Research on GNSS-assisted Visual Dynamic Positioning Method
Aiming at the problem that the moving platform cannot be continuously positioned when the Global Navi-gation Satellite System is intermittently out of look in urban environment,a GNSS-assisted visual dynamic positio-ning method was researched.This method used a camera to shoot three images of the street scene and record the GNSS coordinates of the corresponding positions,and the GNSS coordinate system was the world coordinate system.Firstly the SIFT feature points are extracted from each of the three images separately for three-dimensional recon-struction.And the coordinates of these feature points in the camera coordinate system corresponding to the first im-age and the pose transformation of the camera between these three positions are obtained.The GNSS corrdinates corresponding to the three images are used to obtain the pose transformation between the camera coordinate system and the GNSS coordinate system.Secondly,by shooting an image of this street scene arbitrarily,P3P can be used to calculate the camera pose corresponding to the current image to obtain the GNSS coordinates to the camera plat-form at this time.Finally,the image data of a certain scene is experimented to verify the method,and it's positio-ning accuracy is generally not more than 6 m.
visual positioningGNSS positioningfeature point matchingpose and attitude computationcoordi-nate transformation
万方数据
维普
