Edge Correction Based on Color Variations in Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional edge detection methods in image processing detect not only character and line edges but also edges in photographs, leading to unnatural contours in images.
Innovation Solution
An image processing apparatus that targets pixels, counts identical, similar, and different colored pixels within a window, and determines character pixels based on specific threshold values to differentiate and enhance edge detection for character and line edges while minimizing edge detection in photographs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional edge detection methods are applied to all pixels, then edge detection coverage is improved, but unnatural contours appear in photographs
Solution Approach 1:
The patent applies different edge detection processing to different regions of the image based on local characteristics. Character/line regions undergo intensive edge processing while photograph regions do not, achieving local differentiation of processing intensity to avoid unnatural contours in photographs while maintaining edge detection coverage for text and lines.
Solution Approach 2:
The image is segmented into different types (character/line regions vs. photograph regions) based on color uniformity characteristics. By dividing the image into these segments, the patent applies appropriate edge detection only to character/line regions, resolving the contradiction between comprehensive edge detection and avoiding photograph artifacts.
2Manufacturing precision
If intensive edge processing is applied to all pixels, then edge reproducibility is improved, but processing complexity increases
Solution Approach 1:
The patent implements local quality assessment by evaluating color uniformity in windows around each pixel. Based on this local assessment, only pixels in character/line regions (with high color uniformity) undergo intensive edge processing, while photograph pixels (with low color uniformity) receive minimal processing, thereby improving edge reproducibility without uniformly increasing processing complexity across the entire image.
Data Source
AI summary
An image processing apparatus counts at least one of the number of pixels having an identical color to a target pixel, the number of pixels having a similar color to the target pixel, and the number of pixels having a different color from the target pixel in a target window, and determines an attribute of the target pixel based on a result of the counting.