A.N. Ryubakov - Leading Engineer, VNIIA named after N.L. Dukhova E.V. Egorova - Ph.D. (Eng.), Associate Professor, Department of Telecommunication Systems, MIREA (Moscow). Е-mail: calipso575@gmail.com V.V. Vetrova - Ph.D. (Eng.), Associate Professor, Department of Telecommunication Systems, MIREA (Moscow)

The ability of perception of the external world in the form of images allows a certain authenticity to recognize an infinite number of ob-jects on the basis of acquaintance with a finite number, and the objective nature of the basic properties of the images allows to model the process of their recognition. In general, the problem of recognition consists of two parts: learning and recognition. Training is pro-vided by the demonstration of individual objects, indicating their affiliation to one or another way. As a result the system must acquire the ability to respond to the same reaction to all objects of the same image and different - to all objects of different images. Za training should be the recognition process of new objects, which characterizes the actions already trained system. Automation of procedures of teaching and learning and teaching is a problem pattern recognition.

 

1. Obrazovatelnyjj resurs.http://www.codenet.ru/progr/alg/ai/htm/gl3_1.php.
2. Cunikov A.JU. Obrabotka izobrazhenijj v infrakrasnom diapazone spektra. JAroslavl: Remder. 2011. 60s.
3. Meer P., Mintz D., Rozenfield A., Kim D.Y. Robust regression methods for computer vision // Int. J. Comput.Vision. 1991. V.6. № 1. P.59-70.37.
4. Cunikov A.JU., Egorova E.V., Viktorov A.S., Matveev A.JU., Nefedov V.I. Osobennosti obrabotki teplovizionnykh izobrazhenijj // EHlektromagnitnye volny i ehlektronnye sistemy. 2015. №4. S. 56-60.
5. Pottosin I.K. K obosnovaniju algoritmov optimizacii programm //Programmirovanie. 1973. № 2. S.3-13.