Penentuan Ukuran Citra Minimal Sistem Konversi Aksara Sunda dengan Metode Template Matching Correlation

UUNG UNGKAWA; RACHMAT FAUZI; NYAI ROHAETI

doi:10.26760/mindjournal.v7i2.177-187

Penentuan Ukuran Citra Minimal Sistem Konversi Aksara Sunda dengan Metode Template Matching Correlation

UUNG UNGKAWA, RACHMAT FAUZI, NYAI ROHAETI

Sari

ABSTRAK

Aksara Sunda banyak ditemui di banyak tempat. Untuk itu, dibutuhkan sistem yang dapat menerjemahkan aksara Sunda ke dalam huruf latin. Dalam penelitian ini dibangun sebuah sistem yang mampu membantu orang memahaminya. Penelitian ini bertujuan untuk menentukan ukuran citra minimal yang masih dapat dikenali dengan akurasi di atas 50%. Untuk itu dalam eksperimen ini dibuat tiga ukuran: kecil (5x5 pixel), sedang (10x10 pixel), besar (15x15 pixel). Sistem menggunakan algoritma Template Matching Correlation. Tahapan pertama yaitu proses akuisisi citra digital kedalam sistem, dilanjut dengan tahapan pre-processing. Sistem kemudian melakukan tahapan segmentasi baris, kata, dan huruf lalu dilakukan normalisasi. Citra hasil normalisasi diklasifikasi sesuai dengan label yang memiliki nilai korelasi terbesar. Hasil pengujian untuk citra yang dinormalisasi dengan ukuran besar didapat nilai akurasi sebesar 83,66%, untuk citra dengan ukuran sedang didapat akurasi sebesar 33,66%, dan untuk ukuran kecil didapat akurasi sebesar 8,33%.

Kata kunci: Pengolahan Citra, Aksara Sunda, Template Matching Correlation

ABSTRACK

Sundanese script can be found in many places. For this reason, we need a system that can translate Sundanese script into Latin letters. In this research, we build a system for helping people understand it. This study aims to determine the minimum image size that can still be recognized with accuracy more than 50%. For this reason, we normalize three image sizes: large (15x15 pixels), medium (10x10 pixels), small (5x5 pixels). The system uses the Template Matching Correlation algorithm. The first stage is the process of acquiring digital images into the system, followed by the pre-processing stage. The system then performs line, word, and letter segmentation stages and then normalizes them. The normalized image is classified according to the label that has the highest correlation value. The results for normalized images with large sizes obtained an accuracy of 83.66%, for medium sizes obtained an accuracy of 33.66%, and for small sizes have an accuracy of 8.33%.

Keywords: Image Processing, Sundanese Script, Template Matching Correlation

Teks Lengkap:

PDF

Referensi

Ahmed Abbood, A., Sulong, G., & Peters, S. U. (2014). A review of fingerprint image pre-processing. Jurnal Teknologi (Sciences and Engineering), 69(2), 79â€“84. https://doi.org/10.11113/jt.v69.3111

Amalia, N., Hidayat, E. W., & Aldya, A. P. (2020). Pengenalan Aksara Sunda Menggunakan Metode Jaringan Saraf Tiruan Backpropagation Dan Deteksi Tepi Canny. CESS (Journal of Computer Engineering, System and Science), 5(1), 19. https://doi.org/10.24114/cess.v5i1.14839

Angraheni, N. R., Efendi, R., & Purwandari, E. P. (2017). Pengenalan Tulisan Tangan huruf Hijaiyah sambung menggunakan Algoritma Template Matching Correlation. Jurnal Rekursif, 5(1), 21â€“31.

Arif, Y. M., & Sabar, A. (2012). Sistem Pengenalan Wajah Menggunakan Metode Template Matching. MATICS.

Cristani, M., Olvieri, F., Workneh, T., Pasetto, L., & Tomazzoli, C. (2022). Classification Rules Explain Machine Learning. Proceedings Ofthe 14th International Conference on Agents and Artificial Intelligence (ICAART2022) - Volume 3, Pages 897-904, 3(Icaart), 897â€“904. https://doi.org/10.5220/0010927300003116

Gil, J. Y., & Kimmel, R. (2003). Efficient dilation, erosion, opening, and closing algorithms. IEEE Transactions on Pattern Analysis and Machine Intelligence, 24(12), 1606â€“1617. https://doi.org/10.1109/TPAMI.2002.1114852

Hartanto, S., Sugiharto, A., & Endah, S. N. (2015). Optical Character Recognition Menggunakan Algoritma Template Matching Correlation. Jurnal Masyarakat Informatika, 5(9). https://doi.org/10.14710/jmasif.5.9.1-12

Lamghari, N., Charaf, M. E. H., & Raghay, S. (2016). Template Matching for Recognition of Handwritten Arabic Characters Using Structural Characteristics and Freeman Code. (IJCSIS) International Journal of Computer Science and Information Security, 14(12), 31â€“40.

Nugroho, D. C., Sulistiyo, M. D., & Purnama, B. (2014). Optical Character Recognition Pada Smart Phone Menggunakan Contour Analysis Dan Feature. September.

Powers, D. M. W. (2011). Evaluation: from precision, recall and F-measure to ROC, informedness, markedness and correlation. Journal of Machine Learning Technologies, 2(1). https://doi.org/10.9735/2229-3981

Purnama, A., Bahri, S., Gunawan, G., Hidayatulloh, T., & Suhada, S. (2022). Implementation of Deep Learning for Handwriting Imagery of Sundanese Script Using Convolutional Neural Network Algorithm (CNN). ILKOM Jurnal Ilmiah, 14(1), 10â€“16. https://doi.org/10.33096/ilkom.v14i1.989.10-16

Rahmawati, S. N., Hidayat, E. W., & Mubarok, H. (2021). Implementasi Deep Learning Pada Pengenalan Aksara Sunda Menggunakan Metode Convolutional Neural Network. INSERT : Information System and Emerging Technology Journal, 2(1), 46. https://doi.org/10.23887/insert.v2i1.37405

Srisha, R., & Khan, A. (2013). Morphological Operations for Image Processing : Understanding and its Applications. NCVSComs-13 Coference Proc, December, 17â€“19.

DOI: https://doi.org/10.26760/mindjournal.v7i2.177-187