In late 2016 Terrasolid added Distance as a point attribute in their Fast Binary (FBI) format. This format is the one used for loaded points in TerraScan, so even when processing in LAS you have access to utilizing the attribute, you just need to compute and use it at the same time since it does not get stored in the LAS format. The addition of the distance attribute allowed for the computation and storage of a distance value for each point. By storing the distance value, computed once but used by more than one routine, workflows could be more efficient, particularly for datasets with larger point counts. Over time the distance attribute has contributed to new workflow automation for evaluating and extracting information from point clouds.
Display by Distance was initially used to evaluate the interline fit between overlapping flight lines in a collection to determine if it met specifications such as the ASPRS Positional Accuracy Standards for Digital Geospatial Data. When Compute Distance was introduced a user could now compute the Line average Z and store that value on each point. By storing the value, a user could now not only display the interline differences, but could generate dZ images without needing to recompute the values, and could Classify by Distance to identify areas with point differences that exceed specification.
Compute Distance from the Ground Class made the distance attribute represent the height from ground for each point. Thus, a user is now able to Display by Height Above Ground to readily assess values one may wish to use with other routines, or identify objects, such as buildings or trees over a certain height. In addition, by storing the computed distance value a workflow can become more efficient. For example, change the typical multiple Classify by Height macro steps for segmenting above ground data by height above ground to one Compute Distance and multiple Classify by Distance steps. Since Classify by Height requires computing the height above the TIN in each step versus the Classify by Distance, which just uses the existing attribute, the processing time is now reduced.
In a similar fashion, some processing commands rely on having a distance already computed from ground:
In addition to computing a distance from ground, the Compute Distance command allows for a variety of comparisons. For example, Range 3D allows for a visualization of the point cloud based on its 3D distance from the scanner. This allows identifying the values one may wish to use to remove noise points before running the classification routine. Computing distance to Elements or Wires allows for visualizing and classifying danger points along roads, pipelines, power lines, or rails. Coupled with the Group Classification tools one can identify objects, such as trees or buildings, rather than just points.
Other forms of analysis, such as change detection, can make use of the distance attribute by computing Closest line 3D for wall or tunnel comparisons for mobile or terrestrial laser scanning datasets, or Closest line dz for surface changes between temporal datasets. The ability to assess the differences is limited to how well the matching of the datasets is done.
Recently added to TerraScan is the ability to Compute Distance from the built-in vegetation indices, Normalized Difference and Visible Band Difference. The vegetation index can be used to help distinguish vegetation hits from ground points, especially in photogrammetric point clouds where the real ground is not visible. Smoothen the distance values to reduce noise and then use the distance as a rating in the updated Ground Classification routine to utilize the vegetation index stored as distance as a probability factor for how likely a point is to be ground. This workflow can improve the ground classification result for photogrammetric point clouds. The user can set the Weight higher when more confident the vegetation index separates the vegetated from non-vegetated points.
Another recent addition to TerraScan is the ability to Compute Distance using Road Bumps & Potholes method to streamline the analysis of a road surface from mobile or terrestrial laser scanning data. This new method updates the old method of computing distance to a “design surface” to find the potholes and other road surface deformations.
Thus, the distance attribute provides not only multiple visualization options, but assists with point cloud analysis, and supports efficient processing of point cloud information. When coupled with a wide array of existing classification and vectorization tools users have many options at their disposal. How one utilizes the attribute will change depending upon the workflow and desired results, but the flexibility and utility of the distance attribute has increased as more options have been added to the product. If you haven’t tried some of these new workflows check out our searchable support knowledge base for more information or contact us for available training options.