy according to local image characteristics, improving compression efficiency
2Productivity
If weighting processing is applied to synthesis prediction with parameter determined by adjacent block information, then compression rate and image quality are improved, but processing complexity increases
Solution Approach 1:
The patent implements self-service by deriving the weighting parameter α from existing prediction mode information of adjacent blocks that is already available during decoding. This avoids the need for additional complex calculations or external parameters, allowing the system to automatically adapt to local characteristics using readily available data
Solution Approach 2:
The patent performs preliminary action by determining the weighting parameter based on adjacent block prediction modes before generating the synthesis prediction image. This pre-determination of weights based on neighboring block characteristics enables optimized compression without adding significant processing complexity during the main decoding pipeline
Data Source
AI summary
An image encoding method for encoding an image includes: a prediction image generation step of generating a prediction image of a synthesis prediction by performing synthesis processing of synthesizing a prediction image of an in-screen block copy prediction and a prediction image of an intra prediction different from the in-screen block copy prediction, to an encoding-target block; and an encoding step of encoding a difference between the prediction image generated in the prediction image generation step and a pixel value of an image of the encoding-target block, the synthesis processing includes weighting processing performed to the prediction image of the in-screen block copy prediction and the prediction image of the intra prediction that is different from the in-screen block copy prediction, and a weighting parameter of the prediction image of the intra prediction in the weighting processing is determined by information of an adjacent block adjacent to the encoding-target block.