Pattern Recognition in Biological Systems of Personality Authentication

There are considered some conceptual subjects of construction of mathematical apparatus for pattern
 recognition in the biological systems of personality authentication. Рассматриваются концептуальные вопросы построения математического аппарата распознавания
 образов в биологических...

Повний опис

Збережено в:
Бібліографічні деталі
Дата:2008
Автори: Glushchenko, V.E., Glushchenko, Yu.V.
Формат: Стаття
Мова:Англійська
Опубліковано: Інститут проблем штучного інтелекту МОН України та НАН України 2008
Теми:
Онлайн доступ:https://nasplib.isofts.kiev.ua/handle/123456789/7457
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:Pattern Recognition in Biological Systems of Personality Authentication / V.E. Glushchenko, Yu.V. Glushchenko // Штучний інтелект. — 2008. — № 4. — С. 215-219. — Бібліогр.: 2 назв. — англ.

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
_version_ 1860049031077560320
author Glushchenko, V.E.
Glushchenko, Yu.V.
author_facet Glushchenko, V.E.
Glushchenko, Yu.V.
citation_txt Pattern Recognition in Biological Systems of Personality Authentication / V.E. Glushchenko, Yu.V. Glushchenko // Штучний інтелект. — 2008. — № 4. — С. 215-219. — Бібліогр.: 2 назв. — англ.
collection DSpace DC
description There are considered some conceptual subjects of construction of mathematical apparatus for pattern
 recognition in the biological systems of personality authentication. Рассматриваются концептуальные вопросы построения математического аппарата распознавания
 образов в биологических системах идентификации личности. Розглядаються концептуальні питання побудови математичного апарату розпізнавання образів в
 біологічних системах ідентифікації особи.
first_indexed 2025-12-07T16:59:25Z
format Article
fulltext «Штучний інтелект» 4’2008 215 3G УДК 004.522 V.E. Glushchenko, Yu.V. Glushchenko East-Ukrainian National University named after Volodymyr Dahl, Lugansk, Ukraine vladimir@atoll.lg.ua Pattern Recognition in Biological Systems of Personality Authentication There are considered some conceptual subjects of construction of mathematical apparatus for pattern recognition in the biological systems of personality authentication. Setting up a problem One of major directions regarding increasing safety of information, processed and kept on objects of informative activity is development of the effective automated systems of checking and access control systems on these objects [1]. Efficiency of the indicated systems can be increased on the basis of improvement of the automated biometrical systems of recognition of personality as element of artificial intellect. Under the «biometrical system of recognition of personality» according to [1] is understood the system, providing recognition of man (personality) on the basis of authen- tication and analysis of the psychophysical state of personality, which is especially actually at the decision of task of control and access to the objects of informative activity with the high degree of secrecy. Both physical objects (apartment, building, territory) and informa- tive-telecommunication systems of the different functional purpose can act as an object of informative activity. The systems of recognition of personality are conventionally subdivided into simple, using one feature, and complex, carrying out authentication of personality on combination of features. The distinctive mark of the complex systems of authentication of personality is a neces- sity of realization of processes of measuring and treatment of features, having, as a rule, different nature, and, therefore, different scales of measuring. The system of recognition of personality makes an analysis of biometrical, beha- vioural and medical informative features presented by identified personality and statistical and dynamic images obtained on their basis; description of parameters, characterizing presented features (images); search of informative parameters, sufficient for correct recog- nition; description of images in space of transformable informative parameters, and also comparison of selected informative parameters with those kept in a database and taking decision on referring personality to the certain class (with some error, errors of the first and second kind). The basic task of the recognition of personality is a decision of task of classification. Сonstruction of description of the personality image on combination of features is a complicated task. It is explained by complication of providing necessary terms during measuring of features, by instability of parameters of personality, to strong influence of external factors both on personality and on the measured features. Glushchenko V.E., Glushchenko Yu.V. «Искусственный интеллект» 4’2008 216 3G Presently the most widespread approach of construction of this class of the systems is their presentation as a combination of subsystems, each of which is directed on measuring, processing and parametrization of one concrete feature. A decision-making on combination of features is made on the basis of statistical estimations of the errors levels of the deve- loped biometrical system, obtained on the stage of classification of personality on each of identifying features. The values of errors depend on the fact how effectively the basic structural stages are realized in the system of recognition, namely: high-quality preliminary processing of the measured features, optimum correlation of amount of the analyzed para- meters and their self-descriptiveness, applied decision rule. Decision of similar task using electrocardiogram , voice and iris of eye as features are considered in [1]. As practice showed, for complex biometrical system of recognition of personality, built on the principle considered higher, characterized by the following: necessity of using unique programming