Random-Access Image Compression With Selective Block Decompression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image data compression methods require decompressing entire blocks to access individual values, leading to inefficient data access for rendering units that need only a few pixels, and reducing memory bandwidth efficiency.
Innovation Solution
Compress image data using fixed-length compression algorithms that identify common base information and fixed-length parameters for each value, allowing individual values to be accessed without decompressing the whole block, by dividing image data into two-dimensional blocks and applying different compression algorithms to each block or sub-block based on their characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If traditional block-based compression is used, then data compression ratio is improved, but data access efficiency deteriorates because entire blocks must be decompressed to access individual values
Solution Approach 1:
The patent segments the compressed data structure into hierarchical levels (e.g., 4×4 pixel blocks divided into 4×4 groups of 4 pixels each). This segmentation allows selective decompression of only the necessary segments rather than entire blocks, resolving the contradiction by enabling both compression and efficient random access to individual pixel values or small groups of pixels.
Solution Approach 2:
The patent applies different decompression granularities to different regions or scenarios. By organizing data into fine-grained segments with local base values, the system enables local quality preservation where only accessed regions are decompressed, while maintaining overall compression for unaccessed regions.
2Speed
If memory bandwidth is increased to handle uncompressed data, then rendering speed is improved, but power consumption increases significantly
Solution Approach 1:
The patent implements partial decompression where only the necessary portions of compressed data are decompressed on-demand rather than decompressing entire blocks. This partial action reduces memory bandwidth requirements and power consumption while maintaining rendering speed by ensuring that only accessed data is processed.
3Loss of substance
If compression algorithms are made more complex to achieve better compression ratios, then data size is reduced, but device complexity increases
Solution Approach 1:
The patent uses a nested hierarchical structure where base values are stored at coarser levels and delta values are stored at finer levels. This nesting enables efficient compression by storing only differences from base values, while the hierarchical organization keeps the decompression logic manageable through systematic traversal of the nested structure.
Solution Approach 2:
The patent changes the parameter representation from storing absolute pixel values to storing relative delta values with respect to base values. This parameter transformation achieves better compression ratios by exploiting local redundancy, while the systematic approach to selecting and storing base values keeps the algorithm complexity manageable.
Data Source
AI summary
Methods and compression units for compressing a block of image data, the block of image data comprising a plurality of image element values, the image element values being divisible into at least a first value and a second value such that the block of image data comprises a two-dimensional block of first values, the method comprising: compressing a first data set comprising all or a portion of the two-dimensional block of first values in accordance with a first fixed-length compression algorithm to generate a first compressed block by: identifying common base information for the first data set; and identifying a fixed-length parameter for each first value in the first data set, the fixed-length parameter being zero, one or more than one bits in length; and forming a compressed block for the block of image data based on the first compressed block.