Open AccessIntelligent system for object recognition on optical images using cascade neural networks
Author(s) -
M. V. Shavranskyi,
A. V. Kuchmystenko
Publication year - 2018
Publication title -
naftogazova energetika
Language(s) - English
Resource type - Journals
eISSN - 2415-3109
pISSN - 1993-9868
DOI - 10.31471/1993-9868-2018-1(29)-50-55
Subject(s) - artificial neural network , artificial intelligence , computer science , cascade , object (grammar) , basis (linear algebra) , cognitive neuroscience of visual object recognition , computer vision , pattern recognition (psychology) , machine learning , engineering , geometry , mathematics , chemical engineering
The paper is devoted to increasing the accuracy of the classification of objects on optical images by developing a structure, model and method of teaching the combined neural network and creating on its basis an intelligent image recognition system for tasks of the oil and gas industry - diagnostics, forecasting of emergency situations of technological objects.