Smartphone tech from NEC can picture fakes

The Japan-headquartered technology firm NEC has unveiled a new system that can detect even the most convincing counterfeits by reading microscopic patterns on everything from a luxury purse to a metal bolt using the camera from a smartphone.

The technology can be also be used to trace the origin of mass-produced offerings by reading so-called 'object fingerprints', or three-dimensional surface irregularities, the firm says.

It could also allow a customs official, for example, to snap a smartphone picture of a specific spot on an object which is then instantly matched (or not) to a manufacturers' pre-registered image.

At a demonstration in Tokyo earlier this week, NEC staffers fitted 3D-printed close-up lens attachments, developed by NEC, to standard smartphone cameras. Then they photographed objects, such as the zipper pull tab on a luxury hand bag and the head of a screw, in extreme close-up.

By querying a cloud database of images and related information through an NEC smartphone app, the system ran a pattern-matching check and quickly identified the parts to confirm their authenticity.

"You can identify offspring that come from the same parental mould," explains Toshihiko Hiroaki, assistant general manager at NEC's Information and media processing laboratories. 

The system can be used to find pirated goods, to trace the origin and distribution through the marketplace of authentic goods and to manage components in industrial applications such as maintenance and repair work, making sure they're being used correctly.

A genuine article can be matched with the time and location where it was produced, NEC says.
Hiroaki noted that the trade in counterfeit goods is estimated to reach into the hundreds of billions of dollars a year, and that a fake or defective part could have serious consequences for finished products.

The technology, which can also work with tablet cameras, allows for objects to be identified at low cost and without special processing, according to NEC.

The company says it is looking for partners to commercialise the object fingerprint technology in its 2015 fiscal year, which runs to March 31, 2016.