What's New in Terrasolid Feb 2019

Summary: February each year brings about the changeover period for the Terrasolid software. Since Terrasolid does incremental releases of new features over the course of the year, the first quarter makes for a good time to review the new features and changes that have been made in the software over the past year. These new features are available to any users with a current maintenance on the Terrasolid software or to those looking to purchase Terrasolid products to take advantage of their LIDAR data or produce better datasets. For demos or purchases, please contact sales@geocue.com. 

Terrasolid continues to add capabilities to the products in support of various terrestrial systems including both tripod and backpack mounted systems. Some of these were minor modifications to allow the reading of Leica PTS file format, or the addition of a Cut Long Range utility to remove lower accuracy data where there is redundancy. More involved were the extension of the translate, rotate and rubbersheet tools to allow automatic identification of various control markers types found in imagery derived and laser point clouds. The comparison of the current data locations and the known point locations can then be used by TerraScan to Fit Using Targets and derive the necessary transformations to obtain a georeferenced dataset. 

Matterport Point Cloud Georeferenced to Square Target Control by TerraScan
ZEB REVO SLAM solution georeferenced to ball target control by TerraScan

For mobile and terrestrial laser data, new tools have been added for working with data collected in tunnels: A new Classify Tunnel Surfaces routine for automatically classify the floor, roof and walls; And a new  Vectorize Tunnel Sections for automatically extracting line string elements which are fitted to the cross section provided by the point cloud. 

 

 

Tunnel points automatically classified as floor (orange), walls (cyan) and roof (red) (Left Image) and Tunnel Section Vector Elements (Right Image)

Several display improvements were made to enhance the ability to view various point clouds. In Display Mode, the border setting replaces the “User Depth” setting. Both add black borders to separate the foreground from the background. With the old system one had to often change the setting to get a nice display between datasets or when rotating a dataset. With the new method you don’t have to change the value as often. 10% is pretty good for most datasets from most viewing angles. In addition, a “Use normal” option changes the brightness of points based on the normal direction. Planar surfaces facing the viewer are brighter and the background, such as the inside of walls or the side of an object not scanned, are displayed darkest. This gives the viewer a better understanding of the shapes. It works with coloring modes that start with a list of colors, such as class, dimension, echo, line, etc. but affects only planar surfaces.

 

ZEB REVO collected dataset of office exterior with new borders and "Use normal" options

The use of groups was long ago introduced to start looking at features in a dataset on an object basis rather than a point by point basis. This has expanded the ability to do above ground classification and from that, data extraction where we can pull values from the dataset on a group or object basis for reporting. Examples of such reporting uses would be for vegetation management, development of tree databases, extracting parameters, such as height or centroid, from buildings or other features, etc. To help make the grouping, classification by groups and reporting from groups more accurate it has then become necessary to add a new set of tools to the toolbox called Displayset tools. The Displayset is used in Display Mode to filter which points get displayed in a view. 

 

Points displayed by groups showing individual trees, all points displayed
Displayset of multiple groups
Displayset of only one group

Coupled with the existing Groups tools, Create Point Group and Merge Point Groups, plus new additions, Add Points to Group and Clear Group Fence, the Displayset can facilitate cleaning up the groups and classification. And, when combined with the new Inspect Groups tool one has a structured, semi-automated workflow that easily drives MicroStation, TerraScan and TerraPhoto, and as additional quick action buttons to ensure each object contains the desired points for a properly classified dataset and assure good reporting using the point cloud.

Inspect Groups Managed Workflow Tool

As shown in this presentation at InterGeo in 2018, mobile and high density airborne or UAS LIDAR can be used to improve railroad maintenance and safety. Similarly, LIDAR could be used for road assessments. To help with the assessment and reporting a new Trunk method has been added to the Assign Group tool. This new method can correctly group leaning trees and small trees under tall trees. Furthermore, a trunk method has been added to the Place Tree Cells tool, which results in better metrics when outputting the report. 

Trees grouped using trunk method
Tree cells placed using trunk method

In addition to all the new features during the past year, there have been several speed improvements in the products. These include better load balancing for multiple threads in many computation routines, including Assign Groups, Compute Normal Vectors, Classify Surface Points, and Smoothen XYZ. In addition, a lot of work was done to speed up the point display in view windows and the reading of many text file formats.

The biggest new feature in TerraPhoto was the introduction of Feature Points to the Tie Point workflow. The Feature Points workflow doesn’t require a ground classification, although it may make the user’s work easier. The feature point workflow makes use of the above ground features in the LIDAR data, can find points on building roofs and in vegetation instead of on the ground only, plus, it has systematic execution time. Similar to photogrammetric packages, TerraPhoto now finds interesting spots in images and computes numbers that describe the contrast changes at that location. Numbers describing the location can then be compared between two images to see how similar the location is in the images without needing to directly compare the images. The feature points try to be independent of the direction angle. They rotate everything into the strongest brightness change direction and tries to be independent of the scale and resolution that is looking at that spot. Hence, can have images taken closer and further away at same spot.

Feature Points compared in two images.

One small change to TerraPhoto was the addition of the ability to write a correction curve as a TerraMatch solution file. This has opened a big opportunity to now use photo identifiable features in an automated way to correct for positional and orientation offsets in the trajectory solution of airborne UAS datasets when there are not enough identifiable features in the laser data alone, or if they are time consuming to manually collect. We started doing a similar concept with mobile datasets a few years back and can now use a combination of laser features and photo identifiable features to obtain a better calibration of UAS projects to account for the lower quality IMU and GNSS found on these systems. This is applicable if the collection of the images and the point cloud were done during the same flight and if the initial positioning of the images and the point cloud is based on the same GNSS/IMU solution. In the future, we expect the software will get smarter and be able to decide which data is best suited for each correction.

TerraStereo continues to be enhanced to make the collection of features from laser data in a stereo environment accurate and efficient. One can now use TerraStereo in MicroStation CONNECT Edition to leverage the CE tools for feature collection. In addition, a new set of cross section tools and perspective settings were added this past year. Plus, TerraSurvey is now supported in stereo.

Laser Data in Stereo with Automatic Section Views

As you can see, there have been a lot of new features for your maintenance dollars. And that’s without getting into the many feature enhancements that extend tool uses, and many bug fixes to improve the toolset. For more details see the what’s new in Terrasolid Feb 2019 post on our searchable support knowledge base. For additional information concerning any of these features in Terrasolid please contact the GeoCue Group Support Team at support@geocue.com.