Dynamic Entropy Coding Grouping for Transform Mode
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing entropy coding methods for transform mode in display interface compression standards like VDC-M are inefficient due to fixed grouping schemes that do not account for varying sample value ranges within blocks, leading to increased bit requirements and reduced decoder throughput.
Innovation Solution
Implementing multiple grouping methods in an entropy encoder to form entropy coding groups based on factors such as block position, intra prediction mode, and color component, allowing for dynamic grouping of sample values within similar ranges, thereby optimizing bit representation and decoder efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed grouping scheme is used for entropy coding in transform mode, then the codec complexity is reduced, but the bit representation efficiency deteriorates and decoder throughput is reduced
Solution Approach 1:
The patent implements dynamic grouping methods that adapt to the characteristics of each block of quantized transform coefficients. Instead of using a fixed grouping scheme, the system evaluates grouping factors and selects from multiple grouping methods (first grouping method, second grouping method, third grouping method) based on the specific block characteristics, thereby optimizing decoder throughput without significantly increasing complexity
Solution Approach 2:
The patent changes the grouping parameters dynamically by evaluating multiple grouping factors (such as block position, transform coefficient characteristics) and selecting appropriate grouping methods. This allows the system to adapt the grouping structure to match the actual data distribution, improving bit representation efficiency and decoder throughput while maintaining manageable complexity through systematic evaluation criteria
2Ease of manufacture
If a fixed grouping scheme is used for entropy coding in transform mode, then the implementation is simplified, but the compression efficiency deteriorates due to increased bit requirements
Solution Approach 1:
The system transitions from static fixed grouping to dynamic adaptive grouping, where the grouping structure changes based on the characteristics of each block. This allows optimal grouping for each block type, reducing the number of bits required to represent the transform coefficients while maintaining implementation feasibility through systematic evaluation of grouping factors
Solution Approach 2:
The patent applies different grouping methods to different blocks based on their local characteristics. By evaluating grouping factors specific to each block (such as position in the transform domain, coefficient distribution patterns), the system tailors the grouping strategy to local data properties, thereby improving overall compression efficiency without requiring complete redesign of the entire encoding system
3Productivity
If multiple grouping methods are implemented dynamically, then the bit representation efficiency is improved, but the device complexity increases
Solution Approach 1:
While implementing dynamic selection among multiple grouping methods, the system manages complexity through structured evaluation of grouping factors and systematic selection criteria. The dynamic adaptation is achieved through algorithmic evaluation rather than complex hardware structures, balancing improved compression efficiency with acceptable encoder complexity
Solution Approach 2:
The system manages the complexity of multiple grouping methods by changing parameters systematically based on evaluated factors. Rather than implementing all possible grouping variations, the patent selects from a defined set of grouping methods using evaluation criteria, thereby achieving good compression efficiency while controlling the growth of encoder complexity through parameter-based selection rather than structural complexity
Data Source
AI summary
A system and method of forming entropy coding groups in an entropy encoder operating in a transform mode includes receiving a block of a first number of quantized transform coefficients as a current block of sample values and evaluating the current block of sample values using one or more grouping factors. In response to a determination that the current block meets a first grouping condition, a first grouping method is selected where the first grouping method forms a first entropy coding group with one sample value of a DC transform coefficient. In response to a determination that the current block meets a second grouping condition, a second grouping method is selected where the second grouping method forms a first entropy coding group with at least two sample values, one of the sample values being the DC transform coefficient.