Abstract
Inductive infrared thermography has been proven as an interesting solution for the inspection of surface defects. However, inductive thermography images can be noisy or present large variations in contrast and texture. Combine that with the large variability in surface defects shapes, sizes and types, the defect detection task becomes very complex. Defect detection methods based on convolutional neural network (CNN) proved their efficiency for complex detection tasks. This paper discusses two main approaches of defect detection with CNN : classification and object detection. Detection results are presented along with the advantages and weaknesses of each approach for real-time defect detection
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call