GNSS Application In Visual Navigation And Measurement

Nowadays,satellite and visual navigation have made a big progress,but it could not meet people's needs based on just one kind of sensor with the increasing demands of position and navigation.It is worth to be discussed how to integrate satellite positioning and visual positioning to provide people with more convenient,robust and intelligent service.Starting from satellite navigation and computer vision,at first,this paper introduces code/carrier pseu-dorange point positioning and deduced static and kinematic relative positioning in detail.Also,this chapter introduces the theory of single-image space resection,multiple-image space forward intersection and bundle adjustment in photogrammetry starting from collinearity equation.And then,introduces the basic framework of SLAM(Simultaneous Location and Mapping).Second,this paper studied the application of GNSS of aerial photogrammetry,implement single-image space resection,multiple-image space forward intersection and bundle adjustment in a small region and evaluate the accuracy.In this progress,a method to detect gross errors of sparse ground control points is proposed to solve the problem that convergence cannot be found because of the gross errors.This new method is used in GNSS-aided photogrammetry.A three-dimensional reconstruction of the campus of Shandong University at Weihai was carried out using a drone.The accuracy of the 3d model is verified in playground.And then,design and implement GNSS and monocular camera integrate navigation,convert navigation trajectory between two systems using data smoothing and rotation matrix,and then analyzes coordinate error.Kalman Filter and Bayes Estimation are used to process the data and a better position is got.The positioning accuracy of the moving carrier with short time passing through the occluded area is analyzed emphatically,which can reach the level of meters.Finally,this paper studied the integrated positioning of GNSS and Visual Odometer base on smart phone and complete the experiment.At the beginning,the accuracy of phone positioning is examined.Then,analysis and compared the different period length of trajectory matching.At last,fuse the positioning data of mobile phone GNSS and visual odometer.The combined system's positioning accuracy and reliability is better than single system.