19-04-2011, 10:01 AM
PRESENTED BY:
Abhishek R. Shetty
[attachment=12343]
Fingerprint scanners
The Basic Working
FINGERPRINTS
Built-in, easy to access identity cards
Fingerprints are unique
Ridges on fingers are meant to help us grip things better
Pattern of ridges depends on genetic and environmental factors
Used in crime investigation and security
FINGER PRINT SCANNERS
A fingerprint scanner system has two basic jobs –
1. It needs to get an image of your finger
2. It needs to determine whether the pattern of ridges and valleys in this image matches the pattern of ridges and valleys in pre-scanned images.
Common methods to scan finger – optical scanning and capacitance scanning
The end result of both types are the same, but the images are obtained through different ways
OPTICAL SCANNERS
Consists of a Charge Coupled Device(CCD) – an array of light sensitive diodes(photosites) which generate eletrical signal in response to photons
Each photosite represents a pixel
The Image is processed through an ADC to convert it to digital form
The CCD system actually generates an inverted image of the finger
Darker areas representing more reflected light (the ridges of the finger) and lighter areas representing less reflected light (the valleys between the ridges).
Scanner checks for clarity, sharpness and then checks for a match
CAPACITANCE SCANNERS
Electric Current is used instead of sensors
Sensor is connected to an integrator(built using op-amp)
The finger acts as a capacitor plate, with plate separation being different in the case of a ridge and a valley
To scan the finger, the processor first closes the reset switch for each cell, which shorts each amplifier's input and output to "balance" the integrator circuit.
When the switch is opened again, and the processor applies a fixed charge to the integrator circuit, the capacitors charge up.
The capacitance of the feedback loop's capacitor affects the voltage at the amplifier's input, which affects the amplifier's output.
Since the distance to the finger alters capacitance, a finger ridge will result in a different voltage output than a finger valley.
The scanner processor reads this voltage output and determines whether it is characteristic of a ridge or a valley.
By reading every cell in the sensor array, the processor can put together an overall picture of the fingerprint, similar to the image captured by an optical scanner.
The main advantage of a capacitive scanner is that it requires a real fingerprint-type shape, rather than the pattern of light and dark that makes up the visual impression of a fingerprint. This makes the system harder to trick.
Additionally, since they use a semiconductor chip rather than a CCD unit, capacitive scanners tend to be more compact that optical devices.
ANALYSIS
Overlapping and checking is not an effective method
Most algorithms only compare specific features – minutiae
Typically, human and computer investigators concentrate on points where ridge lines end or where one ridge splits into two (bifurcations).
The scanner system software uses highly complex algorithms to recognize and analyze these minutiae
The basic idea is to measure the relative positions of minutiae
Ex. If two prints have three ridge endings and two bifurcations, forming the same shape with the same dimensions, there's a high likelihood they're from the same print.
To get a match, the scanner system doesn't have to find the entire pattern of minutiae both in the sample and in the print on record
ALGORITHM
Before any algorithm is applied, the image is binarized(converted to B/W)
Then the ridges are made thinner which removes holes and insignificant details
Pixels are analyzed, small regions at a time, say 3 x 3
A very basic algorithm could label a pixel belonging to a ridge as 0 and valley as 1
There are tons of algorithms to match a given print against a database
Any matching paradigm is applied to get a match
MATCHING PARADIGMS
Manual
1. Human experts use a combination of visual, textural, minutiae cues and experience for verification
Image based
1. Utilizes only visual appearance
2. Requires entire image
Texture based
1. Treats print as oriented texture image
Minutiae based
1. Uses relative positioning of minutiae points
2. Very accurate and popular
Pros
Fingerprint scanners are “Who you are” systems
Advantages
1. Physical attributes are much harder to fake than identity cards.
2. You can't guess a fingerprint pattern like you can guess a password.
3. You can't misplace your fingerprints, irises or voice like you can misplace an access card.
4. You can't forget your fingerprints like you can forget a password.
Cons
Optical scanners can't always distinguish between a picture of a finger and the finger itself
Capacitive scanners can sometimes be fooled by a mold of a person's finger
The internet is filled with ways to beat a fingerprint scanner
Sources
How Stuff Works
Wikipedia
Ezine