Last-Coefficient Position Coding With Fewer Decoder Switches
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding standards face inefficiencies in arithmetic coding and decoding of last position information, particularly due to frequent switching between different arithmetic decoding methods, which hinders efficient processing.
Innovation Solution
An image coding method that binarizes and codes the last position information by alternating between context adaptive binary arithmetic coding and bypass coding, placing coded signals in a bit stream such that signals coded by the same method are consecutive, reducing the number of decoding method switches during decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If different arithmetic decoding methods are used for different parts of the last position information, then coding flexibility and adaptability are improved, but the number of decoding method switches increases, reducing processing efficiency
Solution Approach 1:
The last position information is divided into two separate components: significant flag information (indicating whether coefficients are zero or non-zero) and last position information (indicating the position of the last non-zero coefficient). Each component is coded using a different arithmetic decoding method optimized for its characteristics, allowing flexible adaptation while maintaining processing efficiency through clear separation of coding paths.
Solution Approach 2:
The patent implements dynamic switching between different arithmetic decoding methods (context-adaptive binary arithmetic coding and bypass coding) based on the specific characteristics of the data being coded. The decoding method is selected adaptively according to the type of information (significant flag or last position), optimizing processing efficiency for each data type while maintaining overall system flexibility.
2Measurement precision
If context adaptive binary arithmetic coding is used for all last position information, then coding precision is improved, but processing complexity and time increase
Solution Approach 1:
Instead of applying context-adaptive binary arithmetic coding to all last position information uniformly, the patent applies this precise coding method only to significant flag information where it provides the most benefit. For last position information, a simpler bypass coding method is used, reducing processing time while maintaining adequate precision for position data. This partial application of the more complex coding method optimizes the balance between precision and processing speed.
3Productivity
If bypass coding is used for all last position information, then processing speed is improved, but coding precision and compression ratio deteriorate
Solution Approach 1:
The patent applies bypass coding only to last position information where speed is critical and the data characteristics allow for efficient encoding. For significant flag information, context-adaptive binary arithmetic coding is used to achieve better compression ratios. This selective application ensures processing speed is optimized for position data while compression efficiency is maximized for flag data, preventing overall compression ratio deterioration.
Data Source
AI summary
A decoding method decodes last position information indicating horizontal and vertical positions of a last non-zero coefficient in a predetermined order within a current block to be decoded, the current block including plural coefficients. The decoding includes obtaining a bitstream including first, second, third and fourth partial signals, in this order, performing first arithmetic decoding on the first and the third partial signals respectively to obtain decoded first and decoded third partial signals, performing second arithmetic decoding on the second and the fourth partial signals respectively to obtain decoded second and decoded fourth partial signals, the second arithmetic decoding being different from the first arithmetic decoding, deriving a horizontal component of the last position information from the decoded first and decoded third partial signals, and deriving a vertical component of the last position information from the decoded second and decoded fourth partial signals.