A Novel Review on Biometric System Based on Digital Signal Processing | Open Access Journals

ISSN: 2319-9873

A Novel Review on Biometric System Based on Digital Signal Processing

Divya Jyothi Uppari*

Jawaharlal Nehru Technological University, Hyderabad, Telangana 500085, India

*Corresponding Author:
Divya Jyothi Uppari
Jawaharlal Nehru Technological University, Hyderabad
Telangana, 500085, India
Tel: +91 9590807049
E-mail: divyauppari@rediffmail.com

Received Date: 30/07/2016; Accepted Date: 16/08/2016; Published Date: 27/08/2016

Visit for more related articles at Research & Reviews: Journal of Engineering and Technology


Biometric system is a device used to identify the individual based on physiological or behavior characteristics. The unique identification can be measured either through fingerprints, facial, eyelids, iris, sound, Heart beat etc. The use of biometric system used in day to day life. The management of the system is to maintain security regarding their need of work. In the present review we just tried to present how the biometric system works and used in daily life.


Biometric system, Identification, Security


Biometrics is programmed method to verify or recognize for identification of a living person depending on a physical or behavior characters [1]. This system varies depending on the requirement in other terms it can be purpose, procedures and technologies. But it is utilized by us in our daily life in one or the other way. This review explains about the technology of Biometrics and its application in the present society [2,3].

This science is basically depends on the human identification of the humans [4]. Biometrics is machine language that can make identification physically. Authentication is carried out by machine or either by digital appearance. We generally observe that collection of finger prints, DNA, hair analysis could not be a part of this system [5-7]. The collection of data can be manipulated and accuracy can be minimized. The identification of person to person is also difficult. So, Biometric system is useful for the accurate authentication with a quick approach and result [8,9].

As per the advance technologies this is the improvement is carried out by information technology. This is been observed in our daily life as shown in Figure 1.


Figure 1: Biometric system used in daily life.

Biometric system is a technical system which collects data regarding a person (or other biological organism) to identify that person, depending on exact data about unique biological traits which carry out the work efficiently [10-14]. This system would involve ongoing data by an algorithm for exact result, it is connected to identification of a user or other individual. The recognition of the individual can be through finger prints, eye lids, facial, heart beat (advanced) [15-17].


As per the literature in 1870 to identify this humans and notify dates a measuring system is been invented by Alphonse Bertillon. This system is been designed in such a way to measure the diameter of the skull, hand (arm) and length of feet. In 1920, system was been utilized in USA to identify the thieves. During 1880’s it was slowly developed to identify by finger prints and through measurements of face. While in 1960 the Digital signal processing is been introduced it has given a good support to identify humans identity easily, extending this technique measurement of hand is also notified [18-24]. By the government interest this system is utilized to identify the pupils. Not only this in 1980,s verification is carried out by retina and signature. This is improved in 1990s to recognize the Iris.


Components of biometric system

These devices depend on usually different technologies. Biometric system divided into five subsystems: Sensor (as an input), Signal Processing algorithm, Data storage, matching algorithm and Decision process. These subsystems can be considered at once [25-28].


The screening of the input is initiated with the help of sensors. This is to assemble data and translate information into digital format. Sensor play a major role collect data such as finger prints, eye lids, facial recognition etc. depending on the requirement [30,31].

All biometric systems assemble the data at one position but store and process it at other location. System must require data transmission that great amount of data involve, conversion is required to transmit or storage to save bandwidth and space to store. Conversion and transmit can occur before processing signal and store image [32-35].

During the transmission or compressed data to store should be helpful for further use. In the procedure of conversion compression of data is done and expanding may cause quality loss to restore the signal. This can influence the compression ratio. This conversion may be done depending on the signal generated by biometric system. Researcher has designed to convert the unique signal for collected data [36,37]. It may vary depending on the input. For face it is wavelength scalar quantization, facial images are JPEG format and Code excited linear prediction is done to store voice data.

Signal processing algorithm

Algorithm in the designed in such a way to convert the input data in digital form signal. It makes an easy path to match the data already stored in the system. This plays a crucial role that a template is modified according to registered data. Conversion of this data can be helpful to match the stored data [38].

