Image Inspection False Defect Detection via Area Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image inspection methods for printed products face challenges in accurately determining defects due to false position matching, especially in digital printing where generating a master image is inefficient, leading to incorrect defect detection and re-printing decisions.
Innovation Solution
An image inspection system and method that divides the master and scanned images into areas for precise position matching, verifies defect patterns by counting defect amounts, and determines if the defects are false detections caused by position matching failures by comparing them against pre-set threshold values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If position matching is conducted using the same image pattern for repeated drawings, then the inspection process can be automated, but false position matching occurs leading to incorrect defect detection
Solution Approach 1:
The image is divided into multiple areas, and position matching is conducted separately for each area rather than for the entire image at once. This segmentation prevents false matching when the same pattern is repeated across different regions, as each area is evaluated independently with its own defect amount threshold check.
2Measurement precision
If defect detection sensitivity is increased to detect all potential defects, then more defects are identified, but false positives increase due to position matching errors
Solution Approach 1:
The system performs feedback verification by checking whether the defect amount in each area exceeds a predetermined threshold. This feedback mechanism allows the system to distinguish between actual defects and false positives caused by position matching errors, thereby improving the reliability of defect detection while maintaining high sensitivity.
3Productivity
If the entire image is processed for position matching, then processing efficiency is maintained, but false defect detections occur in areas with repeated patterns
Solution Approach 1:
The image is divided into multiple areas for independent position matching and defect amount verification. This segmentation allows the system to maintain processing efficiency by using automated batch operations on divided regions while preventing false defect detections through area-specific threshold checks that identify and filter out false positives from repeated patterns.
Data Source
AI summary
An apparatus for verifying an inspection result includes an inspection result obtaining unit to obtain inspection result indicating defect amount and defect position in a scanned image, from an inspection result of defect judgment of the scanned image with respect to an inspection reference image, the defect judgment being performed through dividing at least one of the scanned image and the inspection reference image and conducting position matching of the scanned image and inspection reference image; and a verification unit to count defect amount occurring to the scanned image, to determine whether a defect pattern occurring to the scanned image matches a pre-set condition corresponding to a false detection condition causable by failure of the position matching, and to determine that defect occurring to the scanned image is a false detection when the counted defect amount is a threshold or more, and the defect pattern matches the false detection condition.