According to Stefan Savage, the computer science professor who led the project, advances in digital imaging have made it so easy that even a low resolution photo from a cell phone offers enough information to decode and reproduce a key.

How Sneakey Works

Using a digital image of a key from almost any angle, Sneakey measures the depth of each cut, strings together this information and spits out a bitting code - typically a five or six digit number - that locksmiths use to make each cut on a blank key. The bitting code, along with the basic key information is enough to make a duplicate key.

The steps in Sneakey's algorithm

1. The user provides point locations on the target key with a reference key as a guide.
2. The system warps the target image into the pose of the reference key and overlays markings of where the bite codes are to be found.
3. The user specifies where the cut falls along each line and the bit depths are decoded by the system into a bitting code.

The UCSD team is not releasing Sneakey to the public, but they point out that anyone with basic knowledge of MatLab and computer vision techniques could easily build a similar system.

While people usually mask credit card numbers in photos before posting them online, they don't realize they should be doing the same thing with their keys.

Something to think about before posting Friday's drinks night photos on Flickr.

UPDATE: Benjamin Laxton contacted us to let us know he had developed the algorithm and written the software for Sneakey, and Kai Wang helped with much of the experimentation. Thanks Ben