Abstract
This paper describes a new method for distinguishing objects on and alongside the road using a scanning laser radar (SLR) and a vehicle-mounted camera. Objects are detected on the basis of SLR digital signal data and the types of objects are distinguished by using their movement vectors and positions relative to white lane markers. For object detection, a new two-step grouping method is proposed for improving detection accuracy. The movement vectors of objects are adjusted properly based on the steering angle and speed of the host vehicle. The proposed method has been validated on a simulator using data recorded at a velocity of more than 40km per hour on Japanese expressways. The expected discrimination of vehicles, delineators and overhead signs in the forward view was successfully achieved in the simulation.
