Pemilihan Parameter Smoothing pada Probabilistic Neural Network dengan Menggunakan Particle Swarm Optimization untuk Pendeteksian Teks Pada Citra

Endah Ekasanti Saputri, Romi Satria Wahono, Vincent Suhartono

Abstract


Teks sering dijumpai di berbagai tempat seperti nama jalan, nama toko, spanduk, penunjuk jalan, peringatan, dan lain sebagainya. Deteksi teks terbagi menjadi tiga pendekatan yaitu pendekatan tekstur, pendekatan edge, dan pendekatan Connected Component. Pendekatan tekstur dapat mendeteksi teks dengan baik, namun membutuhkan data training yang banyak. Probabilistic Neural Netwok (PNN) dapat mengatasi permasalahan tersebut. Namun PNN memiliki permasalahan dalam menentukan nilai parameter smoothing yang biasanya dilakukan secara trial and error. Particle Swarm Optimization (PSO) merupakan algoritma optimasi yang dapat menangani permasalahan pada PNN. Pada penelitian ini, PNN digunakan pada pendekatan tekstur guna menangani permasalahan pada pendekatan tekstur, yaitu banyaknya data training yang dibutuhkan. Selain itu, digunakan PSO untuk menentukan parameter smoothing pada PNN agar akurasi yang dihasilkan PNN-PSO lebih baik dari PNN tradisional. Hasil eksperimen menunjukkan PNN dapat mendeteksi teks dengan akurasi 75,42% hanya dengan mengunakan 300 data training, dan menghasilkan 77,75% dengan menggunakan 1500 data training. Sedangkan PNN-PSO dapat menghasilkan akurasi 76,91% dengan menggunakan 300 data training dan 77,89% dengan menggunakan  1500 data training. Maka dapat disimpulkan bahwa PNN dapat mendeteksi teks dengan baik walaupun data training yang digunakan sedikit dan dapat mengatasi permasalahan pada pendekatan tekstur. Sedangkan, PSO dapat menentukan nilai parameter smoothing pada PNN dan menghasilkan akurasi yang lebih baik dari PNN tradisional, yaitu dengan peningkatan akurasi sekitar 0,1% hingga 1,5%. Selain itu, penggunaan PSO pada PNN dapat digunakan dalam menentukan nilai parameter smoothing  secara otomatis pada dataset yang berbeda.

 


Full Text:

PDF

References


Angadi, S. A. (2010). Text Region Extraction from Low Resolution Natural Scene Images using Texture Features. International Advance Computing Conference, 121-128.

Bai, B., Yin, F., & Liu, C. (2012). A Fast Stroke-Based Method for Text Detection in Video. IAPR International Workshop on Document Analysis Systems, 69-73.

Donald, F. S. (1990). Probabilistic Neural Networks. Neural Network I Pergamon Press pie Original Contribution, 3.

Epshtein, B., Ofek, E., & Wexler, Y. (2010). Detecting text in natural scenes with stroke width transform. Computer Vision and Pattern Recognition (CVPR), 10, 2963-2970.

Huang, X. (2012). Automatic Video Text Detection and Localization Based on Coarseness Texture. International Conference on Intelligent Computation Technology and Automation, 1(2).

Jamil, A., Siddiqi, I., Arif, F., & Raza, A. (2011). Edge-based Features for Localization of Artificial Urdu Text in Video Images. International Conference on Document Analysis and Recognition, 1120-1124.

Ji, R., Xu, P., Yao, H., Zhang, Z., Sun, X., & Liu, T. (2008). Directional correlation analysis of local Haar binary pattern for text detection. IEEE International Conference on Multimedia and Expo, 885-888.

Jorhedi, R, A., & Jasni, J. (2013). Parameter Selection in Particle Swarm Optimization: a survey. Journal of Experimental & Theoretical Artificial Intelligence, 25, 527–542.

Karaoglu, S., Fernando, B., Trémeau, A., & Etienne, S. (2010). A Novel Algorithm for Text Detection and Localization in Natural Scene Images. IEEE Digital Image Computing : Techniques and Applications, 641-648.

Kennedy, J., & Eberhart, R. (1995). Particle Swarm Optimization. 1942-1948.

Le, H. P., Dinh, T. N., Kim, S. H., & Lee, G. S. (2010). Text Detection in Binarized Text Images of Korean Signboard by Stroke Width Feature. IEEE International Conference on Computer and Information Technology (CIT 2010), 1588-1592.

Lee, J., Lee, P., Lee, S., Yuille, A., & Koch, C. (2011). AdaBoost for Text Detection in Natural Scene. International Conference on Document Analysis and Recognition.

Liu, B., Wang, H., & Cheng, X. (2011). Exchange Rate Forecasting Method Based on Particle Swarm Optimization and Probabilistic Neural Network Model. International Conference on Network Computing and Information Security, 288–292.

Meng, Q., & Song, Y. (2012). Text Detection in Natural Scenes with Salient Region. IAPR International Workshop on Document Analysis System, 384-388.

Pan, Y,F., Liu, C., & Hou, X. (2010). Fast scene text localization by learning-based filtering and verification. IEEE International Conference on Image Processing, 2269-2272.

Shi, C., Xiao, B., Wang, C., & Zhang, Y. (2012). Graph-based Background Suppression For Scene Text Detection. IAPR International Workshop on Document Analysis System, 210-214.

Shivakumara, P., & Tan, C. L. (2010). New Wavelet and Color Features for Text Detection in Video. International Conference on Pattern Recognition.

Shivakumara, P., Phan, T. Q., & Tan, C. L. (2011). A Laplacian Approach to Multi-Oriented Text Detection in Video. IEEE Transaction on Pattern Analysis and Machine Intelligence, 33(2), 412-419.

Wang, C., & Wang, H. (2010). Utilization of Temporal Continuity in Video Text Detection. International Conference on MultiMedia and Information Technology, 343-346.

Wen-xu, W. (2011). Fault diagnosis of transformer based on probabilistic neural network. International Conference on Intelligent Computation Technology and Automation.

Yang, H. (2012). An Improved Probabilistic Neural Network with GA Optimization. International Conference on Intelligent Computation Technology and Automation, 76-79.

Zhang, H., Zhao, K., Song, Z,Y., & Guo, J. (2013). Text extraction from natural scene image: A survey. Neurocomputing, 122, 310-323.

Zhao, H., Liu, C., Wang, H., & Li, C. (2010). Classifying ECoG Signals Using Probabilistic Neural Network. WASE International Conference on Information Engineering.


Refbacks

  • There are currently no refbacks.




Journal of Intelligent Systems(JIS, ISSN 2356-3982)
Copyright © 2020IlmuKomputer.Com. All rights reserved.