0

I'm looking for a way to get a quick overview of aerial images taken together with exterior orientation data. Available are horizontal and vertical coordinates of the camera, focal length and also the orientation of the camera at the instant of the exposure. Orientation can be as Roll/Pitch/Yaw or Omega/Phi/Kappa. Also we have available a rough dtm for the mission region.

What I would like is a process to drape the images in correct scale and perspective onto the DTM surface which should result in a somewhat accurate "Quick View Orthophoto". Main purpose are quality checks for example to check for complete coverage of a mission area, sufficient image overlaps, clouds or cloud shadows and so on.

Can anybody point me to a toolset able to do such things?

I'm thinking of maybe using Python in QGIS to process images and metadata into such an project overview.

PolyGeo
  • 65,136
  • 29
  • 109
  • 338
Pete Bo
  • 139
  • 1
  • 5

1 Answers1

1

I developed this Python script that will create a KML file for a directory of DJI drone images. The script will populate the KML file with the lat, long, GPS elevation, altimeter value, pitch roll and yawl of the camera and the aircraft. The KML can be uploaded to Google Earth. By clicking the points in Google Earth will open the image and display the image data. You can also see the full sized version of the image in GE. Maybe you can adapt the script for your particular drone. This script has been tested on Phantom 3 and 4 Pros, and a Mavic.

As far as footprints go, if you use Agisoft you can run this script. Then in chunks, save the polygons out to a shapefile. Each image extent will be a polygon in the final file.

At this point you could convert the KML to a shapefile and join the two based on image name. Then you would have a single shapefile with both the image telemetry an the image footprint.

For the overlap count... Here are some instructions for methods using ArcMap to get the overlap count for multiple polygons in a single shapefile but there are similar tools in QGIS. Here are some links to pure QGIS methods that I have not tested.

GBG
  • 9,737
  • 1
  • 15
  • 44