Defect super-resolution algorithm based on infrared thermal imaging physical kernel

Infrared thermal imaging technology has been adopted for its advantages such as fast detection response, non-contact testing, and applicability to various objects. When testing a target object, this technology presents temperature distribution information of the target object's surface in the form of an image, achieving visualization. However, due to limitations in the hardware system of the thermal imager and the noise generated during the detection process, the resolution of infrared images is relatively low, and the details of the image are not rich enough, leading to limitations in specific defect detection. In this study, a defect super-resolution algorithm based on infrared thermal imaging physical kernel is proposed. The imaging degradation factors of the infrared images are analyzed, and the modulation transfer function of the infrared thermal imaging system is used as the physical prior to generate the underlying blur kernel of the infrared images. The infrared images are then reconstructed using a super-resolution algorithm based on the blur kernel. Obtained experimental results have demonstrated that the proposed method significantly improves the defect detection rate and the overall image quality. The demo code will be updated soon in https://faculty.uestc.edu.cn/gaobin/zh_CN/lwcg/153392/list/index.htm.