基于视频检测的卡尔曼滤波车辆跟踪算法及行为分析
Kalman Filter Vehicle Tracking Algorithm and Behaviour Analysis Based on Video Detection
-
摘要: 采用基于目标检测的运动目标跟踪策略, 首先用时间差分法进行运动目标检测, 获得运动目标的初始位置;然后采用卡尔曼滤波器预测运动目标下一时刻所在的位置, 再利用运动目标检测的结果评估和矫正预测结果, 获得运动目标的准确位置, 并依据跟踪结果进行车辆的行为分析.实验结果表明, 本方法可有效地解决运动目标部分遮挡及短时间全遮挡下的可靠跟踪问题.Abstract: This paper adopts the moving object tracking strategy based on object detection. Moving object detection is carried out using time difference method to obtain the initial position of the moving object, the position of the moving object is predicted at the next moment with Kalman filter; the prediction with the inspection result of the moving object is evaluated and corrected to achieve its precise position. In accordance with the tracking results, vehicle behavior can be analyzed. The experiments demonstrate that this method can provide an effective solution to reliable tracking of moving object partially shielded or fully shielded for a moment.