I’ve started a few fragmented posts as I’ve been working to dial in this sensor I thought it might make sense to see if we can get the DJI P1 mainstreamed in ODM with a more holistic approach. Seems like several people are flying the P1 now and more to follow as this is DJI’s hero photogrammetry payload for the near future.
As far as I can tell, a few things need to happen:
With these three items, I believe ODM could deliver some fantastic results for this sensor and retain more users who test the P1 on ODM and then give up because they didn’t get the results they expect.
While I’m not much of a coder, I’d be happy to test and provide feedback to help move this forward.
Hi Summitbri,
Thanks very much for raising this. It would be great to see the P1 sensor and perhaps the L1 camera support added to this project. I think it would certainly drive take up of the project, especially with the number of these units out there.
Thanks.
Craig
Hadn’t thought about the L1 camera as the lidar dataset is typically going to be processed elsewhere, but good idea to bring this up for photogrammetry-only missions! I’m suspecting the issues are the same. Does the L1 camera tag the photos with rtk data?
I believe it’s all the same data as the P1, so position, angles and accuracy.
So I took a shot at adding the P1 to the OpenSFM sensor data, sensor data detail and calibration files.
Once they are accepted in OpenSFM, how long until they start showing up in ODM docker builds?
My Zenmuse P1 sensor updates have been merged with the mapillary/opensfm repo: sensor_data.json and sensor_data_detailed.json.
Does anyone know how these ultimately get pulled through to the ODM docker build? Or how to create our own build via cloned repo in a way that would pull these through?
@saijain_naib I’m thinking you might have some advice 
Just a quick update, found the repo’s and merged changes to accommodate the P1 sensor, should be in the next ODM update 2.8.9 as that becomes available. I did not add a calibration file so please use your own cameras.json for calibration for now. Looking forward to feedback, in particular if this fixes the challenges with Smart Oblique datasets.