Identify Areas of Pain in Limb Using Thermal Image Processing and Pattern Recognition

Abstract

Pain is the most common cause of going to physicians, and the treatment of pain and elimination of its adverse effects, has always been the focus of physicians. In this paper an efficient method is proposed for detection of the pain location in the lower extremities using thermal infrared images. In the following parts, an intelligent method of fuzzy derivation- based image separation and PCA feature extraction, and SVM classification is introduced that automatically separates certain areas of the body that emit a wavelength of pain, in thermal images. Identification and differentiation of muscle pain in lower extremities and location of abnormal metabolism areas are the most important achievements of this method. In this study, a very fast and non-invasive method is proposed to detect the exact location of pain in the lower extremities, which can distinguish muscle aches and soreness from the false pain in these areas.