DocumentCode
2139323
Title
Oriented Smoothness Aided Harmonic Gradient Vector Flow for Active Contours
Author
Li, Kun ; Wang, Yuanquan
Author_Institution
Tianjin Key Lab. of Intell. Comput. & Novel Software Technol., Tianjin Univ. of Technol., Tianjin, China
fYear
2009
fDate
17-19 Oct. 2009
Firstpage
1
Lastpage
5
Abstract
A new external force called oriented smoothness aided harmonic gradient vector flow (OSAHGVF) is proposed for active contours. This proposed method incorporate the oriented smoothness borrowed from optical flow into the harmonic gradient vector flow (HGVF) model, therefore, there are three terms, i.e., the harmonic term, the oriented smoothness term and the fidelity term, in the model. Two weighting functions are utilized to regularize the effect of the first two terms respectively, aiming at extending the capture range and preserving object boundaries simultaneously. Experiments and comparisons with the gradient vector flow (GVF) are presented to show the advantages of this approach, including superior noise robustness, edge-preserving and C-shape concavity convergence.
Keywords
gradient methods; image segmentation; surface topography; C-shape concavity convergence; active contour; active contours; edge preservation; fidelity term; image segmentation; object boundary; oriented smoothness aided harmonic gradient vector flow; weighting functions; Active contours; Computer science education; Computer vision; Convergence; Diffusion processes; Educational technology; Image motion analysis; Laboratories; Noise robustness; Smoothing methods;
fLanguage
English
Publisher
ieee
Conference_Titel
Image and Signal Processing, 2009. CISP '09. 2nd International Congress on
Conference_Location
Tianjin
Print_ISBN
978-1-4244-4129-7
Electronic_ISBN
978-1-4244-4131-0
Type
conf
DOI
10.1109/CISP.2009.5303489
Filename
5303489
Link To Document