Image Processing Apparatus Show-Through Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing techniques fail to effectively reduce show-through in images, particularly in thin media like magazines and pamphlets, where low-density or medium-density portions on the front surface are overlapped, leading to inadequate show-through deletion due to increased variance values.

Innovation Solution

An image processing apparatus that includes a storage unit for index values, a first unit to obtain the degree of variation, a second unit to obtain brightness values, and a determination unit to correct pixel values based on the obtained values, using a look-up table to differentiate and correct show-through components by comparing variance and average luminance values.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If show-through deletion processing is executed using variance value threshold comparison, then show-through components can be identified and deleted, but low-density or medium-density portions on the front surface that are overlapped with show-through cannot be appropriately deleted because their variance values are increased

Engineering Contradiction:
Improveshow-through detection accuracyVSAvoidshow-through deletion completeness
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent changes the parameter used for show-through detection from variance value alone to a combination of average luminance value and variance value. By introducing average luminance as an additional parameter and creating a two-dimensional evaluation system with threshold lines, the method can distinguish between front surface low-density portions (high average luminance, high variance) and show-through components (low average luminance, low variance), thereby improving both detection accuracy and deletion completeness

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent adds another dimension to the evaluation space by introducing average luminance value as a new axis alongside variance value. This transforms the one-dimensional variance-based detection into a two-dimensional detection system where show-through components and front surface portions can be differentiated based on their positions in the average luminance-variance plane, enabling more accurate identification and deletion of show-through

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Object-affected harmful factors

If background removal function is strongly applied to reduce show-through, then show-through density is reduced, but the density of the image on the front surface is also reduced causing low-density images to be lost

Engineering Contradiction:
Improveshow-through densityVSAvoidfront surface image density
Core Design Contradiction:
Object-affected harmful factorsVSLoss of information

Solution Approach 1:

The patent applies local quality by executing show-through deletion processing selectively only on pixels identified as show-through components, rather than uniformly reducing density across the entire image. By using the two-dimensional evaluation system to identify specific pixels that are show-through components and applying correction only to those pixels, the method removes show-through while preserving the original density and quality of front surface images

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS9736334B2Image processing apparatus method and medium correcting value of pixel of interest in image data using determined amount of correction
Publication Date: 2017.08.15 CANON KK
  • US9736334B2 patent drawing
  • US9736334B2 patent drawing
  • US9736334B2 patent drawing

AI summary

For preferably reducing show-through in an image obtained by reading an original, an image processing apparatus: stores an index value for each of a plurality of mutually differing degrees of variation; obtains a degree of variation of signal values included in a region of interest of a predetermined size including a pixel of interest included in image data corresponding to the read image; obtains a value indicating a brightness of the region of interest; determines an amount of correction of a value of the pixel of interest using the obtained value and an index value corresponding to the obtained degree of variation; and corrects the value of the pixel of interest using the determined amount of correction.