Executive Summary : | Dermatological diseases are the most prevalent diseases worldwide, but diagnosis is challenging and requires extensive experience. Thermal imaging cameras, such as Fluke Ti480 PRO, can detect infrared radiations and produce a thermal profile of the scene. This non-invasive method has numerous applications, including inflammatory pain detection, emotion recognition, and automatic categorization of dermatological diseases using CNN. The proposed research aims to investigate the spatial distribution of RGB components in dermatological thermal images, prepare a database of thermal images of human tissues exposed to different heat, compare the distribution of parameters obtained from visible and thermal images for normal and diseased skin, and automatically categorize dermatological diseases using CNN. The efficiency of texture analysis techniques, classification algorithms, and methods for image analysis will be quantified using the MOS Mean Opinion Score. The proposed project aims to provide a framework for the analysis of dermatological diseases based on infrared images, helping dermatologists in accurate and well-timed diagnosis. |