交通运动目标检测、跟踪及轨迹生成研究

Research on traffic moving object detection,tracking and track-generating

提出一种基于视频的车辆检测,跟踪和轨迹生成算法。该算法由改进的车辆检测方法,快速跟踪算法和新的车辆轨迹生成算法3部分组成。基于区域和车辆间的相互关系,在视频序列中,车辆被视为可自主运动团块。在对该团块实现有效跟踪及获取运动轨迹的基础上,运用相关的数学手段可获得团块其它运动信息。在高速公路上的实验结果表明,该车辆检测,跟踪算法切实可行,轨迹生成技术可用于交通流检测。

Algorithms for video-based detection,tracking and trajectory generation of vehicles is presented.Processing is done by three steps：Improved vehicles detection method,faster tracking algorithm and new vehicles trajectory generation technique.Vehicles are viewed as rectangular blocks with certain active action.The proposed method is based on the correspondences between regions and vehicles,as the vehicles moving through the video sequence.On the basis of effective blocks tracking and obtaining moving trajectory,other moving information could be acquired with related math method.Experimental results from highway scenes demonstrate the effectiveness of the vehicles detection and tracking method,the trajectory generation technique is used for traffic flow detection.