BALINESE PAPYRUS MANUSCRIPT IMAGE SEGMENTATION USING DBSCAN CLUSTERING METHOD

Papyrus Manuscript is handwritten in the form of an ancient script and written using Balinese script. Existence of Balinese Papyrus Manuscript in Balinese’s life is scarce. Today’s young generation almost left Balinese Papyrus Manuscript because of its complexity and difficult to understand. This traditional manuscript used Balinese Letters, which had some unique characteristics and patterns. This unique characteristic could identified using image processing technologies, which presented in this research. Segmentation on Balinese Papyrus Manuscript image is to get Balinese script letter object. The result of segmentation process would be used to helped recognized Balinese Letters by its features, with a computer based application. Segmentation of the Balinese Papyrus Manuscript image used DBSCAN clustering method. This research consist of three steps. First step was image acquisition. The second step was image pre-processing. The last step was segmentation process using DBSCAN clustering method. The average success rate of segmentation process reached 78.8% with 21.2% error rate. The average of excecution time when the testing process reached 27.227 seconds.