High Definition 3D Map Creation Using GNSS/IMU/LiDAR Sensor Integration to Support Autonomous Vehicle Navigation.

Area of Science:

• Robotics and Autonomous Systems
• Geomatics Engineering
• Sensor Fusion

Background:

• Autonomous systems rely on accurate environmental perception using diverse sensors like GNSS, IMU, LiDAR, radar, and cameras.
• Rigorous sensor modeling is crucial to mitigate inherent error sources for reliable integrated solutions.
• Mobile mapping systems and autonomous vehicles share environmental reconstruction goals but differ in sensor grade and operational objectives.

Purpose of the Study:

• To evaluate the mapping performance of autonomous vehicle sensor systems.
• To assess the potential of using only Velodyne sensor data for creating accurate point clouds.
• To determine if centimeter-level accuracy is achievable with this sensor configuration.

Main Methods:

• Focused solely on Velodyne sensor data for point cloud generation.
• Assessed the performance potential of autonomous vehicle sensor systems.
• No other sensor data or fusion techniques were employed in this specific investigation.

Main Results:

• Velodyne sensor data alone is sufficient for generating accurate point clouds.
• Achieved centimeter-level accuracy in the generated point clouds.
• Demonstrated the high-definition mapping potential of autonomous vehicle sensors.

Conclusions:

• Autonomous vehicle sensor systems, specifically Velodyne, possess significant potential for high-definition mapping.
• Centimeter-level accuracy is attainable using only Velodyne sensor data, validating its utility for precise environmental reconstruction.
• This research highlights the capability of cost-effective sensors in autonomous vehicles for advanced mapping tasks.