About the Project

The project involves training your Raspberry Pi for facial recognition of family and friends. Once trained, the system sends email notifications upon identification, offering a practical and secure application for enhanced home security and interactive scenarios.

Configuration

• Install and Compile OpenCV on Raspberry Pi: https://pimylifeup.com/raspberry-pi-opencv/

Notes

• Dataset: train_model.py will analyze photos within the dataset folder. Organize your photos into folders by person’s name. For example, create a new folder named Annie and place all photos of Annie face in the Annie folder within the dataset folder.
• Encodings: train_model.py will create a file named encodings.pickle containing the criteria for identifying faces in the next step.
• Detection Method: using the HOG (Histogram of Oriented Gradients) detection method.

Run source

# get data to train model
python3 headshots.py

# train model
python3 train_model.py

# test model 
python3 train_model.py

Demo

• Setting Raspberry Pi with USB Camera

  

• Get data to train model

• Test model