Hi,
I've been using exiftool to grab metadata from some images taken with a Duo Pro R. I was wondering if geotagging information is currently supported? I've had a look at a hex dump for the image which is attached, as well as the FFF file. I had a quick search, but I couldn't see anything related to this on the forum already.
The GPS sequence is near the end, before the datum which is readable if viewed as ASCII. It looks like a bastardised version of the EXIF GPS tags: https://www.exiftool.org/TagNames/GPS.html (https://www.exiftool.org/TagNames/GPS.html).
The format is:
4 bytes 0x32323030 - "2200" (magic?)
2 bytes longitude (N or S, null terminated)
2 bytes latitude (E or W, null terminated)
4 bytes null
8 bytes longitude in degrees (double precision float)
8 bytes latitude in degrees (double precision float)
4 bytes altitude in metres (single precision float)
32 bytes null
2 bytes "K"+null (speed reference - K/M/N /s)
2 bytes "T"+null ((T)rue or (M)agnetic North)
4 bytes speed
4 bytes heading
Example sequence (this image was taken in Indonesia, presumably just as the drone was taking off):
0x32323030 : "2200"
0x53004500 : "S E "
0x00000000
0xDDABFB6C896102C0 : -2.2976254
0x710CB73217825C40 : 114.0326659
0x14AE4742 : 8.72000026702881 m
0x00000000
0x00000000
0x00000000
0x00000000
0x00000000
0x00000000
0x00000000
0x00000000
0x4B005400 : "K T "
0x00000000
0x85EB513F : 0.819999992847443 km/s
0xCD4CD942 : 108.650001525879 deg
0x00000000
0x57475338 34000000 "WGS84"
Note even though you're given the direction, the decimal coordinates are always referenced to North/East.
It would be useful to see this in the tag output, as it doesn't seem to be picked up at the moment.
Thanks!
Great, thanks! I didn't have any FLIR samples with known coordinates so I hadn't been able to decode this. ExifTool was only decoding the GPSMapDatum from this block.
ExifTool 11.16 (to be released in about a week) will extract the following information from this image:
> exiftool ../pics/FLIR_DuoProR.fff "-gps*"
GPS Version ID : 2.2.0.0
GPS Latitude Ref : South
GPS Longitude Ref : East
GPS Latitude : 2.297368 S
GPS Longitude : 114.032183 E
GPS Altitude : 8.72 m
GPS Speed Ref : km/h
GPS Track Ref : True North
GPS Speed : 0.81
GPS Track : 108.65
GPS Map Datum : WGS84
GPS Position : 2.297368 S, 114.032183 E
- Phil
Edit: Actually, some of my FLIR samples DID have GPS, but I just didn't know it!
Awesome, thanks!
Looking at the Duo Pro R documentation, there are lots of other fields that it could potentially save, depending on what gets sent by the drone. I assume the built-in GPS was used for that flight, which is why we have e.g. no precision information. It'd be interesting to see if that 32-byte block gets filled in if we feed more data into the camera.
Let me know if you need any more files to test against.
Great. Please let me know if you discover anything else. I don't need more GPS FFF samples, but a JPG sample would be nice.
- Phil
The last 4 bytes of the 32-byte block store the GPS DOP (dilution of precision).
- Phil
I think the first 4 bytes (just before the start of your hex dump) are a 4-byte integer that is 1 if GPS is valid, or 0 otherwise. And there is definitely some more information in this record that looks interesting but isn't yet decoded. At offset 0xa4 from the GPSValid flag there is a 6-byte string that has values of 000208, 081210, 020409, 000608, 010408, 020808 and 091011 in my samples. I don't know what this is (it isn't the camera serial number).
Also, there are two doubles at offset 0x78 that are 0 or -1 in all my samples, and two floats at offset 0x88 that are 0 or -1 in all my samples.
I think I'll release version 11.16 today instead of waiting until next week.
- Phil