Let's restore a damaged image

In this exercise, we'll restore an image that has missing parts in it, using the inpaint_biharmonic() function.

We'll work on an image from the data module, obtained by data.astronaut(). Some of the pixels have been replaced by 1s using a binary mask, on purpose, to simulate a damaged image. Replacing pixels with 1s turns them totally black. The defective image is saved as an array called defect_image.

The mask is a black and white image with patches that have the position of the image bits that have been corrupted. We can apply the restoration function on these areas.

Remember that inpainting is the process of reconstructing lost or deteriorated parts of images and videos.

From scikit-image, import the module restoration and the inpaint function.

Take Hint (-10 XP)

Exercise

Let's restore a damaged image

Instructions 1/3