Skip to main content
placeholder image

Skeleton extraction based on anisotropic partial differential equation

Journal Article


Abstract


  • Skeleton extraction is of great significance for image recognition processes. In this paper, a novel skeleton extraction method is proposed based on the anisotropic diffusion of the gradient vector field. Firstly, the gradient vector is calculated for the original image. Then, the gradient vector field is updated by an anisotropic partial differential equation to strengthen the source point or sink point of the image energy. At last, according to the divergence property of the final gradient vector field, the skeleton of the initial image is obtained. The experimental results demonstrate that the proposed method is robust for noise and can give the exact skeleton for images.

Authors


  •   Zhang, Fang (external author)
  •   Xiao, Zhitao (external author)
  •   Geng, Lei (external author)
  •   Wu, Jun (external author)
  •   Xu, Zhenbei (external author)
  •   Wang, Danyu (external author)
  •   Xi, Jiangtao

Publication Date


  • 2015

Published In


Citation


  • F. Zhang, Z. Xiao, L. Geng, J. Wu, Z. Xu, D. Wang & J. Xi, "Skeleton extraction based on anisotropic partial differential equation," Optik, vol. 126, (23) pp. 3692-3697, 2015.

Scopus Eid


  • 2-s2.0-84952316323

Ro Metadata Url


  • http://ro.uow.edu.au/eispapers/5636

Number Of Pages


  • 5

Start Page


  • 3692

End Page


  • 3697

Volume


  • 126

Issue


  • 23

Abstract


  • Skeleton extraction is of great significance for image recognition processes. In this paper, a novel skeleton extraction method is proposed based on the anisotropic diffusion of the gradient vector field. Firstly, the gradient vector is calculated for the original image. Then, the gradient vector field is updated by an anisotropic partial differential equation to strengthen the source point or sink point of the image energy. At last, according to the divergence property of the final gradient vector field, the skeleton of the initial image is obtained. The experimental results demonstrate that the proposed method is robust for noise and can give the exact skeleton for images.

Authors


  •   Zhang, Fang (external author)
  •   Xiao, Zhitao (external author)
  •   Geng, Lei (external author)
  •   Wu, Jun (external author)
  •   Xu, Zhenbei (external author)
  •   Wang, Danyu (external author)
  •   Xi, Jiangtao

Publication Date


  • 2015

Published In


Citation


  • F. Zhang, Z. Xiao, L. Geng, J. Wu, Z. Xu, D. Wang & J. Xi, "Skeleton extraction based on anisotropic partial differential equation," Optik, vol. 126, (23) pp. 3692-3697, 2015.

Scopus Eid


  • 2-s2.0-84952316323

Ro Metadata Url


  • http://ro.uow.edu.au/eispapers/5636

Number Of Pages


  • 5

Start Page


  • 3692

End Page


  • 3697

Volume


  • 126

Issue


  • 23