image-upscaling-api

A simple Python wrapper for the image-upscaling.net upscaling API.

Installation

pip install image-upscaling-api

Documentation

• Full API Documentation: image-upscaling.net/api.html

• Help on Model & Scale Config: get_upscalers_config

• Client ID: 32 digits, 0-9, a-z, A-Z
  You can find your client id here: image-upscaling.net/account
  If you have an account with positive balance, you need to use your linked client id from the account page to use your balance in this api.

Examples

1. Basic File Upscaling

• Uploads an image from disk and saves it back to disk.
• upscaled by 4x using the "plus" model
• The upscaler encodes your image as png but you can save it as any opencv supported extension.

from image_upscaling_api import UpscalerAPI

api = UpscalerAPI(client_id="a9898cb899c1542468b8555e59dfa331")

# Upload (scale: 4, model: plus)
img_id = api.upload("baboon.png", model="plus", scale=4, face_enhance=False)

# Download and save as jpg
api.wait_and_download(img_id, save_path="./baboon_upscaled.jpg")

2. OpenCV & Diffuser

• uses opencv images for upload and download
• uses diffuser model with fixed output resolution (-1 = 4MPx)
• no save_path -> image is returned as opencv image

from image_upscaling_api import UpscalerAPI
import cv2

api = UpscalerAPI(client_id="a9898cb899c1542468b8555e59dfa331")

image = cv2.imread("photo1.jpg")

# Upload with diffuser-lite and custom prompt
img_id = api.upload(image, model="diffuser-lite", scale=-1, prompt="", creativity=0.1)

# Download back into numpy array
upscaled = api.wait_and_download(img_id)
cv2.imwrite("upscaled_photo1.png", upscaled)

Github

Find the source code here: [dercodeKoenig/image-upscaling.net_API]

License

MIT