DocumentCode :
3746431
Title :
Divergence-aided gradient vector flow snakes
Author :
Qiang Li;Tingquan Deng
Author_Institution :
Harbin Engineering University, Harbin, China
fYear :
2015
Firstpage :
443
Lastpage :
447
Abstract :
Gradient Vector Flow (GVF) is an important external force field driving active contours evolution. However, due to difficulties in evolving over the saddle and stationary points of the GVF field, the GVF snake which takes the GVF field as the external driving force suffers from a poor performance of dealing with complex geometries. In this paper, using the divergence of GVF field we construct a new balloon force which can determine the curve evolution when the GVF force is not valid. The modified GVF snakes model with the balloon force has a better capability of dealing with complex geometries. The significant difference between the proposed and existing balloon forces (e.g. constant force) is that the proposed divergence based balloon force has a negative value on the edge of image and so can better prevent the occurrence of boundary leakage. Some experiments on synthetic and real images are presented in this paper.
Keywords :
"Image edge detection","Force","Active contours","Geometry","Image segmentation","Gray-scale","Market research"
Publisher :
ieee
Conference_Titel :
Image and Signal Processing (CISP), 2015 8th International Congress on
Type :
conf
DOI :
10.1109/CISP.2015.7407921
Filename :
7407921
Link To Document :
بازگشت