Edge Correction for Image Inspection Apparatus
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In print bookbinding systems, the difference in edge reproducibility between RIP data and scanned images leads to over-detection or detection failure of image anomalies, as edge portions in scanned images may appear thicker due to dot gain or reading effects, causing increased differences between the two.
Innovation Solution
An inspection apparatus that obtains edge correction parameters to minimize density differences between scanned and reference images, applying these parameters to perform edge correction on the reference image to match the edge reproducibility with the scanned image, thereby reducing over-detection and improving anomaly detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If edge correction is not applied to the reference image, then the reference image maintains its original sharpness, but over-detection increases due to density differences between scanned and reference images
Solution Approach 1:
The patent applies edge correction parameters to modify the density values of the reference image, specifically adjusting edge portions to match the density characteristics of scanned images. This parameter change resolves the density mismatch between reference and scanned images, thereby reducing over-detection while maintaining anomaly detection capability.
2Object-generated harmful factors
If the detection threshold is lowered to reduce over-detection, then false positives decrease, but detection failure of actual image anomalies increases
Solution Approach 1:
The patent performs edge correction on the reference image before the anomaly detection process. This preliminary action pre-adjusts the density characteristics of the reference image to match scanned images, creating a more accurate baseline for detection. This allows the detection threshold to be set appropriately without causing either over-detection or detection failure.
3Object-generated harmful factors
If the reference image is corrected to match scanned image density, then over-detection is reduced, but the original reference image sharpness is compromised
Solution Approach 1:
The patent applies edge correction selectively to specific regions of the reference image, particularly focusing on edge portions where density mismatch occurs. By applying different correction levels to different regions (edges versus interior), the patent reduces over-detection while preserving the overall sharpness and quality of the reference image.
Data Source
AI summary
An inspection apparatus inspects image data obtained by reading a sheet using a reference image of the image data. The inspection apparatus obtains first image data by reading a sheet having lines printed thereon, obtains second image data from print data of the lines, performs edge correction on the lines included in the second image data by applying a plurality of edge correction parameters being different from each other, and determines an edge correction parameter of the plurality of edge correction parameters that minimizes a difference between a density of the second image data subjected to the edge correction and a density of the first image data. In addition, edge correction is performed on the reference image using the determined edge correction parameter, and image data obtained by reading a sheet of an inspection target using the reference image on which the edge correction has been performed is inspected.