top of page

About

Is there anything more frustrating than completing a puzzle and finding out that the last two parts are missing?

Today's technology can complement images with missing pieces even without prior knowledge of the whole original picture.

 

In this project we implemented a puzzle completion application. The application gets as an input the scanned missing puzzle, and completes the missing pieces in it.

The main process consists of two parts:
• Search for the missing image in an image database (DB)
• Completion of an image using an Inpainting algorithm


We found that the most successful method to complete the picture is "Exampler based Inpainting", and we have based our project on an article called "Shift-Map Image Editing".


Finally, we created a friendly graphical user interface that allows to experiment with the implementation of the algorithm. 

bottom of page