Print Inspection Image Alignment for Local Deformation Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing print inspection systems face challenges in accurately and efficiently detecting defects such as stains and color loss in print products due to local displacements between reference and inspection images, which can reduce inspection speed and accuracy.
Innovation Solution
An image processing apparatus and method that generates a reference image by adding local displacements based on deformation information between a calibration document image and a read image, allowing for accurate comparison with an inspection target image to inspect print products.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If position alignment is performed by dividing the entire image into blocks and calculating displacement amounts for each block, then inspection accuracy is improved, but inspection time increases
Solution Approach 1:
The patent divides the image into multiple blocks and performs position alignment independently for each block. This segmentation allows the system to calculate displacement amounts for individual blocks rather than processing the entire image as one unit, improving inspection accuracy while managing computational complexity.
Solution Approach 2:
The patent performs position alignment as a preliminary step before defect detection. By pre-calculating displacement amounts and creating aligned reference images before the actual inspection, the system prepares the data in advance, which improves subsequent inspection accuracy without adding time during the critical detection phase.
2Productivity
If the reference image is generated without considering local displacement, then image processing speed is improved, but inspection accuracy deteriorates
Solution Approach 1:
The patent applies different processing approaches to different regions of the image. Specifically, it calculates local displacement amounts for each block individually and applies position alignment only where needed, rather than uniformly processing the entire image. This localised approach maintains accuracy in critical areas while preserving overall processing efficiency.
Solution Approach 2:
The patent changes the parameters of the reference image by adding calculated displacement amounts to the coordinates of blocks. This parameter modification allows the reference image to reflect actual local deformations in the printed material, improving inspection accuracy while using efficient computational methods to calculate and apply these parameter changes.
Data Source
AI summary
An image processing apparatus obtains first information indicating a local displacement between a first document image and a first read image obtained by reading a print product based on the first document image, generates a reference image obtained by adding the local displacement in accordance with the first information to a second document image and compares a second read image with the reference image to inspect a print product based on the second document image, the second read image obtained by reading the print product based on the second document image. The reference image includes a document region obtained by adding local deformation in accordance with the first information to the second document image, and a blank region surrounding the document region. An image inside the document region is obtained by adding the local deformation regarding the first information to the second document image and deforming the addition result.