Halftone Processing for Thin Line Continuity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional halftone processing methods, such as the error diffusion and dither methods, often result in broken thin lines due to increased blanks between dots, leading to poor line continuity in printed images, especially for bright lines, and require high computational load for line detection and threshold value adjustment.
Innovation Solution
An image processing technique that uses a threshold value matrix to adjust threshold values dynamically based on pixel values, allowing for increased dot formation in areas where no dots are initially present, thereby reducing breaks in thin lines by redistributing threshold values within a predetermined area.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional halftone processing (error diffusion or dither method) is used, then gradation is represented by pseudo gradation using dot density, but thin lines appear broken due to increased blanks between dots
Solution Approach 1:
The patent applies different processing strategies to different regions of the image. Specifically, it identifies thin line regions and applies targeted threshold value adjustments only in those regions, while maintaining standard halftone processing in other areas. This local differentiation resolves the contradiction by preserving line continuity in critical areas without compromising overall image quality.
Solution Approach 2:
The patent dynamically changes threshold values based on local image characteristics. By detecting thin line regions and adjusting threshold values specifically in those regions, the patent modifies the parameter (threshold value) to control dot formation, thereby improving line continuity while maintaining gradation representation in other areas.
2Reliability
If threshold value adjustment is performed to prevent line breaks, then line continuity improves, but computational load and processing time increase
Solution Approach 1:
The patent segments the image processing task into distinct stages: first performing standard halftone processing, then detecting thin line regions, and finally applying threshold value adjustments only in detected regions. This segmentation reduces computational load by avoiding full-image reprocessing and focuses computational resources only where needed for line continuity.
Solution Approach 2:
The patent performs preliminary halftone processing before thin line detection and threshold adjustment. This preliminary action establishes the base dot pattern, and subsequent operations only modify specific threshold values in thin line regions, rather than reprocessing the entire image, thereby maintaining processing efficiency.
3Ease of manufacture
If dither method with blue noise threshold matrix is used, then processing is simpler than error diffusion, but line breaks are more noticeable due to dense and sparse dot distribution
Solution Approach 1:
The patent maintains the simplicity of the dither method by using a blue noise threshold matrix for general processing, but applies local quality enhancement by detecting thin line regions and adjusting threshold values specifically in those areas. This approach preserves processing simplicity while targeting line continuity improvements where most needed.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image processing apparatus is configured to generate a halftone image of an input image by performing quantization using a threshold value matrix in which different threshold values are arranged, wherein the threshold values of the threshold value matrix are allocated to pixels of the input image. The image processing apparatus comprises: a pixel position determination means for determining that a pixel position corresponding to a pixel having a pixel value greater than a predetermined value is a candidate of a pixel position whose allocated threshold value is to be replaced with a new threshold value, wherein the new threshold value is a threshold value allocated to a pixel having a pixel value equal to or less than the predetermined value, a threshold value determination means for determining the new threshold value to be allocated to the pixel corresponding to the candidate, for the candidate in a processing area of the input image, using a threshold value in the threshold value matrix allocated to one of the pixels having a pixel value equal to or less than the predetermined value in the processing area, wherein the processing area is an area in the input image that has a predetermined size smaller than a size of the threshold value matrix, wherein the threshold value determination means replaces the threshold value in the threshold value matrix allocated to the pixel corresponding to the candidate with the determined new threshold value and repeats the determination for each of a plurality of further processing areas being areas of the input image having the predetermined size, and a dither processing means for performing the quantization using the threshold value matrix with the new threshold values.