Analysis of object recognition by microwave illumination
K. Struckman · 2002
Object location and identification continues to be a major problem for robotics operations. Microwave object recognition could play an major role in the solution to many robotics recognition problems. Our original object recognition concept was based on a pattern matching approach employing computer stored replica correlation processing. Advances in artificial neural network processing technology offers an attractive alternative to the correlation processing approach, namely the ability to learn and to adapt to changing work environments. The input to these artificial networks would be the spectrums reflected from objects within the view of the robot. This paper describes the uniqueness of these input vectors as calculated by the complex correlation equation. These numerical results are based on method of moments computations, ESP IV code.>