Old School, New Tech

By Doug Aaberg • September 27th, 2018
Share

There has been a lot of buzz and excitement about surveying with drones, scanners, and mobile LIDAR units over more than a decade and companies worldwide are scrambling to get on top of it. Manufacturers are spending a lot of time and resources in producing both hardware and software for what appears to be one of the fastest growing markets in the land surveying business.

Carlson itself has three different products that deal with drones and point cloud data. I talked about Carlson’s new PhotoCapture program in a previous blog. This one is focused on another product: Point Cloud.

With the release of Carlson 2019, a new slimmed down version of the product called Point Cloud Basic is available and geared perfectly for surveyors like myself—those who want or need to get into the high tech scene of managing data from a cloud but prefer to work in a familiar environment with points, break lines, and simple graphics for the production of a typical plan.

Here I am stepping you through a real project that I completed over this last year using some tools found in both versions of Point Cloud to demonstrate my work flow for producing a final plan. There’s no way I can cover all of the available tools contained in the program nor do I want to minimize someone else’s approach for completing the same tasks, but here’s my “Old School” approach to using some “New Technology.”

Overview

Essentially, whatever hardware and post processing equipment you use, at the end of it all, you will create a point cloud. This cloud contains data in the form of pixels and they usually look quite impressive.

The problem is: what will you or can you do with that kind of data? If your intent is to create a surface model and your resulting contours look like this……

……you are more than likely going to be disappointed.
I frequently field questions just regarding the fact that “contour lines from a drone are ugly.”
The other major problem is that a cloud will contain tens of thousands to millions of pixels, which, in terms of point data, is far too much information. A surface model from that much data will create a TIN file that is completely unmanageable. The above contours are not inaccurate, but most will find them cosmetically unappealing.

Results: It should be mentioned that to obtain accurate data, it is critical that quality instrumentation is used along with properly spaced ground control points, flight overlap, nadir, and oblique imagery, etc.

Point Cloud certainly has tools for thinning out a cloud, densifying contours, and simplifying a TIN. However, I am somewhat of an “old school” kind of land surveyor and tend to approach survey work through my own lens of understanding. I embrace technology and do whatever I can to make myself more efficient and I do not typically push against new products. I do, however; try to find a way to use them in a manner that makes me satisfied with the outcome and confident in my data. Accordingly, I really like Carlson’s Point Cloud program because it allows me to do just that.
The project
To demonstrate my process of producing a plan from cloud data, I must first explain the project itself. In this example, the task at hand is to produce an “As Built” plan of a newly constructed roadway in a residential area. The survey needs to extend to the rear of the constructed lots and needs to confirm roadway grades including top and bottom of curb elevations. The survey was performed via two different methods; a drone flight at approximately 150’ above ground and a mobile LIDAR unit mounted to a vehicle. I need to produce a surface model for preparing a plan and profile sheet as well as a 1’ contour topographic plan for the purpose of identifying conformance with the original design plans. One of the biggest obstacles to a typical on-the-ground survey for this site was access to the house lots which have now been occupied for several years.
My process is to use the drone data for the lot grades and the mobile LIDAR data for the detail within the right of way which requires more detailed information. I will say, as an aside, that in many similar types of surveys, I performed an on-the-ground survey within the right of way as opposed to using mobile LIDAR.

The Process

1.1.1 Creating and Editing the Cloud
One of the unique properties of Carlson Point Cloud and perhaps one of the reasons I really like it is the fact that it runs right within a DWG file. You open or create a new drawing using your typical Carlson setup routine and then launch Point Cloud. All of the work you perform in the cloud can be immediately ported into your drawing. There is no need for importing and exporting using different formats. It is exceptionally well suited to using dual monitors. You can have the cloud open on one monitor and your drawing on the other.

1. First import the cloud file itself. This can be done by right clicking the cloud and importing in several supported formats.
2. Select the drone cloud first.
Point Cloud recognizes any classifications that exist within the cloud.
At this point you can choose to not process different data contained within the cloud. I will keep all classification selected to be able to demonstrate additional tools.
3. Once imported, highlight the cloud and select View, or simply double click on the cloud.


