Printed Image Inspection with Defect Severity Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image defect inspection systems require users to manually inspect defects of varying levels, which is time-consuming and inefficient, as they often include non-problematic defects in the inspection results.
Innovation Solution
An image inspection apparatus and method that digitizes defect features to assign levels, sorts defect pages by these levels, and displays them in descending order, allowing users to prioritize inspection based on severity, reducing the need for manual sorting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the defect detection level is set strictly to prevent overlooking defects, then the reliability of defect detection is improved, but the quantity of defects requiring manual inspection increases
Solution Approach 1:
The patent applies parameter changes by quantizing defect feature amounts into discrete levels (first level, second level, etc.). This transforms the continuous defect detection process into discrete categories, allowing the system to maintain strict detection thresholds while organizing results into manageable groups that reduce manual inspection burden.
Solution Approach 2:
The patent segments detected defects into multiple levels based on quantized feature amounts. By dividing the defect set into hierarchical levels (with higher levels indicating more significant defects), the system enables prioritized manual inspection, where users first examine high-level defects and can stop earlier if satisfactory results are achieved, thus reducing the total quantity of defects requiring manual review.
2Reliability
If all detected defects are displayed for manual inspection, then the completeness of inspection results is improved, but the time and effort required for user inspection increases
Solution Approach 1:
The patent performs preliminary quantization and level assignment of defect features before manual inspection. By pre-processing defects into organized levels with distinct visual characteristics, the system prepares the inspection results in advance, allowing users to efficiently navigate and prioritize their review without spending time on initial defect characterization or sorting.
Solution Approach 2:
The patent applies local quality by making different defect levels visually distinguishable through different display characteristics. High-level defects are presented with visual properties that make them stand out, allowing users to quickly identify and prioritize critical defects without examining every defect in detail, thus reducing overall inspection time while maintaining completeness.
3Device complexity
If a single defect level is used for all defects, then the simplicity of the inspection system is improved, but the ability to prioritize critical defects deteriorates
Solution Approach 1:
The patent transforms the single defect level concept into multiple quantized levels by changing the parameter of defect severity classification. This multi-level parameter system enables prioritization while maintaining relative simplicity through automated quantization algorithms and standardized display rules for each level, balancing complexity and operational ease.
Solution Approach 2:
The patent segments the uniform defect classification into hierarchical levels with distinct visual characteristics. This segmentation provides prioritization capability by allowing users to focus on high-level defects first, while the systematic organization into discrete levels prevents the system from becoming overly complex through clear, rule-based display differentiation.
Data Source
AI summary
An image inspection apparatus includes circuitry. The circuitry inputs a master image, the master image being original data for printing. The circuitry inputs a scanned image acquired by reading, with a reading device, an image printed by an image forming apparatus performing a print job. The circuitry compares the master image and the scanned image to detect a defective portion in the scanned image. The circuitry digitizes a feature amount of the defective portion to assign a level to the defective portion based on the feature amount. The circuitry sorts detected defect pages of the print job each page having the defective portion assigned with the level by the level of the defective portion in the detected defect page. The circuitry displays a list of the detected defect pages in an order having been sorted.