A Comparative Study on Human Spermatozoa Images Classification with Artificial Neural Network Based on FOS, GLCM and Morphological Features

Abstract

This paper aims to evaluate the accuracy of artificial neural network based classifiers using human spermatozoa images. Three different neural network based classifiers are used: Feed Forward Neural Network, Radial Basis Neural Network and Elman Back Propagation Neural Network. These three different classifiers were investigated to determine their ability to classify various categories of human spermatozoa images. The investigation was performed on the basis of the different feature vectors. The feature vector includes first order statistics (FOS), textural and morphological features. The extracted features are then used to train and test the artificial neural network. Experimental results are presented on a dataset of 91 images consisting of 71 abnormal images and 20 normal images. The radial basis network produced the highest classification accuracy of 60%, 75% and 70% when trained with FOS, Combined and Morphological features. When feed forward neural network is trained with GLCM features, a classification accuracy of 75% is achieved.