I’m using a DJI Mini 2 to generate MP4 videos that have GPS coordinates (and other data) in the subtitling. However, if I use the “Video Sequence” option to pull frames out, it doesn’t seem to encode the GPS data in the EXIF data. Is there any way to accomplish this?
Hi, what would also be good is when you inport a flight log that this does not have to be used as an exact position but as a track so to speak.
That RC looks where on this track, the pictures could have been shot. Because in the logs of 2 of my drones, it does not say when which picture was taken.
This function would then be extracted from the MP4 images also very helpful, because you could georeference the individual images and thus meet the scale of the scan better.
I mean video(RC extracts images from the video, so after that it’s also about “normal images”) and normal pictures.
The difference is, he means GPS information “in” MP4.
And I mean general information from a flight log. But You could use those GPS Information in the MP4 like a Flighpath / Flightlog.
Let’s assume the following.
I have pictures without GPS information (removed because of errors or simply not available, e.g. because of MP4) and a flight log which contains no information when the pictures were taken during the flight (Or the GPS Information in the Mp4 Video). *
Then it would probably be interesting to align the images.
And then, for example, align the result with the flight log.
So you would know the GPS coordinates for the images.
And in this process, I don’t think it would matter if this is done for the video or for other images.
Maybe you can go one step further and even align images with GPS coordinates to the flight path after the alignment. So to speak, use the flight path as a control path in the air. Or control point for the images and not for the point cloud.
This would make the function versatile and not “only” for videos.
I hope it is now clearer what I mean
*You could also ignore the information when the images were taken in the flight log and take this flight log as another control, so to speak.
Thank you for the ideas. I will add these to the feature request.
Any chance this has been included in the latest release ??
Hi Mark, this wasn’t included in the latest release and there is no estimation time of this to implement.
It would be extremely helpful to be able to use the GPS data that DJI drones can store in the caption track.
Hi Donald, thank you for your vote. I added it to our database.
This would be a wonderful feature and very helpful to me
This is something that would probably be better suited for a GIS application, in my opinion. What you’re talking about is basing the locations of the extracted frames on the flight log path of the drone: the flight path would be a line feature which includes a time stamp. The video file will have the starting time, which you can then pair up with the flight log feature to “guess-timate” the time of each frame extracted.
So the workflow would be:
- Save the extracted frames as images in a folder,
- Import the flight path log file into a GIS system (ArcGIS, QGIS, etc.) as a “feature layer”
- Break the feature layer you just created into smaller line segments: you’ll want to have the same number of line segments as you do images,
- Join the line new line segments with the images, using a 1-to-1 join (you won’t be able to do a “spatial join” because the spatial information is what you’re trying to establish).
There’s already an option to add flight logs and I have a way to modify my DJI flight logs (SRT files) to work in the software. It’s just cumbersome and does not allow for other information already included in the file.