Shaped Component Anomaly Detection With Blockwise Confidence Intervals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing quality monitoring systems in press shops face challenges such as limited accessibility of quality-relevant component properties during forming, stochastic component position and orientation, pseudo-defects, and high pseudo-reject rates due to environmental interference, which compromise the reliability and efficiency of inline inspection systems.
Innovation Solution
A neighborhood-based anomaly detection method that involves image processing, including foreground/background separation, block division, and confidence interval analysis to distinguish genuine defects from pseudo-defects, while compensating for positional and environmental interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If optical measuring and inspection systems are integrated to monitor component surface after dropping onto press discharge conveyor, then quality monitoring becomes possible, but stochastic deviations in component position and orientation cause interference in comparative inspection procedures
Solution Approach 1:
The component surface is divided into multiple blocks, and each block is compared with corresponding blocks from previous components. This segmentation approach makes the inspection system robust to position and orientation deviations, as each local block comparison is less sensitive to global misalignment than a full-component comparison would be.
Solution Approach 2:
The system performs preliminary capture of reference images of previously produced components before conducting the actual inspection. These reference images are stored and used as comparison baselines, allowing the system to account for normal variations and establish what constitutes a genuine defect versus normal manufacturing variation.
2Measurement precision
If additional handling systems are integrated for precise positioning of test objects, then measurement precision improves, but additional costs increase
Solution Approach 1:
The inspection system uses the component's own geometry and features to perform self-positioning and self-alignment during the comparison process. By comparing local blocks against reference blocks from the same component type, the system automatically compensates for position variations without requiring external positioning mechanisms.
Solution Approach 2:
The patent replaces mechanical positioning systems with an image processing-based solution. Instead of using handling systems to physically position components with high precision, the system uses computational methods to achieve accurate comparison through image block analysis and pattern recognition.
3Area of stationary object
If area-covering inline inspection systems monitor large component surfaces, then monitoring coverage increases, but probability of false detections increases due to oil droplets and contamination
Solution Approach 1:
The inspection system applies different evaluation criteria to different blocks based on their local characteristics. By analyzing each block's distance vector and comparing it with reference blocks, the system can distinguish between local variations caused by contamination and genuine defects, maintaining high detection accuracy across large component areas.
Solution Approach 2:
The system uses feedback from the distance function calculations and block comparisons to dynamically adjust detection thresholds and criteria. By continuously comparing new components against reference images and analyzing the distribution of distance values, the system learns to distinguish normal variations from genuine defects, reducing false detections while maintaining comprehensive coverage.
Data Source
Figure 1
AI summary
In a method for detecting anomalies in formed components, a) at least one first image (1) of an area of a first component is first recorded, and then b) the first image (1) is divided into a plurality of blocks (4). Then, c) a distance value is determined between each block and its neighboring blocks using a distance function (6), and subsequently d) steps a) to c) are carried out for a second image in the same area of a second component, so that e) all distance values to all neighboring blocks of the respective block are summarized in a distance vector. Subsequently, f) a most similar neighboring block is determined for each block from the respective distance vectors, and then g) for the most similar neighboring block, an index of the most similar neighboring block, an average distance to the most similar neighboring block, and a standard deviation of the distance values to the most similar neighboring block are stored for each block.Then, h) a confidence interval is determined for each block using the index, the mean distance, and the standard deviation. Then, i) steps a) to c) are performed for a detection image of a component to be detected. Then, j) it is checked whether the determined distance values for each block of the detection image lie within the confidence interval for the respective block. k) If the respective distance value of the respective block of the detection image lies outside the respective confidence interval, this is evaluated as anomalous.