Perry & Geisler, University of Texas at Austin, Technical Paper CPS-TR-13-001.
Denoising algorithms applied to an image of a bell pepper. Top left: original. Bottom right: de-noised.
For bigger and better digital photos? A free fix is here!
This is Sandra Tsing Loh with the Loh Down on Science.
Digital photos, easy to take, outcome unclear. Maybe the lighting is wrong or the image is fuzzy. Such imperfections, or "noise," can turn a masterpiece into a mess! Most "de-noising" software takes the same approach to every photo. Which is limited.
So psychologists at the University of Texas at Austin took a radically different approach. Inspired by how the human brain interprets visuals in the real world, they wrote a new program that learns as it goes.
Their de-noiser has learned from thousands of snapshots what's noise, and what's not. With every photo analyzed and improved, the computer program gets smarter and better.
They call it “image processing by natural scene statistics.” I call it a gift to friends who grouse about my Instagrams.
You just upload your photos and the program enhances or enlarges them. For free!
If you want to try it out, we've got a link to it on our site. It's called—duh—image processing by natural scene statistics. Try it. You won’t need statistics to see how many of your friends will thank you.
***** For more 90-SECOND SCIENCE FACTS, click here.*****
Follow us on Twitter!