Research of batik image classification based on support vector machine
Qing-Ni Yuan, Jian Lu, HaisongHuang, WeijiPan
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12B) 193-196
Key Laboratory of Advanced Manufacturing technology (Guizhou University), Ministry of Education, Guiyang 550003, China
The digital protection and development of batik is applied in the digital design of the arts and crafts by the digital image acquisition of batik to construct a graph database. Its key technology is the automatic classification of image. In this paper, we use image analysis and recognition technology to image classification recognition of five type of batik: Bronze drum lines, Butterfly lines, Bird lines, Fish lines and Flower lines etc. On the basis of the segment image of batik, we extract the shape and texture feature by Histogram of Oriented Gradient (HOG). Then, we respectively use Support Vector Machine (SVM), Minimum Distance Method and BP Neural Network to classify test. The result shows that the classification recognition ability of SVM is better than the Minimum Distance Method and the BP Neural Network. Therefore, the classification recognition method of the Histogram of Oriented Gradient (HOG) and the Support Vector Machine (SVM) is feasible to the automatic classification of batik image.