Image Encoding Skip Region Selection for Coding Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing entropy encoding and decoding methods are inefficient, particularly as image or block sizes increase, leading to a rapid increase in coding data and suboptimal performance in context-adaptive arithmetic encoding and decoding.
Innovation Solution
The proposed method encodes the position of a reference coefficient within a transform block and selects a skip region based on this position. It then encodes skip region information and determines probability information for binarized transform coefficients, selecting an appropriate probability information table based on the use of a skip region.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If image or block sizes are increased to improve processing efficiency, then fewer blocks need to be processed, but the amount of data to be entropy encoded increases rapidly
Solution Approach 1:
The transform block is divided into multiple regions, with one region designated as a skip region where coefficients are assumed to be zero. This segmentation allows the encoder to process only the non-skip regions, reducing the amount of data that needs to be entropy encoded while maintaining the ability to handle large block sizes efficiently
Solution Approach 2:
The skip region is extracted from the transform block as a separate entity that does not require entropy encoding. By identifying and extracting this region based on the position of reference coefficients, the patent eliminates unnecessary encoding operations for regions that will contain zero coefficients, thereby reducing overall coding data volume
2Measurement precision
If context-adaptive arithmetic encoding is applied to improve compression performance, then encoding precision is improved, but the complexity of selecting appropriate probability information increases
Solution Approach 1:
Different probability information tables are assigned to different regions within the transform block. The skip region uses one set of probability characteristics while non-skip regions use another. This local differentiation allows the system to maintain high encoding precision by using appropriate probability models for each region while simplifying the overall selection process through region-based categorization
3Productivity
If skip region is used to reduce coding data amount, then encoding efficiency is improved, but the complexity of determining skip region and selecting probability information increases
Solution Approach 1:
The skip region is determined in advance based on the position of reference coefficients before the actual entropy encoding process begins. By performing this determination preliminarily and using straightforward criteria (reference coefficient positions), the patent reduces coding data amount efficiently while avoiding complex real-time decisions during the encoding process itself
Data Source
AI summary
The present invention relates to an image encoding/decoding method and device, and the image encoding method or device according to an embodiment of the present invention may encode a position of a reference coefficient within a current transform block to be encoded, and encoding skip region information of a skip region selected on the basis of the position of the reference coefficient. The skip region information may represent whether or not coefficients within the skip region have an identical coefficient value.