Global Sources
EE Times-India
Stay in touch with EE Times India
 
EE Times-India > Interface
 
 
Interface  

Fingerprint identification technology set for use

Posted: 01 Apr 2003     Print Version  Bookmark and Share

Keywords:Embedded 

/ARTICLES/2003APR/B/2003APR01_ID_CT_EMS_TA.PDF

By Douglas McArthur

BiometricsSensorGroupDirector

Fujitsu Microelectronics

America Inc.

Email: dmcarthu@

fma.fujitsu.com

As the cellular phone evolves to

become a mobile terminal that

provides access to personal and

corporate data from any loca-

tion, users need reliable secu-

rity to prevent unauthorized

access. New solid-state sensing

devices provide cost-effective,

reliableandsophisticatedphysi-

cal-access security for mobile

devices such as laptop comput-

ers, PDAs and cellphones.

Two types of solid-state fin-

gerprint sensing devices have

been introduced in the market

that will serve an increasing list

of applications. The first is a

single-touch sensor, which re-

quiresfirmpressureonafinger-

print-capturelocation.Thesec-

ond requires a rapid swipe of a

finger or thumb over the sensor

face, where a unique set of fac-

torsiscapturedquicklyforrapid

analysis and authentication.

Both types of sensors detect

the increased capacitance

present when a fingerprint

ridge is placed above sensing

capacitor pixel electrodes. The

pixelsinthesensorareapproxi-

mately 455m square and are

placed on a 505m grid. The

resolution of the capacitive

pixel array is slightly better

than 500dpi. The base technol-

ogy of the sensors is a standard

single-polysilicon,triple-metal

CMOS process with 0.505m

design rules.

The third layer of metal in-

terconnect, which forms the

capacitor pixel plates, is made

of titanium nitride and is cov-

ered with a layer of silicon ni-

tride, which is only 7000E

thick.Thiscombinationofhard

metal electrodes with an abra-

sive-resistant coating results in

a rugged sensor that can with-

stand years of repeated use.

Biometric finger-scanning

systems are more efficient and

dependable than the older

methods used by law enforce-

ment agencies because they

store only data about specific

points of the fingerprint in-

stead of the full image.

All biometric technologies

work with a four-stage proce-

dure.Thefirstistheacquisition

stagewherethedevicecaptures

abiometricfingerprintsample.

Specific, usually unique, data

are extracted from the sample

and converted into a template

using pre-established math-

ematical formulas or algo-

rithms.Enrollmentverification

extracts a minimum of seven

characteristic-matching points

out of the 30 to 40 points that

exist on an individual's finger.

These include bifurcations and

endings that create a minutia of

relations, which are defined as

distances between and among

the characteristic points.

At the time of enrollment,

the code is stored as a reference

template for subsequent au-

thentication by users. As a user

accesses the system, he or she

sweeps a finger across the sen-

sor area and the live scan tem-

plate is compared with the

reference template. The entire

process is completed in 1 or

2 seconds.

This leads to a match/no-

matchdecisioninwhichthesys-

tem decides whether the live

scan template contains suffi-

cient biometric data matching

the reference template. If not,

it fails, and the user is cleared

for a second sweep.

False rejection rates, in

which a good fingerprint is not

recognized, are no higher than

1 percent. False acceptance

ratesarealmostnon-existent,at

0.01 percent.

Fingerprint detection

Anindividual'sfingerprintcon-

sistsoftightlyspacedridgesand

valleys, which are detected by

the sensor's capacitive array

through the use of a standard

reference discharge current for

each pixel. Each pixel is

precharged to a reference volt-

age and then discharged by the

reference current. The rate of

change of the potential (volt-

age) on the capacitor plate is

proportionaltothecapacitance

seen by the capacitor plate,

through the relation:

A pixel that is beneath (high

C) a fingerprint ridge dis-

charges slowly, whereas a pixel

underneath a fingerprint val-

ley discharges rapidly. This dif-

ference in the discharge rate is

detected through sample

and hold (S/H) circuitry and

converted to an 8bit output. In

this manner, high sensitivity to

fingerprint ridges and valleys

can be achieved and excellent

raw fingerprint images can be

detected.

Complex software algo-

rithms identify fingerprints.

The software captures the raw

fingerprintimage,digitizesthe

image data, extracts the result-

ing minutia template and then

performs a test to see if the ex-

tracted minutia template

matches a candidate template.

