Forum

Author Topic: FLIR DUO R - Thermal orthomoscaic  (Read 6299 times)

jmgc

  • Jr. Member
  • **
  • Posts: 71
    • View Profile
FLIR DUO R - Thermal orthomoscaic
« on: November 28, 2017, 07:42:44 PM »
Hi All,

Anyone one with experience with FLIR DUO R?
Just bought one and I'm getting involved with the file formats and processing workflows.
My objective is to process thermal images with Photoscan, to output thermal orthomosaics.
This camera can capture radiometric JPG images (RGB+Thermal), or an RGB JPG + TIFF Thermal.
I guess Photoscan can't handle radiometric JPG's, if anyone have any experience  I would appreciate some advise.

Any suggestions?
José Miguel Campos
Geospatial Specialist
UAS Operations Manager

SvaneSDU

  • Newbie
  • *
  • Posts: 48
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #1 on: December 05, 2017, 02:32:48 PM »
Hi jmgc
We are presently beginning to use a Vue Pro R, and will also make stitches of the images.
I don't think we will have to keep the radiometric information per pixel in the stitch, but instead "import" this data into the raster map afterwards using a GIS software.
It's a manual approach but using the thermal colorgrading to create vector layers with the corresponding temperature values is how we'll do it.
We have to calibrate all measurements anyway with a handheld thermometer.

Regards
Niels

jmgc

  • Jr. Member
  • **
  • Posts: 71
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #2 on: December 05, 2017, 08:12:07 PM »
Hi Niels,

The FLIR DUO R has a 160 x 120 thermal resolution. This is a problem when trying to obtain an thermal mosaic: the flight path must be unacceptably tight.
My first approach was to use the radiometric jpeg images (1920 x 1080) and somehow retrieve a thermal mosaic from that those images, but from my understanding there's isn't a straight forward way to get the thermal data  from radiometric jpeg unless using FLIR software. I guess this would be the best option, since the greater resolution of the RGB sensor would be enought to Photoscan processing.
Any suggestion on this approach?

Considering that the camera can shoot RGB + Thermal separately, what I'm trying to do now is process the RGB images with Photoscan, and with the model already built, change the image paths to the thermal ones, already scaled to fit the RGB size. The first problem is that the field of view (FOV) is different for both sensors.
I'll be working on this second approach.

Regards,
jmgc




« Last Edit: December 11, 2017, 01:10:02 PM by jmgc »
José Miguel Campos
Geospatial Specialist
UAS Operations Manager

SvaneSDU

  • Newbie
  • *
  • Posts: 48
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #3 on: December 06, 2017, 12:27:19 PM »
Hi
It sounds interesting, I hope you post your experience/result from this approach.
With the Vue Pro, we'll only have the thermal images, though I am not sure if the sensor also includes RGB for extrapolation of the thermal information.
We will then have a standard RGB camera mounted next to the thermal sensor, to provide high-res georeferenced RGB footage. I think I will try to create a composite image using ImageJ, or something like that.

BobvdMeij

  • Full Member
  • ***
  • Posts: 157
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #4 on: February 14, 2018, 11:58:48 AM »
Any news on this? Curious to hear how you're coming along as we may be opting to acquire the (higher res) FLIR DUO Pro R or, alternatively, equip our M210-RTK bird with a X5s/X4s and Zenmuse XTR simultaneously.

jmgc

  • Jr. Member
  • **
  • Posts: 71
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #5 on: March 08, 2018, 03:37:59 PM »
So far we can build the model using the RGB images, and then we use the thermal images with the RGB model to obtain an thermal orthomosaic. This thermal orthomosaic isn't yet perfect because the aligment of the images RGB to the Thermal ins't quite good (originally from the camera). Work must be down to "match" both images.

José Miguel Campos
Geospatial Specialist
UAS Operations Manager

ArnauCM

  • Newbie
  • *
  • Posts: 38
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #6 on: April 10, 2018, 06:34:56 PM »
Hi jmgc,

I will be doing exactly the same you are trying to do and I was curious how was your project doing. Could you achieve good results?

ArnauCM

jan-pieter14

  • Newbie
  • *
  • Posts: 1
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #7 on: January 05, 2019, 01:55:31 PM »
@jmgc

Hello,

Could you explain how exactly you used the thermal images with the RGB orthomosaic to create a thermal orthomosaic?

Best regards,
Jan-Pieter

mr4k_bear

  • Newbie
  • *
  • Posts: 3
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #8 on: March 31, 2019, 10:45:50 PM »
Hi jmgc,

I will be doing exactly the same you are trying to do and I was curious how was your project doing. Could you achieve good results?

ArnauCM

Hi everyone!

We have the same problems with aligning FLIR Duo Pro R images (thermal - 640x512 and RGB - 4000x3000).
In results we need maximum as possible good thermal orthomosaic.
Our images have correct GPS info on EXIF but, for some reason, we can't to align more than 20% of images.

Thanks,
Andriy!

julia

  • Newbie
  • *
  • Posts: 5
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #9 on: May 13, 2019, 11:39:23 AM »
Hi jmgc
We are presently beginning to use a Vue Pro R, and will also make stitches of the images.
I don't think we will have to keep the radiometric information per pixel in the stitch, but instead "import" this data into the raster map afterwards using a GIS software.
It's a manual approach but using the thermal colorgrading to create vector layers with the corresponding temperature values is how we'll do it.
We have to calibrate all measurements anyway with a handheld thermometer.

Regards
Niels

Hi Niels,

Could you provide some more information on whether you managed to produce the temperature orthomosaics this way and exactly how you did it?
I'm considering buying a Vue Pro R so it would be good to know if I can actually produce the temperature orthomosaics.

Thanks!

jmgc

  • Jr. Member
  • **
  • Posts: 71
    • View Profile
Re: FLIR DUO R - Thermal orthomoscaic
« Reply #10 on: February 05, 2021, 02:17:37 PM »
I've opened a new topic on this subject here: https://www.agisoft.com/forum/index.php?topic=13049.0

Thanks
José Miguel Campos
Geospatial Specialist
UAS Operations Manager