利用视觉信息辅助GNSS紧耦合定位方法

A vision information aided GNSS tightly coupled positioning method

针对在复杂地理和电磁环境条件下,全球卫星导航系统(GNSS)信号易受到遮挡或者电磁干扰,往往无法正常地提供导航定位服务的问题,提出1种利用视觉信息辅助GNSS的紧耦合定位方法:将影像中世界坐标已知的特征点视作“视觉伪卫星”,利用共线方程求解相机近似坐标,再反算出相机至物方特征点的距离,与GNSS测距信号联合定位;利用相机进行远距离定位时近似坐标精度不高,通过联合GNSS信号可提升定位精度,另一方面视觉测距信号具有抗干扰的特性,深耦合定位可提升GNSS在复杂环境下定位的可用性。实验结果表明,紧耦合定位算法具有可行性,能够提升定位精度。

Aiming at the problem that under the complex geography or electromagnetic environment,it is liable to the blockage or electromagnetic interference for GNSS signals,which leads to the difficulty in providing normal navigation and positioning services,the paper proposed a vision information aided GNSS tightly coupled positioning method:the features with known coordinates in the image were viewed as the‘vision pseudolite’,and the approximate positions of the camera were determined with the collinear equations,then the distance between the features and the camera was computed to jointly position with GNSS ranging signals;for which,on the one hand,the camera positioning may have poor accuracy for long distance positioning,while its precision could be improved by introducing GNSS signals,on the other hand,the vision ranging signals may resist the electromagnetic interference,thus the tightly coupled positioning could improve the availability of GNSS under complex environment.Experimental result showed that the proposed method would have feasibility with higher positioning accuracy.