This article describes systemically the method of detecting the preceding vehicle based on a monocular camera. The main content is as follows: first, a primary area of interest is found by the lane borderlines that are identified in a camera image, and a likelihood target vehicle is searched by the gray difference between the target vehicle and the background; second, an identifying area of interest is found again based on the area of a likelihood target vehicle, a target vehicle is affirmed by a symmetry character of the vehicle outline and a position of the vehicle symmetrical axis is ascertained; third, the object vehicle is tracked by Kalman forecast principle in the sequence images; fourth, a method of detecting distance in a frame of image is introduced. The calibration of the camera's interior parameters and the results of some experiments are given.
[1]
B. Michaelis,et al.
Robust vehicle detection at large distance using low resolution cameras
,
2000,
Proceedings of the IEEE Intelligent Vehicles Symposium 2000 (Cat. No.00TH8511).
[2]
Frank Dellaert,et al.
Robust car tracking using Kalman filtering and Bayesian templates
,
1998,
Other Conferences.
[3]
Yi Lu Murphey,et al.
DepthFinder, a real-time depth detection system for aided driving
,
2000,
Proceedings of the IEEE Intelligent Vehicles Symposium 2000 (Cat. No.00TH8511).
[4]
Li-bing,et al.
A vision navigation algorithm based on linear lane model
,
2000,
Proceedings of the IEEE Intelligent Vehicles Symposium 2000 (Cat. No.00TH8511).