Chesapeake Times, Vol 9 | April 2022
While working with a customer recently, we had to solve an issue that sidescan processors deal with on a regular basis – My sidescan data has navigation issues and I need to use smoothing to clean up the image, but I need to maintain the positional accuracy. What do I do? In this example project after navigation and heading smoothing, my individual sidescan files look clean but as they overlap in the plan view, the same boulders are offset by +/- 15m which really mucks up GeoTIFF exports.
Since I know I am not in Death Valley and these boulders aren’t moving with each pass, I can assume that none of their positions are the true location and that is where map corrections come in to play. The tool allows you to take marked locations of the same contact and correct them either based on an assumed average or a “ground truth”, if it is available for this survey.
Our first step is locating all the instances where this boulder appears in the Map Corrections tools. You can see that without any “ground truth” to the actual position of the boulder, the tool will make an averaged assumption of where it should go, which is a good improvement, but not ideal.
Because we do have that “ground truth” in the form of a bathymetry grid, we know the true position of this contact, which changes our assumed position considerably.
Once the points are identified, the tool uses a weighted correction on the sidescan files that contain points and, after applied, you can see that not only the contact but other places in our mosaic start to align better.
In an ideal situation, I have now solved the problem that I faced earlier – I was able to smooth my navigation to get a much cleaner image and then improve the accuracy for the contacts! Now because this correction is weighted around the points placed, we recommend creating corrections for as many locations as you can in order to clean up the entirety of the mosaic.
The Map Corrections tool also has the ability to correct Magnetometer Files which is very useful in this example project where an identified anomaly in my magnetometer is due to a shipwreck identified in my sidescan.
