Discussions

mismatch images to CAM messages in georef

Posted by Darrel Maddy on July 1, 2016 at 4:57am

I am sure many of you have had the same issue but I am wondering what strategies you take to minimise this issue.

To illustrate. I have a flight with 312 images (taken with a sony a5100 from a skywalker 1900 and pixhawk fc). When using georef I get the mismatch with the log file recognising 307 images. Now there are spurious images which were taken using 'trigger now' to test before flight, but when I remove these I still have 308 images. I removed the first image and went through the processing (and then on to Photoscan Pro) and the positional accuracy is poor. I suspect the 'off by one' quick fix did not work! I could do the same again by removing the last image but it looks relevant so I am reluctant to remove it. How do I determine which image is missing? I am assuming the camera failed to fire on one occasion - albeit that seems unlikely.

Any suggestions for an efficient way forward would be very welcome.

Thanks

Darrel

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –

Replies

IronDOME April 25, 2017 at 4:47am

Hey Jay

Sounds complicated to use, but how good is this tool?

Thanks

Jay Case said:

Try Swift Tag in the future. It handles cam/photo mismatch scenarios and adds additional features.

http://swiftradioplanes.com/swifttag/
Jay Case July 17, 2016 at 3:38pm

Try Swift Tag in the future. It handles cam/photo mismatch scenarios and adds additional features.

http://swiftradioplanes.com/swifttag/
Moderator

MarioSpeedwagon July 13, 2016 at 3:50pm

No problem, I'm glad it helped.
Will Fellmeth July 13, 2016 at 5:33am

Mario,

Thanks a bunch for posting this video. We just added one more setup to our QC process. Thanks again.

cheers,

Will F

MarioSpeedwagon said:

I don't know if I would word it that way. Mission planner works, but it messes up often enough that I abandoned it.

Here, try this. Seems a little intimidating at first, but go through it a few times and it'll only take you about 3-5 minutes to do it when you're familiar.

Harry Courter said:

So tagging with mission planner just is not a realistic goal?

and I should look for an outside method for geo tagging.
Darrel Maddy July 12, 2016 at 9:10am

Thanks Mario - that is also a very useful video for me. I will try using geosetter also :)
Harry Courter July 12, 2016 at 8:36am

Mario,

That's a good video. Actually very helpful. I will try this and let you know my result.

Thank You.
Moderator

MarioSpeedwagon July 12, 2016 at 7:25am

I don't know if I would word it that way. Mission planner works, but it messes up often enough that I abandoned it.

Here, try this. Seems a little intimidating at first, but go through it a few times and it'll only take you about 3-5 minutes to do it when you're familiar.

Harry Courter said:

So tagging with mission planner just is not a realistic goal?

and I should look for an outside method for geo tagging.
Harry Courter July 12, 2016 at 4:13am

So tagging with mission planner just is not a realistic goal?

and I should look for an outside method for geo tagging.
Moderator

MarioSpeedwagon July 11, 2016 at 6:30am

I know this isn't really the "fix" you're looking for, but just wanted to throw in my 2cents.

I've never had consistent, reliable results when tagging using mission planner. CAM messages OR time offset.

I use geosetter, and results have been much better. After about the 5th time, it doesn't take me any longer than using mission planner, but the results are much better.

Peau Productions / Mapir put out a good instructional video. The only difference is I use "assign exactly as found" with the .bin file.
Harry Courter July 11, 2016 at 5:25am

Darrel:

I have a similar issue, after a short survey grid mission I will have 30 pica and only 12 cam messages. Telemetry logs show the locations of all 30 cam trigger locations but the log files don't come up with the geo ref data.

A few basic issues I'm currently learning.

The time assigned by the pixhawk is ahead of the camera time by 1 1/2 hours. My impression was that the pixhawk acquired the time from the GPS. Do I need to somehow set the time? Or perhaps the time zone?

When I review the log file what setting actually shows the cam trigger data?

I have reviewed all the wiki files I can find, and these items don't seem to be included.

I can post log and telemetry files tonight.

Any help would be much appreciated.