Publication of the journal

The section is currently being updated

In most existing video tracking systems, automatic region of interest extraction is used, utilizing a detector that works with a prebuilt model of the desired object belonging to some pre-known set of classes. This approach significantly limits applicability of the automatic tracking system. In this work, the problem of semi-automatic object extraction for visual scene images has been solved. The possibility of implementation of a model-free approach to object extraction using low-level image segmentation methods was demonstrated. The analysis of drawbacks of the basic methods for image segmentation limiting their application in the developed automatic tracking algorithm of the video stream processing system for unmanned aerial vehicles was conducted. The algorithm for extracting an object in the image by superpixel fusion is proposed. The testing results with real and synthetic images for the proposed algorithm are presented. Experimental studies have shown the possibility to apply the proposed algorithm as part of automatic video tracking system software.
Alexander V. Sotnikov
National Research University of Electronic Technology (Russia, 124498, Moscow, Zelenograd, Shokin sq., 1); “Zelenograd Innovation and Technology Center” JSC (Russia, 124527, Moscow, Zelenograd, Solnechnaya ave., 8)
Evgeny A. Shipatov
National Research University of Electronic Technology (Russia, 124498, Moscow, Zelenograd, Shokin sq., 1)
Andrey V. Shipatov
National Research University of Electronic Technology (Russia, 124498, Moscow, Zelenograd, Shokin sq., 1); “Zelenograd Innovation and Technology Center” JSC (Russia, 124527, Moscow, Zelenograd, Solnechnaya ave., 8)

124498, Moscow, Zelenograd, Bld. 1, Shokin Square, MIET, editorial office of the Journal "Proceedings of Universities. Electronics", room 7231

+7 (499) 734-62-05
magazine@miee.ru