… Pix4D engineers not having one …
If this is the problem, I can provide mine, in about 2 weeks. Of course not for a long time, but if that helps to create a proper rig calibration, I sure will.
@Ina, what do you think?
1 Like
Hello there, Reto that looks like a good idea if Pix4d agrees.
I just finished processing the shared dataset once again and this time these are the results. I started by splitting the dataset into 100images, processed separately step 1 (RGB/grayscale), merge the two and processed step 2 and 3, I am looking to have a RGB ortho and the Reflectance Map.
Results folder: https://drive.google.com/drive/folders/17Vcyj60TnbfRUtfC0Tq2_StFXeU8kNfF?usp=sharing
The Reflectance map was correctly created, but the same did not happen with the RGB Ortho.
Thanks,
Pedro Penedos
It would be really appreciated if Pix4d engineers could look into this, like a proper rig calibration or correct documentation for this sensor.
Hi Pedro!
I think Ina suggested to opt for all prior if the camera parameter optimisation yields more than 5% error (you have 27%).
I would suggest re-running it again.
Best Regards,
Balint
Thanks for the info balint!
Pedro
Hi all,
Thank you for your patience, for the last days I was out of the office and is only now that I could come back to you.
@John Wallcot I am not familiar with the modes of the camera. Since the mapper does identify a different focal length, what I would do in this case is to take a look at the EXIF and see if this information is accurate. It is very often that the focal length is wrongly written or read from FLIR images.
I will transfer this concern to our developer for a more in-depth look. And I will come back with more details.
Meanwhile, could you give me some documentation regarding these modes? I could not find it online.
Is this issue present when you extract the imagery with FLIR Tools or only with the mapper?
@Pedros before merging the project the quality of step 1 of each project has to be assets. Could you send me individual quality reports from each project?
@Reto as mentioned before we do not have the sensor and neither a dataset suitable for the calibration. Unfortunately, we cannot create documentation on workflows we have not tested.
Cheers,
Ina
1 Like
Hi Ina
> we do not have the sensor and neither a dataset suitable for the calibration
I well understand.
That’s why I volunteered to send/bring you the camera to Lausanne, if this is “the missing piece”.
Reto
1 Like
Hi Ina, I was able of processing the thermal reflectance map, but not the RGB Images, did any of you had trouble wit the rgb Orto-mosaicking part?
Link:
https://drive.google.com/open?id=1E9lSvSvWQ4fsxtFJX0WqPDO34tXgK7i_
@Reto while it would be very nice to have the camera here, we cannot risk on bring any harm to your good. The best would be for you to send us the dataset. If possible ideally it would be to have some GCPs as well.
@Pedro I cannot find the quality reports in the shared folder could it be that you have missed to add them?
Sure it is, just inside the folder: 2ortho, Anyway here it is. It looks like has poor overlap but the thermal reflectance map came out ok.
https://drive.google.com/drive/folders/1nhcfB1Mkfb37KsUFFilbT02lmYWgBUui?usp=sharing
Hi Ina
No problem, once I have the FLIR Duo Pro (hopefully within the next 2 weeks) I will capture and provide a dataset according to your needs, including GCPs. I will contact you beforehand to make sure the overall setup fits your needs.
Regards
Reto
@Pedro
Indeed the problem here is the overlap. You can always manually check if the overlap is sufficient by placing two consecutive images side by side. Make sure you verify two successive side and frontal images.
I just got notified that my FLIR Duo Pro is not delivered before 6 or 7 of July 
It’s a pain, FLIR changed their consignee statement which led to yet another delay.
@ Ina, Thanks for the confirmation we will trying to improve the overlap once we have the Sensor and the Gimbal and a consistent flight planning software for our Mseries DJI drones.
@ Reto Weiss, What sensor did you pick? are you flying at lower altitudes (80m) with 80% overlap? With a Multi-copter? What type of terrain are you interested in mapping? My ROI is Solar PV panels but would be nice to successfully test it in any other terrain… Until now i was able to create a Reflectance map from the Radiometric grayscale images but the images did not have enough overlap to build a RGB ortho.
Cheers,
Pedro
Hi Pedro
I will use a FLIR Duo Pro R 640 with 13mm lens, if FLIR ever decides to ship this camera…
I will (also) use a DJI M210.
I will target rather flat grounds, e.g. turf areas like golf courses and yes, I’m looking for RGB orthos as well as reflectance maps. That’s why I plan to use a FLIR DUO and not a Tau. I also think, camera calibration should be much easier with the 12 MPix RGB images rather than IR images only.
Reto
I will also be using a Flir duo pro 640 13mm for the same purposes, but using a DJI S900 for development.
What flight software are planning on using? are you using CAN BUS in your gimbal? We are going to use a Gremsy Gimbal we have.
Pedro
I ordered a dual gimbal from copterworks, which can also take a Rededge-M, besides the FLIR.
First missions I plan to fly with GSP (DJI GS Pro), mapping or waypoint mode. Maybe I will switch to another app later, it depends on what’s available.
Hi! We have also flown another mission (with DJI M600 this time) for agricultural purposes: https://we.tl/WZAznXbxXB No gimball just a downward looking sensor mounted on shock absorbing frame.
In case you are interested please find attached. Processing with optimize for all prior worked well.
Hi Gyula Sólyom
Did you take also RGB images? Or the only ones are the Radiometric ones?
Thanks
Pedro
Never mind i was able to get the RGB ones!
Thanks
Can any one confirm if these are the specs of the RGB sensor?