Single-touch sensors and

sweep sensors differ in size and

cost. Single-touch versions are

large, with a typical array mea-

suring15mmx15mm,allowing

the rapid capture of even the

largest fingerprint or thumb-

print images. This type of sen-

sor is easy to use and provides

rapid transmission of a full fin-

gerprint at 500dpi, which is an

automatedfingerprintidentifi-

cation standard.

These sensors are now being

designed and perfected for fed-

eral, state and regional govern-

ment agencies and police de-

partments for their use in iden-

tification. They are also finding

their way into automotive ap-

plicationsforone-touchkeyless

entry systems and in newly

Fingerprint identification technology set for use

emerging homeland security

applications.

The sensor consists of a 256

x 300 column-and-row con-

figuration of tiny metal elec-

trodes. Each column is linked

to a pair of S/H circuits. The

fingerprint image is recorded

in sequence, row by row. Each

metal electrode acts as a single

capacitor plate while the con-

tacting finger acts as the sec-

ond plate. A passivation layer

on the device's surface forms

the dielectric between the two

plates. Placing a finger on the

sensor creates various capaci-

tive values across the array,

which are determined by

ridges and valleys of the finger-

print and connected into a full

image for authentication.

Sweep sensors represent a

new class of fingerprint-

acquisition devices that re-

quire the user to slide his or her

finger over the device. The ad-

vantages of sweep sensors in-

clude small size (e.g., Fujitsu's

MBF300 measures 3.6mm x

13.3mm) and low cost. These

devices are intended for em-

bedded authentication (secu-

rity) applications for mobile

devices, such as cellphones

and PDAs. Sophisticated im-

age reconstruction software

rapidly captures multiple im-

ages from the sensor at frame

rates approaching 2000fps and

stitches the resulting minutia

pattern together from the indi-

vidual frame data.

Information and authentication

There is no question that the

adoption of portable, low-cost

fingerprint-identification

technology will have an impact

on our lives. For example, in

the near future, a police officer

may stop a suspect in a high-

crime area and ask for a finger-

print sample (instead of an ID

or vehicle registration papers).

Matching engine diagram

The sensor captures a fingerprint and then digitizes the image data and extracts a template of data, or minutia, unique to the fingerprint.

The individual will place his or

her right hand's first, second

and third fingers on a sensor

attached to a wireless PDA.

This will provide immediate

FBI verification of the

suspect's identity and past

criminal offenses.

A less intrusive application

involves the harried cellphone

user. In this case, the

cellphone,whenturnedon,will

ask the user to go through a

quick authentication process.

The user will sweep his or her

finger over the sensor and will

be granted access to all of the

phone's functions. If it is the

wrong user, the phone stays

locked down. If repeated at-

tempts to authenticate do not

work, the cellphone erases

critical information from

memory and powers off.

Anothertypicalmobileappli-

cation, voice-mail, could be-

come even easier to use. After

dialing the voice-mail number,

a single sweep of the finger

could authenticate the user to

thesystem.Fingerprintauthen-

tication can replace the need to

key in any mailbox or personal

identification numbers.

In the automobile of the fu-

ture, users may enter finger-

print identities of family mem-

bers allowed to drive the family

car. The enrollment process

will be simple: each of the au-

thorized drivers places his or

her finger on the sensor and

thensetsthevariousfeaturesof

the car to the driver's personal

preferences. The settings are

then stored in the car's com-

puter memory.

Whenadriverentersthecar,

he or she initiates the authenti-

cation process by placing a fin-

ger on the sensor. In less than a

second, the fingerprint tem-

plate is compared with the

storedtemplates,andamatchto

the driver is made. The finger-

print templates and matching

softwarewillresideinanembed-

ded module in the car. After a

match is detected, all internal

preprogrammedsettingsareap-

pliedtothecar.Themirrors,car

seat, radio stations and climate-

control preferences go into ef-

fect. Speed can also be con-

trolled--not a bad idea if there is

ateenageraround,forwhomthe

maximum speed could be lim-

ited to 55mph.





Comment on "Fingerprint identification technolog..."
Comments:  
*  You can enter [0] more charecters.
*Verify code:
 
 
Webinars

Seminars

Visit Asia Webinars to learn about the latest in technology and get practical design tips.

 

Go to top             Connect on Facebook      Follow us on Twitter      Follow us on Orkut

 
Back to Top