Powder Bed Image Anomaly Detection With Adaptive Thresholding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing anomaly detection methods in powder-bed-based additive manufacturing are not robust enough to detect defects consistently and require manual threshold adjustments, leading to inconsistencies and inefficiencies.
Innovation Solution
A semi-automated anomaly detection method using Principal Component Analysis (PCA) and clustering algorithms to compute adaptive thresholds, allowing for semi-supervised anomaly detection and pattern recognition in image data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If static thresholding-based image processing algorithms are used for anomaly detection, then the process is simple and fast, but the detection consistency and robustness deteriorate due to manual threshold adjustments and image setting changes
Solution Approach 1:
The patent transforms static thresholding into dynamic adaptive thresholding by automatically adjusting thresholds based on image characteristics. The system computes thresholds adaptively for each image or image sequence, allowing the detection parameters to dynamically adapt to varying image conditions while maintaining detection consistency without manual intervention.
Solution Approach 2:
The system performs self-adjustment of detection thresholds by automatically analyzing image characteristics and computing appropriate thresholds without requiring manual operator intervention. This self-service mechanism eliminates inconsistencies caused by different operators setting thresholds manually while maintaining fast processing speeds.
2Adaptability or versatility
If manual threshold adjustment is performed by operators, then flexibility in detecting different anomaly types is improved, but time consumption and operational complexity increase
Solution Approach 1:
The system automatically determines appropriate thresholds for different anomaly types by analyzing image characteristics, eliminating the need for manual threshold adjustment by operators. This self-service approach maintains flexibility in detecting various anomaly types while significantly reducing the time and operational complexity associated with manual threshold setting.
Solution Approach 2:
The system dynamically changes detection parameters (thresholds) based on image characteristics and anomaly types. By automatically adjusting these parameters rather than using fixed manual values, the system achieves both flexibility in detecting different anomalies and efficiency in processing without operator intervention.
3Productivity
If static thresholds are used for anomaly detection, then the processing is efficient and fast, but inconsistencies occur when image settings are changed
Solution Approach 1:
The patent implements dynamic threshold adjustment that automatically adapts to changes in image settings. Rather than using fixed static thresholds, the system computes thresholds that dynamically respond to variations in image acquisition parameters, ensuring consistent detection results across different imaging conditions while maintaining processing efficiency.
Solution Approach 2:
The system incorporates feedback mechanisms where detection results and image characteristics are used to automatically adjust thresholds. This feedback loop ensures that when image settings change, the thresholds are automatically recalibrated to maintain consistent anomaly detection performance without sacrificing processing speed.
Data Source
AI summary
A computer-implemented anomaly detection method in powder-bed-based additive manufacture of a workpiece includes (i) providing an image data set and applying a Principal Component Analysis to said image data to compute a number of image clusters, (ii) applying a clustering algorithm to the analyzed image data and computing respective cluster centroids, (iii) comparing the computed cluster centroids with a set of reference anomaly centroids, wherein based on a match of cluster centroids with the reference, the image data is segmented layerwise into cluster images of a specific anomaly, (iv) transforming the segmented images into a defined color space, such as a Lab color space or greyscale spectrum, and, (v) integrating a pixel information of the transformed segmented cluster images to compute a threshold value for the image data set in order to determine a respective anomaly.