论文部分内容阅读
介绍了一种以目标形心位置和帧间位移测量的多目标跟踪方法。当目标在传感器视场中交叉运动,图像相互重叠时,得到由两个目标形心位置线性组合的融合后的图象测量方程。图像的重叠导致状态估值误差的相互依赖性,利用联合测量耦合滤波器对两个目标进行状态估值,并提出了改进的滤波模型。仿真结果表明,该算法能实现对交叉运动目标进行精确跟踪。
A multi-target tracking method based on target centroid position and interframe displacement measurement is introduced. When the target crosses the sensor’s field of view and the images overlap each other, a fused image measurement equation that is linearly combined by the two target centroid positions is obtained. The overlap of the images leads to the interdependence of the error of the state estimation. The joint measurement coupling filter is used to estimate the state of the two targets and an improved filtering model is proposed. Simulation results show that the algorithm can accurately track the cross-moving target.