移动背景下运动目标检测与跟踪技术研究

Target detection and tracking under moving background

文中分析了移动背景下运动目标检测与跟踪技术。介绍当前移动背景下运动目标检测常用的方法,指出了难点与关键。结合目前的工作,给出了一个移动背景下运动目标检测与跟踪的解决方案,并以人物为跟踪目标进行了实验,较好地实现了移动背景下的全局运动参数鲁棒性估计以及人物目标在不同场景下的跟踪。但是在目标检测时图像边缘白色块的去除和跟踪时目标遮挡的阈值自适应调节等方面有待进一步研究改进。

The target detection and tracking technology under moving background is analyzed in this paper.First,the difficulties and key points of using conventional processing method are explained.Then,a solution for moving ob-ject detection and tracking is proposed.Finally,the solution is tested with human targets,and experimental result shows that the method has a robust estimation of the global motion parameters,and achieves the target of tracking people un-der different scenes.But the current method suffers from white blocks in the edges of the detecting images and adaptive thresholds setting.So further researches on improvement is needed.