This document discusses fingerprint recognition using neural networks. It begins with an overview of fingerprints and their unique patterns. It then describes the components of a pattern recognition system for fingerprints, including image acquisition, edge detection, thinning, feature extraction, and classification. Neural networks are proposed for fingerprint recognition because they can learn from examples and process large amounts of data quickly. Other applications of neural networks discussed include character recognition, image compression, stock market prediction, and more. The document concludes by noting that fingerprints will continue to be a reliable biometric for human identification.
2. Fingerprints
A fingerprint in its narrow sense is an
impression left by the friction ridges of a
human finger.
Their pattern is permanent and unchangeable on each
finger during the whole life time of an individual.
The probability that fingerprints of two individual are
alike is about 1 in 1.9×1015.
According to FBI the accuracy and reliability of
fingerprint scans are correct 99.8% of the time.
4. Pattern recognition System
Image Edge Feature Classifier
Thinning
Acquisition Detection Extractor
Image Acquisition
Converting a scene into an array
of numbers that can be manipulated by a computer.
Edge Detection and Thinning
These are the part of preprocessing step which involves
removing noise, enhancing the picture and, if necessary,
segmenting the image into meaningful regions.
5. Pattern recognition System
Image Edge Feature Classifier
Thinning
Acquisition Detection Extractor
Feature extraction
The image is represented by a set of
numerical “features” to remove
redundancy from data and reduce its dimensions.
Classification
Class label is assigned to the image by examining its
extracted features and comparing them with the class that
it has already learned.
6. Why use Neural Network?
A neural network consists of an interconnected group
of artificial neurons, and it processes information and
help us to find solution.
There is no need to program Neural Network they
learn with the examples.
Neural Networks offers significant speed advantage
over conventional techniques.
7. Other Applications
Character Recognition
The idea of character recognition has become very
important as handheld devices like Palm Pilot are
becoming increasingly popular.
Image Compression
Neural networks can receive and process large
amount of information at once, making them useful in
image compression. With internet explosion and
more and websites using more and more images,
using neural networks for image compression is
worth a look.
8. Other Applications
Stock Market Prediction
The day-to-day business of stock market is
extremely complicated. Stock prices will go up or
down is the result of many different factors. Since
neural network can examine a lot of information, they
can be used to predict stock prices.
Travelling Salesman Problem
Neural network can solve the travelling salesman
problem, but only to a certain degree of
approximation.
Medicine, Security, and Loan Applications.
9. Preprocessing System
The first phase of the work is to capture the
fingerprints image and convert it into a digital
representation of 512×512 by 256 grey levels.
The binary image is further enhanced by a
thinning algorithm which reduces the image
ridges to a skeletal structure.
10. Preprocessing System
After obtaining the binary form of the
fingerprint image, there may be some
irregularities caused by skinfolds and spreading
of ink due to finger pressure, and so on…
The remedy to this problem is smoothing to fill
holes, delete unnecessary points, removing
noisy points and filling necessary missing points.
11. Application of fingerprint
Recognition
The fingerprint recognition system can be easily
embedded in any system. It is used in-
◦ Recognition of criminals in law enforcement.
◦ Used in providing security to cars, lockers, banks, shops.
◦ To differentiate between persons.
◦ To count the individuals.
◦ Drug detection.
12. Criticism
Despite the widespread acceptance of fingerprint
evidence, many question its worth due to a significant
amount of identification mistakes. There is a question of
its reliability and accuracy.
For example, in 2000, an individual was arrested for
murder and was told by police that fingerprint experts
matched his fingerprints to those found at the crime
scene. The individual's attorney hired his own fingerprint
experts, two former FBI examiners, who determined
that absolutely no positive identification took place.
After some post-incarceration legal wrangling, this
evidence was deemed sufficient for an acquittal.
13. Conclusion
For centuries fingerprints have been one of the most
highly used methods for human recognition; automated
biometric system have only been available in recent
years.
The advancement of technology have led to next
generation of fingerprint recognition devices which are
highly reliable and accurate.
Fingerprints have a broad acceptance with the general
public, law enforcement and the forensic science
community.
Hence, they will continue to be used for human
recognition and for new systems that require a reliable
biometric.