Multiple objects recognition system for intelligent vehicle system using 24GHz radar

In the active safety system, the object classification is important to decide the action for passenger´s safety. Using microwave radar as a sensor, we can get the measure about the distance from the sensor, the azimuth angle, and also the signal power of the reflected signal. In this paper, we use these information from 24GHz microwave radar sensor and classify the objects for the active safety system. The K-nearest neighbor, multilayer perceptron, support vector machine and reduced multivariate polynomial model are employed for verification. This paper includes some experiment results for classification among human, vehicle and bicycle.