基于单目运动摄像机的三维物体运动参数估计

针对单目视觉监测系统摄像机和目标物体同时运动时,如何有效测量物体三维运动参数的难题,提出一种基于单目序列图像对插入虚拟视点的算法确定运动物体位姿及运动参数的方法。动摄像机连续采集运动物体的图像序列,根据摄影几何原理对摄像机和运动物体进行分析,结合本质矩阵的特性,估算空间物体运动参数。实验结果表明,该方法能够在为摄像机与目标物体同时运动的情况下,有效地估测出物体三维运动参数,且该方法简单有效,具有较高的精确度和准确性。

Analysis of motion parameters of 3D object based on moving monocular camera

How to measure the 3D motion parameters of the moving object effectively becomes a technical problem, while the camera in the monocular vision system moves in the same time. A method based on visual view inserted in sequence image pair is proposed to determine the pose parameters and the motion parameters of the target object. The moving camera captures the image sequences of the moving object consecutively. According to the theory of photo geometry, the motion relation between camera and object is analyzed. Then, combined with the characterization of essential matrix, the motion parameters of object are achieved to be estimated. The experiment results show that the proposed method can effectively calculate the motion parameters of target object under the situation of camera and object both moving. And the method is simply to realize, meanwhile, it is of high precision and accuracy