Wavelet Image Compression via Tile Segmentation and Modulo Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data compression methods for graphics rendering systems are expensive for on-the-fly operations, lack random access, and are not suitable for real-time applications, especially when transitioning between lossy and lossless compression.
Innovation Solution
A method using wavelet transforms with modulo arithmetic to compress two-dimensional image data, dividing the data into arrays, performing prediction and encoding with entropy schemes, and optimizing for parallel processing to achieve symmetric encoding and decoding costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If general lossy image compression methods (JPEG, GIF) are applied to entire images, then compression ratio is improved, but processing cost increases and random access is lost
Solution Approach 1:
The patent divides the image data into multiple tiles or blocks, allowing independent compression and random access to specific regions. This segmentation enables partial processing rather than entire image processing, reducing processing cost while maintaining compression efficiency.
Solution Approach 2:
The patent applies different compression strategies to different regions based on their importance and characteristics. Critical regions use lossless or higher quality compression, while less important regions use more aggressive lossy compression, optimizing the balance between compression ratio and processing cost.
2Reliability
If lossless compression methods (PNG, ZIP) are used, then data integrity is improved, but processing cost and time increase
Solution Approach 1:
The patent applies lossless compression only to essential regions or data elements, while using lossy compression for other parts. This partial application of lossless compression maintains data integrity where needed while improving overall processing speed through selective compression strategies.
3Ease of operation
If texture compression schemes are used, then random access is improved, but encoding complexity increases significantly
Solution Approach 1:
The patent divides texture data into tiles that can be independently encoded and accessed. This segmentation enables random access to specific texture regions without decoding the entire texture, improving ease of operation while distributing encoding complexity across multiple smaller units.
4Quantity of substance
If Rasmusson et al's RGB compression system is used, then compression capability is improved, but register size requirements increase
Solution Approach 1:
The patent transforms color data from RGB space to alternative color spaces (such as YCbCr or other perceptually optimized spaces) that allow for more efficient compression with smaller register requirements. This parameter transformation enables maintaining compression capability while reducing the volume of data that must be processed and stored.
5Ease of manufacture
If sequential prediction schemes are used, then implementation simplicity is improved, but parallel processing capability is lost
Solution Approach 1:
The patent divides the prediction process into independent tile-based units that can be processed in parallel. Each tile maintains its own prediction context, allowing simultaneous processing of multiple tiles without interfering with each other, thus enabling parallel processing while keeping implementation relatively simple through modular design.
Data Source
Figure 1
Figure 2
Figure 3(a)~3(d)
AI summary
The invention provides a system and method for compressing an electronic image data set. The image is divided into a plurality of arrays, each of which are separately transformed using a wavelet transformation. The resulting wavelet coefficients are then encoded using an entropy encoding scheme to provide a compressed data set.