Hybrid Image Compression via Region Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for compressing raster image processing (RIP) output fail to achieve satisfactory performance due to coarse granularity in selecting compression methods, leading to unacceptable memory and computation overheads, and the challenge of maintaining even quality across pages while ensuring a 'guaranteed-fit' memory size.
Innovation Solution
A hybrid compressor analyzes pixels to identify regions with substantially identical colors, representing these regions through edge data and fill colors, and uses pixel-based encoding for other pixels, ensuring efficient compression and decompression by encoding boundaries and fill colors to achieve high compression ratios without sacrificing quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If lossy pixel-based compression (JPEG, wavelet) is used, then compression ratio is improved, but visual quality deteriorates due to visible artifacts in text and graphics with sharp transitions
Solution Approach 1:
The patent applies different compression methods to different regions of the image: lossless edge-based compression for text and graphic regions with sharp transitions, and lossy pixel-based compression for image regions with smooth transitions. This local differentiation resolves the contradiction by preserving visual quality where needed while achieving high compression where acceptable.
Solution Approach 2:
The patent segments the image into different region types (text, graphic, pattern, image regions) based on characteristics such as colour transitions and spatial frequency. This segmentation enables selective application of appropriate compression methods to each region, resolving the quality-compression ratio trade-off.
2Manufacturing precision
If lossless pixel-based compression (JPEG-LS) is used, then visual quality is improved, but memory requirements worsen as resolution and colour depth increase
Solution Approach 1:
The patent applies lossless compression only to text and graphic regions where sharp transitions require precise preservation, rather than uniformly applying it to the entire image. This local application reduces overall memory requirements while maintaining necessary visual quality.
Solution Approach 2:
By segmenting the image into region types, the patent applies memory-intensive lossless compression only to regions where it is necessary (text and graphic regions), while using more memory-efficient methods for image regions, thus resolving the memory quality contradiction.
3Quantity of substance
If edge-based compression is used, then memory requirements are improved, but compression ratio worsens for natural images with many colours
Solution Approach 1:
The patent applies edge-based compression specifically to text and graphic regions where it achieves excellent compression ratios, while using pixel-based methods for natural image regions. This local differentiation resolves the contradiction by matching the compression method to the content type.
Solution Approach 2:
The segmentation of image regions enables the patent to apply edge-based compression only where it is effective (text and graphic regions with single colours), while using pixel-based compression for regions where it performs poorly (natural images with many colours), thus resolving the compression ratio-memory contradiction.
4Productivity
If hybrid compression methods are used, then compression performance is improved, but device complexity worsens due to multiple encoding/decoding paths
Solution Approach 1:
The patent performs preliminary classification of image regions into different types before compression. This pre-segmentation simplifies the subsequent compression process by determining which encoding method to use for each region in advance, reducing the complexity of real-time decision-making during compression.
Solution Approach 2:
By segmenting the image into distinct region types, the patent creates clear boundaries between areas requiring different compression methods, simplifying the control logic and reducing device complexity compared to continuous or adaptive methods.
Data Source
AI summary
A Hybrid Compressor and method of compressing a tile of pixels within an image are disclosed. An Edge Processor of the Hybrid Compressor analyzes pixel runs of the tile and generates boundary data defining one or more regions comprising pixels of substantially identical color. A Palette Processor of the Hybrid Compressor generates data defining a color value for each region. An Image Processor of the Hybrid Compressor generates a representation of pixels not included in the one or more regions.