The view creating dialog box appears with options on what category and type of view to create. Most common and useful is color.

A view window becomes active containing toolbars for viewing, editing and creating entities. of particular note is the ability to exaggerate the vertical scale making it easier to see detail such as a curb line.

Note: the following steps demonstrating the Bare Earth extraction are only available in the full Point Cloud program.

The original cloud shows houses, trees, cars, and other vertical objects that should not be included in a surface model, so the next step is to filter out such objects.
4. Right click on the cloud and select Bare Earth by Grid.
There are options to set the cell size which relates to the units being used, a pothole depth and curb height which sets values of acceptable vertical height changes.
In this case, the program will evaluate a 2-foot-wide grid.
5. Click

6. When completed, highlight the newly created cloud (by default the name suffixed with “Cleaned” and select View.

The program does a good job of deleting most all of the unwanted objects that would cause vertical spikes in a surface model.
By selecting each cloud, right clicking and selecting Properties, you can examine the differences in size and data before and after the Bare Earth cleaning.

If there are left over pieces, floating pixels or tree remains, they can be manually cleaned out as well.

By setting the vertical exaggeration to a larger scale, you can select entities using the perimeter method and delete or hide those pixels.

1.1.2 Creating a Grid
Keeping in mind that I am only going to use the drone data for the lot grades, I would like to completely remove any data within the right of way from this cloud. I drew a polyline for the right of way of the subject road in my drawing file using typical Carlson commands.


1. In the Project Menu, expand the Processed Data category, right click on Polylines and select Read from CAD
2. In the drawing editor, select the right of way polyline.
The polyline is added to the project data base.

Since the polyline created in the drawing file is at elevation 0 and the site elevations in the cloud range from approximately 160 to 240 US FT, the polyline will be difficult to see and/or make use of. It would therefore be prudent to elevate the polyline above the cloud.
3. Highlight the polyline, right click and select Transform

4. In the Transform Vertices dialog box, click the Plus + button.
5. Set the Offset Z value to 500.
This will ensure the polyline is well above the site and visible for use.
6. Click the Check Mark button.

7. Select the Polyline again, right click and select View

8. In the View Polyline dialog box, select Append to Existing and select the cleaned cloud.
9. Click the Check Mark button.
The polyline is now visible in the cloud viewing screen.

10. From the Action Tab, under Selection Set, select Polyline.
11. From the viewer, select the imported polyline.
The program highlights the selection set in red.
12. Select Delete to remove the selected pixels.

The result is a cleaned cloud without the roadway detail.


13. Once satisfied with the cleaned cloud, from the Project menu, highlight the cloud, right click and select Survey by Grid.

14. Set the Grid values as desired.
In this example I am choosing a 25’ grid and having the program pick the lowest elevation within a .25’ radius. The point code is SG which corresponds to a spot grade in my Field to Finish Code Table.

15. Click

The points are set in the cloud coordinate file.

16. From the Project menu under Processed Data, right click on Coordinate Points and select Field to Finish.
If you do not have a code table already set up, you can just draw the points and the program will use the current point defaults.

The result is a grid of Carlson Points that can easily be used to create a surface model from as well as be utilized in all of your normal work flow functions.
It is important to note that at this point, some projects would be complete. For example, an open pit, golf course or other such sites. However since a drone mostly captures data from above, it may not be suitable for capturing top and bottom of curb elevations. For sites that require adding breaklines or other detailed information, follow the below steps.

1.1.3 Adding Detail from LIDAR
As I mentioned earlier, I could now supplement the above data with an on-the-ground survey, but in this case, I am going to utilize the data from a mobile LIDAR unit.
1. Following the above procedure, import the LIDAR data and create a View.

In this instance, the mobile LIDAR unit produced a cloud with intensity and not color.
The full Carlson Point Cloud version has an option to recolor the cloud from an image file.

