Improved faster R-CNN and adaptive Canny algorithm for defect detection using eddy current thermography

Jianyu Li,Laibin Zhang,Wenpei Zheng

DOI: https://doi.org/10.1063/5.0189084

IF: 1.697

2024-02-01

AIP Advances

Abstract:Eddy current thermography (ECT) is a non-invasive testing method that combines electromagnetic induction and infrared thermography to identify flaws in materials that conduct electricity. However, ECT faces difficulties in accurately locating and classifying defects owing to its low signal-to-noise ratio and complex defect patterns. In this paper, we propose a new method that integrates an improved faster region-convolutional neural network (R-CNN) and an adaptive Canny algorithm to enhance the defect detection performance of ECT. An improved faster R-CNN is a deep neural network that can automatically detect and locate multiple defects in a single ECT image, whereas the adaptive Canny algorithm is an edge detection technique that can identify defect boundaries. The proposed method was tested using a dataset of ECT images with different types of defects. The results demonstrated that our method achieved better accuracy, precision, and speed than existing methods.

materials science, multidisciplinary,physics, applied,nanoscience & nanotechnology

What problem does this paper attempt to address?

The paper primarily addresses the issues in Eddy Current Thermography (ECT) technology for material defect detection, particularly the difficulties in defect localization and classification caused by low signal-to-noise ratio and complex defect patterns. To solve these problems, the authors propose a new method that combines an improved Faster Region Convolutional Neural Network (Faster R-CNN) and an adaptive Canny algorithm. Specifically, the main contributions of this study are as follows: 1. **Improved Faster R-CNN**: This is a deep neural network capable of automatically detecting and localizing multiple defects in a single ECT image. By introducing the Inception-v4 architecture as the feature extraction network, the performance of defect recognition is further enhanced. 2. **Adaptive Canny Algorithm**: This is an edge detection technique that can identify defect boundaries. By employing techniques such as local adaptive threshold determination and non-maximum suppression, this algorithm can better handle noise interference in images and improve the accuracy of edge detection. Through experimental validation, this method demonstrates better accuracy, precision, and speed on different types of ECT image datasets, showing significant advantages over existing methods. In short, this paper aims to improve the efficiency and accuracy of ECT technology in non-destructive testing by integrating machine learning techniques and image processing algorithms, to meet the needs of industrial production and ensure the safety of equipment.

Improved faster R-CNN and adaptive Canny algorithm for defect detection using eddy current thermography

Automatic Recognition and Classification of Eddy Current Testing Signals for Scanning Inspection of Defect in Multi-Layered Structures

Spatio-Temporal 3-D Residual Networks for Simultaneous Detection and Depth Estimation of CFRP Subsurface Defects in Lock-In Thermography

Adaptive rapid defect identification in ECPT based on K-means and automatic segmentation algorithm

An Improved Feature Extraction Algorithm for Automatic Defect Identification Based on Eddy Current Pulsed Thermography

Enhanced Diffusion-Based Analysis for Fast Defect Detection in ECPT Image

A Smart Surface Inspection System Using Faster R-CNN in Cloud-Edge Computing Environment

Eddy Current Array for Defect Detection in Finely Grooved Structure Using MSTSA Network

Two-stream Convolutional Neural Network for Non-Destructive Subsurface Defect Detection Via Similarity Comparison of Lock-in Thermography Signals

Eddy Current Array for Defects Detection Based on Spatiotemporal Self-attention Network

Improvement of Defect Feature Extraction in Eddy Current Pulsed Thermography.

Defect Detection of Composite Material Terahertz Image Based on Faster Region-Convolutional Neural Networks

Defect Image Recognition and Classification for Eddy Current Testing of Titanium Plate Based on Convolutional Neural Network

Method of Imaging and Evaluation for Defect Based on Pulsed Eddy Current

Filtered Selective Search and Evenly Distributed Convolutional Neural Networks for Casting Defects Recognition

Improved Faster R-CNN Algorithm for Defect Detection in Powertrain Assembly Line

AEGLR-Net: Attention Enhanced Global-Local Refined Network for Accurate Detection of Car Body Surface Defects

Defective Surface Detection Based on Improved Faster R-CNN

A Defect Detection Method Based on Faster RCNN for Power Equipment

Evaluation of Defects Depth for Metal Sheets Using 4-coil Excitation Array Eddy Current Sensor and Improved ResNet18 Network

Automatic Defect Classification for Infrared Thermography in CFRP based on Deep Learning Dense Convolutional Neural Network