Polar Coding of RLE Bit Planes for Faster Lossless Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional entropy coding methods, such as Huffman coding, have low throughput and coding efficiency due to their variable-length coding schemes, which result in relatively slow data processing and compression performance in image and video compression applications.
Innovation Solution
A coding method that involves run-length encoding (RLE) data, sorting based on distribution probabilities, mapping, and polar coding to increase compression throughput, utilizing a polar code for parallel compression and implementing fast lossless compression by determining code sequences and decoding vectors based on occurrence probabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If variable-length coding scheme is used in conventional entropy coding, then compression ratio is improved, but throughput and coding efficiency deteriorate
Solution Approach 1:
The patent divides the data into multiple bit planes based on significance levels, and processes each bit plane independently using polar codes. This segmentation allows parallel processing of different bit planes, significantly improving throughput while maintaining compression efficiency through systematic handling of each segment.
Solution Approach 2:
The patent transforms the data representation from variable-length codes to fixed-length polar code words by changing the encoding parameters. Each bit plane is encoded using polar codes with specific length and rate parameters, converting the variable-length output into fixed-length structured data that enables efficient parallel processing and hardware implementation.
2Loss of substance
If variable-length coding scheme is used in conventional entropy coding, then compression ratio is improved, but coding efficiency deteriorates
Solution Approach 1:
The patent segments the compressed data into multiple bit planes and processes each plane independently with polar codes. This segmentation improves coding efficiency by enabling parallel encoding operations and avoiding the sequential processing bottleneck inherent in variable-length decoding, while still achieving effective compression through the combined representation.
Solution Approach 2:
The patent maintains continuous useful action by processing multiple bit planes in parallel rather than sequentially. The independent polar code encoding of each bit plane allows simultaneous computation, eliminating idle time and maintaining continuous productive operation throughout the encoding process.
3Productivity
If parallel polar coding is implemented, then compression throughput is increased, but system complexity increases
Solution Approach 1:
The patent divides the compression system into independent bit plane processors, each handling a specific significance level. This segmentation reduces system complexity by breaking down the complex parallel processing task into multiple simpler, identical modules that can be independently implemented and managed, while still achieving high throughput through their parallel operation.
Solution Approach 2:
The patent uses fixed parameters for polar codes in each bit plane (code length, rate, and structure), which simplifies the implementation complexity compared to variable parameters. The standardized polar code parameters across bit planes enable reuse of encoding templates and reduce the computational overhead of parameter management, balancing throughput improvement with acceptable system complexity.
Data Source
AI summary
The present disclosure provides example coding method, apparatus, and computer-readable storage medium. One example method includes sorting a plurality of pieces of run-length encoding (RLE) data based on their distribution probabilities. The plurality of pieces of sorted RLE data are mapped onto a plurality of pieces of reassembled data. The plurality of pieces of reassembled data are expanded into a plurality of pieces of binary data. A first matrix is generated based on the plurality of pieces of binary data. A code sequence is determined by processing a source signal, where the code sequence comprises a first set and a second set. A first vector is decoded to output a third set, where the third set includes one or more bits reserved after processing the source signal and one or more locations at which a decoding error occurs in one or more bits discarded after processing the source signal.