Video Entropy Coding Layout for CM and Non-CM Replication Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video compression technologies face inefficiencies in entropy coding and decoding due to interleaved processing of copying parameters with and without Context Model (CM) in high-resolution images, leading to reduced throughput.
Innovation Solution
The method involves dividing copying parameters into subparts with and without CM, classifying, merging, and re-sequencing them, and performing entropy coding or decoding on each group separately to avoid interleaving, thereby improving processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If entropy coding is performed sequentially on multiple copying parameters by taking a copying parameter as a unit, then the coding process can handle complex image patterns, but the interleaved entropy coding on bin parts with and without Context Model reduces processing throughput
Solution Approach 1:
The patent segments the entropy coding process by separating bin parts with Context Model from bin parts without Context Model. This segmentation allows independent processing of each type, eliminating the interleaving overhead and enabling optimized processing paths for each segment, thereby improving throughput while managing complexity.
Solution Approach 2:
The patent merges all bin parts with Context Model into one group and all bin parts without Context Model into another group. By combining similar processing requirements together, the system eliminates the need for frequent switching between different processing modes, reducing control complexity and improving overall processing throughput.
2Measurement precision
If copying parameters are processed in positional order of pixel segments, then the decoding accuracy is maintained, but the interleaved processing of CM and non-CM bin parts reduces entropy coding efficiency
Solution Approach 1:
The patent segments the copying parameters into two distinct groups: those requiring Context Model processing and those not requiring it. This segmentation maintains the necessary processing precision for each type while enabling parallel or sequential processing of homogeneous groups, thereby improving throughput without sacrificing decoding accuracy.
Solution Approach 2:
The patent performs preliminary classification of bin parts into CM and non-CM groups before the actual entropy coding process. This preliminary action organizes the data in advance, allowing the subsequent processing to proceed more efficiently without the overhead of interleaved processing, thus improving throughput while maintaining accuracy.
Data Source
Figure 1~1(d)
Figure 2
Figure 3
AI summary
Disclosed are an image encoding and decoding method, image processing device, and computer storage medium, comprising: entropy-encoding a plurality of replication parameters of a current encoding block according to their order after adjustment, to generate a video code stream comprising information of the plurality of replication parameters; said plurality of replication parameters comprising one or more types of replication parameter components. Parsing the video code stream, comprising the information of the plurality of replication parameters, of a decoding block; entropy-decoding said plurality of replication parameters to obtain binary code of said plurality of replication parameters; adjusting said binary code to obtain the values of said plurality of replication parameters; said plurality of replication parameters comprising one or more types of replication parameter components.