By obtaining and feasible to transmit a biometric character, preparation should be done to match between other like measure. This divides the signal-processing four tasks such as segmentation, feature extraction, quality control and pattern matching. Segmentation is a process to find the pattern for biometric system to transmit the signal. For example, a facial recognition [39,40] is made to recognize the boundaries or measures of the face or faces that converted to digital code from image. Sounds can be predicted depending on the wavelength periods, this means the extraction of data is carried out. Extraction of feature is interesting and biometric pattern, segments from the greater signal, covers no repeat alters caused to exhibit, sensor and processes of the system to transmit [41,42].

The altered or destructive elements should be eliminated from the pattern in the biometric system. During the preservation of quality for both duplication and distinct templates can be featured in mathematical form. In extraction a vowel of speaker is recorded depending on the vowels frequency of speech can be estimated in decision making. This data can be correlated with Algorithms and proceed for next feature extraction [43,44]. It is an irreversible system i.e image cannot be extracted if once it is constructed. As per require bandwidth transmission initiated after feature extraction. Once in the data collection quality of system does not vary even though it is modified into good quality [45,46].

The Template is to represent storage data, these features of the same type of a sample. The featured sample is Vector in mathematical notation [47-50], the stored template can also be known as Vector. The construction of template in data storage is Model producing features specific of a particular user. Models and features will be of different mathematical types and structures. Some models can be utilized to recognize voice and face [51-54]. Templates are prepared based on fingerprints, iris, and hand geometry recognition systems.

Enrollment is to place template or model into the database. Once in the database and coordinated with an unique identity by information externally, the biometric enrollment data is as the template or model for the pupil that it is referred. Generally many authentications can be carried by magnetic strip. To identify here the templates are matched calculating the distance in the pattern. In the model and sample of similar person, distances would be rare, be zero, if ever, as there would be always a non-duplicate biometric, result , sensor or transmission- that are related which vary remains after processing [55-64].

Data storage

The collection of registered templates is stored in Data storage. This stored data can been helpful for identification of the task given by the system [65]. It make easy to process to identify the data and generate output to recognize the task given by the system (Figure 2).


Figure 2: Data collection to identify.

Collection of data in Biometric systems is a measure that depending on behavioral/physiological character [66-68]. It can be measured and used for distinctive individual and repeat several times. The troubles to measure and control this variation initiated at data collection subsystem.

User character should be designed for a sensor. The output of a biometric character in the sensor explains a behavior such as psychological component for all method in biometric system [69]. By the behavior components which might change in users, applications and test of the operation. Depending on the following character the output result can be generated.

The measurement of biometric input,

The method to generate output,

Technical conversion of the sensor.

Both duplicate and distinctive measure is negatively influence the change for any of this factor. Every system should maintain the unique standardized character to collect and represent the output. Collection of data should not coincide with data result in the given task. A proper management should be done to correlate with the previous records [70-75].

Matching algorithm

In the system templates are designed in such a pattern that can recognize the data and generate the output. This template can be matched depending on the minutes. Depending on the maximum minutes task can be completed to identify [76-78].

Decision process

This is the final component that gives the output result. After the matching is done depending on the minutes score the results can be made. If is the score is maximum then the result is yes and output is matched. If the score is low then the result is no and displayed as mismatch.

The decision process is implemented by the system from searching the database and analyze by “matches” or “mismatches”. This is done on the basis of similarity (or) distance or either through measurements that are collected from the pattern match, and ultimately makes decision to either accept or reject decision depending on the system [79,80]. As per the system, reject the identity claim of any individual’s pattern is not acquired. Including the attained pattern, the program may confirm the match by a distance less than a fixed value and “accept” a user identity to claim can also depending on the single match or the program can declare a match of a distance lower than a user-dependent, time-variant, or environmentally linked required match in multiple measure for a decision to “accept”. When the number is lower that can be made as false non-matches, against flying the number of false matches [81-83].

Uses of Biometric system

In many of government, businesses and organizations biometric system are used to analyze information regarding individual. These are improved because of security [85-89]. In airports device scan is done based on password such as bio-password to the system, or in data assemble procedure is an example of a biometric system that uses identifying data for a security result. In India Unique Identification Authority of India (UIDAI) has implemented Aadhar card for unique identification number for individuals. In figure 3 biometric systems used for data collection to generate Aadhar card [90-95].


Figure 3: Biometric Systems used in Aadhar cards.

Biometric systems are utilized by the employees to register their attendance. Recently Indusland bank has implemented fingerprints as passwords for transaction. Hope there would be more advancement in the biometric system and useful for individuals [95-100].