Tying Data Together

The HDMS system (Black Box – top of image) mounted under helicopter collecting in Chicago

The HDMS system (Black Box – top of image) mounted under helicopter collecting in Chicago

The LiDAR profession continues to evolve and with evolution comes new solutions, innovation and analytics which causes the provider of data to evaluate the best solution for the application. Merrick & Company uses most every remote sensing technology to provide solutions to most mapping applications.

One of the more challenging aspects of combining technologies is making all the data fit well together. Currently, Merrick & Company is working on several programs that require mobile LiDAR, airborne LiDAR and some sort of remote sensed imagery whether that be hyperspectral, multispectral, thermal, oblique or traditional digital orthophotography. Typically, some level of survey is required for these projects and then all the information needs to be “tied” together to provide the clients the accuracies they require.

In the past year or so the demand for these types of projects has increased significantly. To stay relevant and generate the kinds of revenue the company expects we need to embrace these types of demanding projects. Surprising enough the ability to tie mobile LiDAR to airborne LiDAR can be surprisingly difficult, even with an extensive amount of survey ground control.

The specifications for ground control for must mobile Jobs follow Caltrans or TxDOT specifications and typically require parallel points along a corridor. The spacing of these parallel sets of points varies from 500-foot spacing to 2500-foot spacing depending on accuracy. Most professionals believe these specifications are dated and based on previous mobile technology. Currently, to get 2 to 3 cm vertical accuracy the requirement would appear to be 1500-foot spacing.

Merrick & Company’s High Definition Mapping System (HDMS) which is mounted on a helicopter can achieve 2 to 3 centimeter vertical accuracy with 20 survey points along a 100-mile corridor. This same type of mapping project with mobile would require approximately 400 to 600 survey points. So that being said the use of the HDMS system requires much less control.

What if the mobile was then tied to the airborne LiDAR? This solution is currently being used on several Merrick & Company projects. Typically, this type of project generally requires some form of airborne – sensed data in most cases.

The rational for this method is that mobile LiDAR requires extensive control. There is cost savings to be had as it relates to survey control. The HDMS system provides the necessary control and point density to identify control locations for the mobile data.

Additionally, the trajectory information provided by mobile collection is less accurate than that of an airborne system. Granted, the mobile sensor uses velocity information from the Distance Measurement Instrument (DMI) but there is still GPS cycle slips, loss of lock and elevation masking that causes the GPS be less accurate then the airborne GPS solution because these factors are not an issue for the plane.

The airborne system provides a continuous POS solution over the entire mission and the collection rate of the airborne sensor is significantly more efficient than that of the mobile. The combination of mobile and airborne LiDAR is a very powerful tool for many high accuracy and engineering grade applications including but not limited to transportation, highway safety, autonomous vehicle, airport, infrastructure, planimetric and large facility mapping.

Colorized LiDAR points for a Railroad corridor project that included Mobile LiDAR, Airborne LiDAR, Oblique Imagery and Orthophotography

Colorized LiDAR points for a Railroad corridor project that included Mobile LiDAR, Airborne LiDAR, Oblique Imagery and Orthophotography

Typically, most of these mobile and airborne projects require additional remotely sensed data such as oblique imagery, thermal, hyperspectral, multispectral and medium or large format traditional digital imagery. All these data sources are typically very high resolution and in the neighborhood of 2” pixel resolutions or smaller. These data can be collected on the same platform such as the Merrick & Company’s HDMS or other platforms which will further validate the accuracy of all data sources.

What this means is that there is inherent error in all data sources and an indication of accuracy of one form of data is further validated based on the repeatable accuracy and reference of all the data sources to each other. This is a good indication of the relative accuracy of all the data if it matches up well together and then the absolute accuracy is validated by the reference to the survey data.

This may seem obvious to trained remote sensing professionals, but we find that clients continually question the accuracy of their data because they want to make sure that they are getting what they paid for.

It is obvious there are several ways to get the job done.  The stated accuracies of different technologies don’t always lend themselves to a particular application as well as we would like them to or expect them to. Having an in-depth understanding of the technology and how it performs in different application environments is the key to insuring that a data provider delivers as required.



James (Jamie) Wilder Young CP, CMS-L, GISP is currently a Senior Geomatics Technologist for Merrick & Co. located in Greenwood Village, Colorado. His experience includes all aspects of LiDAR including sensor development, applications development, data acquisition, data processing and project management.

One comment

  • Hi Jamie,

    I have few questions regarding lidar data processing:

    1. I’ve tried 2 types of strips matching: tie line and surface matching. I like to use tie line matching due to strips, time constraints, size and etc. The results also good. However I saw in few articles surface matching is better than tie line matching. However it takes a lot of time. What is your opinion regarding these methods?

    2. Intensity image can be produced in terramodeler. However, if data are captured in different day, then the color of intensity image would be different. How we can reduce the imbalance color of intensity image?

    3. Is there any other way to improve building vectorization besides classification?



Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.