Groups within the context of the point cloud engine are used to segment the point data for either or both of the following reasons.
- Isolate individual scans/setups
- Separate out different sections/parts of a larger point cloud. e.g. vegetation, roads, buildings... etc.
By default when an e57 file is imported, as long as the scans have not been merged in the software that generated the file, the scan will come in with a series of groups. Each represents a different setup. If multiple scans are detected in an e57 the shading will be set to "Group Colour" so that it's easy to see which part of the cloud was observed from each setup. See below.
LAS/LAZ files do not allow for multiple scan locations, so when importing a series of LAS/LAZ files where each file is actually a different scan location, the option to Merge scans should NOT be enabled. That will preserve each scan and allow it to be viewed and manipulated independently of each other if needs be. LAS/LAZ files can, however, have classifications built into them.
In the example above different colours have been applied to the various classifications. These have become groups with the appropriate names e.g. high vegetation, low vegetation, ground... Notice the Groups section in the properties of the point cloud now contains a new section for each group. Thereby allowing easy access to each, and the settings being used to render it.
Notice that alongside each group's name, is a checkbox that controls whether or not the group is displayed. Hence when regrouping data, it's simple to remove unwanted groups from the view just by un-ticking them.
Group Key
If the Point Cloud has groups and the display mode of the chosen groups is set to "group colour" then the below key will be displayed.
Re-Grouping
Once a point cloud has been properly loaded into the 3d Viewer, then additional groups can be created and points transferred into them from other groups. There will always be at least one group, in which case that group would contain the entire scan.
Adding point data to a group can be accomplished via one of three ways.
- Polygon Selection
- Distance from a DTM
- Group by Overlapping Scan
Before any points can be reassigned to a group, the intended group must first exist. If it does not, then new groups can be generated via the Add Group button above the list of groups in the properties of the point cloud.
Adding Points To Groups By Polygon.
This option re-groups points to a new group by drawing a selection polygon around the required points.
Having selected this tool, the tool specific properties are displayed and docked to the far right. The options are as follows.
- In Polygon: If this is enabled then the tool is carried out on points inside the selected area. Conversely, if it is unchecked, then reverse is true.
- Obey Clipping: If this is enabled then the tool is carried out on just those points which are currently inside any clipping box. Conversely, if it is unchecked, then reverse is true.
- Point Cloud: Specifies the cloud to use as the source data.
- Original Group: Allows individual groups or simply all the groups to be available as sources of data for regrouping.
- Target Group: Specifies the group in which the selected points should be reallocated.
Having made a selection onscreen, either right click or press Enter to start the regrouping. Note on large datasets this process can take a while.
Adding Points To Groups By DTM.
This option re-groups points to a new or different group by specifying a distance tolerance from a given DTM.
The basic idea being that a simple DTM is generated over roadways or pavements where traffic and pedestrian noise needs to be removed. This DTM can be created either by manually digitising points in the point cloud where the roads and other surfaces can be picked up. Doing this in enough places allows a primitive grond model to be derived. Alternatively, the gridding tools can be used to drape a grid of levels over the scan, but with the Height Sampling mode set to lowest. This generally creates a good first draft model that is easily edited to remove any unwanted high points, in order to arrive at a decent but still approximate ground model.
Once a representative DTM has been generated it can then be used to re-group data in the point cloud based on the distance of each point from the DTM surface.
Having selected this tool, the tool specific properties are displayed and docked to the far right. The options are as follows.
- Point Cloud: Sets the source data for the tool.
- Model: Sets the DTM to be used.
- Original Group: Allows individual groups or simply all the groups to be available as sources of data for regrouping.
- Target Group: Sets the group to be assigned to all the points within the search parameters.
- Min Distance: Controls the minimum distance from the DTM to start searching. Zero is a valid number as this allows data very close to either side of the DTM to be isolated.
- Max Distance: Sets the furthest distance away that data can be grouped.
- Visible Only: If this is checked then only point data that is visible is eligible to be regrouped.
- Vertical Distance: if this is checked then the vertical distance to the DTM is considered. Rather than perpendicular offset to the surface of a triangle.
