双摄像机系统对移动目标的跟踪

Tracking of Moving Object with Bi-camera System

提出了一种由一个静止摄像机加上一个动态摄像机组成的双摄像机实时跟踪系统.该系统利用两种形式摄像机各自的优点,克服它们自身的不足,实现了对运动目标的实时跟踪.文中给出了双摄像机系统的组成以及功能划分.利用由近似的摄像机投影模型导出的单应性关系,实现两个摄像机图像之间的目标匹配.系统在静态摄像机的图像平面上建立目标的2D运动模型,采用卡尔曼滤波实现目标位置的预测,由单应性关系求出动态摄像机图像平面上对应目标的预测位置,然后计算摄像机动态平台的旋转角度,实现对动态平台的伺服控制.

This paper proposes a hi-camera real-time tracking system which is composed of two cameras： one is a static camera, and the other is a removable camera. The system implements real-time tracking of moving object by making use of the respective advantages and overcoming the disadvantages of the two cameras in the system. The system structure and functional partition of the bi-camera system are presented. The object matching on two image planes of the two cameras is achieved by homography relation educed from an approximate camera projective model. In the system, a 2D motion model is established on the image plane of the static camera to predict the position of the object by Kalman filter, then by using homography relation, the predicted position of object can be obtained on the image plane of the dynamic camera. Then, the rotation angle of the dynamic camera platform is calculated, and the servo control of the dynamic camera is implemented.