提出了一种在复杂环境中车辆检测与跟踪方法。该方法首先分析车辆的外形特点，确定采用从Blob块提取的几何形状参数作为车辆的特征并用于运动车辆的检测; 然后结合Blob分析和CamShift算法提出了车辆跟踪方法。实验结果表明，本文采用的车辆特征参数能够准确地检测车辆，并能准确地区分车辆和行人群体，车辆检测与跟踪准确率分别达到了96.7%和86.7%，证明该方法适用于复杂环境下车辆的检测与跟踪应用要求。

The paper proposes a method of vehicle detecting and tracking in complex environments. Firstly, by analyzing the vehicle external feature, the geometrical shape parameters extracted from its Blob blocks are employed to detect the moving vehicles. Then, vehicle tracking method is presented based on Blob analysis and CamShift algorithm. Experimental results show that the proposed method can detect the target vehicle and distinguish the vehicle and pedestrian accurately. The precision rate of vehicle detecting and tracking achieve 96.7% and 86.7% respectively, which prove that the method is appropriate for the demand of vehicle detecting and tracking in complex environments.