提出一种基于改进的 GVF-Snake 模型(基于梯度矢量流的活动轮廓模型)的车辆跟踪算法.该算法利用帧差法自动获取车辆的初始轮廓,通过改进的 GVF-Snake 模型,从车辆视频流中准确提取出车辆的轮廓.为提高计算速度,改进的 GVF-Snake 模型采用贪心算法逐点迭代,并根据控制点的距离,自适应地增删控制点,以适应车辆目标大小变化.在此基础上,应用预测算法对车辆进行快速准确的跟踪.实验结果验证该算法的有效性. A vehicle tracking algorithm based on improved GVF-Snake model (Active Contour Model based on gradient vector flow) is proposed, which uses frame difference method to automatically obtain the initial contour of the vehicle. The improved GVF-Snake model, In order to improve the speed of calculation, the improved GVF-Snake model adopts the greedy algorithm to iterate point by point and adaptively adds and deletes the control points according to the distance of the control points to adapt to the change of the vehicle target size. , The prediction algorithm is applied to track the vehicle quickly and accurately.The experimental results verify the effectiveness of the algorithm.