Detection and recognition of persons in snowy foliage using active hyperspectral sensing for border control

We demonstrate the use of an active hyperspectral sensor for detection and recognition of persons in snowy foliage. The sensor employs a tunable broadband supercontinuum light source for remote spectrum measurement. An algorithm based on support vector machines is used in real-time to classify the object, from which the supercontinuum laser pulse scattered. We present results for person detection during simulated illegal border cross activities in snowy terrain.