Kee et al., ACM Transactions on Graphics, 2013. Original image credit: NASA.
Fig. 1. Our algorithm finds that the shadows in this 1969 moon landing photo are physically consistent with a single light source. The solid lines correspond to constraints from cast shadows and dashed lines correspond to constraints from attached shadows. The region outlined in black, which extends beyond the figure boundary, contains the projected light locations that satisfy all of these constraints.
Are you ready for CSI: Photoshop?
This is Sandra Tsing Loh with the Loh Down on Science.
Sometimes, you know when a photograph’s been doctored. But other times, a photo’s so expertly altered, the deception’s hidden.
Computer scientists at Dartmouth have a solution. They've developed software that spots fakery undetectable to the naked eye! The software centers on a basic fact: There’s always a direct line between a light source and shadows.
But in photographs, which are two-dimensional, it’s not always easy to match shadow with source. Objects may appear to block light. Shadows may look distorted.
So the software scans digital photographs using geometry-based algorithms. It looks at sections of shadows, and creates wedges pointed to the apparent light source.
In authentic photos, the wedges should intersect. If not? Doctored image!
As a test, the researchers scanned a picture of the 1969 Apollo moon landing. Conspiracy theorists insist that photo is faked. But the software indicated a single light source. So the moon walk really happened!
If only that recent Cabo vacation hadn’t happened. With, unfortunately, the photos to prove it.
***** For more 90-SECOND SCIENCE FACTS, click here.*****
The Loh Down on Science is produced by LDOS Media Lab, in partnership with the University of California, Irvine, and 89.3 KPCC. And made possible by the generous support of the Gordon and Betty Moore Foundation.
Follow us on Twitter!