Edge-Aware Binary Thresholding Without Dither Matrix
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image processing devices using dither matrices for thresholding often result in unnatural patterns near image edges, leading to a loss of image quality due to the formation of finely textured patterns.
Innovation Solution
An image processing device that determines whether to use a dither matrix or not based on specific conditions, setting binary values without the matrix for pixels between edge and reference pixels where pixel values monotonically increase, thereby avoiding the use of dither matrices in areas with intermediate tones near edges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a dither matrix is used for thresholding, then the binary image data can be generated with controlled patterns, but unnatural patterns appear near edges and image quality is degraded
Solution Approach 1:
The patent applies different thresholding methods to different regions of the image. Specifically, it uses a first thresholding method (without dither matrix) for regions containing edges and intermediate tones, and a second thresholding method (with dither matrix) for other regions. This local differentiation eliminates unnatural patterns at edges while preserving the beneficial pattern control of dither matrices in non-edge regions.
Solution Approach 2:
The patent segments the image into different regions based on edge detection and intermediate tone identification. By dividing the image into edge regions and non-edge regions, it can selectively apply appropriate thresholding methods to each segment, avoiding the application of dither matrices in regions where they would create unnatural patterns.
2Stability of the object's composition
If dither matrices are applied to all pixels, then consistent thresholding is achieved, but image quality is lost in critical areas with intermediate tones near edges
Solution Approach 1:
The patent implements local quality by applying different thresholding strategies to different spatial locations. It identifies pixels with intermediate tones near edges and applies a specialized thresholding method (first thresholding method) to these critical pixels, while using the conventional dither matrix-based method (second thresholding method) for other pixels, thus maintaining both consistency and quality.
Solution Approach 2:
The patent introduces dynamic adaptability in the thresholding process by selecting between two different thresholding methods based on the characteristics of each pixel region. The system dynamically switches from the conventional dither matrix method to the edge-aware first thresholding method when intermediate tones near edges are detected, making the thresholding process adaptive rather than static.
Data Source
AI summary
An image processing device includes a controller. The controller specifies an edge pixel based on image data, and sets a target pixel and a reference pixel from among a plurality of pixels. The target pixel is located between the edge pixel and the reference pixel. The controller determines whether a determination condition is satisfied for the target pixel. The determination condition indicates that a pixel value of the edge pixel, a pixel value of the target pixel, and a pixel value of the reference pixel monotonically increase in said order. The controller sets a binary value of the target pixel to a first value without using a dither matrix when the determination condition is satisfied, and sets the binary value of the target pixel to one of the first value and a second value by using the dither matrix when the determination condition is not satisfied.