How to enable the radiometry mode?

[Dominique Beauchamp]

Hello,

I just want to enable the radiometry mode in order to have constant response vs temperature rather than with a slope function of the camera temperature. I'm not seeking for accuracy, just a flat response.

I cannot find the way to activate this mode. I read that it is disabled by default, I would like to enable it.

Thank you!

[sa...@pureengineering.com]

see: https://groups.google.com/forum/#!searchin/flir-lepton/radiometry$20%7Csort:relevance/flir-lepton/uw9XKkOllis/iAsFgb_jRhUJ

[Dominique Beauchamp]

Well, it is written that it is complicated and that a full radiometric mode will be introduced later. But is it possible to enable the "not full" current radiometric mode? Or is this feature completely for the future?

If it is not available, I guess I could read the camera temperature and do the correction by software?

[sa...@pureengineering.com]

I would just do it in software. I think the full radiometic mode will apply to another future module, as the leptons are not up-gradable. 

[Larry Tee]

You didn't answer his main question. His question was

"is it possible to enable the "not full" current radiometric mode? Or is this feature completely for the future?"

Well is it?

My Lepton is version 1.3, and in the documentation version 1.2.3 I see it mentioned

[Larry Tee]

http://www.flir.com/cvs/cores/knowledgebase/index.cfm?view=67336

 
I think I found the answer, and that is, you have to buy 1,000 of them directly from FLIR and agree an NDA.

Thanks FLIR, really great to know that before we made the purchase and embarked on the research angle looking to exploit all the exciting features boasted, we'll totally be able to give the PhD student their valuable time back...

[sa...@pureengineering.com]

I think the main issue with radiometric mode in the current lepton version is that it is not just as simple as enabling it. There are a number of steps that need to be done to calibrate the unit and get it ready for temperature readings. I think the 1000 unit min with nda is because flir would need to assign an engineer to supporting the application to get it to work. 

That said, the basic idea is read the temperature of the core via i2c, as this is the reference point to the raw data from the sensor. 

Then take raw data from several temperatures below and above the current temperate. A calibrated infrared thermopile can be used as a reference reading of a black body. (I heard black painted copper is a good option)  

From that you can build a lookup table or equation to convert raw data to temperature. 

you would need to do this at several different room temperatures as i'm sure the curves change at temperature. 

[Service Marketing]

Hello,

About the LEPTON 2.x/3.x and the radiometric mode.

To date, as you mentioned, it is not straight forward to have a radiometric mode on FLIR Lepton. 

In the meantime, it is possible to do it with the right procedure, hardware & software. For our internal needs, we decided several months ago to develop our own methodology to use the RAW mode of FLIR LEPTON. After several trials of calibration, we finally found a method that gave us good results in both accurate absolute temperature measurement and repeatability of measures. What about a 180mK repeatability?

Icing on the cake, our complete platform (LEPTON, HW (Embedded Linux), SW) is shutterless meaning that we do not use anymore the FLIR shutter to correct temperature drift.

We have plenty of idea to enhance our current platform but we are really interested to know whether or not it could be of interest for your applications.

   

[Fordarm Kim]

Hi, I am interesting about that. How can I get your help to implement on my platform?

[Fordarm Kim]

Users just want to know, how to transfer raw data value to temperature with a good guide.

But the problem is there is no right and straight guide for that. Therefore users are confusing they can do that or not.

For example, turn on sensor and get images is easy step due to there are a lot of source example for that.

But there are a lot of questions about radio metry or calculate accurate temperature from the raw data but there is no any source code for that.

I am wondering it's really high tech region or just complicate procedure issue. I really want to have clear answer.

We can make reliable temperature values with Lepton no matter radio metry enable or not.

If so, how to do that, who can give us this answer?

2015년 9월 24일 목요일 오전 1시 27분 29초 UTC+9, sa...@pureengineering.com 님의 말:

[sa...@pureengineering.com]

To get true calibrated temperatures for the lepton, it currently is not a simple procedure. What you get from the lepton by default is raw un-calibrated sensor data reference to the temperate of the sensor. There is some IP you could get from flir for larger volume applications for radiometry, or you can DIY with some effort. 

[Max Ritter]

The DIY-Thermocam has raw-to-absolute conversion with the FLIR Lepton implemented. It uses a MLX90614 single-point IR sensor as a helper to calculate the calibration coefficients.

Check the project out at www.diy-thermocam.net . The full source code is available on Github, you can find everything inside the folder Thermal/Calibration.h

There is also some discussion on the forum concerning that topic, so you may check that out, too !

By the way: If you already have a FLIR Lepton + Breakout board from GroupGets and want to use the DIY-Thermocam with that HW, just choose "Self-Assembly w/o thermal sensor & board" inside the store ;)