DIY Thermal Imaging Infrared Camera





I'm always looking for new Projects for my physics lessons. Two years ago I came across a report on the thermal sensor MLX90614 from Melexis. The best one with just 5° FOV (field of view) would be suitable for a selfmade thermal camera.

To read out the temperature I use an Arduino. In the internet you can find many descriptions about reading out the datas (f.e.

What you have to do creating a whole thermal picture is to change the alignment of the sensor like the electron beam in an old TV. Those z-tracks can be realised with a two-servo-mount.

Here you can find help, how to control servos with an arduino:

So you will need:

Teacher Notes

Teachers! Did you use this instructable in your classroom?
Add a Teacher Note to share how you incorporated it into your lesson.

Step 1: The Structure

The thermal camera just consists of the arduino uno, which is reading the temperature and controlling the two servos. The algorithm is quite simple: Read the temperature and go one servo step further ...

To start the measurement I use a button. With the program teraterm you can read the data: x, y, temperature

Those three rows are saved as a file, which can be finally visualized with the freeware gnuplot.

Step 2: The Results

Here you can see some thermal Pictures (cooktop, naked human body [me ;-)], candle)

They consist of 40x40 Pixels but it's up to you, which number of pixels you program. The more Pixels the longer the exposure takes. You can try to minimize the exposure time for on Pixel, but it will still last a certain time..

Maybe you want to take a look at my other Projects:

Thank's for your time ;-)

Be the First to Share


    • Made with Math Contest

      Made with Math Contest
    • Multi-Discipline Contest

      Multi-Discipline Contest
    • Robotics Contest

      Robotics Contest

    16 Discussions


    2 years ago

    How far is range? I though maybe a point and shoot temp sensor like in a hardware store. And if I am running an IR flashlight will it affect the sensor? Thanks for help.

    1 reply

    Reply 2 years ago

    thermal Imaging infrared cameras don't have a Limit concerning their range. The only Problem is, that the detected area increases with increasing distance. For example you can measure the temperature of the sky...

    The opening-angle is at least 5° with those MLX-sensors...

    IR-light could indeed affect the sensor, because it's measuring the incoming IR-light...


    3 years ago

    Good. Something worth a try ! I suspecr you learned more of what is shown in the IR pic.


    4 years ago on Introduction

    I'm thinking that maybe a modified omnidirectional barcode scanner with an ir sensor and laser will somehow be similar to this. With a faster software (I didn't take a look about the max data rates of the sensor o the arduino, sorry) the image will be much better and detailed.

    If i could i would try this, but now it's complicated and i don't have the control over arduino required to make something like this.

    Good instructable!!


    4 years ago

    Nice, was looking at doing something similar with Lidar for long distance 3D scans.
    How long did these exposures take (per pixel/per image)

    1 reply

    Reply 4 years ago on Introduction

    At the Moment I scan 4 Pixel/second. Therefore you'll have to wait 400 seconds for a 40x40 Image. But I'm sure, that you can increase that...

    I/you should take a look at the datasheet of the melexis sensor ;-)


    4 years ago on Introduction

    Fantastic build! Many times I've seen discussion about trying this, it is awesome to see you results!


    4 years ago on Introduction

    I had wondered if this was possible. You did an amazing job.