Arithmetic Entropy Coding With Multi-Context Single-Cycle Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional arithmetic coding methods for image processing, such as those used in JPEG 2000, require high hardware resources and have low processing efficiency due to complex parallel processing structures.
Innovation Solution
A system and method for arithmetic entropy coding that processes at least two contexts in a single clock cycle using a single hardware logic, by obtaining coding parameters from lookup tables, determining probability interval values and lower limits, and performing normalization operations to improve efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional arithmetic coding method is used, then coding accuracy is maintained, but hardware resource consumption increases and processing efficiency decreases
Solution Approach 1:
The patent divides the arithmetic coding process into multiple independent processing units, each handling specific contexts. Multiple contexts are processed in parallel within a single clock cycle, segmenting the overall coding task into concurrent sub-tasks that reduce total processing time and hardware resource requirements.
Solution Approach 2:
The patent pre-calculates and stores probability values and coding parameters in lookup tables before the actual coding process. This preliminary preparation allows the coding operation to proceed with simple table lookups and basic arithmetic operations, significantly reducing the computational complexity and hardware resources needed during real-time processing.
2Speed
If parallel processing structure is implemented, then processing speed increases, but hardware resource consumption increases
Solution Approach 1:
The patent designs processing units that can handle multiple contexts using the same hardware resources. The same arithmetic coding logic and probability calculation units are reused across different context processing, allowing parallel processing of multiple contexts without proportionally increasing hardware resource consumption.
Solution Approach 2:
By pre-computing and storing probability values in lookup tables, the patent eliminates the need for complex real-time probability calculations during parallel processing. This preliminary action allows multiple processing units to operate in parallel using simple, resource-efficient operations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to systems and methods for coding. The methods may include receiving at least two contexts including a first context and a second context. The methods may include, for each of the at least two contexts, obtaining at least one coding parameter corresponding to the context from at least one lookup table. The methods may further include determining a probability interval value corresponding to the context based on a previous probability interval value and the at least one coding parameter. The methods may further include determining a normalized probability interval value corresponding to the context by performing a normalization operation on the probability interval value. The methods may further include determining a probability interval lower limit corresponding to the context based on a previous probability interval lower limit and the at least one coding parameter. The methods may further include determining a normalized probability interval lower limit corresponding to the context by at least performing the normalization operation on the probability interval lower limit. The methods may also include outputting at least one byte corresponding to the context based on the normalized probability interval lower limit.