GeoTalks Express – Session 16 Questions & Answers

The sixteenth of Blue Marble’s GeoTalks Express online webinar series entitled What’s New in the Global Mapper Lidar Module v22, was conducted on October 14th, 2020. During the live session, numerous questions were submitted to the presenters. The following is a list of these questions and the answers provided by Blue Marble’s technical support team.

 

​Does the software take advantage of Multi-Core CPU or GPU processing to improve speed?

Many analysis and lidar functions in Global Mapper are multithreaded. By default, Global Mapper will typically use one fewer than the cores available when running multithreaded processes. In the newer versions of Global Mapper, there is a setting in Configuration > General > Advanced to set the maximum number of cores used for these processes. 

​Global Mapper does not utilize an available GPU for processing. A stronger or dedicated GPU will help to improve rendering ​in the 3D view and assist with some aspects of the Pixels to Point process. 

 

Do the terrain paint functions edit the original DEM or generate a new DEM file with the alterations?

Any changes made to terrain layers with the terrain painting tools in Global Mapper will be local to the workspace.  Saving the Global Mapper workspace will save the edits to the terrain in the workspace (*.gmw) file, but the source terrain file will not be edited. To save the altered terrain to an external file you will need to export the elevation data layer(s). The edits made to the terrain will be written into the export file. 

 

We would like the capability to extract bridges (class 17). Is that a possible future option in GM?

Global Mapper does not currently support an automated bridge classification or extraction from point cloud data. Using methods for manual classification, you can select and choose to assign points to class 17 to represent bridge features. 

Using the Perpendicular Path Profile view you can perform custom feature extraction for lines and areas. This workflow allows you to digitize/extract specific features in 3D based on your point cloud data. 

 

If you painted a road with the Smooth Terrain option, would this cause the road to “follow” the existing terrain, but still remain cross-axially flat (no sideways slope)?

 

The Smooth operation in the Terrain Painting tool will update the cells in the painted area to reflect an average elevation based on the values in the specified box size neighborhood for each cell in the brush area. So painting a road with the smooth operation will follow the terrain since the new values are derived from the terrain, but will not flatten the terrain to an even height so that it is cross-axially flat. 

 

You can use that to fix an elevation defect by voiding the defect then filling gaps. 

Yes, a great application for the terrain painting tool is to remove anomalies in the terrain by combining multiple operations in a single area. 

 

Do you also make cut and fill volume statistics with the Terrain Paint Tool?

The Terrain Painting operations do not produce cut and fill or volume calculations as you edit the terrain. However, after altering the terrain with the new Terrain Painting tool you can calculate volumes and/or cut and fill statistics for the edited layer. 

If you want find the volume between the unedited and edited terrain data, load two versions of the original terrain into Global Mapper, edit one layer, and then calculate the volume between the surfaces.

 

Is the feathering function on this similar to daylight to existing?

The feathering in the Terrain Painting tools created a softer slope from the edited terrain area to the existing terrain surrounding the edited area. This helps to prevent vertical or steep walls connecting the existing terrain to the edited pixels. 

The feathering size, like the brush size, is set as a number of grid cells. The actual slope of the feathered area is determined by the feathered area size, and the elevations of the edited and existing terrain. 

 

Can you create an exact width and slope for this terrain paint option?

There is not an option to specify a slope for an area when editing the terrain. To create a terrain area with a specific slope you would need to generate an area feature describing your desired slope and area. With that area feature, you can then generate a flattened site based on the elevations of the area feature.

 

Any help button on this tool to explain all those things you just mentioned?

Details on the Terrain Painting tool and options can be found in the Global Mapper knowledgebase.

 

Can you use an existing shape file for Terrain painting? 

The Terrain Painting tool does not have the option to use existing vector features to define the area to edit. The tool is designed to manually edit the terrain.

Other tools in Global Mapper, like those to generate terrain or create a flattened site plan allow you to use existing area features to create new terrain data. 

 

Can the Lidar Module edit/reclassify individual points in the point cloud in a 3D view?

Yes, you select point cloud returns in the 2D, 3D, or Path Profile view and manually classify them in the Lidar Module. The Manual Classification toolbar includes some buttons to classify more common classes, however, by choosing to edit the selected features, or using the Change Lidar Class tool you can select any class for manual classification. 

 

How does the building classification work with buildings under vegetation or very large buildings? 

The new building classification method looks at the points in a 3D space, as opposed to a gridded 2D space, and therefore does a better job identifying building points near to or under vegetation. With this new method, there is also an improvement to the classification across larger flat building roofs. 

 

Can classification be run in a batch process?

Yes, the automatic classification tools, including building and tree classification, can be run for only one point cloud, or across many point cloud layers. 

Point cloud classification can also be scripted through Global Mapper script if you prefer to perform the classification without viewing the data in Global Mapper. 

 

Would this smoothing tool work well in reducing noise in a roadway without losing the integrity of the data?

The Smooth operation in the Terrain Painting tool will update the cells in the painted area to reflect an average elevation based on the values in the specified box size neighborhood for each cell in the brush area. The smoothed values are changed from the original elevations, but since they are derived from surrounding terrain some integrity is retained. 

 

What would I change to if I have a very dense data set, like 250 points per meter? Would I change the cluster higher or lower?

The parameters set for the building and tree classification, as well as any automatic classification or extraction tool in Global Mapper, are dependent on the data being used. The Minimum Cluster Size parameter in the segmentation non-ground classification method sets the minimum point count for a cluster of identified points to be considered as a building or tree feature. For denser data where there are more points per meter you would likely want to increase the Minimum Cluster Size as there would be more points available and identified in clusters as opposed to fewer points being identified in sparser data. 

 

In the Terrain Painting tool, is feathering added to brush size?

The feathering distance is added around the brush that indicates the pixels to be edited. When using the Terrain Painting tool, a red circle around the cursor represents the brush size and a large blue circle represents the feathering area. 

 

What is the best accuracy that we can achieve in a classification? Is it possible to make a change detection in mm accuracy when comparing several epochs of measurements?

 

The classification accuracy depends on the data and the parameters used for classification. The automatic point cloud classification tools in Global Mapper do a great job classifying a point cloud, but if you want to detect small changes between datasets, visually inspecting, manual classification, and cleanup will be needed. 

 

In the Pixels to Points tool report, Is there an accuracy component? (RMSE error) of camera and GCP locations?  

The Pixels to Points log file generated when running the process includes detailed logging related to the control points. The log includes triangulation, transform, and fit information for the control points.

 

​Can the overlap tool provide a grid with elevation change between clouds?

The Find Overlap tool in Global Mapper will only show where the point cloud layers overlap. To create a grid quantifying the change in elevation between the layers you would need to first generate a separate elevation grid layer for each point cloud. With the generated elevation grid layers you could then use the Combine/Compare Terrain tool to generate a new gridded layer demonstrating the elevation change between the grids derived from your point cloud layers. 

 

If you apply the “Smooth terrain” option, then you close the workspace and then reopen it, is it still possible to use the “Revert to original heights” option?

Yes, if you load an elevation layer into Global Mapper and edit it with the Terrain Painting tools, save and then reopen the workspace, you should be able to revert the data to the original values from the file with the Terrain Painting tool. 

Leave a Reply

Your email address will not be published. Required fields are marked *