Identification of X-ray Weld Defects under Artificial Intelligence Framework

Abstract

In view of the need of automatic detection of weld defects, an automatic extraction and classification algorithm for welding defect features based on convolution neural network is proposed. The algorithm directly takes the preprocessed weld images as the input and the welding defect type as the output, effectively avoiding the adverse effect of artificial identification subjective experience on the detection results. The experimental results show that the welding defect identification technology based on convolution neural network has a good identification rate and can provide an important reference for the research of welding quality detection.