FFT-based Image Mosaicing of Infrared Thermal Wave Images
基于傅里叶变换的红外热波图像拼接*

In the field of infrared thermal wave nondestructive inspection, changes among infrared thermal wave serial images mainly include offset and rotation of small angles. Based on these features, the technology of phase correlation in FFT is used to determine the angle and offset within the possible rotation angle scope. Therefore, image registration is realized, and the images are mosaiced. The experiment result shows that serial images with little overlapping can be mosaiced efficiently.