Evaluation of certain signs and their collectively for individual recognition objects

Viktor Yevetskyi

Abstract


The article discusses approaches to possibility estimation of various features for the individual object recognition. On the basis of the proposed recognition parameter, which allows to estimate informatively of feature, as well as the use of assumptions about normal distribution of features and Neyman-Pearson criterion the formulas were obtained to assess the result of using of different feature groups for the individual object recognition. As an example the results of calculation the probabilities of object correct recognition for variety combinations of several features are given in the article.

Keywords: object recognition, the individual object recognition, object attribute, feature space, recognition parameter, the probability of correct recognition, the probability of false detection.


References


Fainzilberg, L. S. (2010), Matematicheskie metody otcenki poleznosti diagnosticheskikh priznakov [Mathematical methods for evaluating the usefulness of diagnostic features], Osvita Ukrainy Publ., Kyiv, 152 p.

CHaban, L. N. (2004), Teoriia i algoritmy raspoznavaniia obrazov [Theory and algorithms of pattern recognition], MIIGAiK Publ., Moskow, 70 p.

Yevetskyi, V.L. (2012), Povyshenie effektivnosti zashchity dannykh za schet biometricheskoi identifikatcii polzovatelia [Improvement of the efficiency of data protection by biometric identification], Information technology and security, No. 2 (2), pp. 14-16.




ISSN 2411-1031 (Print), ISSN 2518-1033 (Online)