Most people use credit cards today as an alternative to purchasing items with cash. However, since these cards often carry a high credit limit, credit-card counterfeiting is a world-wide problem costing millions of dollars. Conventional methods of identifying credit cards through inspection of security features, such as numbers, the owner's name, or a hologram, can fail to detect a counterfeit card.
An optically nondestructive and data-non-intrusive system would greatly enhance credit card identification before transactions take place and could reduce monetary losses. We have demonstrated such a system with a hyperspectral imaging-based credit card verifier.
Since the microscopic detail of the embossed hologram on a counterfeit credit card is not exactly the same as the genuine one, we create angular-sensitive color spectra from the embossed hologram through a diffraction effect by positioning broadband light sources (e.g., LEDs) at different angles. (See diagram below.)
Because of the reflective grating characteristics of the hologram, the broadband light is diffracted into its corresponding spectrum in a two-dimensional space. By using a 2D digital color camera situated at a fixed angle of θcam, the diffracted 2D spectrum is captured and registered as red, green, and blue bands. At this stage, we use the smart feature to analyze the red, green, and blue spectral images via a mathematical neural network.
One prototype uses three off-the-shelf broadband white-light LEDs and a 320 x 240-pixel digital color camera. The electronic controller and processing unit is composed of an 8-bit microcontroller for controlling the operation of the LEDs and a typical notebook for analyzing all spectral images.
Photonics-based credit card verifier structure.
Si:broadband light sources.
This prototype shows a red display on counterfeit credit cards with a very low false rejection rate (FRR) of 0.39%. With genuine credit cards, the prototype shows a green display with a low FRR of 1.20%.
A second prototype, a compact, lower-cost version, is now commercially available.
Both systems are now used in two of the largest banks in Thailand.
S. Sumriddetchkajorn and Y. Intaravanne, "A credit card verifier structure using diffraction and spectroscopy concepts
," Proc. SPIE 7003, p. 700318, Strasbourg, France (2008). doi:10.1117/12.779148
S. Sumriddetchkajorn and Y. Intaravanne, "A hyperspectral imaging-based credit card verifier structure with adaptive learning," Appl. Opt. 47, pp. 6594-6600 (2008).
S. Sumriddetchkajorn and Y. Intaravanne, "Non-intrusive hyperspectral imaging-based credit card verifier with histogram analysis approaches," Proc. AsiaSense, pp. 122-128, Bangkok, Thailand (July 2009).
S. Sumriddetchkajorn and Y. Intaravanne, "Data-nonintrusive photonics-based credit card verifier with a low false rejection rate," Appl. Opt. 49, pp. 764-771 (2010).
- SPIE Senior member Sarun Sumriddetchkajorn is director of the Intelligent Devices and Systems Research Unit (IDSRU) at the National Electronics and Computer Technology Center in Thailand.
- Yuttana Intaravanne is an assistant researcher at IDSRU's Photonics Technology Lab.
Have a question or comment about this article? Write to us at email@example.com.