- Personal project
- Runs locally
- Aug–Sep 2022
Grow Garden
Draw a digit by hand; a CNN recognizes it and grows your virtual plant.
- My role
- Solo project.
- Key result
- Recognizes hand-drawn digits and turns them into plant growth in an interactive app.
- Technologies
- Python
- CNN
- OpenCV
- Skills shown here
What I owned
- Data preparation
- Modeling
- Application/API
In [1]:Problem and constraints
Make digit recognition something you can see and play with.
In [2]:Data and preparation
Handwritten digit images for training. The user's drawings are cleaned up with OpenCV before prediction.
In [3]:Architecture and modeling decisions
A CNN classifier, with OpenCV preprocessing so hand-drawn digits look like the training data.
In [4]:Implementation and integration
A drawing canvas, then OpenCV preprocessing, then the CNN's prediction, then the plant grows based on the recognized digit.
Out[4]:Results and evidence
- Recognizes hand-drawn digits and turns them into plant growth in an interactive app.
In [5]:Limitations, failures, and lessons
- Unusual handwriting can be misread.