Image registration in PCB Fault Detection Based on Infrared Thermal Imaging

In this paper, the problem of image registration in PCB (Printed Circuit Board) fault detection based on infrared thermal imaging is studied. In order to improve the quality of PCB infrared thermal image, a novel preprocessing method is presented in this paper, which can guarantee the effect of image registration. Due to the large difference between PCB infrared thermal image and common image, few existing image registration studies are dedicated to PCB infrared thermal image registration. Therefore, this paper compares a variety of registration algorithms to find suitable PCB infrared thermal image registration algorithms for different detection methods. The result shows that the registration method based on mutual information is more accurate and time-consuming, and it is suitable for thermal imaging sequence detection; the registration method based on SIFT feature is more efficient, but has lower precision, and it is suitable for the thermal imaging differential detection method.