Batik is a decorative cloth used widely in the indonesian culture. indonesian batik also is very much diverse and consists of many different patterns, colors, and textures. in the course of further research on indonesian culture heritage, we are trying to develop an automatic batik classification system using the characteristics and features contained in each batik patterns, we conduct expriments to differentiate various batik patterns using two well-known image processing features, namely the log-gabor feature vector and the generalized hough transform. building on previous work, we attempt to account for variations in scale and orientation. our expriments are able to show good accuracy, particularly using the generalized hough transform. for the log-gabor filter, further exprimentation is required to determine the optimal configuration of scale and orientation parameters in consturing the filter bank.