Real-Time Optimal States Estimation with Inertial and Delayed Visual Measurements for Unmanned Aerial Vehicles.

Area of Science:

• Robotics
• Computer Vision
• Sensor Fusion

Background:

• Accurate motion estimation is critical for Unmanned Aerial Vehicle (UAV) applications.
• Existing methods often rely solely on visual odometry, which can be susceptible to drift and environmental conditions.

Purpose of the Study:

• To develop a robust and computationally efficient motion estimation solution for UAVs.
• To integrate data from an Inertial Measurement Unit (IMU) and a monocular camera for improved accuracy.
• To achieve real-time 3D position and translational speed estimation.

Main Methods:

• A two-step approach involving visual localization and multisensory data fusion.
• Utilizing IMU attitude information within Kalman filter equations for enhanced localization.
• Implementing a multi-rate delay-compensated optimal estimator based on the Kalman filter.
• Optimizing the estimator for onboard, real-time computation.

Main Results:

• The proposed method effectively fuses IMU and monocular camera data.
• Accurate estimation of 3D positions and translational speed was achieved.
• Experimental validation on a quadrotor system demonstrated superior performance compared to other methods.

Conclusions:

• The developed multisensory fusion approach significantly enhances UAV motion estimation accuracy.
• The real-time, computationally efficient design enables onboard implementation.
• This method provides a reliable solution for critical UAV navigation tasks.