A real-time segmentation scheme for continuous color images

Author

Kuo, Chin-Hwa ; Wang, Tay-Shen

Author_Institution

Dept. of Comput. Sci. & Inf. Eng., Tamkang Univ., Tamsui, Taiwan

Volume

2

fYear

2001

fDate

6-9 May 2001

Firstpage

385

Abstract

A real-time segmentation scheme for continuous color images is presented in this paper. The proposed scheme consists of two main steps: (1) seed searching and region growing, (2) region-based change detection. A new color representation model, RBG-Ellipse, is proposed. This color model is similar to the HSI representation. However, the transformation between RGB and RGB-Ellipse is linear. Therefore, we are able to take advantage of noise tolerance processing as well as the efficiency in dealing with color difference computation. By using the proposed segmentation scheme, we implemented applications, (1) intelligent networked visual monitoring system and (2) user interface for distance learning to highlight the value of the proposed scheme. Users can view the results through our web site, http://www.can.tku.edu.tw

Keywords

distance learning; image colour analysis; image segmentation; user interfaces; HSI representation; RBG-Ellipse; color difference computation; color representation model; continuous color images; distance learning; intelligent networked visual monitoring; noise tolerance processing; real-time segmentation; region growing; region-based change detection; seed searching; user interface; Color; Colored noise; Computer aided instruction; Image edge detection; Image segmentation; Intelligent networks; Monitoring; Motion estimation; Object segmentation; User interfaces;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 2001. ISCAS 2001. The 2001 IEEE International Symposium on

Conference_Location

Sydney, NSW

Print_ISBN

0-7803-6685-9

Type

conf

DOI

10.1109/ISCAS.2001.921088

Filename

921088