PENGENALAN AKSARA BALI TERCETAK PADA CITRA DOKUMEN MENGGUNAKAN MULTI LAYER PERCEPTRON DENGAN PEMBELAJARAN BACKPROPAGATION
| Main Authors: | , I MADE GEDE SUNARYA, , Drs. Agus Harjoko, M.sc., Ph.D |
|---|---|
| Format: | Thesis NonPeerReviewed |
| Terbitan: |
[Yogyakarta] : Universitas Gadjah Mada
, 2012
|
| Subjects: | |
| Online Access: |
https://repository.ugm.ac.id/100112/ http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=56916 |
Daftar Isi:
- Balinese character are a vital means to know the culture of Bali in depth, because Balinese character serves as a means to write a wide range of religious issues, customs, literature and so forth. Balinese character in use is not always written sequentially like Latin character. In the formation of a word or phrase in Balinese character, the characters can be combined at the top, bottom or side of a core character. This study aims to develop a prototype application to recognize the characters on the image of a printed Balinese text and know the level of recognition accuracy of the characters on the printed Balinese text. The method used for the recognition is Backpropagation Neural Network (ANN-BP). Implementation of the algorithms in the development of Balinese character recognition prototype software is able to recognize characters on printed Balinese text with error on the training images of 3.43% and 10% in testing image.