Analysis of color features performance using support vector machine with multi-kernel for batik classification

Batik is a sort of cultural heritage fabric that originated in many areas Indonesia. It can be traced back to different parts Each region, particularly Semarang Central Java, Indonesia, has its design. Unfortunately, due lack knowledge, not all residents recognize the types batik. Therefore, this study proposed an automated method for classifying batik was classified into five categories according method: Asem Arang, Blekok Warak, Gambang Semarangan, Kembang Sepatu, and Semarangan. required analyze color features based on space develop discriminative since able differentiate these patterns. Color were produced RGB, HSV, YIQ, YCbCr spaces. Four kernels used feed Support Vector Machine (SVM) classifier: linear, polynomial, sigmoid, radial basis functions. The experiment carried out using local dataset 1000 images classes (each class contains 200 images). A cross-validation test with k-fold value 10 performed method. In each SVM Kernels, results showed achieved accuracy 100% by utilizing YIQ space, which reliable throughout tests.