Research On Pedestrian Navigation Methods Via Inertial Measurement Unit With The Aid Of GNSS And Visual Odometry

In modern navigation campaigns,inertial navigation system(INS)is very attractive and has many merits for integrating with global navigation satellite system(GNSS).A typical GNSS/INS integrated system,besides bridging the GNSS navigation gaps with inertial sensors in many changeling environment.it can efficiently overcome the drawbacks of inertial drifts and biases with the aid of GNSS.However,the integrated navigation accuracy is strongly dependent on the inertial instruments.With the increasing demands of seamless navigation,visual odometry(VO)is drawn intensive attentions.It generates the accurate and consistent egomotion estimation and can also continuously correct the INS error in case GNSS signals are absent.Such,it naturally becomes a good auxiliary for improving the overall performances of navigation system.This paper deals with the pedestrian navigation problem of IMU with the aid of GNSS and vision sensor.The main contents are as follows:1.We develop behaviors classification base distance measuring system for pedestrians via a foot-mounted inertial sensor.Two key issues are studied:step detection and step length determination.A time-and frequency-domains joint step detection method based on the behaviors identified is proposed.And a two-segment functional based step length model is established to adapt both walking and running behaviors.2.We propose a visual-inertial odometry based on the point and struct lines in man-made building enviroments.Our algorithm optimizes the prior information from marginalization,the residuals of IMU measurement model and the point and structline features re-projection residuals.3.We deals with the integration problem for GNSS/INS integration navigation with the aid of vision sensor and develop a vision-aided GNSS/INS fusion method based on the sequential Kalman filter structure.With the aid of VO,the performance of GNSS/INS is significantly improved.Especially for these GNSS challenging enviroments,corrected by monocular camera,INS can endure a much longer duration without GNSS signals.