Image Encoding Residual Rearrangement for Transform Skip Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding methods using motion compensation prediction with multiple reference images face decreased encoding efficiency when the transform skip mode is applied, as orthogonal transform processing is not performed, leading to inefficient entropy encoding.
Innovation Solution
An image encoding device and decoding device that evaluate the similarity between reference images on a pixel-by-pixel basis, rearrange prediction residuals based on this evaluation, and encode them in a manner that prioritizes positions with lower similarity, thereby improving encoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If orthogonal transform processing is skipped in transform skip mode, then processing complexity is reduced, but encoding efficiency deteriorates because entropy encoding cannot be performed efficiently
Solution Approach 1:
The patent changes the scanning order parameter for entropy encoding based on the similarity characteristics of reference images. When transform skip mode is used, instead of applying fixed scanning orders, the system dynamically selects scanning orders (e.g., horizontal, vertical, diagonal) that match the directional characteristics of prediction residuals, enabling efficient entropy encoding without orthogonal transform
Solution Approach 2:
The patent introduces dynamic selection of scanning orders based on evaluated similarity characteristics between reference images. The scanning order is not fixed but adapts to the content characteristics, allowing the system to optimize entropy encoding efficiency for each block independently based on local image characteristics
2Measurement precision
If multiple reference images are used for motion compensation prediction, then prediction accuracy is improved, but the amount of data to be encoded increases
Solution Approach 1:
The patent applies different scanning orders to different regions or blocks based on local similarity characteristics. Each block is evaluated independently and assigned an appropriate scanning order, allowing efficient encoding of significant coefficients while compressing less important areas, thus reducing overall data quantity while maintaining local prediction accuracy
Solution Approach 2:
The patent extracts and prioritizes encoding of significant coefficients (non-zero prediction residuals) while compressing or skipping less significant data. By identifying and focusing encoding resources on important coefficients first, the system reduces the total amount of encoded data while preserving essential image information
Data Source
AI summary
An image encoding device encodes a block-based target image. The image encoding device comprises, in a transform skip mode in which orthogonal transform processing of the target image is skipped, a motion compensation predictor configured to generate a prediction image corresponding to the target image by performing motion compensation prediction using a plurality of reference images, an evaluator configured to evaluate a degree of similarity between the plurality of reference images on a pixel-by-pixel basis, a subtractor configured to calculate prediction residuals each indicating a difference between the target image and the prediction image on a pixel-by-pixel basis, a rearranger configured to rearrange the prediction residuals based on a result of evaluation by the evaluator and an encoder configured to encode the prediction residuals rearranged by the rearranger.