and hardware facilities on the stages of measuring, processing, paramet- rization of each feature; uniqueness of apparatus for depiction of the personality images; uniqueness of the knowledge base for description of images, algorithms of classification and authentication of personality. That is why, presently, one of the actual directions of increasing efficiency of the biometrical systems of recognition the personality of this class is development of new approaches to description of images and apparatus of their aut- hentication. The purpose of the article is presentation of mathematical apparatus for creation algorithms of pattern recognition, based on conception of decision-making in state space with using knowledge about the structure of this space. Description of basic material The fundamental concept of the offered apparatus is a concept “pattern”, under which are understood description of object, situation or phenomenon with the help of the preset system of features. Recognition represents process of decision-making on the basis of entering informa- tion about recognizable pattern, knowledge about the structure of space of descriptions and procedural knowledge of finding a co-ordinate decision in this space. This process has a polycyclic nature and will realize procedure of advancement and verification of hypotheses. The fig. 1 shows the chart of the intellectual complex system of personality recog- nition built on the basis of principles offered by authors. We will consider the main provisions of the conception of construction the complex biometric systems of the personality recognition offered by authors. Description of patterns The pattern is described with the help of the great number of features À. Connection between the pair of features a,b A, described by the followings binary relations: a) relation of equivalence (a-b):  = {(a,b) : z(a)= z(b)}; (1) b) relation of strict preference (a,b):  = {(a,b) : z(a)> z(b)}, (2) where z(i) is a value of feature of i in the preset scale. The description of pattern obtained in such a way represents linear quasiorder and has the appearance of line, reflecting organization of features A. Recognition of patterns in biological systems of authentication of personality «Штучний інтелект» 4’2008 217 3G The relation of equivalence is generated by breaking up of great number A as follows: a and bA are referred to one class, if ( ,b)  . Such description of pattern can be presented as follows : R = {k1, by 2, ..., to m}, (3) U m i ik 1 =A, (4) where ki is a class of breaking up of A, k i  k j = , i j. The classes of breaking up of R can be numbered so, that relation  coincides with the relation of succession. Thus, we obtain the well-organized breaking up. In the case when in breaking up of R there are no equivalent elements, i.e. all classes of breaking up are one-element, then relation  forms a linear order. Great numbers of all relations of linear quasiorder on A from the geometrical point of view form the space of linear quasiorders (SLQ), great numbers of linear orders are space of strict order (SSO). Ranging of elements of great number A we will interpret as points of these spaces. We will designate through QLM space of linear quasiorders with carrier A, A = M, where M is an amount of objects in a great number A. Let R = {R 1, R 2, ..., R m} and P = {P 1, P 2, ..., P m} are two breaking up of great number of À. Matrices r = 1 , , jiM jir ,  = 1 , , jiM ji are matrices of connection of ele- ments in breaking up of R and P respectively. Figure 1 – A chart of the intellectual complex system of personality recognition Let             ),( if,0 ~),( if 1, ),( if,1 , ji ji ji e ji (5) Then distance of dx (R, P) between the points of R and P is determined on a formula: .),( 1, ,,   M ji jijix prPRd (6) Decision- making Recognition Knowledge base Forming of appearance Scanning Parameter 1 Preliminary treatment Scanning Parameter 2 Preliminary treatment Scanning Parameter 3 Preliminary treatment Scanning Parameter n Preliminary treatment . . . . . . . . . Glushchenko V.E., Glushchenko Yu.V. «Искусственный интеллект» 4’2008 218 3G Consequently, dx(R,P) is the amount of the bitwise non-coincidence of elements of mat- rices r and p. Function of selection By the function of selection of the co-ordinate point in SLQ will call three: (QL, DQD, C:DQL), (7) where QL is an initial great number of possible descriptions of recognizable images QLQA, D is some part of great number of QD of all subsets of multitude QL, C:DQL is such relation, which for any subset of G from QL, D(G) T, T is a co-ordinate point. As a co-ordinate point of multitude the point T is taken, for which is adopted the just condition:  n i i n i i TTT 11   (8) Using apparatus of geometrical approach to finding group decisions, there were developed standard procedures of finding co-ordinate point in SLQ, which made basis of multitude M of grammar W and procedural knowledge of the biometrical system of per- sonality recognition. Researches of properties of different functions of selection with taking into account of admissible types of producing multitude G showed that the functions of selection for po- ints 1 S YXG met condition of inheritance, that is ,1G ,' GG  ,1' W )'()( GcRGcR  (9) and to the condition of omission: ,1G ,' GG  ,1' G ,)(:' RGcG  ).