Coated Surface Defect Assessment With U-Net And SVM
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated systems for assessing coated surfaces struggle to distinguish and characterize different types of surface defects, relying on manual work and human expertise, limiting the ability to correlate defects with coating formulation and application techniques.
Innovation Solution
An automated system using a combination of machine learning algorithms, specifically a convolutional neural network (CNN) and support vector machine, to recognize and quantify various surface defects like pinholes, blisters, and craters, integrating a database for training and providing quantitative and qualitative assessments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated systems are used to assess coated surfaces, then productivity is improved, but the ability to distinguish and characterize different types of surface defects deteriorates
Solution Approach 1:
The assessment system segments the defect detection process into multiple specialized components: a CNN-based system for detecting and classifying defect types (pinholes, blisters, craters, seeds), and a separate analysis system for characterizing defect properties. This segmentation allows each component to specialize in specific tasks, maintaining high precision while improving overall productivity through automated parallel processing.
2Ease of operation
If automated machine learning systems are used, then ease of operation is improved, but device complexity increases
Solution Approach 1:
The system introduces an intermediary database layer that stores pre-trained machine learning models and defect type definitions. This intermediary layer simplifies the user interface and operation while managing the complexity of multiple ML algorithms, training datasets, and classification rules in the background without requiring users to directly interact with the complex system architecture.
3Measurement precision
If multiple defect types are assessed simultaneously, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The system dynamically adapts its assessment capabilities by loading and activating specific machine learning models based on the coating type and defect types being assessed. Rather than running all possible detection algorithms simultaneously, the system dynamically selects and configures the appropriate subset of models, maintaining high detection precision while reducing computational complexity and resource requirements.
Data Source
AI summary
The invention relates to a method for providing a system for assessing a coated surface with respect to a type set containing at least one type of surface defect that can occur on the surface, to such a system for assessing a coated surface, to a measuring device for acquiring an image of a coated surface including such a system, and to a method for assessing a coated surface using such a system. The system can use a convolutional neural network (CNN), in particular the so-called U-Net architecture, to recognize the at least one surface defect in an image provided to the system. Moreover, the system can use a support vector machine algorithm to provide, based on the recognition results, quantitative and/or qualitative information about the depiction of the at least one surface defect in the image provided to the system.