Real-time Indoor Navigation of UAV Based on Visual Delay Compensation

For the visual navigation system of unmanned aerial vehicles (UAV), it is of importance to estimate the relative time delay between camera and inertial measurement unit (IMU) because the image data with undetermined delay cannot meet the synchronous requirement with other sensors of UAV indoor navigation. In this paper, an image delay estimation is designed a method that can effectively estimate. By using extended Kalman filter (EKF), the fusion between the IMU data and the visual data is performed with delay compensation so that the real-time pose and velocity of UAV can be effectively estimated. Simulation and experimental results conducted on a UAV prototype show that the delay can be accurately estimated so that the control performance of real-time indoor navigation is obviously improved on the UAV prototype.