ReAding KEeper (RAKE) is a reading assistant. It is made of a Raspberry Pi and a Camera attached to a reading lamp.
The camera is filming the book you read and feeds the video to the Raspberry.
We then use Mediapipe's Hands to detect hands and recognize hand gestures with a Tensorflow model.
-
Take a screenshot of an interesting passage in the book when you want it
-
The current gesture to take a photo is the following :
IMPORTANT : It should be done with the LEFT hand.
The gesture is not recognized on the right hand in order to minimize the risks of taking a capture when turning a page with your right hand.
- Detect when you turn a page to do statistics on your reading like speed.
- Take a screenshot gesture
- Take a screenshot OK
- Take a screenshot only of the zone of interest NOT OK
- Detect page turning
- Setup Raspberry camera
- Deploy code on raspberry
- Attach camera to lamp
- Compile data into a mobile app
The project runs with Python 3.7 ONLY
- (Optional) Setup a Python virtual environment
python3 -m venv ./venv
source venv/bin/activate
- Install required packages
pip install --upgrade pip
pip install mediapipe opencv-python tensorflow
- Run the app (main.py)
python main.py