Support Website Contact Support Blog

Help! No geolocation metadata in photos

Sorry, this is a noobie question. I used a trial-licensed copy of Pix4dMapper for Mac about a month ago and successfully processed a small set of aerial photos taken with a GoPro Hero4 Black camera. I pretty much left all software settings at defaults and it worked great, making a nice ortho map for me.

I just tried running a new set of photos, taken with the same camera etc., on both the cloud platform and the updated Mac software. Both seem to be complaining that the project can’t be processed because the photos lack geolocation data. When I examine a photo file’s metadata, it seems to be true: the GPS data field is empty.

But this data was missing when I ran my project last month and it worked anyway. Anyone know why? Or how I can fix this?

Here are some error messages for the above:

One more thing: The cloud-based app “successfully” completed the project, but only 7 of the 86 photos were used. The log file seems to say the others were excluded:

[2016.12.08 16:24:50][3%RAM][18%CPU][Warning]: Load ortho: skip image <GOPR1675.JPG>: image is not calibrated

you can still run a project without geotagged images i think your having a different issue here

Does the cloud processing option accept untagged photos? Since it takes significantly longer (and with poorer results) to process untagged images, I wouldn’t be surprised if they disabled it for cloud processing. 

 

Also-

I’ve had this issue when using an S100 and internal camera GPS, as well as when I did something wrong during geotagging in mission planner. 

 

I follow Peau Production’s instructional video for geotagging with geosetter, if needed. 

Hello everyone, 

 

@Steven, did you notice any problem with the GPS while acquiring the images?  
When geotags are inaccurate the software will discard them automatically.
As Ramiro mentioned, you will still be able to process your project even without image geolocation.  If not all the images are calibrated, then you might need to double check the quality of the images and the overlap. 

You might find this article useful. 

 

@Derrick, you can process on the Cloud even without having image geolocation, this is fully supported. 
Indeed it might take some more time as it would take when processing on the desktop. 

We encourage you to give it a try and do not hesitate to share your experience with us! 

 

Best Regards, 

 

 

 

 

hello everyone 

I met a same problem here, i need to use template Ag Multispectral, but it not allow me to use it without geolation data.

and I use sequoia camera to take images, can anyone know how to use it? 

@Wang Yu Min: the Ag Multispectral processing options template uses the “Alternative” calibration method which requires geolocation of images and that is optimized for flat areas like fields.

If you do not have image geolocation, you could change the calibration method to “Standard”. This should work in most cases as well, although it is not optimized for flat areas.  

Yes same error here. Somehow our flight log got corrupted in our E-Bee and we were not able to geotag the images in the Sensefly software after the fact by connectin the drone and using the flight manager then processing the survey in Pix4D Mapper as per usual.  I would be OK with just an un-referenced stitched up photo mosaic in this particular case however when i process the photos only 30 are included in the results. Seems the program wasn’t able to match the rest.  I tried processing using the time stamp, thinking it would be able to find the neighbouring images and locate the tie points that way but neither approach works. Basically, if the images are not geotagged and there are no recognizable landmarks, processing doesn’t work. 

Hi Jakub,

 

The  al ternative calibration method is optimized for aerial nadir images with accurate geolocation and low texture content and for the relatively flat terrain. 

From your description, it looks like you are mapping such kind of area hence having geolocation does help the reconstruction.

However to better understand your current results and if possible to give you a solution could you post here your quality report?

 

Thank you,

Ina