提出利用双目视觉系统检测六自由度大载荷实验平台的位姿。首先,采用时空联合的分割算法提取运动目标并由运动目标上不共线三点描述当前时刻运动平台的位姿;然后,根据运动平台负载大、惯性大的特点,对相邻时刻的位姿变化加入补偿,以此估计平台下一时刻的位姿。实验结果表明,该算法在较小的允许误差范围内能够较好地跟踪运动平台,准确地估计下一时刻的位姿。
This paper put forward an algorithm to measure the position and orientation of the 6-DOF(the six degrees of freedom) large load test platform based on the stereo vision.Firstly,segmented the moving test platform from the background based on temporal and spatial union information and expressed its position and orientation by coordinates of three feature points which were not in a same line.Then,according to the platform's features,large load and large inertia,forecasted its next position and orientation by adding compensation to the current and previous position and orientation change.Experimental results demonstrate that this algorithm can calculate the current position and orientation of the 6-DOF platform,and can predict the next position and orientation accurately.