Mo-norvana is a novel efficient framework, which powers the core of LidarTools for processing mobile lidar data efficiently. Our standalone and light-weight software carries three key features including various functions designed to help users manage, process, and analyze their mobile lidar data.
Data/Metadata Management
Sensor Trajectory Reconstruction
Mo-norvana is able to reconstruct the sensor trajectory from an unorganized mobile lidar point cloud. The trajectory includes information such as location, orientation, and speed, so that the users can have more insight of their data. The trajectory along with other information can serve as metadata of mobile lidar point cloud for data management.
Easy Data Annotation
Once the trajectory is reconstructed, Mo-norvana is also able to align the data to the scan patter and render a 2D scan panorama. The scan panorama can directly serve as a thumbnail of a scan for a quick preview of the data. It is also a lot easier to make data annotation on a scan panorama for communication or management purpose.
Upcoming
- Data Indexing
Data Quality Assessment
Per-point Uncertainty
The uncertainty for each point in lidar point cloud is different and varies with multiple factors (e.g., range and incidence angle) which can only be computed with the sensor trajectory available. While the global accuracy of the data can be assessed by ground control points, Mo-norvana estimates the local uncertainty rigorously to assess the data quality quantitatively on a case by case basis. Additionally, the per-point uncertainty can be used for merging geospatial data from different sources and the accuracy of product can be properly reported.
Upcoming
- object-based cloud alignment
Feature Extraction
Quick Segmentation
Mo-norvana has segmentation function that can group local points based on their geometric attributes. Millions of points can be segmented within minutes such that users can handle each segment instead of each individual points, which significantly reduces manual tasks that sometimes can take hours. The point cloud can then be easily categorized into different classes with more advanced information extracted from each segment.
Road Surface
Road surface extraction in Mo-norvana is built upon the point cloud segmentation result and take advantage of the context information in the sensor trajectory. The road surface detection built in Mo-norvana does not solely rely on a well-defined road boundary (e.g., curbs, markings). Thus, in addition to the lanes, the results also include shoulders, ramps, driveways, and other area that a vehicle has access to.
Upcoming
- Road Markings
- Powerlines
- Poles
- Signs
- Curbs
- Barriers
- Buildings
- Bridges
- …