Amalgamation of Image Features for Medical Image Classification with Ensemble of Classifiers

COVID 19 disease happened to be one of the most infectious diseases in the 21st century. This widespread disease was very huge where people from almost all parts of the world were victims. This paper presents an idea of amalgamation of various image features for COVID 19 detection and classification. Four most significant statistical image features namely, Maximally Stable Extremal Regions (MSER), Histogram of Oriented Gradients (HOG), Scale Invariant Feature Transform (SIFT), and Local Binary Pattern (LBP) are extracted from the chest X ray images. Contribution of each feature in accurate classification and detection of covid is analyzed initially with classifiers namely Naive Bayes (NB), Support Vector Machine (SVM), and Decision tree (DT). Simulation results are also obtained to evaluate how ensemble of classifiers comprehend a specific set of features in classification. Through numerous experimental findings, it has been discovered that amalgamation of all four features to categorize the images using ensemble of classifiers yields best results with the classification accuracy of about 97.5%.