Intelligent Welding Defect Detection Model on Improved R-CNN

ABSTRACT

The quality of welding is directly related to the performance and life of the welding structure. The nondestructive testing technology is mainly used for welding defect defection, while the X-ray testing technology can directly and reliably reflect the shape, location and size of defects. For X-ray images, manual inspection is used at present, and it is easily caused by subjective factors, such as professional level, which may lead to low accuracy. This paper focuses on establishing an end-to-end automatic detection model of X-ray welding defects to improve the accuracy and efficiency of detection based on a deep learning algorithm. Considering the feature information of welding defects, this paper improves on the basis of Faster R-CNN and uses the deep residual network Res2Net to enhance the original backbone network to improve the feature extraction ability. And the weighted feature fusion module is studied, which combines the high-level semantic information and the low-level high-resolution edge detail information to predict the feature map of each layer respectively, to improve the detection performance, especially for small targets. The experimental data show that this method can effectively improve the accuracy and efficiency of welding defect detection.

KEYWORDS:

• Welding defect detection
• Deep learning
• Improved Faster RCNN

Data Availability Statement

The data used to support the findings of this study are available from the corresponding author upon request.

Additional information

Notes on contributors

Yongbin Chen

Yongbin Chen received his doctorate in engineering from the Guangdong University of Technology in 2019. Now he is a postdoctoral researcher at the Guangdong University of Technology. His main research interests are image processing, machine vision and augmented reality. Email: [email protected]

Jingran Wang

Jingran Wang is a graduate student at the Guangdong University of Technology. His main research interests include deep learning and image processing. E-mail: [email protected]

Guitang Wang

Guitang Wang received his BSc degree in 1986 from the Guangdong University of Technology; where he is a professor and master supervisor at the Guangdong University of Technology. His main research interests include instruments science and technology machine vision. E-mail: [email protected]