Machine Learning for Utility detection based on Photogrammetric and Lidar-based data
It is suitable for those interested in photogrammetry, Lidar point cloud, and machine learning
Utilities form the essential infrastructure of a contemporary society, fulfilling fundamental requirements for both residents and industries. To streamline the construction and upkeep of these vital facilities, it is imperative to accurately map and record the location and characteristics of underground utilities. While traditional mapping methods rely on conventional surveying techniques, demanding highly skilled human resources and yielding only 2D representations, our commitment extends to embracing innovative technologies. We are actively exploring alternative solutions that leverage a more sophisticated and user-friendly toolset, integrating new technologies into our working procedures to enhance efficiency and precision.
Machine Learning for Utility Detection: The utilization of photogrammetric processing and Lidar measurement represents a pivotal advancement in utility data collection. A central focus within this realm pertains to extracting essential information from point clouds for utility registration and assessment purposes. Research initiatives that delve into the development of automated methods for deriving valuable insights from these data sources are highly valued and encouraged. In this study, we will deploy a machine learning-based strategy for the automatic detection of utilities using RGB images and Lidar point clouds.