Will We Always Need Model Extraction from LiDAR Data?

Last week I attended the Florida Survey and Mapping Society conference at the Marriott Sawgrass, Jacksonville. During an excellent dinner with friends who are also TopoDOT users, the question was asked, Ted, what will you do when LiDAR data just passes into all CAD programs and extracting models isnt necessary?

Its a good question, although I dont entirely accept the premise. With two kids in college and retirement around the corner, its definitely worth careful consideration as I admittedly have quite a bit riding on the answer. So I just thought Id share my thoughts with you.

I can start by stating my agreement with the point clouds rolling into CAD premise of the question. TopoDOT was designed and has been very successful at achieving just that objective. While the majority of TopoDOT customers are still at the data production level, i.e. companies with airborne, mobile and/or static LiDAR systems, an increasing number of customers are entirely on the consumption end of data.

These are design, engineering, construction departments receiving data from either inside or outside their companies. So TopoDOT has been very successful at pulling this data from the survey/mapping operations directly into the CAD environments. This phenomenon was not only foreseen, but an objective of the overall TopoDOT strategy.

The larger question regarding the necessity of extraction is actually more difficult to answer. In short, I foresee levels of model extraction consistently increasing and decreasing in unforeseen ways. One cant always have it both ways, but I think Im standing on firm ground. Allow me to explain.

Lets start with the decrease in requirements for model extraction. Lets consider the relatively simple problem of power line clearances. A construction company might ask for a survey company to extract the vectors modeling the power lines from the LiDAR data. They will need this information to measure clearances for various field operations. Since they dont really know exactly where and when theyll need this information, they request all power lines be modeled. This modeling is a rather time consuming and expensive process.

Contrast this over-modeling scenario with the approach of engineering companies who pull LiDAR data into their CAD environment. In about 1 minute, the engineering or construction department can find the LiDAR data needed, extract a cross section and measure the clearances directly without modeling as shown in the TopoDOT screenshot. This approach totally avoids the time and effort associated with power line modeling. This allows LiDAR data to flow downstream to decision-makers in design, engineering, and construction clearly reducing cost and increasing productivity.

So in this example, my friend was correct. Model extraction is cut to a minimum as is the corresponding TopoDOT usage. However Im not all that concerned. For every instance of decreasing model extraction, I could come up with a counter-example for increasing model extraction and general measurement activities.

Consider these next two examples. The first screenshot is a typical topography model extracted in TopoDOT. Note the information contained in just this small section of data. Break lines are extracted and identified by name, level, feature, etc. Cells representing different types of roadside furniture are extracted, clearly marked and perhaps saved in a database. Extracted elevation coordinates and break lines comprise the topography. This small example is rich in information describing this street corner and does so with just a minimum of data. So while the point cloud and calibrated image contained this information, the downstream operations need the intelligence associated with the model represented as efficiently as possible.

A second example demonstrates the clear need for extracting complex measurements from LiDAR data in an easily interpreted way. A recent project in Sanford, Florida shows an historic schoolhouse. Cross sections of point clouds reveal a bulge in the side of the wall. However automatic measurement tools in TopoDOT sample the wall, extract measurements and color-code them based on distance from vertical. Cross-sections of data start to tell the story, but the color-coded CAD elements representing distance quickly provides a more comprehensive insight into what is actually happening.

In conclusion, I agree that expensive and time-consuming over-modeling will be avoided and productivity greatly enhanced as downstream operations increase their direct use of LiDAR data for measurements, profile assessment, etc. This will decrease much of the up-front modeling activities significantly.

On the other hand, LiDAR data is rich in information. There will always be a need to extract models with intelligence, interpret data in new and complex ways, all across an increasing number of applications and scenarios. So Ill stand by my statement and well continue to do our best to give our customers the tools they need to exploit this information rich data.

About the Author

Ted Knaak

Ted Knaak started his professional career as a control systems analyst for General Electric Space Systems Division through the eighties. In 1990, he moved to Florida where he was exposed to lidar technology at Schwartz Electro-Optics serving as marketing/sales and program manager. Mr. Knaak founded Riegl USA in 1993 and has since been actively involved with lidar and laser scanning technology. In 2011 he founded Certainty 3D, a company focused on data processing software and technology solutions for the laser scanning industry. Certainty 3D’s primary product is TopoDOT™ a point cloud/calibrated image processing program for extracting CAD models from LiDAR data. Mr. Knaak holds MSEE and MSME degrees from Brown and Drexel universities.
Contact Ted Article List Below