()'( GcGc  (10) If S R S R XXLG 11  , 1LG , the class of functions of selection for LG possesses with property of co-ordination: G , LGz  , )()()( zGczcGc  (11) therefore, the task of search of a co-ordinate decision in SLQ on the preset presentation of G in SLQ comes to selection and using procedures of finding point, meeting condition (8), on the basis of procedural knowledge of the system. Because the process of searching of a co-ordinate decision in SLQ does not depend on concrete data domain (DD), then for recognition of the preset multitude of objects the adjustment of system for concrete DD is needed. In this case the relevant subspace is selected in SLQ which is used as space of search of decisions for preset DD. In this space the classes Q = {Q1, Q2. … Qm} are selected concering descriptions of recognizable objects, standard descriptions of these objects are formed, which are considered as centers of relevant classes, the maximum and minimum points of each class are determined. The maximum point of class of Ql, Ql = {Q1, Q2. … Qt}, is the point Rl MQl, for which justly Pot Rl M  Pot Ri, ti ,1 , and minimum is a point of Rl mQl, Pot Rl m  Pot Ri, ti ,1 . The distance from the center of class of Rl to Rl M is called the distance to the high boundary of class of Ql, distance from Rl to Rl m – is the distance to the low boundary of class of Qi. Recognition of patterns in biological systems of authentication of personality «Штучний інтелект» 4’2008 219 3G The combination of classes of objects descriptions forms the space model of sear- ching decisions for concrete DD. Knowledge about the structure of SLQ allows to use only three characteristics in a database system during description of class: description of the class center and value of distances to the upper and lower of its boundaries. Conclusion The distinctive feature of the offered method consists in introducing “universal”, in certain sense, space of pattern descriptions for pattern classification on minimum of distance. There are developed standard procedures of decision-making on the basis of input data and knowledge about space of pattern description. Basic provisions of the offered approach will be considered and detailed in next papers of authors. Literature 1. Темников В.А., Пономаренко Л.В. Система распознавания личности как основа повышения эффек- тивности систем контроля и управления доступом // Вісник Східноукраїнського національного університету ім. Володимира Даля. – 2006. – № 9 [103] – C. 64-69. 2. Глущенко В.Е. Результаты исследований взаимного расположения образов в пространстве описаний с целью формирования знаний интеллектуальной системы диагностирования // Вісник Східноук- раїнського національного університету ім. Володимира Даля. – 2002. – № 6 [52] – C. 82-87. В.Е. Глущенко, Ю.В. Глущенко Распознавание образцов в биологических системах идентификации личности Рассматриваются концептуальные вопросы построения математического аппарата распознавания образов в биологических системах идентификации личности. В.Є. Глущенко, Ю.В. Глущенко Розпізнавання зразків в біологічних системах ідентифікації особи Розглядаються концептуальні питання побудови математичного апарату розпізнавання образів в біологічних системах ідентифікації особи. Статья поступила в редакцию 21.07.2008.
id nasplib_isofts_kiev_ua-123456789-7457
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
issn 1561-5359
language English
last_indexed 2025-12-07T16:59:25Z
publishDate 2008
publisher Інститут проблем штучного інтелекту МОН України та НАН України
record_format dspace
spelling Glushchenko, V.E.
Glushchenko, Yu.V.
2010-03-30T14:08:32Z
2010-03-30T14:08:32Z
2008
Pattern Recognition in Biological Systems of Personality Authentication / V.E. Glushchenko, Yu.V. Glushchenko // Штучний інтелект. — 2008. — № 4. — С. 215-219. — Бібліогр.: 2 назв. — англ.
1561-5359
https://nasplib.isofts.kiev.ua/handle/123456789/7457
004.522
There are considered some conceptual subjects of construction of mathematical apparatus for pattern
 recognition in the biological systems of personality authentication.
Рассматриваются концептуальные вопросы построения математического аппарата распознавания
 образов в биологических системах идентификации личности.
Розглядаються концептуальні питання побудови математичного апарату розпізнавання образів в
 біологічних системах ідентифікації особи.
en
Інститут проблем штучного інтелекту МОН України та НАН України
Системы принятия решений, планирования и управления. Информационная безопасность интеллектуальных систем
Pattern Recognition in Biological Systems of Personality Authentication
Распознавание образцов в биологических системах идентификации личности
Розпізнавання зразків в біологічних системах ідентифікації особи
Article
published earlier
spellingShingle Pattern Recognition in Biological Systems of Personality Authentication
Glushchenko, V.E.
Glushchenko, Yu.V.
Системы принятия решений, планирования и управления. Информационная безопасность интеллектуальных систем
title Pattern Recognition in Biological Systems of Personality Authentication
title_alt Распознавание образцов в биологических системах идентификации личности
Розпізнавання зразків в біологічних системах ідентифікації особи
title_full Pattern Recognition in Biological Systems of Personality Authentication
title_fullStr Pattern Recognition in Biological Systems of Personality Authentication
title_full_unstemmed Pattern Recognition in Biological Systems of Personality Authentication
title_short Pattern Recognition in Biological Systems of Personality Authentication
title_sort pattern recognition in biological systems of personality authentication
topic Системы принятия решений, планирования и управления. Информационная безопасность интеллектуальных систем
topic_facet Системы принятия решений, планирования и управления. Информационная безопасность интеллектуальных систем
url https://nasplib.isofts.kiev.ua/handle/123456789/7457
work_keys_str_mv AT glushchenkove patternrecognitioninbiologicalsystemsofpersonalityauthentication
AT glushchenkoyuv patternrecognitioninbiologicalsystemsofpersonalityauthentication
AT glushchenkove raspoznavanieobrazcovvbiologičeskihsistemahidentifikaciiličnosti
AT glushchenkoyuv raspoznavanieobrazcovvbiologičeskihsistemahidentifikaciiličnosti
AT glushchenkove rozpíznavannâzrazkívvbíologíčnihsistemahídentifíkacííosobi
AT glushchenkoyuv rozpíznavannâzrazkívvbíologíčnihsistemahídentifíkacííosobi