The Java character is one form of writing of Javanese culture in addition to using the alphabet that we often encounter in some relics that are in the special region of Yogyakarta (DIY). Characteristics of Java's character is very complex, so it is a bit difficult to understand due to it's not familiar.
Technological developments are demanding to be able to build an application that can help the domestic or international travelers to identify a Java character in which there is historical information from several existing relics at DIY. The main objective of this research is to handwriting recognition in which the application can convert Java's character to alphabets. Identification methods in this research use of digital image processing, namely the extraction of characteristics by method diagonal based feature extraction. The method can recognize the characteristics of each Java character, Especially the basic character in the Java character which consists of 20 types of letters.
Based on the results of the study known that by using the method of extracting the characteristics of image taking positions and the form of writing is to determine the process of calculating the extraction of these characteristics. Each character in the main, character uses 3 different sets of character, so the total is 60 data sets used. The application's accuracy rate reaches 75% in order to properly convert Java character to the alphabet. From 20 data testing, get 15 correct data and 5 wrong data.