Image Smoothing for Halftone Show-Through Elimination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques struggle to accurately eliminate show-through effects in scanned documents, particularly when the halftone dot frequency is around one-half of the reading sampling frequency, leading to inconsistent results within the same halftone dot region.
Innovation Solution
An image processing apparatus that includes a reading unit, a smoothing unit, a calculation unit, a determination unit, and a correction unit to perform brightness correction based on the degree of variation in signal values within a target region, effectively reducing show-through by calculating variance and applying correction amounts to improve image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If show-through elimination processing is performed on low-density portions using variance thresholding, then show-through components are reduced, but halftone dot regions with frequency around Nyquist are incorrectly processed due to inconsistent variance values
Solution Approach 1:
The patent applies smoothing processing as a preliminary step before variance calculation to preprocess the image data. This smoothing operation uniformizes the variance values across halftone dot regions, ensuring that regions with the same halftone dot density have consistent variance values regardless of their position or the reading optical resolution. This preliminary action prevents the inconsistent classification that would otherwise occur in regions where halftone dot frequency is around the Nyquist frequency.
2Object-affected harmful factors
If overall image density is decreased to reduce show-through, then show-through visibility is reduced, but low-density images on the front side may disappear
Solution Approach 1:
The patent implements local quality by performing show-through elimination processing selectively based on variance values calculated for different regions. Instead of uniformly decreasing overall image density, the system calculates variance values for each pixel's surrounding region and applies show-through elimination only to pixels with variance values below the threshold, indicating show-through components rather than legitimate low-density image content. This localized approach preserves front-side low-density images while eliminating show-through.
Solution Approach 2:
The patent changes the parameter used for show-through detection from direct density values to variance values calculated from smoothed image data. By transforming the detection parameter to variance (which measures local intensity variation), the system can distinguish between show-through components (low variance) and halftone dot patterns (high variance), enabling selective processing that removes show-through while preserving legitimate image content including low-density regions.
3Measurement precision
If reading optical resolution is increased to resolve halftone dots accurately, then halftone dot detail is improved, but show-through elimination becomes more difficult due to Nyquist frequency effects
Solution Approach 1:
The patent applies smoothing processing as a preliminary step before variance calculation to preprocess the image data. This smoothing operation uniformizes the variance values across halftone dot regions, ensuring that regions with the same halftone dot density have consistent variance values regardless of their position or the reading optical resolution. This preliminary action prevents the inconsistent classification that would otherwise occur in regions where halftone dot frequency is around the Nyquist frequency.
Data Source
AI summary
An image processing method includes reading a document on which a halftone-processed image is printed, performing smoothing processing on image data acquired by reading the document, setting a target region of a predetermined size to the smoothed image data and then calculating a degree of variation in signal values of a plurality of pixels included in the target region, determining a brightness correction amount for the target region based on the calculated degree of variation and the signal values of the plurality of pixels included in the target region, and performing brightness correction on the target region in the image data acquired by reading the document, based on the determined brightness correction amount.