Single Pass Recompression for Image Memory Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image compression methods in Raster Image Processing Systems face challenges in efficiently managing memory resources, leading to inconsistent image quality and multiple rounds of recompression to achieve target memory sizes, especially when using hybrid lossy and lossless compression strategies.
Innovation Solution
The Single Pass Recompression (SPR) method determines a target recompression size by adjusting the mix of edge and image data compression methods, allowing for one additional compression pass to meet memory requirements, ensuring optimal image quality and efficient memory utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If JPEG compression with quantization is used to reduce memory size, then memory usage is reduced, but image quality deteriorates due to information loss
Solution Approach 1:
The patent performs segmentation and classification of image regions before compression, identifying flat regions that can be losslessly compressed. This preliminary action allows the system to preserve important image information while still achieving memory reduction through selective compression strategies.
Solution Approach 2:
The patent applies different compression strategies to different regions of the image: lossless compression for flat regions and lossy JPEG compression for image regions. This local quality approach ensures that important areas maintain high quality while less critical areas use more aggressive compression to reduce memory usage.
2Quantity of substance
If multiple rounds of recompression are performed to achieve target memory size, then memory requirement is satisfied, but processing time increases
Solution Approach 1:
The patent performs region segmentation and classification before compression, and calculates estimated compressed sizes in advance. This preliminary analysis allows the system to determine the appropriate compression strategy and quantization level before actual compression, avoiding multiple trial-and-error recompression cycles.
Solution Approach 2:
The patent uses feedback from the segmentation and classification stage to guide the compression process. By estimating the compressed size based on region characteristics and selecting appropriate quantization levels accordingly, the system can achieve target memory sizes in a single compression pass rather than through multiple recompression attempts.
3Quantity of substance
If quantization level is increased to reduce compressed image size, then memory usage decreases, but compression ratio prediction becomes difficult
Solution Approach 1:
The patent performs preliminary segmentation and classification of image regions, and estimates the compressed size of each region based on its characteristics. This preliminary estimation allows the system to predict the overall compression ratio and select appropriate quantization levels without needing to perform actual compression first.
Solution Approach 2:
The patent changes the quantization level parameter based on the characteristics of different image regions. By adjusting the quantization level for flat regions versus image regions differently, the system can control the overall compressed size while maintaining predictability through the preliminary estimation process.
4Loss of information
If hybrid compression strategy is used to maintain image quality, then compression ratio is reasonable, but memory management becomes complex
Solution Approach 1:
The patent segments the image into flat regions and image regions, allowing different compression strategies to be applied to each segment. This segmentation simplifies the management of hybrid compression by providing clear boundaries between regions that require different treatment, reducing the complexity of overall compression strategy management.
Solution Approach 2:
The patent applies different compression qualities to different regions: lossless compression for flat regions and lossy compression for image regions. This local quality approach maintains overall image quality while simplifying memory management by clearly defining which regions use which compression method, making the hybrid strategy easier to manage.
Data Source
AI summary
Disclosed is a method of compressing an image to be stored in a memory to satisfy a memory requirement. A size of a region having a uniform colour in the image is determined. The determined size of the region is compared with the candidate values of the region size threshold. A value is selected from the candidate values as the region size threshold based on the comparison between the estimated data amounts to satisfy the region size threshold and the memory requirement. The edges constituting the region which satisfy the selected value of the region size threshold are compressed losslessly.