Support Website Contact Support Blog

[Android] Pix4Dcapture with Bebop 2

New thread about Pix4Dcapture Feedback and Experience on Android.


Hello Pix4D members and Android users,

We are glad to announce a new version of Android Pix4Dcapture.

Please use this post to share your experience with other users and us.
We are looking forward to your feedback!

Pix4D Support Team

Just tried this two times and the BB2 seems to have followed the course and taken the photos correctly but had errors both times after uploading to Pix4D. From the error report there are two sections with exclamation marks: “Images: median of 974 keypoints per image” and “Georeferencing: yes, no 3D GCP”. I cannot see what I can do to improve the results because the BB2 followed the route and took the pics perfectly.

On Android v2.1.1 Pix4d capture seems to think that the battery life of the Parrot Bebop 2 is 10 minutes, with larger missions reporting “too large”, despite the battery life of the Bebop 2 being 25 minutes.

Because of its low weight (< 0.5 kg), the Bebop 2 is a very interesting drone. However, mainly because of the limited camera quality, I honestly doubt that it can be used for serious mapping projects.

Any hands-on experience so far?
Can anyone provide an orthomosaic, based on images taken by a Bebop 2?

Thanks a lot.

Hi all,

@ Alan
We are glad to read about your positive experience with the app and the Bebop 2.
Regarding the subsequent processing,

  • To increase the median number of keypoints detected on the picture, we suggest that you have a look at this article:
  • If you do not have GCPs, then it is normal to have a yellow warning sign and the message “no 3D GCP”. The sign will turn green if you measure and import GCPs’ coordinates, meaning that the georeferencing is very accurate.

@ Gregor
There was indeed a bug with the time estimation of the grid mission. This should be solved in the latest version 3.0.1.
Did you have the opportunity to try it?

@ Reto
Based on our experience, we shared some recommendations for the Bebop 2:
There is a dataset and the results that you can download:


Hi Julie

Thanks for the good sum up.

I completely agree with your recommendations/conclusions for the Bebop 2. IMHO, the camera is not powerful enough to create serious 3D-models or even high-quality orthomosaics. Thanks to the fixed super wide-angle camera/lens, the Bebop 2 is really lightweight, but the drawback on image quality is more than obvious. For a video shared on youtube this might be sufficient, but not for serious or even professional work, I suppose.

It would have been too nice to be true – a < 0.5 kg drone able to capture high-quality images.

New thread about Pix4Dcapture Feedback and Experience on Android.