基于激光与视觉的车辆防撞技术的研究

提出了一种基于视觉与激光测距相结合的车辆防撞技术.首先采用改进的Hough变换对车道标志线进行检测与提取,根据车辆底部的阴影特征搜寻车辆可能出现的位置；利用边缘检测排除非车辆区域；然后结合车辆在图像中的宽度,利用激光测距传感器对前方车辆动态扫描,计算出两车的车距和相对速度；最后用卡尔曼滤波算法连续跟踪检测到的车辆；实验表明:该方法能够实时有效地检测前方车辆,在高精度汽车防撞系统中具有一定的意义.

Approach to Front Vehicle Detection and Tracking Based on Visual and Laseri

A vehicle collision preventing technology based on laser and visual is presented.At first the improved Hough transform is used to detect and extract the two lanes and the sign of underneath shadow is employed to generate hypothetical position.Edge detection is used to eliminate some non-vehicles areas.With consideration of the width of vehicle in the images,the distance and relative speed of two vehicles are scanned and calculated dynamically by laser sensor.Finally,Kalman filters are used to track the candidate vehicle continuously and the result of the experiment has shown that the method provides a robust approach,which can effectively detect the front vehicles in real time significantly in vehicles collision avoidance system with high accuracy.