Automated classification of objects through ultrasound signature analysis

The objective of this research is to develop a method for automated detection of a class of objects with similar physical features through the analysis of ultrasound signatures produced by those objects when acoustically excited.

As a step in the development of such a system, a machine has been designed and built, which is capable of detecting and identifying three specific objects.

The design of the recognition subsystem is presented. This subsystem uses the discrete Fourier transform to analyze and pattern recognition techniques to classify the acoustic signals. Development of the design is given from a pattern recognition point of view.