In the last decades, the visual navigation system has been investigated by many researchers as an aided navigation system for Inertial Navigation System (INS) in the Unmanned Aerial Vehicles (UAV).In this research, for improving the INS errors a new approach based on feature tracking 4 Piece Queen Storage Bed algorithm is used.In this approach, in order to estimate the feature points in the current image, the INS states, the feature points of the previous image and dynamic equations are used.Also, in this approach, for improving the estimation Stereo Systems of terrain points, the outlier estimated feature points delete.Furthermore, in this article, for improving the altitude error, a barometer is used by the mentioned vision navigation system.
The simulation results illustrate the desirable accuracy of the vision system and barometer observations in the update step of Extended Kalman Filter (EKF) and remarkable performance of integrated navigation system for calculating the UAV navigation parameters.