基于雷达和视觉技术的车辆跟踪窗口

针对基于视觉技术的运动检测鲁棒性差、计算量大的缺点,提出一种融合雷达与视觉信息建立车辆跟踪窗口的方法。采用动力学模型描述车辆运动,通过基于雷达量测的扩展卡尔曼滤波(EKF)估计,确定目标初始位置,然后根据灰度信息自适应调整跟踪窗口的中心位置及尺寸,快速而准确地建立起以目标形心为中心且适合目标大小的跟踪窗口,缩小了后续图像处理区域。三维场景仿真实验证明,该融合方法能快速有效地建立起目标的跟踪窗口。

Vehicle Tracking Window Based on Radar and Vision Fusion

Motion detecting system based on vision technology easily results in the problem of enormous computation cost.To enhance system performance,a new method is presented for the installation of vehicle tracking window by integrating radar and vision information.For the mobile case,a dynamic model of vehicle motion is adopted,and the moving target's coarse position is identified through radar-based extended Kalman filter (EKF).Then both the center and the size of the tracking window are adaptively adjusted according to the image intensity data,so that a vision tracking window centering on the target centroid with its size suitable for target is established rapidly and accurately.The window identifies the working area for following image processing,and greatly reduces computation cost.The 3-D simulation experiment confirms that the proposed fusion method can efficiently establish the moving target's tracking window among multi sampling interval.