Cell Detection with Gabor Filter-Based Features in Histopathologic Images

Abstract

In this study, it was aimed to perform cell detection in histopathologic images which contains ground truth information. Firstly, in the patches which taken from the image, it is aimed to find out whether there is a certain frequency content in certain directions by using Gabor Filter. Features are extracted from the patches. Thus, training and test data sets were created. Then, using the training data sets, k-nearest neighbors, support vector machine, and random forest trained classification methods were used for classification. The success of classifications were observed with using test dataset and the results were presented.