I.I. Troitskiy – Ph.D.(Eng.), Associate Professor, 

Department «Information Security», Bauman Moscow State Technical University

E-mail: iitroickiy@mail.ru

M.A. Basarab – Dr.Sc.(Phys.-Math.), Head of Department «Information Security», 

Bauman Moscow State Technical University

E-mail: basarab@bmstu.ru

E.V. Glinskaya – Senior Lecturer, 

Department «Information Security», Bauman Moscow State Technical University E-mail: glinskaya-iu8@rambler.ru

When analyzing images and sound, the problem often arises of recognizing a discrete signal with K gradations in additive noise. In addition, in some cases, when there is a strict restriction on the range of variation of the additive mixture of signal and noise, we can assume that the noise has a uniform distribution law.

Currently, expressions are obtained for the probability of discrete signal recognition in additive normal and uniform noise for two signal gradations, and posterior probabilities of signal recognition by K gradations are also determined.

Previously, an analytical approach to increasing the probability of discrete signal recognition against additive noise of a similar type was considered by the authors and was based on the use of auxiliary channels with correlated noise.

It is proposed to use the amount of information according to Shannon as an indicator of the recognition efficiency of a discrete signal having K gradations in the presence of additive noise with a uniform distribution.

As an indicator of the efficiency of recognition of a discrete signal having K gradations, it is proposed to use the amount of Shannon information. It is shown that the information approach has the greatest sensitivity to the noise distribution law. The maximum value of the amount of information is analytically determined depending on the values of a priori probabilities of the appearance of signal gradations.

1. Gruzman I.S., Kirichuk V.S., Kosykh V.P., Peretyagin G.I., Spektor A.A. Tsifrovaya obrabotka izobrazhenii v informatsionnykh sistemakh: Ucheb. posobie. Novosibirsk: Izd-vo NGTU. 2002. 352 c. (in Russian)
2. Duda R., Khart P. Raspoznavanie obrazov i analiz stsen. M.: Mir. 1976. 510 s. (in Russian)
3. Troitskii I.I., Basarab M.A., Matveev V.A. Opredelenie veroyatnosti raspoznavaniya diskretnogo signala v additivnom shume dlya dvukh kanalov peredachi informatsii. Elektromagnitnye volny i elektronnye sistemy. 2014. T. 19. № 11. S. 40−44. (in Russian)
4. Troitskii I.I., Basarab M.A., Matveev V.A. Opredelenie veroyatnosti raspoznavaniya diskretnogo signala v additivnom shume dlya n kanalov peredachi informatsii. Elektromagnitnye volny i elektronnye sistemy. 2016. № 6. S. 66−70. (in Russian)