In-Vehicle Passenger Occupancy Detection Using 60-GHz FMCW Radar Sensor

Modern autonomous driving vehicles are equipped with a number of sensors to perceive the surrounding environment and enable automation of various systems. Especially, the passenger occupancy detection system can be utilized to detect a child left unattended in a parked vehicle, and efficiently manage the energy inside the vehicle. In this study, we propose a method of detecting the occupancy of passengers inside a vehicle using multichannel 60-GHz frequency-modulated continuous-wave (FMCW) radar. The received signal from the radar is converted into a range-angle map, and clutter suppression is performed to eliminate reflections from stationary objects. Then, by applying classification algorithms, such as the support vector machine (SVM), multilayer perceptron (MLP), and convolutional neural network (CNN), various arrangements of passengers inside a vehicle are identified. The classification results demonstrated that the proposed method can detect the location and number of passengers with an accuracy of 97.68%.