Very interesting.
So far i only use m2p with pix4dcapture and photoscan for some photogrammetry with old building for cloud points import to use in architecture, but would be interested to find out more, how make results even more precise etc. So, this data is imported to photoscan to correct lens curvature etc, right? Is this not autocorrected via photo import allready?
Here you will maybe find some info, this guy test few prosumer drones for photogrammetry work and re-done test for m2p after exif update was done, so it works correctly:
you can use english subtitles. He also have some other great videos.
As i remember, m2p takes 3rd place in accuracy, after p4p and parrot. But i must say, i was impressod wirth resoults. Now, i only wish they make droneharmoony app with 3d routing cheaper, as i don't need all that pro stuff..
Great blog, by teh way, must read the content about this whole geodata stuff..
izotop Posted at 1-7 14:38
Very interesting.
So far i only use m2p with pix4dcapture and photoscan for some photogrammetry with old building for cloud points import to use in architecture, but would be interested to find out more, how make results even more precise etc. So, this data is imported to photoscan to correct lens curvature etc, right? Is this not autocorrected via photo import allready?
Hi Izotop,
Metashape applies an automatic camera calibration on image alignment as part of the standard workflow. If a separate calibration is carried out with the same lens/camera on a controlled target array (chessboard) a higher precision can be achieved.
From the manual:
"After you have saved the calibration parameters for the lens, you may proceed with the workflowsteps in a separate chunk for the actual image set captured by the same camera and lens. To protect the calibration data from being refined during Align Photos process one should check Fixcalibration box on the Initial tab for the chunk with the data to be processed. In this case initialcalibration values will not be changed during Align Photos process"
What I'm keen to know is if there is any variation in these values at different distances/ exposures.
In the video clip no GCPs were used. This might save time but for mapping it is essentail to use GCPs. It is well known the P4Pro has a better GPS system than the Mavic.
I'd like to know more about what the minimum density of GCP should be for the Mavic, in theory 2 should provide a base scale but I have not tested beyond this yet.
aha, thnx for clear this out. for me, mapping etc is not that important, but would like to achieve as much high object precision as possible via photogrammetry (this is where more precise lens calibration probably does help, also).
izotop Posted at 1-8 05:32
aha, thnx for clear this out. for me, mapping etc is not that important, but would like to achieve as much high object precision as possible via photogrammetry (this is where more precise lens calibration probably does help, also).
You are correct, lens calibration will alpply to all types of phtogrammetric data.
lvl.4
