Texture-Pattern-Adaptive Partitioned Block Transform for Image Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image encoding methods using 2D separable DCT or DWT struggle with ringing artifacts around edge orientations other than horizontal and vertical, leading to deteriorated visual quality, and require more than N 1-D transforms, which complicates hardware implementation and encoding efficiency.
Innovation Solution
Applying a texture-pattern associated invertible mapping to accumulate pixels within partitions of an image block, allowing for N 1-D transforms and enabling efficient encoding and decoding by adapting to multiple strip and non-directional texture patterns, and entropy encoding data for texture pattern reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If 2D separable DCT or DWT is used for encoding image blocks, then energy compaction and redundancy exploitation are improved, but ringing artifacts appear around edge orientations other than horizontal and vertical, deteriorating visual quality
Solution Approach 1:
The image block is divided into multiple partitions based on texture patterns (e.g., horizontal, vertical, diagonal, or non-directional patterns). Each partition is processed separately with appropriate 1-D transform directions, allowing the method to adapt to different edge orientations and texture characteristics within the same block, thereby reducing ringing artifacts while maintaining energy compaction.
Solution Approach 2:
The transform approach becomes dynamic by selecting different partitioning strategies and transform directions based on the detected texture pattern. Instead of applying a fixed 2D separable transform, the method adapts the transformation process to match the local image characteristics, optimizing both energy compaction and artifact reduction for each partition.
2Object-affected harmful factors
If direction-adaptive partitioned block transform is applied to address ringing artifacts, then visual quality is improved, but more than N first 1-D transforms or more than N second 1-D transforms are required, complicating hardware implementation
Solution Approach 1:
Different regions (partitions) of the image block are processed with locally optimized transform parameters. By identifying texture patterns and creating partitions that align with these patterns, each partition can use a standardized set of N 1-D transforms in appropriate directions, reducing the total number of transforms needed while maintaining adaptive performance for different local characteristics.
3Device complexity
If texture-pattern-adaptive partitioning with invertible mapping is applied, then the maximum number of required 1-D transforms is limited to N, but complex texture patterns require more sophisticated mapping and partitioning strategies
Solution Approach 1:
The texture pattern is detected and the invertible mapping is established before the actual transform process. This preliminary analysis allows the system to pre-determine the optimal partitioning strategy and transform directions, enabling the subsequent processing to proceed with a fixed maximum of N 1-D transforms per partition while still adapting to complex texture patterns.
Solution Approach 2:
The invertible mapping acts as an intermediary step that transforms the original pixel arrangement into a reordered arrangement that aligns with the detected texture pattern. This intermediate representation facilitates efficient partitioning and transform application, bridging the gap between complex texture patterns and the simplified N-transform constraint.
Data Source
AI summary
The invention is related to encoding an image block of an image using a partitioned block transform. The inventors recognized that applying a texture-pattern associated invertible mapping to the pixels of a first partition, said first partition resulting from partitioning said image block according to a current texture pattern with which said texture-pattern associated invertible mapping is associated, allows for limiting the maximum number of required first 1-D transforms to not exceeding a number of columns in the image block as well as limiting the maximum number of required second 1-D transforms to not exceeding a number of rows in the image block, also. Achieving limitation of maximum required 1-D transforms enables more efficient implementation on hardware and improves encoding performance but also allows for further partitions according to texture patterns which comprise at least one of multiple strips, texture patterns with highly unsymmetrical pixel distribution and non-directional texture patterns.