Texture Compression With Iterative Color Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing texture compression methods, such as ASTC and BC, require complex encoding and decoding processes and significant memory, making them unsuitable for embedded systems with limited processing power and storage.
Innovation Solution
A method involving iterative updating of a color map and generating indexes for pixels, allowing for efficient texture compression with lower complexity, flexibility, and faster encoding/decoding, suitable for embedded systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If classic texture compression methods (ASTC, BC, ETC) are used, then image compression quality is improved, but device complexity and memory requirements increase significantly
Solution Approach 1:
The image is divided into multiple blocks, and each block is processed independently using a simplified compression algorithm. This segmentation allows the system to avoid the complex global optimization required by ASTC and BC methods, reducing computational complexity while maintaining acceptable compression quality for embedded systems
Solution Approach 2:
The patent uses a simplified color map representation with limited entries (e.g., 4x4 color map for 16x16 blocks) instead of the complex hierarchical structures used in classic methods. This disposable, lightweight approach trades some compression quality for significantly reduced computational requirements, making it suitable for embedded systems
2Manufacturing precision
If classic texture compression methods (ASTC, BC, ETC) are used, then image compression quality is improved, but memory usage increases significantly
Solution Approach 1:
The patent extracts only the essential color information needed for compression by creating a simplified color map from the original block. This extraction process removes redundant data and complex metadata required by classic methods, reducing memory usage while preserving the most important visual information
Solution Approach 2:
The patent changes the representation parameters by using a fixed-size color map with limited entries and simple index tables, rather than the flexible but memory-intensive parameter sets used in ASTC and BC. This parameter simplification significantly reduces memory requirements for embedded systems
3Manufacturing precision
If classic texture compression methods (ASTC, BC, ETC) are used, then compression capability is improved, but processing speed decreases due to complex encoding/decoding
Solution Approach 1:
The patent performs preliminary action by pre-calculating the color map for each block during encoding and storing it in a compact format. This preliminary preparation enables extremely fast decoding by simply looking up colors in the pre-computed map, reversing the typical trade-off where complex encoding yields simple decoding
Solution Approach 2:
The patent inverts the traditional approach by using a fixed, limited-size color map rather than trying to represent all original colors. This inversion simplifies both encoding and decoding operations, achieving faster processing speeds while maintaining acceptable compression capability for embedded applications
Data Source
AI summary
The present disclosure provides a system and method for effectively compressing an image with low bandwidth and memory requirements while retaining a high quality. The method includes obtaining a block of an image including a plurality of pixels, obtaining a preliminary color map, iteratively updating the preliminary color map to generate an updated color map, generating indexes for the plurality of pixels, wherein each of the plurality of pixels has an index pointing to a color in the updated color map, determining an average color corresponding to colors with same indexes, replacing the color in the updated color map by the average color to generate an averaged color map, and storing the averaged color map and the indexes of the plurality of pixels as a representation of the block of the image.