Image Block Approximation Using Widened Neighborhood for Video Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced video coding architectures face challenges in maintaining image quality while controlling encoding and decoding computation time and error resilience, particularly due to increased data complexity and hardware limitations.
Innovation Solution
A device comprising a working memory and a driver that analyzes digital image data streams, using a 'matching pursuit' algorithm to approximate current image blocks by linear combinations of already decoded or encoded blocks from a selected neighborhood, reducing critical data and enhancing robustness and compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If advanced coding architectures are used to compress more faithfully at comparable bit rate, then image quality is improved, but decoding and encoding computation time increases substantially
Solution Approach 1:
The patent segments the neighborhood of a block into two distinct parts: an approximation neighborhood (immediate surrounding blocks) and an approach base (widened neighborhood blocks). This segmentation allows the decoding process to first quickly reconstruct the approximation neighborhood, then use it to derive the current block, thereby reducing the computational burden compared to processing the entire neighborhood uniformly.
Solution Approach 2:
The patent performs preliminary decoding of the approximation neighborhood (immediate surrounding blocks) before decoding the current block. By having the approximation neighborhood decoded first and stored in memory, the current block can be derived more efficiently using this pre-decoded information, reducing overall computation time while maintaining image quality.
2Loss of information
If advanced coding architectures are used to improve compression, then data represents a condensation of information, but transmission errors tend to degrade the restored image
Solution Approach 1:
The patent applies different decoding strategies to different parts of the neighborhood: the approximation neighborhood (immediate surrounding blocks) is decoded with higher priority and stored in working memory, while the approach base (widened neighborhood) is used to derive the current block. This local differentiation ensures that critical nearby blocks are more robustly decoded, improving error resilience while maintaining compression efficiency.
Solution Approach 2:
The patent decodes and stores the approximation neighborhood (immediate surrounding blocks) in advance in working memory before decoding the current block. This preliminary action creates a buffer of reliable reference data that can compensate for potential transmission errors in the current block's data, thereby cushioning against error degradation while maintaining efficient compression.
3Manufacturing precision
If the size of encoded files is increased to improve image quality, then encoding rate increases, but hardware limitations require the speed to be limited
Solution Approach 1:
The patent uses partial action by selectively decoding only the approximation neighborhood (immediate surrounding blocks) with high priority and using it to derive the current block, rather than uniformly processing all neighborhood blocks with equal detail. This partial approach maintains image quality while reducing the overall encoding rate to fit hardware limitations.
Data Source
Figure 1
Figure 2~3
Figure 4~6
AI summary
A device for decoding a digital image stream, comprises: * a work memory (42) for data of already-decoded image blocks, * a driver (40) mounted so as to analyse an incident stream of digital image data, and devised so as to determine data to be decoded of a current block of complete data or data-parameters type, so as to store the complete data in the work memory (42) in the guise of data of already-decoded image blocks, and to call a decoding unit (48) with the data-parameters, * the decoding unit (48) reacting to the receipt of data-parameters by: - selecting certain already-decoded image blocks, forming an approximation base, which neighbours the current block according to a chosen rule, - calculating an approximation of the current block, on the basis of a linear algebraic function of the data of the approximation base, and - storing in the work memory (42), for the current block, data, derived from this approximation, of already-decoded image blocks.