Concealed Weapon Detection

Abstract

In the last years, there has been a renewed interest in security applications designed to detect potentially dangerous concealed object carried by an individual. In particular automatic detection and classification of concealed weapons is a fundamental part of every surveillance system.Until now merely all the research in image processing for Concealed Weapon Detection has been focused on millimeter wave imagers and X-ray imagers with very little work done in the microwave range. The main objective of this thesis is to develop robust novel image processing algorithms for detection and classification of concealed weapon. In particular, the developed algorithms are specifically tailored to work with microwave radar images. The algorithms shall also perform efficiently with a low false alarm rate in a reduced contrast envinroment such as the one of microwave images. Depolarization Analysis and SIFT Analysis which are two novel algorithms for concealed weapon detection and classification in the field of 3D high resolution microwave radar imaging are presented in this thesis research project.