Point cloud data are changing the way researchers and practitioners conduct their studies and activities across various fields. Being the more descriptive and intuitive way to represent the world in 3D, point cloud are fascinating insiders due to the great availability of sensors and methods to produce them. Both photogrammetry and laser-based systems have reached a comparable and outstanding degree of maturity. The latter, LiDAR, as an active and accurate remote sensing technique, is extensively adopted in several domains including land use/land cover classification, urban modelling, smart cities, infrastructures’ inspection, and environmental applications like forest inventory and water management systems. Meanwhile, considering its proximal compartment and adaptability to many settings such as terrestrial, mobile or airborne (namely TLS/MLS/ALS), it represents one of the most used measuring systems. The ever-reducing costs, the increasing easiness of use and the increasing demand of understanding our surroundings, made such solutions unavoidable like never in the past. However, the output data (i.e., point clouds) are complex and there are still several bottlenecks hampering their full exploitation. For examples, they have large volume, variable densities, unstructured geometries, and are often discontinue and incomplete. Thus, there is a need to perform cutting edge researchers to improve the body of knowledge and to fulfil a thorough understanding of point clouds, including registration, feature extraction, semantic segmentation, modelling, large-scale computing, visualization and so on.
This Article Collection is focused on the above-mentioned issue, and it welcomes both theoretical and applicative papers dealing with LiDAR point cloud processing, understanding and interpretation that make extensive use of artificial intelligence methods. The article collection aims to highlight the importance of this topic, being a showcase of the most recent advances in LiDAR point cloud processing, understanding and interpretation. The contributions should match with the following topics, but not limited to:
- Semantic Segmentation of proximal and remotely sensed data
- Point Cloud co-registration
- Real Time processing, mapping and positioning (e.g., SLAM based systems)
- Data fusion
- Environmental applications (e.g., forestry inventory)
- Scene understanding
- Synthetic Data
