Home > Archive > 2017 > Volume 7 Number 4 (Aug. 2017) >
IJMLC 2017 Vol.7(4): 89-93 ISSN: 2010-3700
DOI: 10.18178/ijmlc.2017.7.4.626

A Level Set Method Based on Bayesian Risk for Textured Image Segmentation

Yao-Tien Chen

Abstract—A level set method based on the Bayesian risk is proposed for textured image segmentation. First, the original texture image is converted into an orientation image. Second, based on Bayesian risk formed by false-positive and false-negative fraction in a hypothesis test, the level set evolution functional is deduced. Third, to prevent the propagating curves from generating excessively irregular shapes and lots of small regions, curvature and gradient of edges in the orientation image are integrated into the functional. Finally, to extract the boundaries of texture targets, the Bayesian level set equation is derived from the functional and then orientation image is regarded as input of the Bayesian level set equation. Compared with other segmentation methods for texture image, the proposed Bayesian level set approach relies on the optimum decision of pixel classification; thus the approach has more reliability in theory and practice. Experiments show that the proposed approach can accurately segment the texture image; moreover, the algorithm is extendable for multiphase segmentation.

Index Terms—Texture image, level set method, Bayesian risk, hypothesis test.

The author is with the Department of Applied Mobile Technology, Yuanpei University of Medical Technology, HsinChu, 30015, Taiwan (e-mail: ytchen@mail.ypu.edu.tw).


Cite: Yao-Tien Chen, "A Level Set Method Based on Bayesian Risk for Textured Image Segmentation," International Journal of Machine Learning and Computing vol. 7, no. 4, pp. 89-93, 2017.

General Information

  • E-ISSN: 2972-368X
  • Abbreviated Title: Int. J. Mach. Learn.
  • Frequency: Quaterly
  • DOI: 10.18178/IJML
  • Editor-in-Chief: Dr. Lin Huang
  • Executive Editor:  Ms. Cherry L. Chen
  • Abstracing/Indexing: Inspec (IET), Google Scholar, Crossref, ProQuest, Electronic Journals LibraryCNKI.
  • E-mail: ijml@ejournal.net

Article Metrics in Dimensions