Defect Evaluation System Using Region-Specific Severity Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current printer defect detection systems reject pages with defects that are acceptable if they occur in less noticeable regions, leading to unnecessary reprints and increased costs.
Innovation Solution
A system and method that use a trained classification model to evaluate defects based on their severity and location within an image, allowing for different acceptance criteria in various image regions, reducing the reject rate and costs by considering the significance of image regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a fixed threshold area is used to detect defects, then defects are detected consistently, but acceptable defects in less important regions are rejected unnecessarily
Solution Approach 1:
The patent applies local quality by dividing the image into multiple regions with different significance levels (e.g., high-significance regions like faces and low-significance regions like backgrounds). Different defect thresholds are applied to different regions, allowing the system to accept defects in less important areas while maintaining strict standards for critical regions. This resolves the contradiction by making defect acceptance criteria location-dependent rather than uniform across the entire image.
2Reliability
If all defects meeting the threshold are rejected, then print quality is maintained, but materials and time are wasted on unnecessary reprints
Solution Approach 1:
The system evaluates defects based on their location within the image, applying different acceptance criteria to different regions. Defects in high-significance regions (e.g., human faces, main subjects) lead to rejection, while defects in low-significance regions (e.g., backgrounds, borders) are accepted. This selective approach maintains print quality where it matters most while avoiding waste of materials on acceptable prints.
3Device complexity
If a fixed threshold area is used for defect detection, then the detection process is simple, but the system cannot distinguish between acceptable and unacceptable defects
Solution Approach 1:
The patent segments the image into multiple regions based on significance (e.g., using object detection models to identify faces, subjects, and backgrounds). Each segment is then evaluated with appropriate defect thresholds. This segmentation approach enhances defect acceptability assessment without requiring complex changes to the overall detection system, as it builds upon existing defect detection capabilities.
Solution Approach 2:
Different defect thresholds and evaluation criteria are applied to different image regions based on their significance. High-significance regions use strict thresholds while low-significance regions use lenient thresholds. This local differentiation improves measurement precision regarding defect acceptability while maintaining relative simplicity in the detection process.
Data Source
AI summary
A system and method evaluate defects in printed images. A target image, which has been captured of a printed image, is processed to identify defects, where present, which do not occur in a source image from which the printed image was generated. A trained classification model predicts a defect class for respective regions of the target image, each of the defect classes being drawn from a predefined set of defect classes. For at least one of the identified defects, a measure of severity of the defect is determined, such as a size of the defect. A decision on the acceptability of the printed image is made, based on the measure of severity of the at least one defect and the predicted defect class of a respective one of the regions in which the defect occurs.