Introduction In a number of business transactions and access to privileged information, reliable and accurate verification of people is extremely important. Forensic science labs and identification units for criminal investigations routinely use fingerprints. More recently an increasing number of civilian and commercial applications like welfare disbursement, cellular, phone access, laptop computer login are either using or actively considering to use fingerprint based verification because of the availability of inexpensive and compact solid state scanners as well as its superior and proven matching performance over other biometric technologies.
Automatic verification methods based on physical biometric characteristics such as fingerprint or iris can provide positive verification with a very high accuracy. However the biometrics-based methods assume that the physical characteristics of an individual used for verification are sufficiently unique to distinguish one person from another. Identical twins have the closest genetics-based relationship and therefore the maximum similarities between fingerprints are expected to be found among identical twins.
Pattern Recognition The concept of pattern is universal in intelligence and discovery. For example, we perceive the colored lines on the walls of the caves at Lascaux, France, painted in prehistoric times, as mammals of prey. The patterns in biological data contain knowledge, if only we can discover it. Discrimination of signal patterns allows personal identification by voice, handwriting, fingerprints, facial images, and so on, as well as the recognition of speech, written characters, and scenes in images. It also includes the identification of military targets based on radar, infrared, and/or video images. Patterns exist in high-frequency electromagnetic scans of body chemicals and other organic chemicals, including DNA. The concept of classification involves the learning of likeness and differences of patterns that are abstractions of instances of objects in a population of non-identical objects. The associations between patterns and their causes are the bricks from which the wall of scientific knowledge is built.
Recognition Versus Classification
Humans recognize the familiar faces in a crowd, characters and words on the printed page, the different types of voices , favorite melodies, the scent of perfumes and fruits, pattern of weave in cloth, the shape of leaves, contextual meaning in word phrases, and so forth. Other mammals are also excellent recognizers. The senses preprocess signals such as sound or light waves that have been modulated-that is, transformed in some fashion by interaction with an object that impressed information on them. The preprocessed-modulated signals are then mapped into a decision that equates to recognition to recognition of the objects. Such processing detects subtle differences in the modulation of the signals to perform recognition. A pattern will be taken to be primitive here in that we agree on its meaning without being required to define it. When it is determined that an object from a population P belongs to a known subpopulation S, we say that pattern recognition is done. The recognition of an individual object as a unique singleton class is called identification. Classification is the process of grouping objects together into classes according to their perceived likeness or similarities. The subject area of pattern recognition includes both classification and recognition and belongs to the broader field of machine intelligence-that is, the study of how to make machines learn and reason to make decision, as do humans.
After having read in an entire sentence, the machine may be in a final state, in which case we say the machine accepts the sentences; or it may be in any other state, in which case we say that the machine rejects the sentence. If the machine accepts a sentence, then it classifies it as belonging to the class that the machine recognizes.
Fundamentals Information processing has always been under study through the history of science. Recently pictorial information processing has become increasingly important. The availability of digital computers and massive amount of pictorial data in all fields has made picture processing one of the major topics of current research. Formal language theory, which has an inherently hierarchical structure, seems to be suitable for picture analysis. Biometrics And Basics Of Fingerprint
Till date the research conducted in automatic fingerprint processing, concentrated on description of fingerprint impressions by determining the location of ridge endings, bifurcations, enclosures etc. Some of the researchers attempted to use optical techniques in fingerprint classification. But the drawbacks are noise and the requirement of numerous matching filters. At Cornell Aeronautical Laboratory (CAL) in addition to fingerprint reader, minutiae location and ridge direction i.e. mainly pattern matching is concentrated. There was also another proposal with the use of a sampling matrix where each sample square contains the direction of the predominant slope of the ridges passing deciding whether or not any two patterns are same through it. Later technique for smoothing the sampling matrix and extracting the global structure of patterns was introduced. Then syntactic approach as one of promising approaches for picture processing was proposed.
Dermetoglyphics
Dermetoglyphics is an interesting field in anthropology. Usually Dermetoglyphics analyses the total palm area and it is taken for comparison in different individuals. The analysts usually considered one area labeled thenar/first interdigital area. There are also second, third and fourth interdigital areas found in the distal palm in the region of the heads of metacarpal bones. Configurations encountered in the interdigital areas are loops, whorls, vestiges and open files. To get full understanding of fingerprint patterns, the definition some technical terms is in order.
Pattern Area The area of fingerprints containing the cores, deltas and the ridges is called a pattern area. Generally, a pattern area of loops and whorls can be easily defined while it is difficult, and sometimes impossible, to define arches.
Type line
Two inner most ridges confining the pattern area are called type lines. The type lines must start parallel, diverge and surround, or at least tened to surround, the pattern area. The immediate outside ridges is taken as the continuation of type lines. Bifurcation: Dividing of a ridge into two or more ridges are defined as bifurcation. Divergence: The spreading apart of two parallel or nearly parallel ridges are known as divergence.
Focal points:
Core and delta, two elements in the pattern, are used for classification and are called focal points.
Delta:
The point on or in front of the ridge closes to the divergence center of the type-lines is defined as delta point. Delta could be any element such as a dot, a short ridge, a bifurcation, the end of a ridge, and the intersection of two ridges.
Following rules are used in selecting the delta in the case of several possible choices: The delta may only be located at a bifurcation whenever the bifurcation opens towards the core. When there is a choice between a bifurcation and another type of delta, the delta on the bifurcation should be selected. When there are several possible delta points satisfying the definition, the point closest to the core is selected. No delta may be located in the middle of a ridge running between the type lines toward the core. It may be located at the nearest end of the ridge only.
Core:
The core, which is approximately the center of loop fingerprints, is located with in or on the inner most recurve. Whorls and arches do not have any cores.
Ridge count:
The number of ridges intervening between the delta and the core is called ridge count. Conclusion
Thus the research on Finger Print is very interesting and here in this article we have discussed what is recognition and classification, what are all the parameters for matching the finger prints, that is the fundamentals of finger print recognition.