- Height Mode: Controls how the height calculation considers the point data. To isolate just the points close to the DTM for instance, the Min Distance would be set to 0 and the Max distance to say 0.05.
- All - Data is considered no matter where it sits in relation to the DTM.
- Above - Data has only considered if it is above the DTM surface.
- Below- Data has only considered if it is below the DTM surface.
When all the values are set appropriately, then hit Apply. Depending on the size of the data set, it could take a little while to accomplish.
Before any re-grouping
After re-grouping and setting the new Noise group above the road to Intensity.
After re-grouping and turning the Noise off.
Adding Points To Groups By Overlapping Scans.
This option re-groups points to a selected group by looking for data that does not appear in overlapping laser scans. Roads and other permanent features should exist in all the overlapping scan data. However passing cars and pedestrians will not occupy the same space from scan to scan and can, therefore, be filtered out and moved into a selected group, which can then be turned off.
Having selected the overlapping scan tool, its specific properties are displayed and docked to the far right. The options are as follows.
- Point Cloud: The point cloud to filter.
- Noise Group: The target group for any data filtered out.
- Min Group Count: The minimum number of overlapping groups/scans that data must appear in for it to be kept and not re grouped.
- Box Size: The cell size to test. The entire extents of the cloud are broken down into cells of this size before testing for overlapping data. The lower the value the better the result, but the more data will probably be removed. A good starting value is 0.02M or 2cm.
- Visible Only: With this option enabled, only visible data (points not already hidden) are available to be tested.
- Apply: Initiates the filtering.
Before
After Filtering and turning the Noise group off.
Adding Groups By Distance From A Feature
Having selected the intended feature to group about, grouping scan data via a distance from a given feature has two primary modes.
- Absolute: where data is groped given its overall distance from the selected feature. This effectively selects data radially about the string. This mode also allows for Capping distances from the ends of the feature to be identified as well. So the selection can be capped, or allowed to go out past the length of the feature by the distance specified.
- Directional: Where data is grouped given a perpendicular offset from the given feature. This allows data to be boxed either to one side or about the intended string. Offsets are required Left and Right as well as Up and Down.
The rest of the parameters are much the same as before. Both Source and Target groups need to be selected. Once the settings have been made and the preview agrees with expectations, press Apply to move the selected points to the new group.
Adding Groups By Colour
This option allows you to group any part of a scan by its colour. First select Group By Colour from the Point Cloud Tools section. After you have selected this tool its specific properties will appear to the far right-hand side of the n4ce window:
- Point Cloud: The point cloud to filter.
- Original Group: Allows individual groups or simply all the groups to be available as sources of data for regrouping.
- Target Group: Sets the group to be assigned to all the points within the search parameters.
- Pick Colour: This option allows you to select the specific colour you would like to isolate.
- Shade Cloud: Shows a preview of what colour you'd be selecting when hovering over the point cloud with pick colour enabled.
- Hide Other Colours: When picking the colour it will hide the colours which arent being selected.
- Apply: Groups the specific isolated colour from the point cloud.
Before
After with the trees being selected
Adding Groups By Intensity
This option allows you to group any part of a scan by its intensity. First select Group By intensity from the Point Cloud Tools section. After you have selected this tool its specific properties will appear to the far right-hand side of the n4ce window:
- Point Cloud: The point cloud to filter.
- Original Group: Allows individual groups or simply all the groups to be available as sources of data for regrouping.
- Target Group: Sets the group to be assigned to all the points within the search parameters.
- Pick Intensity: This option allows you to select the specific intensity you would like to isolate.
- Reset Intensity: If the intensity you select is incorrect select reset intensity and reselect.
- Shade Cloud: Shows a preview of what Intensity you'd be selecting when hovering over the point cloud with pick intensity enabled.
- Hide Other Intensities: When picking the Intensity it will hide the Intensities which arent being selected.
- Apply: Groups the specific isolated intensities from the point cloud.
Before when no specific intensity has been selected
After (Trees for the intensity)
Comments
0 comments
Please sign in to leave a comment.