[Pix4Dcapture Android 3.8.4] Feedback and experience

The app is not running properly on Samsung Galaxy Tab A8 after the recent updates.

Is there any compatibility issues ? I 'm unable to connect to my Phantom 4 pro with that tablet after the recent updates.

Hi Suri,

I think this is not an isolated case

Same thing happens to my friend, He calls me this morning asking what can I do since Pix4Dmapper does not work with his P4 Pro after the update. Hope there will be fix soon. 

Hi PiX4D capture developers,

I’m looking for 360° sphere option for DJI drones. I have found Hangar 360 but it’s only for IOS. Is it possible to develop something like that for Capture preferably on Android? It might also be cool that flight plan would support creating array or predefined positions (waypoints) of more 360° spheres. fusion of the images could be part of the Mapper or it could be used to create detailed 3D models.

With best regards,

Rok

@Roy
At the moment, it is not possible to fly higher than 500 meters AGL. Note that this can create connectivity issues to trigger pictures and ensure a sufficient overlap for optimal reconstruction. I can however report your suggestion.

@Suri, Roy
For optimal experience with Pix4Dcapture, we recommend to proceed with basic checks. This article helps users to troubleshoot app issues in most cases. If it persists, let us know so we investigate further.

@Rok
This would be an interesting new flight plan, but could you please give more information about the applications? What do you need to model exactly? What would be the benefits of a sphere vs. cylinder image acquisition?

Hi Julie,

The main purpose of the sphere would be just visual effect like this https://viewer.hangar.com/o04zDqj2 or google streetview. In this case only 1/2 spheres would be captured. This could be used for visual inspection since ground based cameras sometimes suffer from obstructed FOV. I guess it would be nice additional feature of PiX4D mapper/viewer/cloud (sphere merging). What do you think?

Hi,

I’ve tested a double grid mission with a phantom 4 advanced (latest versions of software/firmware). The mission performs flawlessly however I’ve noticed that Pix4Dcapture keeps switching back the image ratio to 4:3 on the images captured, significantly cropping the native 2:3 resolution of the sensor, therefore reducing the swath width (about 11%)… Is there a way to bypass this limitation ?
Thanks.

@Rok
I am not sure to completely understand, but I believe you are interested in the 3D model in which you could navigate. Our software is able to generated such a model with colors (step 2. Point Cloud and Mesh) after the initial processing of the images. In this case, the acquisition plan that I would recommend is the double grid over the area of interest. If you need more information about the different missions, check this article.

@Colin
Pix4Dcapture automatically set the image resolution to full meaning 4000x3000 (4:3) 12 MP. You cannot change this option, as we always recommend to process full resolution images to generate optimal results and get the most precise reconstruction.

About flying Pix4Dcapture with the Phantom 4 Advanced, you can provide maybe more information about you proceeded in this community thread.

Hi Julie,

 

I know about 3D models and raycloud capabilities of PiX4D mapper but I was looking to actually generate 360° sphere photo from 20 overlapping images taken from one point in all directions (except the zenith). Taking of photos with optimum overlap to cover whole sphere could be tricky if done manually, so it would be of great help if PiX4D capture could automate the procedure. And maybe not only for one sphere but for array of spheres or spheres along corridor… 

Thanks.

Hi Rok,

I better understand what you want to achieve. Actually Pix4D software performs image processing, meaning extracting keypoints from 2D data to match and project them in 3D in order to reconstruct the spatial reality. As a side note, our software can do this with 360° images.

However, it is not possible to stitch images as it not what Pix4D core algorithms perform. Stitching means 2D data are matched to create another 2D image but that is not faithful to reality as space is not taken into account (the results have distortions).

Hope it clarifies my previous answer :).

Hi everyone, please post any technical issue as a new post in the  Image acquisition  topic. This will enable us to more efficiently follow-up with you. If there are any recommendations or feedback for Pix4Dcapture in general, feel free to post in the topic  Product feedback  with Pix4Dcapture in the title. 

I will close this post. Thank you!