A LiDAR and IMU Integrated Indoor Navigation System for UAVs and Its Application in Real-Time Pipeline Classification

Area of Science:

• Robotics and Automation
• Computer Vision
• Sensor Fusion

Background:

• Accurate vehicle mapping and localization are challenging, especially for unmanned aerial vehicles (UAVs) due to altitude variations and complex motion dynamics.
• Existing ground robot localization methods face difficulties when applied to UAVs in unknown indoor environments.

Purpose of the Study:

• To propose a robust and efficient indoor mapping and localization solution for UAVs.
• To integrate real-time pipeline classification within the proposed navigation framework for industrial applications.

Main Methods:

• Utilizing low-cost Light Detection and Ranging (LiDAR) and Inertial Measurement Unit (IMU) sensors for data acquisition.
• Employing a point-to-point scan matching algorithm with a horizontally scanning LiDAR for planar position estimation.
• Using a vertically scanning LiDAR for accurate altitude estimation relative to the floor.
• Fusing data from both LiDARs using a Kalman filter for precise 3D position determination.
• Implementing a novel pipeline classification method based on radius estimation, region of interest (ROI) selection, and directional histograms.

Main Results:

• Demonstrated accurate planar and altitude estimation for UAVs in indoor environments.
• Successfully fused sensor data to achieve robust 3D localization.
• Validated the real-time classification of pipelines within the mapped environment.
• Experimental results confirmed the feasibility of the integrated navigation and classification system.

Conclusions:

• The proposed system offers a robust and efficient solution for indoor UAV mapping and localization.
• The integration of real-time pipeline classification enhances the system's utility in industrial plant engineering.
• The use of low-cost sensors makes the solution economically viable and practical.