PAUT S-Scan Classification for Defect vs Diffuse Reflection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing phased array ultrasonic testing (PAUT) methods face challenges in accurately distinguishing between defective areas and diffuse reflection areas in S-scan images due to varying skill levels of testers and increased diffuse reflection signals, affecting defect detection accuracy.
Innovation Solution
A system and method that utilizes pre-trained classification models to classify S-scan images from PAUT, involving an input unit, conversion unit, extraction unit, and classification unit to identify and differentiate between defective and diffuse reflection areas by analyzing signal characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If S-scan imaging is used in phased array ultrasonic testing, then two-dimensional cross-sectional views can be obtained, but diffuse reflection signals increase and defect detection accuracy varies depending on tester skill
Solution Approach 1:
The patent replaces manual visual inspection and human judgment with an automated AI-based classification system. The system uses deep learning models (first trained model for defect candidate detection, second trained model for area classification) to automatically distinguish between defective areas and diffuse reflection areas, eliminating the dependency on tester skill level while processing S-scan images
Solution Approach 2:
The patent introduces an intermediary processing layer between the S-scan image acquisition and final defect detection. The conversion unit converts S-scan images to original images, the extraction unit extracts signal characteristics from column data, and the classification unit uses trained models to classify areas. This intermediary processing chain filters out diffuse reflection signals and identifies true defects systematically
2Area of stationary object
If multiple probes are used to receive ultrasonic signals, then comprehensive coverage is achieved, but diffuse reflection signals increase
Solution Approach 1:
The patent extracts and isolates the harmful diffuse reflection signals from the comprehensive S-scan images obtained using multiple probes. The extraction unit specifically targets column data corresponding to potential defect areas and extracts signal characteristics that can distinguish between true defects and diffuse reflections, allowing the system to maintain comprehensive coverage while filtering out false signals
Solution Approach 2:
The patent applies local quality analysis by examining signal characteristics at specific locations (column data at each pixel position) rather than treating the entire image uniformly. The extraction unit selects representative data from column data and extracts peaks at specific positions, allowing the system to differentiate between local defect characteristics and diffuse reflection patterns at each location
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables rapid and accurate classification of defective and diffuse reflection areas, enhancing defect detection precision and reducing reliance on human skill variability.
Implementation Method 1
Ultrasonic testing is a method of detecting discontinuity in a test specimen by transmitting ultrasonic waves into the test specimen and is a method of measuring a position, size, etc. of a defect by comparing the amount of energy reflected from the discontinuity within the test specimen
Implementation Method 2
there is a limitation that a diffuse reflection signal increases depending on the surface and material of the test object
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The present disclosure relates to a system and method for area classification and a defect detection system using the same. The system includes an input unit configured to receive an S-scan image from phased array ultrasonic testing (PAUT) using at least one probe; a conversion unit configured to detect at least one defective candidate using a pre-trained first model and convert the S-scan image into an original image; an extraction unit configured to acquire column data of the defective candidate, select representative data, extract at least one peak by overlapping the representative data in an index direction, and extract signal characteristics; and a classification unit configured to input the signal characteristics into a pre-trained second model to classify each defective candidate as a defective area or a diffuse reflection area.