Contour-based Image Rectification

A small library to rectify masked images based on the mask contour. Given an image and the corresponding binary mask:

the library extracts the mask area and unwarps it based on the mask contour.

Usage

VS-Code Devcontainer

We highly recommend to use the provided Devcontainer to make the usage as easy as possible:

• Install Docker and VS Code
• Install VS Code Devcontainer extension ms-vscode-remote.remote-containers
• Clone the repository

  git clone git@github.com:FelixHertlein/contour-based-image-rectification.git

• Press F1 (or CTRL + SHIFT + P) and select Dev Containers: Rebuild and Reopen Container
• Go to Run and Debug (CTRL + SHIFT + D) and press the run button, alternatively press F5

Data

Mount your data folder to /workspaces/contour-based-image-rectification/data/YOUR_DATA_DIRECTORY.

Exemplary Usage

Load the image and the mask

from contour_based_image_rectification import load_image

image = load_image("../data/warped_document_multifold_color.jpg")
mask = load_image("../data/warped_document_multifold_color_mask.jpg")
mask = (mask > 128).any(axis=2)

Rectify the image

from contour_based_image_rectification import rectify

image_rectified = rectify(image=image, mask=mask, output_shape=(500, 1000))

Note, that the most processing time is required for estimating the 4 corner points of the mask. If you already have corner points, you can provide them to the rectify function:

from contour_based_image_rectification import rectify

corners = [(428, 520), (1552, 787), (1482, 1370), (270, 1087)]

image_rectified = rectify(image=image, mask=mask, output_shape=(500, 1000), corners=corners)

Affiliations

License

MIT License as declared by the licence file.