tkleyton / glassifier-api

Flask API that returns a stylized version of an image.

glassifier-api

Flask API that returns a stylized version of an image.

Usage

The provided requesting tool in request_glassifier.py opens an image path provided by the user, formats it and communicates with the server API. The returned object is a Pillow.Image object, which can be saved and/or displayed to the user.

See the mock code or # Example to see an example usage.

The API is hosted at https://glassifier-api.herokuapp.com.

Running in local server

To run the api in a local server, simply clone this repository and run app.py.

The server should be running at http://localhost:5000.

Installing dependencies

Before you run the server, make sure to have the requirements installed on your machine.

You can install from the requirements.txt file:

$ pip install -r requirements.txt

or manually:

$ pip install Flask-RESTful Pillow sklearn pandas requests

Alternatively, to run only the request tool:

$ pip install Pillow numpy requests

Example

from request_glassifier import request_img_glassifier


api_url = 'https://glassifier-api.herokuapp.com/api/glassifier'

my_img = 'milu.jpg'
filtered_image = request_img_glassifier('img/'+my_img)
if filtered_image:
    filtered_image.save('output/filtered_'+my_img)
    filtered_image.show()