PHOTOMOSAIC

Final Project for NYU Large Scale Web Applications taught by Professor Yair Sovran
Group Members: JD Choi, Zachary Lin, Melissa Lopez, and William Chen

Project Objective

To build a Django application with an image classifier that allows the user to:

1. Have their images classified by dominant color and stored in our file system
   
2. Upload a source image and generate photomosaic from it (using images already in our system)
   

Core Technologies Used

• Google Cloud Platform --> (deployment)
• Google Cloud Storage --> (image storage)
• Python Imaging Library (PIL) --> (image classification + mosaic construction)
• NumPy Library --> (image classification)

Code Organization

/mysite --> boilerplate Django app files
/myvenv --> virtual environment files
/photomosaic --> application (including image classification by color + mosaic code)

Image Classification + Mosaic Construction Files

• color_id.py
  • Returns a color classification tuple of the form: ( 'color_name' , (R, G, B) )
• get_mosaic.py
  • Returns "mosaic map" representing the colors of tiles needed for the final photomoasic
• get_tile_requirements.py
  • Normalizes size and aspect ratio of source image + tile images
• collage.py
  • Constructs final moasic based on mosaic map