Image Evolution with Python DEAP

Evolutionary algorithm written in Python using DEAP and Pillow to evaluate an image and mutate an array of polygons over generations to gradually increase fitness; the algorithm will generate a final image and graphed statistics after the set number of generations.

Requirements

• Python3.6+

Setup & Run

git clone git@github.com:James-02/image-evolution.git
cd image-evolution
python3.6 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python main.py 0

Add a new image to test against

• Add a 200x200px .png image to the media directory
• Update config.json's "Image" parameter:

"Image": {
        "TARGET_IMAGE": "media/target-image-1.png"
}

Run Configurations

python main.py <code>
        0 - Reference algorithm
        1 - Solution Representation
        2 - Offspring Generation
        3 - Selection Function

Example Timelapse

Timelapse was created externally using images generated by this algorithm