CABAC Escape Coding With Capped Prefix Length
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video data compression and decompression systems face inefficiencies in entropy encoding, particularly with the CABAC technique, which results in suboptimal bit usage and increased decoding complexity due to the exponential-Golomb coding method, leading to high worst-case escape code lengths.
Innovation Solution
The proposed solution caps the prefix length of escape codes at a maximum value dependent on the bit depth of the data values, using a truncated-unary coding scheme and exponential function mapping to reduce the suffix length, thereby minimizing the worst-case escape code length to 32 bits, similar to HEVC version 1 standards.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If exponential-Golomb coding method is used in CABAC technique, then coding performance is maintained, but worst-case escape code length increases and decoding complexity increases
Solution Approach 1:
The escape code is segmented into two distinct parts: a prefix portion encoded using truncated-unary coding and a suffix portion encoded using exponential-Golomb coding. This segmentation allows the prefix to be limited in length (capped at maximum prefix length), thereby bounding the worst-case code length while maintaining the exponential-Golomb coding performance for the suffix portion.
Solution Approach 2:
The invention changes the encoding parameter structure by introducing a maximum prefix length parameter that caps the length of the prefix portion. This parameter modification transforms the unbounded exponential-Golomb coding into a bounded hybrid code, reducing worst-case escape code length while preserving coding efficiency through the exponential-Golomb suffix.
2Reliability
If exponential-Golomb coding method is used in CABAC technique, then coding performance is maintained, but worst-case escape code length increases
Solution Approach 1:
The escape code is divided into a prefix portion with bounded length (using truncated-unary coding) and a suffix portion (using exponential-Golomb coding). This segmentation ensures that the total escape code length is controlled by the maximum prefix length parameter, preventing unbounded code length growth while maintaining coding performance through the exponential-Golomb suffix.
Solution Approach 2:
By introducing and enforcing a maximum prefix length parameter, the invention transforms the parameter structure of escape code encoding. This parameter change bounds the worst-case escape code length while preserving the exponential-Golomb coding performance characteristics in the suffix portion.
Data Source
AI summary
A data encoding method includes encoding an array of data values as data sets and escape codes for values not encoded by the data sets, an escape code including a prefix portion and a non-unary coded suffix portion having a length, in bits, dependent upon a value encoded by the prefix portion according to a relationship such that, for at least some values encoded by the prefix portion, the length of the non-unary coded suffix portion is greater than the length, in bits, of the prefix portion.