Algorithm pulls images of internet to fill in the blanks

by James

Algorithm fills in the blanks
Taking advantage of the millions, perhaps billions of photographs posted on the World Wide Web, James Hayes and Alexei A. Efros of Carnegie Mellon University have developed a search and replace algorithm which examines a photograph for missing pieces and then goes out on the internet to find similar photographs to fill in the blanks.

The algorithm examines a photograph for missing sections, bad lighting, even gaps taken out from editing the photograph and then completes the image with a choice of photographs found to match lighting and sharpness. And the coolest part is, it’s all automatic. The only thing the user need do is choose the matching piece of the puzzle they like best.

Then the algorithm patches the holes in a fashion that in seamless and invisible.

This development can help victims of such disasters as Hurricane Katrina try and put their ruined photographs back together, give historians another tool to peek back into the past of a given area they are studying photographically or allow heartbroken sweethearts to remove the source of their pain and still keep the memory of an experience.

Here’s hoping some nefarious shadow government isn’t using it to fudge some photographic evidence.

Write a review or comment

You may use <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong> in your comment.

Top Categories
Latest Posts
Subscribe to Newsletter