The idea of this project was to create a virtual pen and apply a handwritten recognition model in the backend to recognize handwritten digits from a Raspberry Pi camera.
I have adapted the code created by the amazin people at LearnOpenCV: Create a Virtual Pen And Eraser with OpenCV to predict the written images.
Run conda env create
to create the virtual environment with the requirements from requirements.txt
.
Following the steps from the mentioned tutorial, run the following lines:
python step1.py
- remember to press the 's' key to save your color range image.
python step2.py
python step3.py
and now you can play and predict!
Please open an issue if you have ideas on how to improve this! Have fun!