LiDAR-Enhanced Real-Time Tree Position Mapping for Forestry Robots

This article evaluates the effectiveness of an autonomous robot in creating a tree pose map using both simulated and experimental environments, validating its potential for forestry applications.This work also demonstrates the implementation of FastSLAM on a four-wheeled differential-drive robot, integrating real-time tree detection and tracking through LiDAR-based point cloud data.An algorithm is proposed to generate a map showing both the robot's path and detected tree positions during movement.Performance Metrics Analysis (PMA) revealed a high True Acceptance (TA) rate, confirming accurate tree position estimation.Experimental results validated the algorithm's reliability, showcasing strong distance accuracy with minimal discrepancies between actual and estimated positions.These findings highlight the system's potential for advancing forestry management through precise robotic navigation and mapping.