What is Portable Face Recogniser?
The Portable Face Recogniser is a device that can remember people’s names & faces by training an AI Face Recognition Model. It also uses the same model to recognise their faces later.
The device uses face images on the fly to recognise a person or to create calculated face models (a bunch of numbers) in a yml file, but it does not keep face images at all.
It also works without the Internet and can be powered by a mobile battery. It has an audio jack as well as Bluetooth speaker connectivity to greet people with a robotic voice.
Background
You see a person walking towards you. You know that person but quite remember the name or how you met that person. I would love to see a device in the future can assist in this situation.
While I was dreaming about the future, I made a toy version to explore the concept and had some fun at the same time.
After I finished making the device though, I realised a lot of educational values in this project. That led me to document this project as match as I can to make this repeatable for anyone who interested to try making it.
About the Tutorial
This tutorial takes you from the most basic setup of Raspberry Pi to low-intermediate Python programming to make your own Portable Face Recogniser. In between you’ll also use:
- Buttons with Raspberry Pi
- LCD with a Raspberry Pi
- Database with Raspberry Pi
- Bluetooth speaker with Raspberry Pi
- Face detection with OpenCV
- Face model training with OpenCV
- The trained model to recognise face with OpenCV
- Crontab to autorun Python program
Tutorial Index
Reference
If you want to learn more about OpenCV with Raspberry Pi, I recommend visiting Adrian’s website below:
https://www.pyimagesearch.com/
I learned a lot from his website and I still refer back occasionally.