Phantom Vision for Archaeology

[kevin_edw2002]

Hi Stefan,

Sorry - but was unable to respond to your PM directly (not allowed to as a new user, apparently!).
I generated the contour plots by using Photoscan (pro version) to export a DEM as a geotiff. I then imported this into QGIS as a raster layer and used the Raster > Extract > Contour function to generate the contours, which are saved as a separate .shp file. I haven't attempted to use ground controls so far, as my total station is currently out of action. I will be doing an evaluation sometime over the next couple of weeks using a combination of surveyed ground controls, photogrammetry and aerial imagery; I'll post the results when (and if!) I get them.

Best,

Kevin

 

[stefan_fvt]

Thanks for your reply Kevin. Below is the PM Kevin is answering if someone would like to follow along.

Hi,
I saw your post in the Archaeolgy thread and sending you this PM since I don’t want to hijack the thread. I’m trying to do pretty much what you are doing with your Phantom. I am experimenting with an additional camera for straight down shots to get a little bit higher resolution of the orthophotos from Photoscan.

I was wondering how you get the elevation curves in to your DEM as seen in the last picture in your post. This looks amazing and is something I would like to achieve! Is this done with Photoscan or do you use any additional software? Do you use any ground control points?

Hi Stefan,

Sorry - but was unable to respond to your PM directly (not allowed to as a new user, apparently!).
I generated the contour plots by using Photoscan (pro version) to export a DEM as a geotiff. I then imported this into QGIS as a raster layer and used the Raster > Extract > Contour function to generate the contours, which are saved as a separate .shp file. I haven't attempted to use ground controls so far, as my total station is currently out of action. I will be doing an evaluation sometime over the next couple of weeks using a combination of surveyed ground controls, photogrammetry and aerial imagery; I'll post the results when (and if!) I get them.

Best,

Kevin

 

[nasa.bastina.tg]

Hi,

I've also been experimenting with the Vision for use in archaeological applications. I have been getting usable results using the stock DJI camera and a spacer to enable me to get a near-perpendicular view of the ground surface. My current workflow involves making a series of aerial traverses across a site or area and capturing video footage at 1080p at 30fps and narrow FOV (90 degrees). Once the flight has been completed, I import the video into Photoshop and then export the relevant portions as an image sequence (typically at a rate of 1 frame per second, which provides plenty of redundancy). The image sequence is then imported into Photoscan for processing (alignment, point cloud generation, meshing and DEM).

I've attached a few images showing a couple 3D models of a rock outcrop and a large water tank, together with a digital elevation model/orthophoto/contour plot of the latter. As you can see, there is very little distortion and plenty of detail.
Best,

Kevin

Can you please explain getting DEM with Photoscaner. What do you need for ortorectifying images? Do you need locations of every frame (picture)? I am begineer in this.