Lidar-based catenary data processing
AI factory: Monitoring and accessing asset condition using AI and point cloud data.
Goal
Assess the condition of catenary system.
Lidar data processing has been implemented in various stages in the AI Factory pipeline. Process of stitching together multiple scan is termed as registration and a part of raw point cloud generation. Pre-processing involves removing unwanted data points and down-sampling for further processing. Cropping is performed to remove regions that are not required, clustering is used to create collections of points as per features and segmentation is separation of these groups as individual assets. These separated assets in case of catenary processing are the masts and individual wires. Classification of theses individual assets is performed through identification and labelling as tension wire or reinforcement conductor etc. The data hence extracted is used for development of visualization or modelling purposes.
Sponsor: Trafikverket, Infranord, Norrtåg,Transitio
Researchers: Ramin Karim (PL), Amit Patwardhan
Duration: 2021-22
Updated: