Image enhancement segmentation Indonesian's Batik based on fuzzy C-means clustering using median filter

摘要

This research aims at improving the quality of segmentation on the algorithm Fuzzy C-Means (FCM) which is used to classify the patterns of `Batik' with the background or base color so that two clusters can be achieved by processing using the median filter. To calculate the performance, this research utilized two patterns of Batik which will be clustered with the algorithm FCM, and then being compared with the combination of the median filter using the FCM so that both MSE and PSNR can be achieved. The Batik 1 Cluster 1 with MSE value =248.2652, PSNR = 24.1816, cluster 2 value MSE = 21399.2931, PSNR =4.8268, batik 2, cluster 1 value MSE = 20979.7281, PSNR=4.9128, cluster 2 value MSE = 1672.3475, PSNR =15.8975, using the median filter, batik 1 cluster value MSE = 0.0050, PSNR = 23.0419, cluster 2 value MSE = 0.2917, PSNR =4.9258 whereas batik 2 cluster 1 value MSE = 0.2917, PSNR = 13.8054 and batik 2 cluster 1 value MSE = 0.2917, PSNR = 5.3508.