Інформаційна технологія лінгвістичної індексації картин на основі нейронної мережі

Abstract

The master’s degree work proposes the information technology of linguistic indexation of pictures based on the back propagation neural network. The subject area of linguistic indexation of pictures is considered and the methods of processing and analysis of pictures are determined. The stages of the information technology of linguistic indexation of the pictures, as well as its program realization in the object-oriented programming language of high level C #, using the UML technology, under the Windows operating system are developed. The developed program of linguistic indexation of pictures corrects many shortcomings of modern indexing systems, namely, it has an increased reliability of indexing of pictures (85.5%) compared to well-known programs of similar functional purpose (72.8%).