摄影测量在FAST反射面测量中的应用研究

FAST is the largest telescope we built in the worlld. We use several Total Stations do parallel measurement to measure 2,225 reflector nodes at the same time, and it needs about 90 minutes to finish one time static measurement. The measuring efficiency is low. The project plans to use the CRRS measuring system which is developed based on the principle of photogrammetry as the foundation,.choose the gaps characteristics between the reflector panels as the basis, to identify the location of the reflector nodes through the methods of edge detection and feature extraction. After that, the photo of each same node which is photoed by different cameras is used to do precision matching. Finally, the global positions of each node is calculated by using the method of 3 d reconstruction,and the high precision static measurement is fufilled. At the same time, the effective movement tracking algorithm is researched and applied to dynamic tracking measurement. The study of this topic will improve the efficiency of measurement greatly, and to improve the effective observing time of FAST telescope.

FAST是我国建造的世界上最大的望远镜。对于FAST反射面2225个节点的测量采用多台全站仪并行测量的方式，完成一次静态测量需要大约90分钟，测量效率低。本项目拟以基于摄影测量原理研制的CRRS测量系统为基础，以反射面单元面板之间的缝隙特征为依据，通过边缘检测和特征点提取等方法，识别出反射面各节点的位置；将CRRS测量系统中三台异源相机上对应的同一反射面节点进行精准匹配；最后通过三维重建的方法，解算出所有反射面节点在全局坐标系下的位置，实现高精度的静态测量。并寻找有效的运动跟踪算法，应用于动态跟踪测量。本课题的研究能极大地提高测量效率，以提高望远镜的有效观测时间。

FAST是我国建造的世界上最大的望远镜。FAST的反射面测量系统，采用多台全站仪对2225个节点并行测量的方式，完成一次静态测量需要大约90分钟，测量效率低。本项目拟以基于摄影测量原理研制的CRRS测量系统为基础，以反射面单元面板之间的缝隙特征为依据，通过边缘检测的方法，识别出反射面各节点的位置；并通过三维重建的方法，解算出所有反射面节点在全局坐标系下的位置，实现高精度的静态测量。本课题研究的方法能为FAST反射面测量新技术的开发提供依据。. 本课题通过对边缘检测算法的研究，完成以Canny算子为基础的算法对FAST反射面面板照片的节点提取，节点提取的准确率在80%以上。在FAST现场进行实验，所得到的重构精度可达到 =19.332mm和 =12.278mm，实验效果已达预期，能实现该课题中测量设备的所有功能。在精度方面可以进一步提高。