Content for Meta Index Content not found!
Laser Scanning a Vessel:  What's the Point (Cloud)?
By
Brandon A. Mattox, Brian N. Brookshire Jr., Abby E. Parish and Allen Burks, Applied Science Group, NCS SubSea
Following the photogrammetry experiment we conducted in our March Gist, the Applied Science Group had the opportunity to test a Faro X330 laser scanner courtesy of Andy Holroyd at HTS Advanced Solutions. After a short tutorial in the HTS lab, we were ready to scan. At the NCS office we laid out our “scan plan” and went to work testing out some very high resolution scans on a GPS buoy commonly used by NCS on the P-Cable tripoint (Figure 1).


Figure 1. Brian and Allen getting ready to acquire the first scan of a tri-point buoy.

The plan included taking five scans around the buoy at approximately 40 minutes per scan.  The scans were taken at the corner points of an imaginary rectangular box encompassing the buoy, with an extra scan taken directly in front of the buoy.  The scans for two of the diagonally opposing corners were taken with the scanner in a relatively high position (above the top of the buoy), and the scans for the other two diagonally opposing corners were taken with the scanner in a relatively low position (below the top of the buoy). Target spheres were placed around the buoy to ensure excellent scan overlap and ease in processing later. There are two significant settings when scanning with the X330, the resolution and the quality. Increasing the resolution adds more measurements to the survey; increasing quality slows down the measurements and allows for greater distances to be captured as well as decreasing noise captured in the point cloud.

After capturing the scans we registered, or combined, all five scans into an extremely dense point cloud using the FARO Scene software and Autodesk ReCap. Both resulted in excellent point clouds (Figure 2).


Figure 2. Viewing the registered point cloud in Autodesk ReCap.

With a small sample project under our belt, we looked back towards our ultimate goal of scanning a vessel at dock, something we see as adding a lot of value as a marine survey company. We had been warned at HTS that scanning anything moving, or scanning on top of something moving, would add layers of complication during the scan registration process. We were offered extra target spheres to help in the registration process (Figure 3).


Figure 3. A set of FARO spherical targets.

With warnings in mind we decided to scan a vessel not from the dock, but on top of a vessel in various locations. This ensured the vessel would remain motionless during each scan, but the surroundings, such as the dock, shoreline, etc., would move and become noisy. We traveled down to Freeport, to the R/V Brooks McCall. Our main goal was to survey the antenna array, so we came up with our scan plan and placed the target spheres around the vessel. We planned to do eight scans around the vessel at approximately 11 minutes per scan (Figure 4).


Figure 4. Preparing to take a scan on the R/V Brooks McCall.


Figure 5. Measuring offsets using Autodesk ReCap.

In order to successfully register the eight scans from onboard the vessel, we had to carefully consider the parameterization of the registration algorithm to account for the complexity and motion of the vessel.  With a little trial and error, we were able to produce an excellent combined point cloud.    Our primary goal with the vessel scanning test was to confirm that laser sans can be used as a safe and accurate method of measuring antenna mast offsets without requiring tedious, manual (tape) measurements while working at height, or sparse total station shots (Figure 5).  Our secondary goal was to confirm that laser scanning for overall vessel offsets and layout planning was feasible given such a complex environment (Figures 6 and 7).  On both accounts, we did confirm that laser scanning is a promising alternative to the tape measure and total station.


Figure 6. Placing objects on the vessel is a snap. Will your new winch fit on the deck, will an engine part fit through the hatch?


Figure 7. A greyscale image of the registered scans.