High-Accuracy Positioning in GNSS-Blocked Areas by Using the MSCKF-Based SF-RTK/IMU/Camera Tight Integration

The integration of global navigation satellite system (GNSS) single-frequency (SF) real-time kinematics (RTKs) and inertial (INS) has the advantages low-cost low-power consumption compared to multiple-frequency GNSS RTK/INS system. However, due vulnerability signal reception, application SF-RTK/INS is limited in complex environments. To improve positioning accuracy GNSS-blocked environments, we present a tight based on BDS/GPS SF-RTK, measurement unit (IMU), monocular camera. In such system, multi-state constraint Kalman filter (MSCKF) adopted integrate pseudo-range, phase-carrier, measurements, visual data tightly. A wheel robot dataset collected under signal-blocked conditions used evaluate its performance terms position, attitude, run time, respectively. Results illustrated that presented model can provide higher position those provided by visual-inertial Moreover, average running time presents potential method applications.