DETEKSI PLAT NOMOR KENDARAAN PADA UNIVERSITAS XYZ MENGGUNAKAN METODE MSER (MAXIMALLY STABLE EXTREMAL REGIONS)

Andhika Wisnu Widyatama; Kusrini Kusrini; Sudarmawan Sudarmawan

doi:10.46764/teknimedia.v3i2.63

Andhika Wisnu Widyatama Universitas Amikom Yogyakarta
Kusrini Kusrini Magister Teknik Informatika, Universitas Amikom Yogyakarta
Sudarmawan Sudarmawan Magister Teknik Informatika, Universitas Amikom Yogyakarta

DOI: https://doi.org/10.46764/teknimedia.v3i2.63

Keywords: Vehicle, Object Detection, MSER

Abstract

The increasing number of students at XYZ Yogyakarta University every year is balanced with the number of vehicles used, it is necessary to pay attention to developing efficient intelligence and also a safe transportation system. Several studies of plate recognition methods have shown promising performance, but some methods may have weaknesses in situations of plate position and orientation variations, lighting, background, and non-plate objects. This research uses the Maximally Stable Extremal Regions (MSER) method which is a keypoint search method based on the extremal region. The MSER method has been identified as one of the best region detectors due to its resistance to changes in viewing angle, scale, and exposure, as well as sensitivity to image blur. From the results obtained a detection rate of 80% can detect vehicle number plates well.

References

[1] https://www.bps.go.id/indicator/17/57/1/jumlah-kendaraan-bermotor.html
[2] Darma Putra (2010) Pengolahan Citra Digital. 1st edn. Edited by Westriningsih. Yogyakarta: C.V ANDI OFFSET.
[3] Gou, C. et al (2014) ‘Based on ELM’, pp. 217–221
[4] Mikolajczyk, K. et al. (2005) ‘A comparison of affine region detectors’, International Journal of Computer Vision, 65(1–2), pp. 43–72. doi: 10.1007/s11263-005- 3848-x.
[5] Fraundorfer, F., & Bischof, H. (2005, June). A novel performance evaluation method of local detectors on non-planar scenes. In Computer Vision and Pattern Recognition-Workshops, 2005. CVPR Workshops. IEEE Computer Society Conference on (pp. 33-33). IEEE.
[6] Bahtiar, A. (2016, December). Sistem Deteksi Nomor Polisi Mobil dengan Menggunakan Metode Haar Classifier dan OCR guna Mempermudah Administrasi Pembayaran Parkir. Jurnal of Informatika and Technology, 4(1), 40-46
[7] Rainarli, E. (2020) ‘Maximally Stable Extremal Regions and Naïve Bayes to Detect Scene Text’, in IOP Conference Series: Materials Science and Engineering. IOP Publishing Ltd. doi: 10.1088/1757-899X/879/1/012106.
[8] Neumann, L. and Matas, J. (no date) A method for text localization and recognition in real-world images.
[9] https://id.wikipedia.org/wiki/Pelat_nomor
[10] J. Arief, F. Utaminingrum, and Y. A. Sari, “Penentuan Jumlah Karakter pada Plat Nomor Kendaraan dengan menggunakan Selective Ratio Bounding Box,” J. Pengemb. Teknol. Inf. dan Ilmu Komput., vol.2, no. 1, pp. 256–262, 2018.
[11] T. A. Wijaya and Y. Prayudi, “Implementasi Visi Komputer dan Segmentasi Citra Untuk Klarifikasi Bobot Telur Ayam RAS,” ISSN 1907-5022, vol. 2010, no. Snati, pp. 1–5,2010.