Науковий вісник НГУ

Dispersion aggregated objects images skeletonization

User Rating:  / 1

Details

Category: IT technologies

Last Updated on Tuesday, 19 February 2013 10:00

Published on Friday, 16 November 2012 16:30

Hits: 4278

Authors:

Т.А. Ruzova, Cand. Sci. (Tech.), O. Gonchar Dnepropetrovsk National University, Senior Research Scientist of Problem Research Laboratory of Streams, Dnipropetrovsk, Ukraine

Abstract:

Skeletons making is extremely important step in the problem of segmentation elements in dispersed formations. Therefore, the correctness of skeleton representation, matching the structure of examined object and skeleton rezistanse to form distortion caused by noises is essential. The main problem of usual methods is receiving skeletons which are not thin enough (thikness is more then one pixel). Also, they are sensitive to boundary noises.

The purpose of the research is designing of the metod of aggregated elements in dispersed formations skeleton making resisting boundary noises and represented by a set of straight line segments between the branch nodes, sorted from external nodes to internal.

Suggested method allows to build skeletons of high complex objects. Among its advantages are the absence of unnecessary points and rezistanse to boundary noises allowing to avoid continuity violation and object unnecessary refinement.

Method includes next steps: base skeleton making by Zhang-Suen method; exclusion of unnecessary details; pole determination and representation skeleton as a set of branches between poles; skelet improvement in order to represent branches as straight line segments; arranging the branches according to location relative to center.

Obtained results will be useful for designing methods and algorithms for segmentation images of complex objects and aggregated formations.

References:

1. Гонсалес Р. Цифровая обработка изображений. / Гонсалес Р., Вудс Р. – М.: Техносфера, 2005. – 1072 с.

Gonzales, R.C. and Woods, R.E., (2005), Tsyfrovaya obrabotka izobrazheniy [Digital Image Processing], Tekhnosfera, Moscow, Russia, 1072 p.

2. Рогов А.А. Некоторые методы классификации и поиска в электронной коллекции графических документов / Рогов А.А., Кириков П.В. // Труды 12-й Всероссийской научной конференции „Электронные библиотеки: перспективные методы и технологии, электронные коллекции“.– Казань: 2010.– C. 409–414.

Rogov, A.A. and Kirikov, P.V. (2010), “Some methods of classification and search at graphics documents electronic collection”, Materials of 12th All-Russian Sci. Conf. “Electronic libraries: perspective methods and technologies, electronic collections”, Kazan, Russia, pp. 409–414.

3. Ласло М. Вычислительная геометрия и компьютерная графика на С++ / Ласло М. – М.: Бином, 1997. – 301 с.

Laszlo, M. (1997), Vychislitelnaya geometriya i kompyuternaya grafika na C++ [Computational Geometry and Computer Graphics in C++], Binom, Moscow, Russia, 301 p.

4. Местецкий Л.М. Непрерывная морфология би-нарных изображений. Фигуры, скелеты, циркуляры. / Местецкий Л.М. – М.: Физматлит, 2009. – 288 c.

Mestetskiy, L.M. (2009), Nepreryvnaya morfologiya binarnykh izobrazheniy. Figury, skelety, tsyrkulyary [Continuous morphology of binary images. Figures, skeletons, circular], Fismatlit, Moscow, Russia, 288 p.

5. Batchelor, B.G. and Whelan, P.F. (1997), Intelligent vision systems for industry, Springer-Verlag, 457 p.

6. Темнов К.А. Алгоритм скелетизации растрового монохромного изображения / Темнов К.А., Кириллов А.В. // Информационно-измерительные и управляющие системы. – 2009. – №8. – C. 45–49.

Temnov K.A. and Kirillov, A.V. (2009), “Raster monochrome images thinning algorithm”, Informatsion-no-izmeritelnie i upravlayu shchie sistemy, no.8, pp. 45–49

7. Демидович Б.П. Краткий курс высшей математики / Демидович Б.П., Кудрявцев В.А. – М.: АСТ, Астрель. – 2001. — 656 с.

Demidovich, B.P. and Kudravtsev, B.A. (2001), Kratkiy kurs vysshey matematiki [Short course in high mathematics], АSТ, Аstrel, Moscow, Russia, 656 p.

2012_01_ruzova
2013-02-18  335.52 KB  1552

• < Prev
• Next >