Aplikasi Pembelajaran Pengenal Aksara Bali Menggunakan Metode Template Matching

Abstract

This research aimed to design and develop an application identifier balinese character learning that can be used to facilitate for people to recognize text characters Balinese more efficiently in term of time and place. This application is developed by using the template matching method. Input of this application is the writing of user using a digital pen, touchpad or mouse and the output is the image of Balinese character that has the smallest error value. Implementation of this template matching method produces a software called Balinese Character Matching. This application was developed using the PHP programming language which is a web programming language that can be used by many users without the
limitations of space and time. In this application there are two main processes namely preprocessing and recognition by template matching methods. Preprocessing process consists of grayscale, thresholding and size normalization. In the size normalization conducted the search process limits, cropping, scaling, and labeling of 1 and 0. Testing program conducted by admin (person whose data is stored in the database) and user (person whose data is not stored in the database). For admin who test the program obtained more accurate result with an average of 95.45%, while the result of testing on the user obtained an average of 86.25%.