期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2018
卷号：96
期号：13
出版社：Journal of Theoretical and Applied
摘要：This paper presents an approach to word alignment selection by Fourier descriptors that were used together with a neural network for image recognition. Word alignment selection is an important problem of statistical machine translation. The recognition of correct word alignment images is a special case of shape recognition. There are various ways of studying image contours experimentally, and we choose the Fourier method of descriptors, which is proved to be effective and easy to implement. The key implementation options and advantages of the method have been considered. From the given information of the contour and the method of its comparison with the references, an algorithm of word alignment selection has been developed. We also set some threshold conditions for more accurate learning of contours and common patterns.
