Automated classification system for object detection using multi-LiDAR sensor network

This paper presents an automated classification system for object detection in multi-LiDAR sensor networks. The system leverages a deep learning model for an initial coarse classification and subsequently employs feature-based extraction of secondary classes from point-cloud data. Additional labeling and deep learning-based processing on these secondary classes facilitate a more fine-grained classification. The efficacy of the presented approach is demonstrated with real-world data.