Bayer Image De-noise Processing via Multi-directional Pixel Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing circuits for Bayer pattern images, especially those using low-end hardware resources, often result in poor image quality and noise near edges due to oversimplification, necessitating a novel method for enhancing image quality and reducing noise.
Innovation Solution
A method involving a processing circuit with a detection module that calculates absolute differences and averages pixel values across multiple directions to generate detection values, and a de-noise module that performs noise reduction based on these values, ensuring high image quality even with low-end hardware.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional image processing circuits use low-end hardware resources, then material costs are saved, but image quality deteriorates and noise occurs near edges
Solution Approach 1:
The image processing is divided into multiple directional processing stages (first direction, second direction, third direction) with different processing strategies applied to different directions. This segmentation allows the system to achieve high image quality through multi-directional analysis while maintaining cost-effectiveness by using selective processing rather than uniformly complex operations across all directions.
Solution Approach 2:
Different processing methods are applied to different spatial locations and directions: directional gradient calculations are performed for edges in specific directions, while other areas use simpler processing. This local quality approach ensures high image quality is maintained in critical regions (edges) while using fewer computational resources in other areas, resolving the contradiction between cost and quality.
2Ease of manufacture
If conventional image processing circuits are oversimplified to save costs, then manufacturing costs are reduced, but noise near edges increases
Solution Approach 1:
The system performs preliminary directional gradient calculations and edge detection before final image processing. By预先 calculating gradients in multiple directions and identifying edge locations in advance, the system can apply noise reduction only where necessary (near detected edges) while maintaining other areas with simpler processing, thus reducing overall computational cost while eliminating edge noise.
Solution Approach 2:
The patent applies full multi-directional gradient processing only to regions where edges are detected, while using simplified processing for the remainder of the image. This partial action approach ensures that noise reduction is applied exactly where needed (at edges) without unnecessarily complicating the processing of the entire image, thereby controlling manufacturing costs while eliminating edge noise.
3Manufacturing precision
If multi-directional pixel value calculations are performed, then image quality is improved, but computational complexity increases
Solution Approach 1:
The computational process is segmented into distinct directional passes (first direction, second direction, third direction), each handling specific gradient calculations. This segmentation allows the complex multi-directional processing to be broken down into manageable stages, improving image quality through comprehensive directional analysis while controlling computational complexity through structured organization and selective application of processing methods.
Data Source
AI summary
A method for performing de-noise processing includes: with regard to each direction of a plurality of directions, summing up absolute values of differences between a plurality of sets of first pixel values around a target pixel of an image to generate a first detection value, and with regard to each direction of at least a portion of the directions, selectively averaging at least one set of second pixel values around the target pixel to generate a second detection value; sorting a plurality of pixel values around the target pixel and generating a third detection value accordingly; and with regard to a specific direction of the directions, performing de-noise processing on the target pixel according to at least the former two of the first detection value, the third detection value, and the second detection value. An associated apparatus is also provided.