Image Processing Techniques in Thermal and Non-thermal Images

Abstract

Digital thermal imaging (DTI) is a passive, noninvasive, non–ionizing, complimentary diagnostic and real time monitoring technique that enables visualizing and quantifying changes in skin surface temperature. A thermal camera can detect temperature variations in the body, as minute as within 0.1 °C. The image highlights the temperature scale by showing variation of color from red to blue. The created color patterns depending on the prevailing temperature that identifies the image, with or without proper light setting. The computer analyzes the color data and calculates its surface temperature. The various image acquiring modalities like X-ray, MRI, PET is used for extracting information for the diagnosis of abnormality and for after treatments. This paper identifies suitable methods like contrast stretching and morphologic function which are the key steps in the enhancement of thermal diabetic foot images. The images are acquired, filtered and enhanced using the PSNR (Peak Signal to Noise ratio). The results obtained show that contrast stretching is giving better PSNR for thermographic diabetic foot image whereas morphological processing is superior in performance in case of non-thermographic diabetic foot image enhancement.