Automatic Segmentation of Cell Images by Improved Graph Cut-Based Approach
Cell segmentation provides an opportunity to reveal object of interest from the background of an image. In the traditional graph cut segmentation approach, the user initiates the segmentation process by selecting pixels for foreground and background. However, one of the problems of traditional graph cut is that it is time consuming, especially on a large dataset. Thus, we propose a fully automatic technique for cell segmentation on graph cut to automate the selection of sample foreground and background pixels. In order to achieve this, a combination of two methods namely Otsu thresholding and kmeans clustering algorithm is explored. The Otsu thresholding and the k-means provides an initial cell segmentation, creating a platform to automatically select sample foreground and background pixels initiating the graph cut segmentation. Experimental results on two public datasets suggest promising results.
K. O. Oyebode and J. R. Tapamo, "Automatic Segmentation of Cell Images by Improved Graph Cut-Based Approach", Journal of Biomimetics, Biomaterials and Biomedical Engineering, Vol. 29, pp. 74-80, 2016