Adaptive Entropy Coding Tables for Transform Coefficient Reordering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video coding and decoding methods lack a universal entropy coding/decoding method that can adapt to various scenarios, particularly in efficiently reordering and processing quantized transform coefficients for effective storage and transmission.
Innovation Solution
Adaptive methods using tables to determine target symbols based on occurrence probabilities and apply different transforms and scanning orders to convert data between prediction residues and coefficient arrays, with processing circuits to manage these operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a fixed scanning order is used for converting quantized transform coefficients, then the coding/decoding process is simple and fast, but the efficiency of grouping non-zero coefficients is suboptimal for different video content scenarios
Solution Approach 1:
The patent implements adaptive scanning orders that dynamically adjust based on the statistical properties of the video content being encoded. The scanning order is no longer fixed but adapts to different scenarios, allowing optimal grouping of non-zero coefficients for each specific video sequence or scene, thereby improving coding efficiency without requiring complex real-time analysis
Solution Approach 2:
The patent changes the scanning order parameter adaptively based on content characteristics. By modifying this single parameter (scanning order) according to the statistical properties of different video content, the system achieves improved compression efficiency across diverse scenarios without fundamentally changing the coding architecture or adding significant complexity
2Adaptability or versatility
If a universal entropy coding/decoding method is implemented to handle various coding scenarios, then the adaptability improves, but the complexity of determining optimal coding strategies increases
Solution Approach 1:
The patent develops a universal entropy coding/decoding framework that can handle multiple coding scenarios through adaptive table updates. The same core coding infrastructure is used across different scenarios, with the ability to adaptively update probability tables and scanning orders based on content characteristics, providing versatility without requiring separate coding paths for different scenarios
Solution Approach 2:
The patent incorporates feedback mechanisms where the coding system monitors the statistical properties of the video content and adapts the entropy coding tables and scanning orders accordingly. This feedback-driven adaptation allows the universal coder to optimize its performance for different scenarios automatically, reducing the need for manual configuration or complex decision logic
Data Source
AI summary
A method of converting first data into second data includes: determining a target symbol corresponding to the first data by utilizing a determining unit; and generating the second data corresponding to the determined target symbol. One of the first data and the second data is a syntax element, and the other of the first data and the second data is an entropy-coded result of the syntax element. The target symbol corresponding to the first data is determined adaptively according to occurrence probability of candidate syntax element values of the syntax element.