Neural network and K-nearest neigbor algoritm methods comparison for automatic detection of embryo cleavage stages

Abstract

Abstract. This paper represents an embryo cleavage-stage classification algorithm. There are used statistical feature extraction methods(entropy, invariant moments and principal components analysis) and two classification methods: Classification with training(neural networks) and classification without training(K-nearest neighbor algorithm). The main problem of this work is detection of early embryo cleavage stages. The aim is to adapt the proper classification method. There are introduced proposed methods in this article. The proposed method is checked by experiment, and the results is introduced. It is expected that this method will work well in video sequences.