FUZZY TARGETED COLOR C-MEANS CLUSTERING FOR SEARCHING THE IMAGE AFTER THE SAMPLE

  • Алексей Владимирович Гороховатский Kharkiv National University of Economics named after Simon Kuznets

Abstract

The modifications of classical FCM clustering to solve the image searching problem after the sample are proposed. The procedure of cluster stabilization, which makes clustering independent on the starting initialization, is considered. The criteria of anticipatory clustering stop in comparing to the sample are proposed; the experimental research has confirmed the possibility to reduce the processing time up to 50% without quality degradation.

Author Biography

Алексей Владимирович Гороховатский, Kharkiv National University of Economics named after Simon Kuznets

Candidate of Technical Sciences, Associate Professor

Published
2019-03-02
How to Cite
Гороховатский, А. (2019). FUZZY TARGETED COLOR C-MEANS CLUSTERING FOR SEARCHING THE IMAGE AFTER THE SAMPLE. Electrotechnic and Computer Systems, (20(96), 88-96. Retrieved from http://eltecs.opu.ua/index.php/journal/article/view/1628
Section
Computer Systems, Networks and their Components