Skin Cancer Detection based on Terahertz Images by using Gabor filter and Artificial Neural network

Abstract

Over recent years, the technologies of terahertz system which represents optically-driven technologies are dramatically expanding and developing. The moneymaking applications of these technologies are now beginning to become known. Terahertz pulse imaging (TPI) is non-ionizing and has least effects on the human body, it has very large absorption due to water and metals highly reflect terahertz radiation, combining the terahertz and the pattern recognition which related to multivariate statistical tools lead to potentially provide a rapid and non-invasive method to diagnose and detect diseases. The proposed system has the ability to detect skin cancer tissue by using image processing, where images are pre-processed (resize, median filter), then the important information are extracted from the images in vector format by using Gabor features. Simulation results are efficient with less consumption time of the cancer detection using ANN algorithm with accuracy of 94.117%.