Residual Entropy Coding Using Run-Length and Huffman Stages
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding technologies require a low complexity, simple, and fast entropy coding scheme for residual data compression, particularly in hybrid backward-compatible coding systems, as contemporary methods like CAVLC and CABAC are computationally expensive and inefficient for enhancement streams.
Innovation Solution
A method involving run-length coding followed by Huffman coding is applied to residual data, utilizing the unique characteristics of zero values and data frequency to encode residual data efficiently, with fixed-length symbols and flags to facilitate decoding and further entropy coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If contemporary entropy coding methods (CAVLC/CABAC) are used for residual data compression, then data compression capability is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent segments the residual data into runs of consecutive zero values and non-zero values, applying different coding strategies to each segment. This segmentation simplifies the coding process by focusing on the statistical characteristics of each segment type, achieving compression without the complexity of full CAVLC/CABAC implementations.
Solution Approach 2:
The patent changes the coding parameters by using fixed-length codes for run-length values and variable-length Huffman codes for non-zero residual values. This parameter adjustment optimizes the balance between compression ratio and computational complexity, avoiding the exhaustive context modeling of CABAC while maintaining better compression than simple RLE.
2Loss of information
If contemporary entropy coding methods (CAVLC/CABAC) are used for residual data compression, then data compression capability is improved, but processing speed decreases
Solution Approach 1:
The patent performs preliminary run-length encoding to identify and compress sequences of zero values before applying Huffman coding to non-zero values. This preliminary action exploits the common characteristic of residual data (many zero values) to reduce the amount of data requiring complex entropy coding, thereby improving processing speed.
Solution Approach 2:
The patent extracts and separately codes the run-length information from the residual data, removing the burden of processing all data values through complex entropy coding. By taking out the zero-value runs and coding them efficiently with fixed-length codes, the system achieves faster processing while maintaining compression effectiveness.
3Device complexity
If simple run-length coding is applied to residual data, then computational complexity is reduced, but data compression effectiveness decreases
Solution Approach 1:
The patent merges run-length coding with Huffman coding in a two-stage process. The first stage (RLE) reduces data volume by compressing zero runs, and the second stage (Huffman) further compresses the remaining non-zero values and run-length information. This combination achieves compression effectiveness close to CAVLC/CABAC while maintaining lower computational complexity.
Solution Approach 2:
The patent uses run-length encoding as an intermediary step between the residual data and the final Huffman-coded bitstream. This intermediary process transforms the residual data into a format with fewer symbols and more predictable statistics, making the subsequent Huffman coding more effective and achieving better overall compression than either method alone.
Data Source
AI summary
There is provided a method of encoding a video signal, the method comprising: receiving an input frame; processing the input frame to generate at least one set of residuals data, the residuals data enabling a decoder to reconstruct the input frame from a reference reconstructed frame; and, applying a run-length coding operation to the set of residuals data, wherein the run-length coding operation comprises generating a run-length encoded bytestream comprising a set of symbols representing non-zero data values of the residuals data set and counts of consecutive zero values of the residuals data set. In certain embodiments the method comprises apply a Huffman coding operation to the set of symbols. A method of decoding is also provided as well as apparatuses and a computer readable medium.