Quality Control of Derived LiDAR Data

  1. The September issue of PE & RS has an important paper on the work being done by Dr. Habib and others.
  2. A mathematical methodology is being proposed to assess the quality of derived lidar data and the associated system parameters.
  3. The goal of this important research is to produce a set of standards for judging the quality of the derived lidar data.

septpers

The current issue of PE & RS has a very important paper by Dr. Habib and others entitled, “Error Budget of LiDAR Systems and Quality Control of the Derived Data”. The authors point out that although LiDAR systems have been widely adopted for a number of applications, there is a lack of standard methodologies for the QA of the systems and QC of the derived data. In most cases the preferred QC method being used today is to compare the derived positions with ground control points. Not only is this expensive, but it is not accurate enough to verify the horizontal accuracy.

The paper includes a discussion of the basic lidar mathematical model and potential error sources. From there a tool for evaluating the quality of the derived data, as well as the system parameters is discussed. Essentially the methodology is based on developing a mathematical model for comparing elements in overlapping strips. The basic concept being that if all the system parameters were properly calibrated and aligned that the elements would be in exactly the same position. In the real world this is not the case, but the model can be applied to identify the quality of the derived data.

The research is ongoing, but initial results are quite promising. The goal of the work is to develop a set of standards and specifications that can be used to actually accept or reject the deliverables. I think this is really important and valuable work that could lead to an important contribution to the growth of the LiDAR industry.

Congratulations to all those involved.

This entry was posted in Education, Quality, Research, Standards, Technology and tagged , , , , , . Bookmark the permalink.

2 Responses to Quality Control of Derived LiDAR Data

  1. Harold Rempel says:

    Checking the horizontal accuracy of LiDAR has been a heated topic for a while. I would have to agree that using discreet ground survey points is not a good approach for checking the horizontal. Aside from the obvious cost of this approach, additional error is present from the mere fact that the exact location of the pulse reflection could be ~ +/- 1/2 of the diameter of the pulse on the ground. You can also add in additional error if the survey point is (improperly placed)near or on a slope, break in terrain, grass, highly-reflective surface, etc. Lidar data simply cannot be checked in the way that we check orthophotography. On a couple of projects that I worked on for a German client, they established control “patches” of 100 discrete points each to check the quality of the lidar calibration. These surveyed patches had to be located on flat, open terrain (usually on a soccer pitch) and in the overlap area between flight lines. In essence, this gave them a grid of survey points which (in 3D) represented a plane against which they could compare the lidar. Some horizontal acuracy could be derived from this but it was mostly vertical accuracy and calibration they were concerned about.

    If the project area contains significant horizontal features that show up well in the lidar and the lidar is dense enough, linework collected using the lidar could be compared against an orthophoto for placement. Of course, one would have to then factor in the stated H accuracy of the orthophoto. I tried this in an area that had well-defined paint stripes on an airfield that could be mapped off of the lidar and the results were promising. I am looking forward to finally seeing some standards adopted here; hopefully in the near future.

Leave a Reply

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


*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>