Intra Image Coding With Gradient-Guided Angular Mode Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image coding methods face inefficiencies due to the high computational complexity of determining optimal coding modes in intra prediction, particularly with the increased number of angular-based directional modes in advanced encoders like H.266, which affects coding efficiency.
Innovation Solution
An image coding method that analyzes gradient histograms of coding units and sub-blocks to determine consistency in dominant gradient directions, selecting candidate coding modes based on rate-distortion costs to reduce unnecessary calculations and improve efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the number of angular-based directional coding modes is increased to improve prediction accuracy, then the coding precision is improved, but the computational complexity increases
Solution Approach 1:
The patent divides the coding unit into multiple sub-blocks and calculates gradient histograms for each sub-block separately. This segmentation allows the system to analyze local gradient characteristics without needing to evaluate all angular coding modes across the entire coding unit, thereby reducing computational complexity while maintaining coding precision through localized analysis.
Solution Approach 2:
The patent performs preliminary gradient histogram analysis and consistency determination before selecting candidate coding modes. By pre-calculating gradient directions and identifying consistent regions, the system narrows down the search space of coding modes beforehand, avoiding the need to compute rate-distortion costs for all possible angular modes and thus reducing overall computational complexity.
2Device complexity
If gradient histogram analysis and consistency determination are performed to reduce candidate coding modes, then the computational complexity is reduced, but the coding time increases due to additional calculations
Solution Approach 1:
The patent extracts only the dominant gradient direction from each sub-block's gradient histogram and compares it with the coding unit's dominant gradient direction. By taking out only the essential gradient direction information rather than processing complete histograms, the system reduces the amount of data to be processed while still achieving effective candidate mode filtering, thus balancing computational complexity reduction with coding time efficiency.
Solution Approach 2:
The patent performs gradient histogram analysis and consistency determination only for regions where it will significantly impact mode selection. By applying partial analysis to sub-blocks with high consistency and using simplified criteria for others, the system avoids excessive calculations while still achieving sufficient complexity reduction, optimizing the balance between computational load and coding time.
Data Source
AI summary
An image coding method, which relates to the field of image processing technologies, and in particular to the fields of video compression and image encoding/decoding technologies, includes: partitioning an image to be processed into a plurality of coding units; determining, for any coding unit, a gradient histogram of the coding unit and respective gradient histograms of a plurality of sub-blocks of the coding unit, according to gradients of pixels in the coding unit; determining consistency between a dominant gradient direction of each of the plurality of sub-blocks and a dominant gradient direction of the coding unit according to the gradient histograms; determining a set of candidate coding modes from a plurality of angular coding modes according to the consistency; and determining a target angular coding mode to encode the coding unit, according to rate-distortion costs of the angular coding modes in the set of candidate coding modes.