A Subspace Signal Processing Technique for Concealed Weapons Detection

Concealed weapons detection is one of the greatest challenges facing national security nowadays. Recently, it has been shown that each weapon can have a unique fingerprint, which is a set of electromagnetic (EM) resonant frequencies determined by its size, shape, and physical composition. Extracting the resonant frequencies of each weapon is one of the major tasks of any detection system. In this paper, we model the reflected signal from each object as a summation of sinusoidal signals, each at certain frequency equal to one of the object's resonant frequencies. Using this model, we propose a detection approach that is based on a modified version of the multiple signal classification (MUSIC) algorithm. We show by simulations that each object can be represented using a two-dimensional vector, which consists of its two major resonant frequencies.