Ghost photographers, the ones that haunt great shots, can be vanquished, thanks to a new algorithm that removes reflections caused by glass.
Researchers at Google Research and the Massachusetts Institute of Technology (MIT) developed the algorithm, which can be used to clean up shot marred by reflections. The technique can be automated and processing only requires a single image.
To eliminate the ghost of the person behind the camera, the algorithm looks for duplicate, slightly off-set images such as those caused by double-pane windows or thick glass, according to YiChang Shih, lead author of the paper.
"With [double-pane windows], there's one reflection coming from the inner pane and another reflection from the outer pane," says Shih. "But thick windows are usually enough to produce a double reflection, too. The inner side will give a reflection, and the outer side will give a reflection as well."
However, the technique is relatively ineffective for thin and single-pane windows, Shih says. When the glass doesn't generate a double reflection, the outdoor image and the one reflected from indoors combines to form a picture that's too interweaved to untangle, he says.
"If A+B is equal to C, then how will you recover A and B from a single C? That's mathematically challenging," says Shih. "We just don't have enough constraints to reach a conclusion."
The technique is most promising for amateur photographers, who are less likely to jury-rig the setup professionals often use to mute reflections in their photos. So instead of wrapping dark cloth around their camera lens, which is more difficult to do with point-and-shoot cameras than the SLR, photographers can clean out the reflections in their images much like they'd use a sharpen or a noise-filter tool in Photoshop.
"The ideas here can progress into routine photography, if the algorithm is further robustified and becomes part of toolboxes used in digital photography," says Shih.
The algorithm may have bigger implications beyond cleaning up those accidental selfies. The technique could also be used in robotic vision, helping to reduce the confusion of caused by imagery reflected by glass -- still, that won't stop robots in the future from catching a glimpse of themselves in mirrors and becoming self-aware.
