Material discrimination using bispectral signatures

One goal of investigative signal processing techniques is to discriminate between types of materials composing an object. This paper explores a method for material discrimination and characterization using bispectral signatures acquired from an object actively probed with acoustic pulses. The mathematical foundations of the bispectrum are presented, and the proposed technique is tested with an ultrasonic apparatus. Results indicate that at ultrasonic frequencies this technique provides signatures with the potential of discriminating between classes of materials such as plastic, aluminum, and rock.