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：2016
卷号：85
期号：1
出版社：Journal of Theoretical and Applied
摘要：Batik is one of the inherited high-valued artwork. Batik is a way of making clothes using fabric coloring technique. This technique uses �malam� to avoid staining unintended parts of the fabric. The term �Batik� can also be referred to a cloth that is made with coloring technique, possesses certain motifs as well as special characteristic. Batik�s motif pattern recognition requires initial image processing step called as image segmentation. The purpose of image segmentation is to divide the image into several regions based on feature similarity including grayscale level, texture, color, and motion. In this research multiscale segmentation methods, HMTSeg, is used to recognize the pattern of batik�s motif. Images of Batik that have been used in this work are images of Batik originated from Jambi region, Pekalongan region, and Yogyakarta region. Batik�s motif pattern recognition steps include image pre-processing, image segmentation, and pattern recognition using Euclidean distance based method. Result shows that this method gained 80% of accuracy in recognizing batik�s motif. In conclusion, HMTSeg is a good segmentation method to recognize pattern of batik based on the texture.
