Moving Target Classification with a Dual Automotive FMCW Radar System Using Convolutional Neural Networks

Abstract

Autonomous systems are currently trending with the radar taking a major role. Due to the advantages like compact sizes and having high performance, radars are used in advanced driver assistance systems (ADAS) such as the adaptive cruise control (ACC). Since passenger safety has a high priority, it is necessary to detect and classify targets in the near-field environment of the vehicle. Existing approaches use methods such as data fusion with additional sensors. This paper describes a classification process of moving targets using a dual automotive radar system and convolutional neural networks (CNN). Due to the detailed resolution of the radar and the great adaptability of the CNN a high classification probability has been achieved in first measurement trials.