Vision / inertial integrated navigation method for quadcopter based on EKF state observer

Abstract

Quadrotor UAVs are widely used in the field of power line inspections because of their advantages such as high accuracy, strong adaptability, and strong obstacle surmounting capabilities. However, in practical applications, the autonomous navigation of the UAV still has problems such as poor anti-interference ability, insufficient accuracy and still dominated by manual control. Visual navigation is difficult to obtain sufficient information or to track feature points when there is insufficient lighting, sparse features, and large maneuver. These problems can reduce the accuracy of visual navigation. The position estimation error directly obtained by integrating the IMU data of the inertial navigation unit will gradually increase with time. A vision / inertial integrated navigation method for a quadcopter UAV based on loosely coupled extended Kalman filter algorithm is proposed in this paper. Design state observer for UAV based on extended Kalman filter. The visual/inertial integrated navigation algorithm is simulated in gazebo. Finally, an experimental platform is set up to verify the vision / inertial integrated navigation algorithm experimentally. The overall position calculation error meets the positioning accuracy requirements during power line inspections.