Variable Bit Image Compression via Pixel Difference Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image data compression methods, especially lossy compression techniques, result in artifacts and require significant computational effort, while lossless methods struggle with efficiency and parallelization, particularly in high-resolution image processing for applications like digital cameras.
Innovation Solution
A method that converts pixel values into difference values using estimation rules, encoding these difference values with binary words of varying lengths, allowing for parallel encoding and decoding, thereby achieving high compression with minimal artifacts and efficient processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of stationary object
If lossy compression methods are used to achieve high compression factors, then storage space is reduced, but image quality deteriorates with artifacts and blurriness
Solution Approach 1:
The patent changes the parameter of bit allocation from fixed to variable based on pixel value frequency. By analyzing that certain pixel values occur more frequently than others, the method allocates fewer bits to frequent values and more bits to rare values, achieving compression without loss of image quality information
Solution Approach 2:
The patent performs preliminary analysis of pixel value frequency distribution before encoding. By pre-determining which pixel values occur most frequently, the compression method can optimally allocate bit lengths for each pixel value, enabling efficient compression while preserving all original image information
2Volume of stationary object
If variable bit encoding is used to achieve compression, then data volume is reduced, but processing speed decreases due to poor parallelizability
Solution Approach 1:
The patent segments the image data processing into independent pixel value encoding operations. Each pixel value is encoded independently based on its frequency category, allowing parallel processing across multiple pixels simultaneously. This segmentation enables variable bit encoding to be parallelized effectively
Solution Approach 2:
The patent changes the encoding parameter from sequential variable bit processing to parallel categorical encoding. By grouping pixel values into frequency-based categories with predetermined bit allocations, the method enables simultaneous encoding of multiple pixels while maintaining compression efficiency
3Volume of stationary object
If frequency domain transformation is used to compress image data, then compression is achieved, but computational effort increases significantly
Solution Approach 1:
The patent extracts the essential compression mechanism from complex frequency domain transformations. Instead of using DCT or wavelet transforms, it directly exploits the statistical property of pixel value frequency distribution, achieving compression through simple counting and categorical encoding without requiring computationally intensive mathematical transformations
4Manufacturing precision
If high spatial resolution is used to improve image quality, then detail accuracy increases, but data volume increases considerably
Solution Approach 1:
The patent changes the data representation parameter from uniform fixed-bit encoding to variable-bit encoding based on pixel value frequency. By allocating different bit lengths to different pixel values according to their occurrence frequency, the method efficiently represents high-resolution image data with reduced volume while preserving all original information
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for compressing image data is described in which a difference value is calculated for a given pixel. This difference value corresponds to a predetermined relation between the pixel value and the estimated pixel value of the given pixel and is represented by a first binary word of a predetermined length. The first binary word is encoded by at least one second binary word of a predetermined length. The first length is compared to the second length to determine whether the first length is shorter than the second length. If the first length is shorter than the second length, the first binary word is completely encoded by a single second binary word; otherwise, the first binary word is completely encoded by multiple second binary words. Furthermore, a method for decompressing image data is described, which is the reverse of the compression method.