Coding Artifacts Removal Using Original Pixel Values
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image coding standards incorrectly determine target pixels for removing coding artifacts, leading to blurring of object edges and decreased coding efficiency due to sequential processing of horizontal and vertical edges.
Innovation Solution
A method that determines target pixels for coding artifacts removal based on pixel values of adjacent pixels that have not been filtered, allowing for simultaneous processing of horizontal and vertical edges to avoid misidentifying object edges as filtering targets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If horizontal and vertical edges are sequentially processed for coding artifacts removal, then the processing can be completed in order, but target pixels are wrongly determined and object edges are blurred
Solution Approach 1:
The patent applies preliminary action by determining all target pixels for both horizontal and vertical edges before performing any filtering operations. This is achieved by using a target pixel determination unit that identifies pixels requiring filtering based on gradient calculations, and then using a pixel value holding unit to preserve original pixel values. This allows subsequent filtering to be performed on accurately identified target pixels without the interference of sequential processing artifacts, thereby preventing object edge blurring while maintaining coding efficiency.
2Device complexity
If coding artifacts are removed from sequentially processed edges, then processing is simplified, but the probability of degrading image quality increases
Solution Approach 1:
The patent applies the copying principle by creating and maintaining copies of original pixel values through the pixel value holding unit. When determining target pixels for vertical edges, the system references copied pixel values from the holding unit rather than using already-filtered horizontal edge pixels. This ensures that target pixel determination is based on original, unaltered data, preventing false identification of object edges as coding artifacts. The copying mechanism preserves data integrity while enabling simultaneous processing of multiple edges, thus improving image quality reliability without significantly increasing processing complexity.
Data Source
AI summary
A method of removing coding artifacts in first and second processed edges that are processed edges in an image signal and in different directions includes: determining, based on at least pixel values of first adjacent pixels adjacent to the first processed edge, a first target pixel from which coding artifacts are to be removed for the first processed edge (S11); removing coding artifacts from the determined first target pixel (S13); determining, based on at least pixel values of second adjacent pixels adjacent to the second processed edge, a second target pixel from which coding artifacts are to be removed for the second processed edge (S12); and removing coding artifacts from the determined second target pixel (S14), wherein at S12, the second target pixel is determined based on pixel values of the second adjacent pixels from which the coding artifacts are not removed at S13.


