Forum

Author Topic: thermal images multi camera system process  (Read 3156 times)

monviso

  • Newbie
  • *
  • Posts: 9
    • View Profile
thermal images multi camera system process
« on: February 05, 2021, 11:21:09 AM »
Hi,
I am trying to process with the multicamera system two folder containing thermal images and RGB from XT2 camera. If I ran the process using the original r-jpg and rbg together all seems to work (I have in the photos window the two images coupled one small and one large). Does anyone know what metadata allow the two images to be considered "from the same camera"?
I have wrote a script in R to convert r-jpg into absolute temperature tif raster and edit its metadata in order to have the necessary information (e.g. GPS position) to import the tif in metashape but when I try to process them together with rgb, I can't see them coupled. Any help is really appreciated. Thanks

JoseASanchez

  • Newbie
  • *
  • Posts: 5
    • View Profile
Re: thermal images multi camera system process
« Reply #1 on: February 15, 2021, 01:38:30 PM »
Hi,

I'm working on the same issue. Any advance?.

Regards

monviso

  • Newbie
  • *
  • Posts: 9
    • View Profile
Re: thermal images multi camera system process
« Reply #2 on: February 26, 2021, 12:36:30 PM »
Hi JoseASanchez,

yep, managed to build an orthomosaic but completely abandoning the two camera approach. Using the calibrated RJPG (now .tiff files), I have stitched them independently by subset them in three different chunks ( I have in tot 800 images). Fore some reason, not completely clear to me, when the images are divided into chunks metashape is able to align them (while processing all the images together fails). The alignment process is complete after iteratively running the alignment several time (as in https://www.agisoft.com/forum/index.php?topic=8554.0 ) and adding GCP for each chunk. I have subsequently aligned the three chunks using GCP (marker). Consider that in the calibrated .tiff  are written, using exiftool, the original metadata of the rJPG about centroid coordinates, altitude and yaw, pitch, roll degree. I guess without these metadata the alignment would fail more likely (but depending on your images resolution..)

Happy to give further details if needed.
Matteo

JoseASanchez

  • Newbie
  • *
  • Posts: 5
    • View Profile
Re: thermal images multi camera system process
« Reply #3 on: May 07, 2021, 09:32:22 AM »
Hi Matteo,

Thank very much for your clarification. I'm following the same approach using directly RAW data in tiff format extracted from RJPG header. Build the ortho and then, I make the temperature conversion. However, I'm interested in two camera approach and build the multi camera ortho automatically. It is supposed that XT2 is synchronized in the same way as for example ALTUM and metashape works perfectly with it.

Please, someone from metashape could provide support to work with XT2 automatically and build the orto in the same way that altum.

Thanks

madu

  • Newbie
  • *
  • Posts: 2
    • View Profile
Re: thermal images multi camera system process
« Reply #4 on: May 07, 2021, 04:59:01 PM »
Hi,

in my case using a sensefly Duet-t what works for me is to leave the rgb images in the folder and create a subfolder inside with the thermals.This way agisotf understands that the one outside are the masters and the subfolder the slaves.

On the other hand, could you please share the script to convert the images to .tif with temperature values?

Best regards

MichalDV

  • Newbie
  • *
  • Posts: 25
    • View Profile
Re: thermal images multi camera system process
« Reply #5 on: April 29, 2022, 07:06:34 PM »
Hi,

in my case using a sensefly Duet-t what works for me is to leave the rgb images in the folder and create a subfolder inside with the thermals.This way agisotf understands that the one outside are the masters and the subfolder the slaves.

On the other hand, could you please share the script to convert the images to .tif with temperature values?

Best regards

For me it do not recognize Sensefly duet-t as multicamera this way. What is your exact naming for folders?

Paulo

  • Hero Member
  • *****
  • Posts: 1037
    • View Profile
Re: thermal images multi camera system process
« Reply #6 on: April 30, 2022, 11:41:27 AM »
Hello all,

I am also trying to process a Duet-T dataset consisting of 277 RGB images (Soda) and 277 thermal Flir RJpeg images. When I load the images using from folder and multicamera layout, I cannot get 1 chunk containing the RGBs as master and thermals as Slave so as to process the data set as a multicamera. I only get 2 chunks rgb and thm. This is unfortunate as it forces to process the 2 folders one after another and thus does not take into account the advantage of a multi camera synchronized shot. Does support have any plans to improve this process?

PS. to avoid having to align the RGB chunk and then transfer a few foto identifiable CPs from rgb to thm chunk to georeference, with a script we can transfer the camera reference coordinates from rgb to thm chunk and then straight away align the thermal chunk with this camera georeference (in this case RTK) so very precise)....Then the ortho is generated  colored according to band 4 formula
B4*0.01- 273.15 with thermal palette

However, when coloring the dense cloud it only takes the min  and max colors from palette (this case black and yellow)...seems to be a bug see last attachment
« Last Edit: April 30, 2022, 02:26:28 PM by Paulo »
Best Regards,

Paul Pelletier,
Surveyor

Paulo

  • Hero Member
  • *****
  • Posts: 1037
    • View Profile
Re: thermal images multi camera system process
« Reply #7 on: April 30, 2022, 10:30:16 PM »
Dear friends,

another option to be able to process the RGB Soda imagery with the thermal absolute temperature band in a Multi Camera setup would be the following:
  • export the 2nd plane of thermal Rjpeg as 16 bit single band tif imagery;
  • combine the soda RGB and 16 bit imagery exported above in multi camera layout with Soda RGB in plane 1 Master and 16 bit thermal imagery in plane 2 Slave
  • continue with align photos, generate depth maps, DEM, Ortho with Raster Transform formula B4*0.01 - 273.15 for thermal display of results
For this process, a script could be developed, resulting in something similar to following screen...
« Last Edit: May 01, 2022, 04:24:25 AM by Paulo »
Best Regards,

Paul Pelletier,
Surveyor