Classifier ensemble for an effective cytological image analysis

Study of Image Analysis Algorithms for Segmentation, Feature Extraction and Classification of Cells

“…The two-step approach employs the convex Bayesian functional and the convex region-scalable fitting energy functional, and requires only two subsequent steps. In the first step, authors use the region-scalable fitting energy functional, which can cope (Matula et al 2009), (Liao et al 2015), (Dogantekin, Avci, and Erkus 2013), (Stoklasa, Majtner, and Svoboda 2014), (Filipczuk, Krawczyk, and Woźniak 2013) Region-based segmentation These techniques operate iteratively by grouping together pixels which have similar values. The watershed transform is a region-based segmentation technique.…”

“…(Kong et al 2011), (Tonti et al 2015), (Abeysekera et al 2014), (Filipczuk, Krawczyk, and Woźniak 2013) with spatially local mean intensity values in comparison to the region-based Chan-Vese functional and the Bayesian functional which are based on global mean intensity values. In the second step, authors use the Bayesian functional which allows dealing with inhomogeneities at the border of cell nuclei.…”

“…The proposed classifier is k-NN based and was implemented using MESSIF framework (Batko et al 2008). In comparison with SVM, the advantages of k-NN based approach in the implicit support of multi-class division and (Dogantekin, Avci, and Erkus 2013), (Karkanis et al 2003), (Gertych et al 2015), (Ashcroft et al 2011), (Stoklasa, Majtner, and Svoboda 2014), (Nanni and Lumini 2007), (Filipczuk, Krawczyk, and Woźniak 2013) Non-probabilistic: support vector machine, K-nearest neighbor and linear discriminant analysis.…”

“…(Filipczuk, Krawczyk, and Woźniak 2013), (Ashcroft et al 2011), (dos Santos et al 2015, (Kayaaltı et al 2014), (Glezakos et al 2010), (Keltch, Lin, and Bayrak 2014), (Sagonas et al 2013) Boosting Joint boosting, gentle boosting, adaptive boosting.…”

“…An example of classifier ensemble is presented in (Filipczuk, Krawczyk, and Woźniak 2013) applied in cytological images. Authors propose a segmentation method based on the combination of adaptive thresholding in grayscale and pixel classification in color space using one of four clustering algorithms: K-means, fuzzy C-means, Gaussian mixture model (Dempster, Laird, and Rubin 1976), and competitive neural networks.…”