V. S. Titov, M. I. Trufanov, S. I. Korostelev, A. V. Medvedev

At present devices for recognition of objects images are used in different spheres of man activity. In this paper we present the device for object recognition. The main idea of the method for objects recognition, realized in the giver device [3], is not the use of a set of object plane projections on the image, as a standard, obtained from different points of observations of the object, but the use of its 3D representation in the vector form. Each time a number flat images is generated until there may be found coincidence. Solution of the giver problem by the way of direct sorting of all the variants leads to greater machine time consumption, that is why it is suggested to use preliminary processing as a result of which a set of identifications is separated on the basis of typical features (such as location in the space, dimensions), reducing complexity of recognition. There are also present the structure and principles of functioning of the device. A benefit of the developed device is considerable reduction of the quality of required memory and simplification of teaching process. In this case teaching is reduced to adding of characteristics of the object 3D model into the library, while in present systems of technical view it is necessary to form standards library, taking into account position of the object in the space, and at parametric approach to teaching it is necessary to use complex mathematical instrument. This device may be used in industry, at organization of converter machining of parts and also in other fields where it is required automated identification of different objects