Detection mechanism for vehicle collision avoidance using mmWave Radar

In recent years Advanced Driver Assistance Systems (ADAS) is growing at a very fast pace. This is to meet future requirements like automated parking, driver assistance, self-driving cars, etc. Collision avoidance is an essential requirement for any ADAS. Towards this, Radars are required to detect and track objects (moving or static) around the ego vehicle. This paper proposes a detection mechanism using a 77GHz mmWave Radar sensor in on-road vehicle driving scenarios. A real experiment was carried out by mounting a 77GHz mmWave Radar on the bonnet of the car. The Radar detections contain both moving and static objects in front of the ego vehicle. The Radar data was processed to get the Range Vs Relative Speed map and CFAR detection map. The targets of interest are only close-by objects having negative relative speed. The convention adopted here is: a negative relative speed indicates that the ego vehicle is approaching the target or vice versa and a positive relative speed means the target is moving away from the ego vehicle. Zero relative speed means both the ego vehicle and the target vehicle are moving at the same speed. Distance between the ego vehicle and target, and the relative speed between them are to be considered for avoiding a collision. In addition, this experiment also contains a camera to capture the ground truth information. In this experiment, the ego vehicle was moving at a speed of around 30 kmph and intentionally maintained a safe distance from the other on-road vehicles for safety. This paper proposes a method for the detection of targets in front of the ego vehicle. However, targets need to be detected and tracked on all sides of the vehicle. Further accurate tracking and angle estimation are also key requirements to build a robust collision avoidance system.