Sains Malaysiana 38(4)(2009): 577–587 

 

 

Pengesanan Nombor Plat Kenderaan Menggunakan Alkhwarizmi Gugusan dan Kelancaran Jarak Larian (GKJL)

(License Plate Detection Using Cluster Run Length Smoothing Algorithm (CRLSA))

 

Siti Norul Huda Sheikh Abdullah & Khairuddin Omar*

Pusat Teknologi Kecerdasan buatan (CAIT)

Fakulti Teknologi dan Sains Maklumat, Universiti Kebangsaan Malaysia

43600 UKM Bangi, Selangor D.E. Malaysia

 

Marzuki Khalid, Rubiah Yusof

Pusat Kecerdasan buatan dan Robotik (CAIRO) Fakulti Kejuruteraan elektrik

Universiti Teknologi Malaysia, Jalan Semarak, 54100 Kuala Lumpur, Malaysia

 

Received: 20 May 2008 / Accepted: 3 November  2008

 

ABSTRAK

 

 

Pengecaman nombor plat (PNP) kenderaan telah dikaji secara intensif di kebanyakan negara. Berdasarkan perbezaan jenis nombor kenderaan yang digunakan, keperluan suatu sistem PNP adalah berlainan bagi setiap negara. Di dalam makalah ini, suatu sistem PNP automatik dicadangkan bagi kenderaan Malaysia dengan nombor plat piawai berdasarkan pada pemprosesan imej, penggugusan, pengekstrakan ciri dan rangkaian neural. Perpustakaan pemprosesan imej telah dibangunkan dalam satu pembangunan yang dirujuk sebagai Pelantar Pembangunan Sistem Penglihatan (PPSP). Tapisan penajam, tapisan minimum, tapisan median dan tapisan homomorfik telah digunakan di dalam proses pembaikan imej. Selepas penggunaan pembaikan imej, imej ditemberengkan menggunakan analisis blok, profil-profil imbasan garisan mendatar, penggugusan dan pendekatan alkhwarizmi kelancaran jarak larian untuk mengenal pasti lokasi nombor plat kenderaan. Secara keseluruhannya setiap imej dijelmakan menjadi objek-objek blok dan maklumat-maklumat penting seperti jumlah blok, lokasi, tinggi dan lebar, dianalisis bagi tujuan latihan gugusan dan pemilihan gugusan terbaik dengan blok terbanyak. Alkhwarizmi cadangan dipanggil pendekatan Alkhwarizmi Gugusan dan Kelancaran Jarak Larian (GKJL) digunakan untuk mencari lokasi nombor plat pada kedudukan yang betul. GKJL terdiri daripada dua cadangan alkhwarizmi berasingan, iaitu alkhwarizmi cadangan pengesan sisi menggunakan imej hasil topeng kernel 3×3 dan ofset 128 skala kelabu, dan hasil imej tersebut diambangkan untuk mengira Kelancaran Jarak Larian (KJL). Kedua teknik ini memperbaiki teknik gugusan dalam fasa penemberengan. Untuk menilai keberkesanannya, tiga eksperimen berasingan telah dijalankan. Jadual analisis kesilapan dibina berdasarkan kepada tiga eksperimen tersebut. Prototaip sistem mempunyai ketepatan melebihi 96% dan cadangan-cadangan untuk penambahbaikan sistem turut dibincangkan.

Kata kunci: Pengecaman nombor plat kenderaan; penggugusan; alkhwarizmi kelancaran jarak larian; nilai ambang

 

ABSTRACT

Vehicle license plate recognition has been intensively studied in many countries. Due to the different types of license plates being used, the requirement of an automatic license plate recognition system is different for each country. In this article, an automatic license plate recognition system is proposed for Malaysian vehicles with standard license plates based on image processing, clustering, feature extraction and neural networks. The image processing library is developed in-house which is referred to as Vision System Development Platform (VSDP). Sharpen filter, Minimum filter, Median Filter and Homomorphic Filter were used in the image enhancement process. After applying image enhancement, the image is segmented using blob analysis, horizontal scan line profiles, clustering and run length smoothing algorithms approach to identify the location of the license plate. Thoroughly each image is transformed into blob objects and its important information such as total bumber of blobs, location, height and width, are being analyzed for the purpose of cluster exercising and choosing the best cluster with winner blobs. A new algorithm called Cluster Run Length Smoothing Algorithm (CRLSA) approach was applied to locate the license plate at the right position. CRLSA consists of two separate proposed algorithm which applied proposed edge detector algorithm using 3×3 kernel masks and 128 grayscale offset, and the resulting image is thresholded in order to calculate Run Length Smoothing Algorithm (RLSA), which has shown to improve the clustering process in the segmentation phase. Three separate experiments were performed to analyse its effectiveness. From those experiments, analysis of error tables were constructed. The prototyped system has an accuracy of more than 96% and suggestions to further improve the system are also discussed.

Keyword: License plate recognition; clustering; run length smoothing algorithm; thresholding

 

REFERENCES

 

Aksoy, M.S., Cagil, G. &. Tucker, A.K. 2000. Number- plate recognition using inductive learning, Robotics and Autonomous Systems 33: 149-153.

Al-Badr, B. & Mahmoud, S.A. 1995. Survey and bibliography of Arabic optical test recognition, Signal Processing 41: 49-77.

Barosso, J., Dagless, E.L., Rafeal, A. & Bulas-Cruz, J. 1997. Number plate reading using computer vision, Proceedings of IEEE International Symposium on Industrial Electronics 3: 761-766.

Chang, S.L., Chen, L.S., Chung, y.C. & Chen, S.W. 2004. Automatic license plate recognition, IEEE Transaction Intelligent transportation system 5: 42-53.

Emiris, D.M. & Koulouriotis, D.E. 2001. Automated optic recognition of alphanumeric content in car license plates is semi-structured environment, Proceedings of International Conference on Image Processing 3: 50-53.

Fisher, J.L., Hinds, S.C. & Damato, D.P. 1990. A rule-based system for document image segmentation, Proceedings of 10th International Conference on Pattern Recognition 1: 567-572.

Intel Corporation. 2007. Open Source computer Vision (OpenCV) Library, http://www.intel.com/technology/computing/ opencv/ [tarikh akses: 6 Julai 2007]

Nagy, G. 1968. Preliminary investigation of techniques for automated reading of unformatted text, Communication ACM 11: 480-487.

Nijhuis, J.A.G, ter Brugge, M.H, & Helmholt, K.A. 1995. Car license plate recognition with neural network and fuzzy logic, IEEE Proceedings of International Conference on Neural Networks 5: 2232-2236. IEEE.

Trier O. D., Jain A. K., & Taxt T. 1996. feature extraction methods for character recognition-a survey. Journal of Pattern Recognition 29(4): 641-662.

Tamagama, M., Miyamoto, K. & Nagano, K. 1991. Vehicle license-plate recognition by image analysis. Proceedings of IEEE International Conference on Industrial electronics, Control and Instrumentation 3: 1734-1738. 28 Oct-1 Nov 1991.

Wong, K.Y., Casey, R.G. & Wahl, F.M. 1982. Rule based for text, horizontal solid back lines, graphic and halftone images, vertical solid black line. Document Analysis system. IBM Journal of Research and Development 26(6): 647-657.

Ying, D., Nanning, Z., Xining, Z. & Guorang, X. 1988. Automatic recognition of province name on the license plate of moving vehicle. Proceedings of 19th International Conference on Pattern Recognition 927-929, 14-17 Nov 1988.  

  *Corresponding author; email: ko@ftsm.ukm.my

previous