Use of a Clustering Method and Fuzzy Logic for Classification of Tissue Images

A. Yosif, M.

	Use of a Clustering Method and Fuzzy Logic for Classification of Tissue Images
Tikrit Journal of Pure Science
Article 1, Volume 16, Issue 3, 2011, Pages 268-277
Author
M. A. Yosif
Abstract
In this study, a clustering method algorithm known as K-means together with fuzzy logic have been used to classify tissue images where an algorithm is predicted for this purpose. The algorithm is used to make clusters to the data (colour values) which exists in the tissue images that differ in their composition in order to see components that are not seen in original images before applying the algorithm on them. The algorithm has been applied to many images where the Lymphocyte image represents the Lymphocyte cells in agranular white blood cells, and the image Eosinophlis belongs to acidic cells of granular white blood cells, whereas the Sympathetic ganglion were neuron Sympathetic cells. The reason for choosing this algorithm is its simplicity and flexibility in use, and it gives more precise classification results as compared to algorithms of other clustering methods. A set of (5-16) type of classifications are given as input for each of the three tissue images, from these other classifications are obtained, [this gives images of wider visibility and faint structures can be seen better] then, we used Fuzzy techniques to determine the confidence degree depending on the values 0 and 1.

Statistics Article View: 198 PDF Download: 43