Development of an AI-LiDAR Software (CTI) for Automated Detection of Risk Trees in Electrical Corridors

This study introduces CTI Corridor Tree Inspect, an innovative software solution for automated detection of hazardous trees in electrical corridors using LiDAR data, tailored to the challenges of Cameroon's dense forest environments. Integrating drone-acquired LiDAR point clouds with artificial intelligence techniques, including DBSCAN clustering for vegetation segmentation, the software enables precise identification of trees posing risks to power lines. Developed in C# with HelixToolkit for 3D visualization, the prototype processes datasets of up to 10 million points, achieving a mean F1 score of 0.89 and processing times under 60 seconds per km. Tested on a 5 km pilot corridor, it detected 142 trees, with 34 classified as high-risk, demonstrating 89.7% detection precision. The system generates actionable PDF reports with geospatial metrics, facilitating proactive maintenance for utilities like ENEO. While limitations include sensitivity to dense foliage and lack of species classification, this research advances vegetation management in tropical settings, offering scalable, cost-effective tools for infrastructure resilience. Future enhancements could incorporate machine learning for adaptive segmentation and multi-temporal monitoring.   Keywords: LiDAR, Drones, Artificial Intelligence, Vegetation Management, Electrical Corridors, Tree Detection, Cameroon   Received Date: August 21, 2025               Accepted Date: September 13, 2025             Published Date: October 01, 2025 Available Online at ijsrisjournal.com