Journal of Medical Systems

, Volume 34, Issue 1, pp 35–42

Application of K- and Fuzzy c-Means for Color Segmentation of Thermal Infrared Breast Images

Original Paper

DOI: 10.1007/s10916-008-9213-1

Cite this article as:
EtehadTavakol, M., Sadri, S. & Ng, E.Y.K. J Med Syst (2010) 34: 35. doi:10.1007/s10916-008-9213-1


Color segmentation of infrared thermal images is an important factor in detecting the tumor region. The cancerous tissue with angiogenesis and inflammation emits temperature pattern different from the healthy one. In this paper, two color segmentation techniques, K-means and fuzzy c-means for color segmentation of infrared (IR) breast images are modeled and compared. Using the K-means algorithm in Matlab, some empty clusters may appear in the results. Fuzzy c-means is preferred because the fuzzy nature of IR breast images helps the fuzzy c-means segmentation to provide more accurate results with no empty cluster. Since breasts with malignant tumors have higher temperature than healthy breasts and even breasts with benign tumors, in this study, we look for detecting the hottest regions of abnormal breasts which are the suspected regions. The effect of IR camera sensitivity on the number of clusters in segmentation is also investigated. When the camera is ultra sensitive the number of clusters being considered may be increased.


Digital infrared thermal imagingColor segmentationBreast cancer detectionK-meansFuzzy c-means

Copyright information

© Springer Science+Business Media, LLC 2008

Authors and Affiliations

  1. 1.Electrical and Computer Engineering DepartmentIsfahan University of TechnologyIsfahanIran
  2. 2.School of Mechanical and Aerospace Engineering, College of EngineeringNanyang Technological UniversitySingaporeSingapore
  3. 3.Adjunct NUH Scientist, Office of Biomedical ResearchNational University Hospital of SingaporeSingaporeSingapore