Activity Feed › Discussion Forums › Photogrammetry, LiDAR & UAS › Accuracy of Drone Surveying
Good points have been raised on the processing of the example I provided. In order to provide a clear example of accurate photogrammetry, here is a project that does not have GCP tied in at all. In other words, a drone was flown with RTK GPS then separately, control points were shot in for comparison in the photogrammetry project.
@outsideoffice I am continuously amazed at the horizontal accuracy of my DJI M300 with RTK with the L1 and P1 payloads. I dial into the MODOT network VRS and fly. I have a field crew get some check shots if they don’t already have visible control points on the site. The drone is always right on or maybe 0.2′ at worst. Vertical accuracy is always different depending on the height of my flight and what datum they used. I just move my surface up or down to match.
@jflamm Curious if you are using MoDOT VRS check shots to verify the MoDOT VRS controlled photogrammetry? Does your control and check shots come from the same dataset, i.e collected at the same time? Try burning in some static control, say two sets of 4 hour observations, then use a total station to collect check shots (twice as accurate as your dataset), that would be my way to verify accuracy, otherwise you may be “percisely” wrong. But then again relativity may be the only need for your workflow and not true accuracy.
@stlsurveyor I’ll completely agree with you on that but that’s overkill for what I’m doing. Usually it’s a preliminary surface topo of a site or grade check, basin topo in a subdivision that was established long before I started buzzing around. I’m at Sterling these days, they keep things pretty simple here.
- Posted by: @outsideoffice
here is a project that does not have GCP tied in at all
Good try making your point about control tie vs. check. Unfortunately for you it is invalid as your image clearly shows you are using this point for a 3D GCP, and NOT a Control Check.
Change the setting and try again.
Every time I see somebody in law enforcement using survey equipment my blood pressure spikes! They all get a crash course on how to run a TS, DC or other tools but have no clue what the numbers mean except that they see a number in front of their face and take it as gospel.
These people are using the equipment to depict crime and accident scenes and ultimately convict people based on evidence they documented. Should they not be experts that are able to testify as to every aspect of the quality and accuracy of the data they have collected and depicted?
Everything an officer says in court is by definition true and accurate. No need to confuse them by training. /snark
.
Log in to reply.