Study on a tight integrated navigation and positioning algorithm of GNSS/visual observation

The Global Navigation Satellite System (GNSS) signals are susceptible to occlusion or electromagnetic interference in weak GNSS environments, which may seriously affect the reliability, continuity and accuracy of navigation and positioning. To solve this problem, this paper studies a tightly integrated navigation and positioning method with GNSS and visual combined observations. Firstly, with the image data collected by the camera, the ORB-SLAM2 open source platform is used to obtain the visual position result increment, and then combined with the GNSS pseudo-range observation data using Kalman filter for combined positioning solution. Validation is conducted with measured GNSS pseudo-range observation data and image data, the test results show that the algorithm can not only effectively improve the continuity and accuracy of navigation and positioning in weak GNSS environment, but also maintain continuous navigation when the number of satellites is less than four.