To search, Click below search items.


All Published Papers Search Service


Automated Classification of Galaxies Using Transformed Domain Features


M.A. Mohamed, M.M. Atta


Vol. 10  No. 2  pp. 86-91


Classifying galaxy information is one of the most important challenges for astronomers as it can provide important clues about the origin and evolution of the universe. In this paper; the performance of ten artificial neural networks (ANNs) based classifiers was evaluated and tested, based on a selected set of features. These features were extracted in frequency and wavelet domain; and then divided into three categories: (i) Fourier transform based; (ii) Cosine transform based, and (iii) wavelet transform based. The number of features in each category was determined empirically. The results showed that; (i) the support vector machine provided the best results with Fourier based features; (ii) the Jordan/Elman network (JEN) provided the best results in cosine and wavelet based features. In general, the cosine transform based features with JEN classifier provided the best results among all transformed based classifiers; about 0.45718% error in classification. The data set was taken from standardized catalogue from Zsolt Frei website.


Hubble Sequence, Discrete Cosine Transform (DCT), Discrete Wavelet Transform (DWT), Artificial Neural Networks (ANNs)