In the process of infrared imaging, due to the imaging mechanism, the image quality inevitably suffers from degradation phenomena such as low definition and noise interference. In order to improve the usability of thermal infrared image data and facilitate subsequent image processing, it is necessary to enhance the useful signal in the image, suppress the interference signal in the image and make other necessary adjustments, all of which belong to the category of image preprocessing. In addition, for the human body detection system that combines thermal infrared and visible light, the fusion of the two types of images is also the basis for subsequent applications, which can be regarded as a kind of preprocessing.