X.T. Yan, Z.L. Shang

Pages: 27-38

Abstract
In this paper, a vehicle forced lane change behavior detection algorithm based on machine learning is proposed. Collect the vehicle's forward looking monocular video image and determine the reasonable region of interest; The intention and process of vehicle forced lane change are analyzed by video image features, and the vehicle forced lane change motivation model is established; Use Kalman filter in machine learning to track vehicles; Calculate the centroid coordinates of the target vehicle and the distance of the lane line, and determine whether the vehicle in front has a forced lane change behavior in combination with the lane change motivation model. The experimental results show that this method has higher accuracy in detecting vehicle lane change behavior.
Keywords: machine learning; forced lane change of vehicles; behavior detection; areas of interest; grayscale; kalman filtering