Comparative Study for Classification of WBC Segmented images using Bag of Words and WEKA

The representation of the image in a meaningful and easy-to-analyze way can be done with the help of image segmentation. Diagnosis, treatment planning, tumour localization, WBC count, Leukocytes classification etc. can only be performed after image segmentation. For accurate diagnosing of disease, correctly classified leukocytes, and its subclass are required. Whether the segmented image cells are belongs to benign or malignant are carried out by best classification method. In this paper comparative study of two types of classification techniques like WEKA and Bag of Words have evaluated. Geometrical, textural, and statistical features of different images are extracted and applied in classification algorithms. WEKA is made up of machine learning (ML) techniques hence multi level based classification developed using the different classifiers like Lib SVM, Naïve Bayes, J48, Zero R, PART and Random Forest classifiers are considered. The Bag-of-Words (BoW) model is used to construct the feature vectors from densely extracted local features, such as dense scale-invariant feature transform (SIFT) and Speeded Up Robust Features (SURF). The two types of algorithm are tested using 120 test images and 311 training image sets with many image using MATLAB code. Performance indices and classification accuracy of the two classifiers are compared and check the comparative results of the BoW model and WEKA classifiers.