Professional Documents
Culture Documents
Abstract—ATM (Automated Teller Machine) systems have personal authenticity and to detect impostors is growing in the
become instant cash providers for people around the globe and present scenario of global security concern. Development of a
have increased in vast numbers in the past few years. At the biometric system for personal identification.
same time, occurrences of robbery have also increased in the Face recognition has been evolving as a convenient
present. The existing security in these systems is lagging behind biometric mode for human authentication for more than last
in providing fool-proof security. This paper presents a secure two decades. Basically Face Recognition is the process
approach in rendering money with the help of a combination of
through which a person is identified by his facial image. With
face recognition and SIM (Subscriber Identity Module)
identification technologies. The system recognizes the face of the the help of this technique it is possible to use the facial image
user with the help of image processing and GSM modem is used of a person to authenticate him into any secure system.
to send SIM identity to the server database. The system is also Wavelet provides powerful signal analysis tools, which is
equipped with voice annunciation to provide a user-friendly widely used in feature extraction, image compression and de-
approach. noising applications. Wavelet decomposition is the most
widely used multi-resolution technique in image processing.
Keywords—ATM, GSM, LPC2148, Face-Recognition, SIM ID, Due to the excellent time-frequency localization
FAR and FRR characteristic, Wavelet transform provide powerful signal
analysis tools [19]. Images have typically locally varying
statistics that result from different combinations of abrupt
features like edges, of textured regions and of relatively low-
I. INTRODUCTION contrast homogeneous regions. While such variability and
Hundreds of new techniques of image recognition have spatial non-stationary defies any single statistical
been developed over the last few years and are waiting for characterization, the multi-resolution components are more
actual applications. Everyday actions are increasingly being easily handled. Wavelet transform can be performed for every
handled electronically, instead of with pencil and paper or scale and translation, resulting in continuous wavelet
face to face. This growth in electronic transactions has transforms (CWT), or only at multiples of scale and
resulted in a greater demand for fast and accurate user translation intervals, resulting in discrete wavelet transform
identification and authentication. Access codes for buildings, (DWT). Since, CWT provides redundant information and
banks accounts and computer systems often use PIN's for requires a lot of computation, generally DWT is preferred.
identification and security clearances. Using the proper PIN Having processed a face and extracted the features, these
gains access, but the user of the PIN is not verified as arrest ored or transmitted as a facial code (face template),
authentic. This paper proposes the advantage of using face which can be as small as 84 bytes (Visionics). For each
recognition technique for verification of genuine user. The representation type, a distance or similarity measure is defined
information related to the customer is given to the bank that allows ‘similar’ faces to be determined. Much of the art in
including his face image, and then bank will provide an SIM biometrics is in the design of a model of the biometric data
ID to the user. The SIM ID acts like atm card for the customer and, given a scheme for extracting the model parameters as a
for money transaction. Hence from this system the card holder representation of the data, in creating a similarity measure that
only can take money from the bank thus enabling reliability correctly discriminates between samples from the same person
among user. This system proposes two approaches. First is to and samples from different people. As with any biometric
provide authentication of the user which is performed by the system, some threshold on similarity must be chosen above
SIM ID and ARM controller. Second approach is recognition which two face images are deemed to be of the same person.
of user identity by face recognition algorithm. Altering the threshold gives different False Accept and
False Rejection Rates trading the one off against the other
depending on the security level required. This is a trade-off
II. SPECIFICATION OF FACE between convenience and security: user-friendly matchers
RECOGNITION have a low false reject rate, while secure matchers have a low
false accept rate.
The importance of utilizing biometrics to establish
principles. The RISC instruction set and related decode
mechanism are much simpler than those of CISC designs.
III. ARM CONTROLLER
ARM architecture has a unique combination of features
The ARM (Advanced RISC Machine) architecture is a
that makes ARM very popular embedded architecture today:
32-bit RISC processor architecture developed by ARM
Limited that is widely used in a number of embedded designs. • ARM cores are simple compared to most other
Because of their power saving features, ARM CPUs are general-purpose processors.
dominant in the mobile electronics market, where low power • A typical ARM chip contains several peripheral
consumption is a critical design goal. Today, the ARM family controllers, a DSP and some amount of on-chip
accounts for approximately 75% of all embedded 32-bit RISC
memory.
CPUs, making it one of the most widely used 32-bit
• ARM ISA and pipeline designs are aimed to
architectures in the world. ARM CPUs are found in all corners
minimizing energy consumption.
of consumer electronics, from portable devices to computer
peripherals. The Fig.1 shows the 32-bit microcontroller • ARM architecture is highly modular: the only
growth. mandatory component of an ARM processor is the
integer pipeline. All other components, including
caches, MMU, FP unit and other co-processors are
optional.
VII. CONCLUSION
Implementing face recognition technique in the banking
domain verify the originality of the user identity and only
authenticated user is allowed to use the SIM ID based
automated teller machine.
In the future, we plan to increase the robustness of the
proposed method, especially to improve comfort ability of
face recognition