Mobile LiDAR Point Density Part 2:

In the previous article, a discussion of LiDAR density was introduced with an overview of Aerial LiDAR and Static Scanning resolutions for general comparison. Part I of this article series laid the foundation for additional discussion in this article regarding Mobile LiDAR collection variables that influence resultant point densities. These include:

Vehicle speed, measurement frequency, and scanner rates;

Angle of incidence and the impact of flat surfaces;

Distance to target and effect of rotated sensors; and

Collection paths and moving obstructions.

When Baker acquired our Optech LYNX Mobile Mapper three years ago, there wasnt a program or formula for calculating the point density of a resultant cloud based upon specified system settings. I would have loved to have been provided something similar to: X miles per hour + Y laser frequency + Z scanner rate = Required Point Density, but as is the case with emerging technology, these items arent always available. As important as each of the collection settings are, client-mandated point densities are typically loosely defined, misunderstood, and now appearing in ever increasing Mobile LiDAR specifications more about this in a future article. But when you embrace an emerging technology, many lessons are self-taught and learned through rigorous testing which is precisely what we did.

Controlled Variables

Those influences on density which are most easily managed include the vehicle speed, laser frequency and scanner rate (number of lines). These are variables that can be defined during proper mission planning and applied by the operators in the field to achieve the desired results. Quite often, vehicle speed is dictated by localized field conditions such as speed limit, other vehicular traffic, traffic control devices or deviations due to real-time positioning metrics. However, experience aids in developing an educated guess as to the average speed needed for a collection, which further assists in setting the other criteria.

The laser frequency is the number of pulses/measurements per second. Todays mobile sensors are capable of capturing 50,000 to 500,000 measurements per second, or more, per sensor. Although the collection rates are staggering, it doesnt necessarily produce better data, just more of it. The projects requirements, features to extract, or the grid density of the desired Digital Terrain Model, may all permit collection at a lower frequency to effectively produce the desired results, as opposed to capturing unnecessary larger datasets that will need to be decimated.

Most mobile sensors operate using a rotating mirror to achieve 360-degree coverage. Referred to as the scanner rate, this is another user-defined variable with typical values ranging from 80 to 200 revolutions per second. Using the anticipated vehicle speed and the scanner rate, you can readily calculate the line spacing produced by a single sensor.

In order to better understand the relationships between each of these three variables and the resulting captured data, Baker preformed numerous controlled tests to compile and analyze results. Multiple collections were performed where the speed limit was adjusted by 5 mph per scanning pass; the laser frequency was varied between 75kHz, 100kHz, and 200kHz; and the scanner frequency was set at 80Hz, 120Hz, 160Hz, and 200Hz. The results were reviewed to identify the variations each combination had on point density, feature identification, corridor extent, and file sizes; enabling us to establish a defensible foundation and methodology that is applied consistently on new projects.

The image represents 3 passes along the same stretch of a 2 lane residential road. The data was collected at 25 mph with a scanner rate of 120 Hz. The three images show results of collecting at laser frequencies of 75 kHz, 100 kHz and 200 kHz from left to right.

Angle of Incidence/Flat Surfaces

The premise of LiDAR is that an active light source measures the range to an object due to the reflection of the original signal, also known as backscatter. The angle at which the original signal strikes an object, as well as the material of the object, (visualize the angles and materials of stealth aircraft), affects the ability of the sensor to effectively capture information. On wider surfaces, down intersecting roadways, or across a highway median, the amount of data collected is diminished due to the angle at which the signal strikes the surface. In an effort to improve performance, many firms have resorted to raising the height of the sensor heads. It is understood that data captured further from the vehicle is less accurate for various reasons and is often filtered out, however the data can diminish substantially within a short distance from the vehicle due to the orientation of the lasers themselves again, another topic for future discussion.

The image shows how data diminishes on flat surfaces over just 100 from the sensor head. The data is from a single sensor traveling down the centerline of the runway. However the more points are returned on the pavement striping versus the concrete further showing how the material also influences returns.

Distance to Target

Using a scanning mirror, LiDAR sensors measure pulses about a rotational axis. Therefore, the laser frequency and scanner rate affect the angle between sequential pulses or measurements. By calculating the angle between pulses, we can also calculate spacing of points along the scan line at a given distance. Having an understanding of the density of the point cloud at various distances assists in determining which types of features will be discernible in the dataset.

Collection Paths & Moving Obstructions

In almost all instances, Mobile LiDAR surveying is restricted to a drivable surface (land or water). During a typical roadway collection, each travel lane, ramp, overpass and intersection are captured to develop a seamless dataset while minimizing obstructions. While you cannot guarantee that other vehicular traffic will not pose problems with respect to density or data voids, it is strongly recommended that multiple passes be performed. The utilization of a dual sensor system does yield an additional a measure of backup or redundancy, especially when you encounter curious motorists that feel it necessary to follow really close as they attempt to figure out what that strange looking contraption is mounted on the roof of your vehicle.

As many State DOTs are adopting Mobile LiDAR and developing specifications, there is ever increasing language regarding point spacing, data voids and densities; making it paramount to understand your capabilities, and establish best practices for collecting data to mitigate potential issues.

About the Author

Stephen Clancy

Stephen Clancy... Mr. Clancy is a Florida licensed Professional Surveyor and Mapper as well as a Certified GIS Professional with an extensive background in LiDAR, GPS and traditional surveying and mapping. He holds a Bachelor of Science degree in Geomatics from the University of Florida as well as three years of post-bachelorette coursework in the fields of Geomatics, Urban Planning, Geography and Geophysics. In addition to serving in various capacities in surveying and GIS-related activities, Mr. Clancy also has 5 years of university teaching experience in the fields of Geomatics, Photogrammetry and GIS. Mr. Clancy has a diverse and broad background in the Geospatial Sciences and most recently has been charged with the technical management of Baker’s Mobile LiDAR system.
Contact Stephen Article List Below