@jaja6009 - Thank you for sharing. I'm most interested in the subject of your article, particularly given this juncture of Parrot's decision to stop supporting its Anafi, and your own evaluation of its effectiveness for mapping. Would it be okay to get a copy of it?
Given that the Anafi is no longer being promoted by Parrot, can you tell me if the Anafi still receives firmware updates, and if so, is there a link to the firmware's history?
The second question for you, knowing that the Anafi has been discontinued, as well as all of its various issues chronicled on this and other forums, would you still have your article's target audience of public safety entities buy one?
If you could not buy a new Anafi, which sUAV would you buy for mapping?
I'm betting you're already keenly aware of this, but GCPs are necessary for many reasons, and your article's target audience will want to do so in order to utilize the vast collections of spatial datasets tied to the NSRS. I've written about this subject over several years, the most recent was from last year. The PDF file was too large to attach to this message. If you're interested, my article on the
The Vetting of Phodar 20210425.pdf is posted here:
drive.google.com
When I am done the paper I will give you a copy. I am still processing all the data.
Criteria
Drones: Anafi Thermal, Mavic 2 Enterprise, Mavic 2 Enterprise Advanced, Phantom 4 Pro with PPK kit
Map: Out of Box with only checkpoints, Map with GCPs and Check Points - Processed on Agisoft Metashape
Then Out of Box Map with just Check Points, and then Map with GCPs and Check Points processed on Drone Deploy.
I must walk back the elevation statement on the Anafi. It too had very bad elevations and this was my fault for not catching it. The relative elevations are good though.
My first dataset was on Pix4D, but I have since moved over to Agisoft Metashape and am rerunning the data.
The Anafi generated more keypoints and matched keypoints than both the Mavic 2 Enterprise and Enterprise Advanced, but less than the Phantom 4 Pro V2 with PPK. The Anafi has the smallest CMOS sensor of all the drones so this was a little bit surprising, but you have to factor in that the Anafi probably compresses the JPEG better from the DNG file than DJI's cameras. And this was with the 16mp images as the Anafi can no longer use the 21mp images while using Pix4D Capture and this is a shame.
The Anafi under Pix4D Capture estimated then flew the correct height to get the desired 1.5 cm/pixel GSD as the Mavic twins were both off big time.
Part of the paper assumes that most people will just let Pix4D or Metashape use the default settings and will not change the elevation expected for the images. Of the drones tested, the Phantom 4 Pro V2 and the Enterprise Advanced actually report their elevation as above the ellipsoid and not relative to the takeoff point. But if you do not tell the software this you get wildly inaccurate elevations. The Anafi is no different as it is reporting its elevation as 44 meters over the takeoff point but the default Pix4D setting is height over the EGM96 geoid.
If you want a solid, tested mapping drone you cannot go wrong with the Phantom 4 RTK. Its proven and can do both RTK and PPK and their new single oblique flight line at the end of a flight makes it so you most of the time do not need GCPs, just Check Points.
I just bought a Phantom 4 RTK and will test it soon. I was using a custom PPK kit on a Phantom 4 Pro V2, but the added weight really took away flight time, but the data was very good.
Thanks for the link to the article.
If you don't mind I'd love to ask you a question or two in PM.