DocumentCode :
2835021
Title :
Image Segmentation based on Improved Gradient Vector Flow
Author :
Zhang, Hong ; Feng, Lingzhi ; Mu, Ying ; You, Yuhu
Author_Institution :
Image Process. Center, Beihang Univ., Beijing
fYear :
2008
fDate :
Aug. 29 2008-Sept. 2 2008
Firstpage :
765
Lastpage :
769
Abstract :
This paper presents a novel gradient vector flow calculation method and introduces the center position of initial contour line to segment target exactly. The method utilizes the center position of initial contour line and image features to calculate external energy; the contour line can search for and estimate the real edge no matter where the center position of initial contour line is. Experimental results show the method proposed can weaken influence from the center position of initial contour line, enhance the stability of iterative process and exactly segment interested target.
Keywords :
edge detection; gradient methods; image segmentation; contour line; edge estimation; gradient vector flow; image segmentation; iterative process; Active contours; Active noise reduction; Image edge detection; Image processing; Image segmentation; Noise figure; Solid modeling; Stability; Surface fitting; Working environment noise; active contour model; gradient vector flow; image segmentation; implicit method;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Science and Information Technology, 2008. ICCSIT '08. International Conference on
Conference_Location :
Singapore
Print_ISBN :
978-0-7695-3308-7
Type :
conf
DOI :
10.1109/ICCSIT.2008.177
Filename :
4624971
Link To Document :
بازگشت