How to Use a Thermal Camera with a Raspberry Pi

Feb 2, 2021

20 minutes

There are several IR camera sensors now available to electronics hobbyists that allow you to measure thermal radiation. All it takes is the addition of a microprocessor to turn an off the shelf sensor into a full-fledged thermal camera.

Keep in mind the resolution offered by these sensors is much lower than what you would find on a commercially available thermal camera. They are still sufficient enough to act as a person or animal detector, an automatic door sensor, or a contactless thermometer.

The Raspberry Pi offers adequate processing power and with its GPIO pins an easy way to connect sensors. Scripting languages such as python also run on the Pi and can be made to interface with the sensor. This combination gives a flexible testbed for creativity.

What else you'll need

This guide was written using a Raspberry Pi 3, though any Pi newer than the first generation Raspberry Pi should use the same GPIO pinout and the sensor would connect the same. The operating system used for this guide is the latest version of Raspberry Pi OS with Python 3 and Git installed.

Raspberry Pi 3	×	1
Adafruit AMG8833 IR Thermal Camera Breakout	×	1
Jumper wires, female-to-female	×	4
MLX90640 IR Array Thermal Imaging Camera Module	×	1

Howchoo is reader-supported. As an Amazon Associate, we may earn a small affiliate commission at no cost to you when you buy through our links.

Cost and ease of use were considered most when selecting a sensor for this guide. To save yourself the trouble of fabricating your own breakout circuitry and writing your own libraries it is best to make use of what is already available.

Companies such as Adafruit and Sparkfun produce breakout boards and supporting libraries for sensors such as the AMG8833and the MLX90640.

Panasonic manufactures the AMG8833, which is capable of sampling an 8x8 grid of 64 individual temperature readings up to 10 times per second. It can detect temperatures ranging from 0 to 80 degrees Celsius.

The MLX90640 is made by a company called Melexis and while costlier it is superior compared to the AMG8833. It has a resolution of 32x24 and can sample 768 temperature readings up to 64 times per second, detecting temperatures as low as -40 degrees Celsius and as high as 300 degrees Celsius.

At the time of this writing, the MLX90640 breakout board is sold out with no option to backorder, so I will demonstrate the AMG8833 breakout board produced by Adafruit, as well as the python library they provide.

Used here

See all

Check Price on Amazon

Regardless of which supplier you source the sensor from it will likely come with header pins that are not pre-soldered. For our purposes, it will be necessary to solder them in place.

Something the breakout boards for the AMG8833 and the MLX90640 have in common is the I2C protocol they use to communicate with the Raspberry Pi. The protocol uses one wire for bi-directional serial data transmission and another wire to supply a clock signal.

To avoid damaging the sensor or the Pi perform the following with the Pi powered off.

Use one of the jumper wires to connect the Vin pin on the sensor to pin 1 on the Pi.
Connect the GND pin to pin 9
Connect the SDA to pin 3
Connect the SCL to pin 5 using the remaining wires.

A fresh install of Raspberry Pi OS has the I2C interface disabled by default. It will be necessary to turn it on.

Open a terminal window on the Raspberry Pi and run the following:
```
sudo raspi-config
```
Select Interface Options then I2C.
Select Yes to enable the interface.

Next, we will install a python library that will do the heavy lifting of turning the signal input from the sensor into useful data. If you are using an MLX90640 based sensor then you may have to check with the manufacturer for the appropriate library. For the AMG8833 based sensor, we will install a library provided by Adafruit.

In a terminal window run the following:

sudo pip3 install adafruit-circuitpython-amg88xx

This will install the library necessary for interfacing with the sensor.

The following libraries are only needed to run the graphical example code provided by Adafruit. Run the following commands in a terminal:

sudo apt-get install python-dev libatlas-base-dev

sudo pip3 install scipy pygame colour

The first command installs dependencies for NumPy, which is a component of scipy.
Scipy is a data library that will be used in the example code to interpolate pixel data.
Pygame is used primarily for creating games, but here it will be used to render sensor data to the screen.
Colour is a library that simplifies the manipulation of color data.

Your browser doesn't support HTML5 video.

We are almost ready to see some output.

We will fetch the example code with this command:

git clone https://github.com/adafruit/Adafruit_CircuitPython_AMG88xx

Change to the example code directory:

cd ~/Adafruit_CircuitPython_AMG88xx/examples

Finally, run the example:

python3 amg88xx_rpi_thermal_cam.py

You should see a pygame render window that shows a color representation or "heat map" of the sensor data. Cooler temperatures are represented with blue and green hues while warmer temperatures are shown with yellow, orange, and red.

In this configuration, the sensor and the Pi are acting as a low-resolution thermal camera. Take a look at some of the other scripts provided in the example folder to see how you can work this sensor into your own code and your own projects. Like what about an awesome setting up a Raspberry Pi security camera that uses thermal imaging?!

NEXT UP

Odin (182)

Nov 29, 2021

Everyone knows that it’s possible to build just about anything out of LEGO blocks, one of the most imaginative toys ever invented, but now the ability to go high-tech with your LEGO collection just went to a whole new level with a new integration for the Raspberry Pi computer: the world’s smallest full computer and one of the most versatile pieces of hardware around. The new device is called the “Build HAT” (“HAT” stands for Hardware Attached on Top) an is designed to connect with LEGO® Technic™ motors and sensors. The HAT fits any Raspberry Pi with a 40-pin GPIO header and lets you control up to four LEGO® Technic™ motors and sensors from the LEGO® Education SPIKE™ Portfolio. The whole purpose of LEGO’s advanced building platform is to excite interest in STEAM (Science, Technology, Engineering, Arts, and Math) in a young audience, but the incredible power afforded by this array of products means that any LEGO enthusiast can create a whole plethora of robotic wonders. With the addition of the Raspberry Pi Build HAT, the ease of programming (through the HATs accompanying Python library), as well as the extended range of parts and sensors that connecting the Pi allows as builder to use, makes this extremely exciting for those who want to take their LEGO engineering project to hitherto unheard of dimensions.