Deep Adversarial Network for CFRP Thermal Imaging Debond Diagnosis

In this paper, a novel joint loss GAN is proposed in Thermography Nondestructive testing. The proposed joint loss function incorporates both the modified CGAN loss and the penalty loss where this strategy significantly improves performance of defect detection. The results show that the proposed method can better capture the defect features to improve the detection accuracy. In order to verify the effectiveness of the proposed method, experimental studies have been carried out for inner debond defects on both regular and irregular shaped CFRP specimens. A comparative analysis has been undertaken between the proposed method and the state-of-the-art deep semantic segmentation algorithms. It is being proved to be suitable for the end-to-end detection Thermography diagnosis system.