Improving Collection and Data with Carlson Precision 3D

By Steve Cummings • June 27th, 2019
tags:    •    •    •    •    •    •    •    •    •    •    •    •  

An idea possibly foreign to civil engineers and surveyors, who are correct by definition when they place their stamps, comes from management science, specifically W. Edwards Deming. He invented the concept of continuous improvement and applied it to manufacturing systems (and everything else).

Those of you who are old enough will remember when “Made in Japan” indicated a shoddy product of poor materials guaranteed to break. Younger people cannot imagine a Japanese product that is not of the highest quality, lasting for many years. Deming was the guy who made this happen. TQM and Six-Sigma management systems now used widely in the U.S. come directly from his work.

In a nutshell – Measure and Improve.

Over and over. Forever.

For us in the industry, though, until recently there was nothing to improve. Whether we collected data with a chain and a theodolite or a BRx6+ GNSS Receiver, it’s essentially the same data used to produce the same deliverable, certified as correct within our reported error. Equipment manufacturers, including Carlson, are continuously improving products for accuracy and speed of capture but the data captured is the same.

However, now we have new sorts of data collected by satellites, planes, drones, and scanners using cameras and lasers and sonar, some or all of it supplied by others. Is it accurate? Who knows? What would Deming do? Measure and Improve!

We flew downtown Maysville without ground control points. We should find the horizontal locations pretty close because the drone’s GPS is plus or minus a meter or two and it pluses and minuses to very close using lots of images with lots of pixels for best-fit statistics. Elevation is a different story. The drone uses one barometric pressure sensor to determine elevation. Thus, the vertical location of your drone cloud is plus or minus a whole lot.

So, we shot some survey data (measure). Now we need to improve.

Carlson Precision 3D 2019 allows users to import survey data, points, polylines, surfaces, point clouds, both traditional LIDAR and aerial drone survey data, and more from a wide variety of programs and entities to create usable 3D surfaces. It can be used to vertically shift your cloud to a more accurate position using the Correct Elevation with Survey Data command. In the first image below, I selected just one survey point and applied the new filter. The cloud was raised 1.2 feet.

Using the same filter, I select a lot of survey points and run the command.

I get a slightly different answer. What we’ve done is to add more points (measure) to an RMS best fit analysis. This analysis appears in the Output Window, listing the delta between the survey point and the closest point in the cloud before the shift, then it calculates the average delta and the standard deviation. The shift is made and the individual deltas, the average delta, and the standard deviation is listed.

Shifting the cloud vertically improves the average accuracy of a cloud with respect to survey data.

You might not want to shift it. Just see what the average error is or study where exactly the largest errors are. CTRL-Z returns the cloud to its original position.

We next apply our concept of continuous improvement by merging surveyed surface areas with the aerial survey. Here we have a surface created from an aerial state lidar tile as well as some survey grid points and polylines.

We select the closed polyline representing the top-of-bank and the open polyline representing the bottom-of-bank then select the Add Breakline command. The command first processes the closed polyline—in this case removing all the interior points—then it processes the interior polylines. The interior and exterior point spacing adds points to polylines without changing any slopes or vertices. I selected 1.25’ spacing to closely match the point cloud density in order to get the optimum blend between the surveyed ditch and the surface created from the lidar.

The result, shown below, is the lidar surface improved with the survey data. Note that I did not Smooth or Correct Elevation with Survey Data so you could see the error of the state lidar as small black vertical planes along the top-of-bank polyline. This command is also perfect for setting the edge-of-water boundary—eliminating any noise from the water area of the photogrammetry.

For more on W. Edwards Deming, I recommend “The Essential Deming – Leadership Principles from the Father of Quality,” edited by Orsini.

For improving the quality of your own productivity, I recommend getting more sleep, walking more, and Carlson Software!

Click here for more about Carlson Precision 3D 2019. Questions? Comments? Demos? or 606-564-5028

– Steve

Fill out this form and give me a topic that you’d like to see or read!

« Previous | Home | Next »