2. Highlight the LIDAR cloud, right click and select Adjust Color.
3. Select Recolor by Image and select the Check Mark button.

The program supports several file types.
4. Select the image file of the site from the drone capture, and click

Once recolored, you can see that the Lidar cloud contains very clear imagery of the curb line as well as utility structures and other features that are needed as part of the survey.

To define the curb lines, centerlines and other features, I am going to simply set Carlson points as if I were walking through the street performing an on-the-ground survey, a kind of “virtual survey.” Through this process, I can also pick up multiple check points as I go to use as elevation verification.

In this case, these are brand new curbs and are very consistent in height and width. This lends itself perfectly to use a template for collecting the curb line data. This is a feature contained within the Field to Finish code table.

For those of you who are not familiar with using a Carlson code table, there is much more material explaining the following process than can be shown in this blog. The following is a short overview of creating a code specific to this use.

5. From a Field to Finish Code Table (FLD file), create a unique code for a curb line

6. Set the entity type to a 3D polyline and select the Linetype Tab.


7. In the Linetype Tab, select Set and create a template.
8. Select Edit and enter the desired slopes, height and width for the curb line.
In this example, I created a simple vertical curb ½ foot wide with ½ foot reveal.

For more information on creating templates, see previous blog:
http://www.carlsonsw.com/archives/sticky-codes-fixed-parameters-and-templates/


9. Back in Point Cloud, in the Action Tab, select Create Point.
10. Select Code.

11. Set to the correct FLD file and select the code for the curb line.
Note the important options that are now active.

• The Snap Type contains options for controlling which pixels will be selected.
Cloud picks whatever pixel is selected while Low will select the lowest elevation value in the specified radius.
Low and high edge are for curb lines when you are not using a template. Etc.

• Starting Point Number
• Point description
• Start and end linework
• Close linework
• Start a curve

I then literately begin surveying as if I were in the field walking the street using codes that I am very familiar with.

There are also other features to create breaklines and other geometry such as drawing a polyline using the High/Low snap where it finds both the top and bottom of curb as you go:

I just happen to be very comfortable setting points that I know I can easily edit later on. It is well worth your while to review and experiment with all of the other tools contained in this program. This blog would become a book if I went through all of them.
12. As explained above, process the curb lines through Field to Finish.

The curb template will create a 3 dimensional polyline for all template ID points identified.
13. Similarly, set points and create break lines for additional detail such as centerlines, sidewalks, sewer structures, and other information as you would for any survey using codes from your existing FLD file.
14. Again, process the data through Field to Finish.

All point and break line data will be added to the drawing file.

At this point, I am ready to create a surface file using Carlson Triangulate and Contour and selecting all relevant data.

In the end, I end up with a surface model that can be easily dealt with. If I do find extraneous points that did not get removed from the cloud, I can edit the TIN, remove points, add points, supplement the data with on-the-ground survey information, etc.

Summation

I realize that this is not the end of it all. There is still a lot of detail that needs to be drafted that can be picked up from the cloud or from the image files which are amazingly clear coming from a drone. This also does NOT replace an on-the-ground survey nor does it limit the involvement of the surveyor in the mapping process. It is, however, another very useful tool in the tool chest.

What I have demonstrated is not magic by any means, but I have used it on many surveys with great success and with a cost benefit. If you haven’t had a chance to step into the world of surveying through the use of point clouds, then I would encourage you to download a free 30-day trial of Carlson’s Point Cloud and give it a try. I have found that the use of drones has become one of those technologies that once I got used to having it; I struggle to do without it.

Download Doug's Field-to-Finish Guide

Carlson’s 2019 User Conference in Maysville, KY
Consider coming for some classes next spring

Doug
Let’s Grow Together.
Douglas L. Aaberg, PLS
Survey Product Manager
P)617-393-2300×419
daaberg@carlsonsw.com

« Previous | Home | Next »