Capture 14bit Image from Lepton 2.5 using GetThermal

470 views
Skip to first unread message

Paul S

unread,
Apr 24, 2018, 3:09:08 PM4/24/18
to Flir Lepton
Hi, I have managed to get a live stream on to the screen using GetThermal and a Lepton 2.5 in a PT2 board on a Pi3  - now I want to capture the 14bit  Radiometric image so I can do some analysis in Excel - is there a simple way to do this?

Kurt Kiefer

unread,
Apr 24, 2018, 3:34:31 PM4/24/18
to Flir Lepton
There's no prescribed way in GetThermal at this time, but it would be a great feature to have!

Barring making changes to GetThermal itself, if your workflow permits shutting GetThermal down, you could switch over to a different capture method like in https://github.com/groupgets/purethermal1-uvc-capture

pt1-v4l2-grab could be modified to change the capture type, and you'd have to change some code to write out csv instead of pgm. uvc-radiometry.py could be modified as well, if you're more comfortable in python

Mana Saedan

unread,
Apr 24, 2018, 8:06:25 PM4/24/18
to Flir Lepton
Just curious,  what is the minimum firmware version of PT1 that can get radiometric data from lepton 2.5. My PT1 firmware is 0.99.2. I believe that it does not support radiometric yet. Since when I used uvc-radiomatry.py with Y16 capture mode, the data only return me 8 bit long for each pixel. Could anyone confirm the version that works with Y16 capture mode?

Kurt Kiefer

unread,
Apr 24, 2018, 8:24:55 PM4/24/18
to Flir Lepton
Prior to 1.0.0, there was a separate firmware for Y16 (e.g. 0.99.2+Y16) that had to be loaded to get actual radiometric data out of the Y16 channel, but the tradeoff was that normal non-radiometric imaging wouldn't work (which meant that standard means of capture wouldn't work as expected). Starting with the 1.0.0 release, the same firmware works for both radiometric and non-radiometric modes and all leptons. If you don't have a programming adapter, just email GroupGets support and they will help you load new firmware on there one way or another.

Paul S

unread,
Apr 25, 2018, 5:28:53 AM4/25/18
to Flir Lepton
Thanks Kurt,

I'm also interested in understanding why the FFC function and the AGC features are fixed  - and yet they are available in the MacOS version (although if I change the MacOS version to 14Bit, all I get is noise from the signal.....) - for my application I need to can the AGC having done an FFC so I can get a reproducible set of values for analysis - for a quick and dirty test I'm going to find a PGM to CSV converter and see if that works enough make it worth putting effort into code as my code skills are close to zero at best.

Paul S

unread,
May 10, 2018, 11:54:48 AM5/10/18
to Flir Lepton
I am at a loss - largely due to a complete inability to understand how to install / run  https://github.com/groupgets/purethermal1-uvc-capture on my Pi - is there an idiots guide to this somewhere? or a set of simple command lines? My ability is not high - for example dont say "just eat lunch" I need to be told where the knife and fork are, and to use the knife to cut and fork for lifting etc....apologies but there are people like me out here in the dark, still trying to evolve....

Kurt Kiefer

unread,
May 10, 2018, 12:04:07 PM5/10/18
to Flir Lepton
Paul,

You wouldn't install purethermal1-uvc-capture, it's just a collection of scripts and examples. You can run them in place, "git clone https://github.com/groupgets/purethermal1-uvc-capture.git" somewhere on your pi, and "cd purethermal1-uvc-capture/python".

There are instructions in the readme for how to get uvc-radiometry.py running, including how to install libuvc (the library that these scripts depend on):

Kurt

Paul S

unread,
May 11, 2018, 12:09:51 PM5/11/18
to Flir Lepton
Thanks Kurt,

I have installed the scripts, and installed libuvc, but running the script ./uvc-radiometry.py gives me an error - "could not find libuvc!"  I have moved the libuvc directory to various sub-directories to see if that helps (including the /python directory within purethermal1-uvc-capture), but keep getting the same error - is there a specific location it needs to be installed in? 
I am not running GetThermal during this process, and I think that the libuvc sub-directory within GetThermal is probably the same version?

Paul

Kurt Kiefer

unread,
May 11, 2018, 12:20:07 PM5/11/18
to Flir Lepton
Yes, the same one should work. I have mine sitting in the python directory of the examples (the same directory I'm executing the code from). A system-wide install should work fine as well, so it's odd that this doesn't work. One thing to note is that the way ctypes loads the library it expects the library name on linux to be exactly "libuvc.so", is that the case on your system?

Paul Sacker

unread,
May 11, 2018, 12:27:15 PM5/11/18
to Kurt Kiefer, Flir Lepton
Hi Kurt - within the purethermal1-uvc-capture/python/libuvc/build directory there is a file called libuvc.so  is that what you mean?


-- 
You received this message because you are subscribed to the Google Groups "Flir Lepton" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flir-lepton...@googlegroups.com.
Visit this group at https://groups.google.com/group/flir-lepton.
To view this discussion on the web visit https://groups.google.com/d/msgid/flir-lepton/6b2a26f0-d4b0-4ca8-bc4a-57f175987d78%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply all
Reply to author
Forward
0 new messages