FINDERPRINT RECOGNITION IN INEXPENSIVE BIOMETRIC SYSTEM

Abstract

Nowadays a fingerprint recognition is the most developed biometric identification method. Each person has the unique papillary pattern which makes the identification possible. Typically, the algorithms parameters of the fingerprint recognition are the end of the line of the papillary patterns, their line branching and single points of the fingerprints. The features of the papillary pattern are converted into an unique code which is uniquely associated with the fingerprint. These fingerprint codes are stored in the database.

The advantages of the method are a quiet simple procedure of scanning fingerprints and a low cost of scanning devices.

The disadvantages are the sensitivity to tiny damages caused by cuts and scratches. Many scanners have a lot of issues with dry skin and skin of the old people. The number of “impassable” people can vary from less than percent up to the tens of percents for cheap scanners. However, modern fingerprint scanners are equipped with temperature sensors, force sensitive resistors, etc., which enhance the protection of the system from tampering.

Human and technical factors, hardware implementation of the biometric devices, recognition algorithms and technical realization of the system are very important for the process of fingerprint recognition. The human factor, that is, the method of attaching (pressing) a finger to a biometric scanner is an important reason for both reducing the speed of the recognition of the fingerprints and the false triggering. In particular, the registration of the characteristic points for the formation of a qualitative fingerprint pattern with oblique (crooked) finger attachment to the biometric scanner increases the speed of identification of a person.

The hardware implementation of biometric devices depends on the quality of the biometric sensor and the hardware device of the system. Particularly, when resolution is greater than or equal to 500 dpi, a fairly high quality fingerprint image can be obtained, followed by converting it into a digital model by characteristic points. The hardware platform with the built-in processor determines the speed of matching the fingerprint to the fingerprint pattern contained in the memory.

It was created an inexpensive biometric system of person identification by fingerprint recognition on the optoelectronic scanner and microcontroller Arduino base. We have obtained fingerprints in BMP-format and handled by improved algorithm. Work of system was tested in practice.

On the base of description technology it may be created the system of security defence, digital door lock for Smart House a system of work time calculation for office.

In our article it was given wiring scheme of optical scanner and microcontroller Arduino Uno. Also it was shown the block scheme of working system of fingerprint identification of person wich consist an optical system of scanning, butter of images, two buffers of convolution and library of templates.

Recognition time of person identification was changed from one to three sec in dependency of amount of fingerprint library.

Key words: fingerprint recognition biometric system, microcontroller, scanner, algorithm