Authors : R.A. Tomakova, S.A. Filist and A.I. Pykhtin
Abstract: Intelligent classification system for complex structured images designed to highlight the image of a given segment class. It belongs to the training systems of classification. The development urgency of complex structured image classification in intelligent systems occurs in the processing of images from cameras UAV used for navigational purposes in the absence of an artificial earth satellite or the analysis of images in real time by the operator. In this case, there are high demands not only for the quality classification of objects in images, but also to speed identification and classification of the investigated image segments. The contradiction between the effectiveness (classification accuracy) and computing resources offered solved by a distributed computing system (network). The method is built on the idea of hybrid trekking to data processing and aggregation of “weak” classification algorithms in the “strong” algorithms.
R.A. Tomakova, S.A. Filist and A.I. Pykhtin, 2017. Development and Research of Methods and Algorithms for Intelligent Systems for Complex Structured Images Classification. Journal of Engineering and Applied Sciences, 12: 6039-6041.