基于单目视觉和激光测距仪的位姿测量算法

Position-pose measurement algorithm based on single camera and laser range-finder

基于单目手眼相机和激光测距仪,提出了一种尺寸未知的空间矩形平面的位姿测量算法。该算法不需要知道矩形平面的4个顶点的物体坐标,只需要知道它们的图像坐标、激光点的图像坐标和激光测距结果,就能够计算出尺寸未知空间矩形平面在相机坐标系下的位姿,并且计算出矩形平面的尺寸。通过建立单目手眼相机和激光测距仪的数学模型,对该算法进行了验证。实验结果表明,该算法是有效的,可以应用于机器人对空间物体的跟踪、定位以及抓取。

An algorithm is proposed to measure the position-pose of unknown sized rectangle based on single handeye camera and laser range-finder. The algorithm does not need to know the object coordinates of four apexes of the rectangle. From their image coordinates, the image coordinate of the laser point and the measurement result of the laser-finder, it can calculate the position-pose of the rectangle in the camera coordinate system, and obtain the size of the rectangle. The mathematical models of the single hand-eye camera and laser range-finder are built to verify this algorithm. Experimental results show that the algorithm is effective, and can be applied in robots for tracking, locating and grasping objects.