Abstract: The temperature distribution of the pipe surface shows a difference as it prsents a distinctness of heat resistance and emission ratio between the pipes with and without defects, on which it can be relied to carry out infrared nondestructive testing. The quality of infrared image is poor because of its background noise and illegible edges. As the dimension of the defects is usually small, the traditional testing method greatly depends on the personnel as well as his or her experiences, thus the possibility of missing and mistakes cannot be avoided and positioning the defects is becoming very difficult. The digital signal analysis and process of the temperature distribution data within the defect heat-affected regions is researched. The numerical method,such as Fourier transform, power spectrum analysis, polynomial regression, LS regression, middle value filter, Butterworth filter, etc are adopted. The defects identification and positioning software based on LabVIEW is developed.