Reading Temperature Values From Lepton 3

[Nipun Perera]

I am working on a project involving FLIR Lepton 3.

I have run into an issue where I cannot read the temperature from the camera.

I tried to calibrate the camera by taking images of known temperatures and mapping 

them to pixel values.

Is it necessary to calibrate the camera in this way or is it possible to directly output the temperature

value from the code itself?

I would appreciate if someone could help me in this.

Cheers

[TomH]

There are plenty of posts in the user groups. Please read some of those related closely to your project environment such as: Pi, Beagle boards, etc. There are example codes as well specifically for Lepton 3.

Good luck!

[Nipun Perera]

Hi Tom,

I tried to use lukevanhorn/Lepton3 github repository. In that I used control.c to obtain the FPA temperature. But the value is always 135576. Do I need to change the endianness using the utility function given since I am using an RPi?

Thanks 

[TomH]

Hi Nipun,

Sorry, it's hard for me to know your '135576' from where and how it's obtained. The control.c is not for data streaming from SPI channel. The code in the file is for setting configuration and getting parameter to and from the camera via TWI channel.

Check out the post:

https://github.com/TradeCrafter/lepton3

This is the basic set of code (for raspberry pi) that works for me. Please read the Lepton_thread.cpp file in it so you can see how to read the data stream from SPI and how to extract the raw pixel value from the data packet. I'll be glad to help you out.

There are too many examples in the user groups and github that could cause confusions... I asked and directly got reply from Lepton so that I can get more accurate info and compare to what the datasheet and app note show rather than making assumption from some of those observed by individual users.

Right now I'm working on the accuracy of the temperature reading. It's not simple.

Tom

[dmonie...@gmail.com]

Hi Tom, I'm also working on temperature accuracy and I'm using a telemeter to get the distance between the Lepton 3 and the aim, so I can apply some corrective gain to pixel values. I'm not entirely satisfied with this solution, because in my scene I can have different objects at different distances, so the result might be wrong. Are you also using a telemeter ?

Dani

[Nati Barchilon]

Hi Dani , im working on somthing similar to what your are doing would like to consult via mail or skyp if possible . 

Blessings Nati

בתאריך יום רביעי, 7 בפברואר 2018 בשעה 19:39:58 UTC+2, מאת dmonie...@gmail.com: