Performance comparison and evaluation of vari-ous segmentation methods

  • Authors

    • Vikashini Venkatesh Christ (Deemed to be University)
    • Praveen P U
    2018-05-03
    https://doi.org/10.14419/ijet.v7i2.9687
  • Multithreshold, Watershed, Normalised Cut.

  • Image segmentation is the most important method in the concept of image processing. It helps in analyzing the image accurately in many applications. It is generally used to assign or name, a label to individual pixels in an image, so that labels with similar name share common features. These related pixels result in same color, texture, or intensity. It also helps in identifying lines, curves and objects. These kinds of results help in different applications in the field of medical imaging, 3D constructions, etc. There are different kinds of segmentation methods already available for such applications. This paper briefs and compares three different types of segmentation methods like multithreshold method, watershed method and normalized cut method. It is compared based on computational time, complexity and number of clusters of the different methods used in the image.

  • References

    1. [1] Manjula. KA, â€Role of Image Segmentation in Digital Image Pro-cessing for Information Processingâ€, IJCST, vol 3,issue 3,pp.312-318,May-June 2015.

      [2] Sharma D.,â€Image processingâ€, Course-ECE-411, Technical pa-per.

      [3] https://en.wikipedia.org/wiki/Image segmentation.

      [4] Yanowitz S.D., Bruckstein A.M.,â€A new method for image sege-mentationâ€, pp.82-95, February 4, 1998.

      [5] Morse B.S., Brigham Young University, “Thresholdingâ€, 1998-2000, technical paper.

      [6] Ferrari S., “Image segmentationâ€, pp.1-22, 2011, technical paper.

      [7] Amoda N., Kulkarni R.K., “Image segmentation and detection using watershed transform and region based image retrievalâ€, IJETTCS, vol 2, Issue 2, pp.89-94, 2013.

      [8] XUE Guoxin, LIU Qiang, XUE Pei, “Method of watershed with threshold decision for multiscale tra c image segmentation, International Symposium on Computational Intelligence and Designâ€, pp.186-189, 2010.

      [9] Beucher S., Lantuejoul C., “Use of watersheds in contour detectionâ€, pp.2.1-2.12, technical paper, 1979.

      [10] Shi J., Malik J., cuts and image segmentation, pp.1-52, technical paper.

      [11] Xiaoli Xu, Zhimao Lu, Haiyan Li, “Color Image Segmentation Based on Watershed and Ncut of Improved Weight Matrixâ€, IEEE, pp.681-686, 2011.

      [12] Mikulka J., Gescheidtova E., Bartusek K., and Nespor D., “Comparison of Segmentation Methods in MR Image Processingâ€, Progress in Electromagnetics Research Symposium Proceedings, pp.429-432, 2012.

      [13] Vaia Machairas, Etienne Decenciere, Thomas Walter, “Waterpixels: Superpixels based on the watershed transformationâ€, IEEE, 2014.

      [14] Qian Zhu, Liqiu Jing and Rongsheng Bi, “Exploration and Im-provement of Otsu Threshold Segmentation Algorithmâ€, IEEE, 2010.

      [15] Sahil Narang, Kishore Rathinavel, “An implementation of Efficient Graph-Based Image Segmentationâ€, pp.1-4.

      [16] Kun Yang, Cai-Xia Deng, Yu Chen, Li-Xiang Xu, “The denoising method of threshold function based on waveletâ€, IEEE, 13-16 July, 2014.

      [17] Xiie Dong Yang and Vipin Gnpta, “An Improved Threshold Se-lection Method for Image Segmentationâ€,IEEE, 1993.

      [18] https://www.sciencedirect.com/science/article/pii/S0165168409004423

      [19] http://www.bioone.org/doi/abs/10.3394/0380 1330(2007)33

      [20] Najmuzzama Zerdi, Dr. Subhash Kulkarni,Dr. V.D.Mytri,†Detailed image segmentation using normalised cuts and weighted coefficientâ€, IJEEE,April 2014.

      [21] Wenbing Tao, Hai Jin, and Yimin Zhang, “Color Image Seg-mentation Based on Mean Shift and Normalized Cutsâ€, October 2007.

  • Downloads

  • How to Cite

    Venkatesh, V., & P U, P. (2018). Performance comparison and evaluation of vari-ous segmentation methods. International Journal of Engineering & Technology, 7(2), 663-666. https://doi.org/10.14419/ijet.v7i2.9687