Syuhaila, Farah Qoonita (2020) Sistem Deteksi Plat Nomor Otomatis Menggunakan YOLO V3 Dengan Framework Darknet. Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
02311640000120_Undergraduate_thesis.pdf Download (2MB) | Preview |
Abstract
Otomatisasi sistem dalam pengawasan transportasi memiliki peran penting untuk pendeteksian pelanggaran yang terjadi di jalan raya. Pengawasan kendaraan yang melakukan pelanggaran salah satu diantranya bisa dilakukan dengan cara pengidentifikasian plat nomor dari kendaraan dengan standar SNI. Tahapan dalam pendeteksian plat nomor kendaraan, dimulai dari lokalisasi plat nomor kendaraan lalu dilanjutkan dengan pembacaan karakter pada plat nomor. