Image Block Sorting for Inter-Block Correlation Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image compression techniques do not effectively utilize inter-block correlations between digital images, leading to inefficient storage and transmission of large numbers of similar images.
Innovation Solution
A method that identifies image blocks from digital images, extracts their features, sorts them based on average feature differences, and compresses them using inter-block correlations, generating key blocks and block-difference data to reconstruct the images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional compression methods are used without considering inter-block correlations, then the compression process is simpler, but the compression efficiency is lower and more storage space is required
Solution Approach 1:
The patent divides images into multiple image blocks and processes them separately while maintaining their correlation relationships. Each block is compressed individually after sorting, allowing parallel processing while capturing inter-block similarities. This segmentation enables efficient compression by treating manageable portions independently while preserving overall image structure.
Solution Approach 2:
The patent performs preliminary sorting of image blocks based on feature differences before compression. By pre-organizing blocks in an ordered sequence that maximizes similarity between consecutive blocks, the compression process can exploit these pre-established relationships more effectively. This preliminary arrangement reduces the computational burden during actual compression while improving efficiency.
2Loss of information
If image blocks are sorted based on average feature differences, then inter-block correlations are better utilized, but the sorting process requires additional computational resources
Solution Approach 1:
The patent extracts key image features from each block and uses only these extracted features for sorting purposes, rather than processing entire blocks. This extraction of essential characteristics reduces the computational burden of sorting while still capturing the necessary similarity relationships between blocks for effective compression.
Solution Approach 2:
The patent changes the parameter used for sorting from raw pixel data to averaged image features, which reduces the dimensionality and computational complexity. By transforming the sorting criterion to operate on condensed feature representations rather than full block data, the energy requirement for sorting is reduced while maintaining compression effectiveness.
3Quantity of substance
If more image blocks are compressed using inter-block correlations, then the compression ratio improves, but the reconstruction complexity increases
Solution Approach 1:
The patent uses a copying approach where compressed blocks are reconstructed by copying reference block data and applying difference adjustments. Instead of complex reconstruction algorithms, the system copies base block patterns and modifies them based on stored difference data, simplifying the reconstruction process while maintaining high compression ratios.
Solution Approach 2:
The patent applies partial compression strategies where not all blocks require full reconstruction complexity. By selectively applying different reconstruction levels based on block importance and correlation strength, the system achieves high overall compression while avoiding excessive complexity in reconstructing every single block with maximum fidelity.
Data Source
AI summary
A method for compressing images includes identifying a set of image blocks from a plurality of digital images. For each image block from the set, image features of the image block are extracted. The set of image blocks is sorted into an ordered sequence of image blocks based on an average image feature difference between sequentially sorted image blocks. The set of image blocks is compressed into compressed image data based on inter block correlations between image blocks in the ordered sequence.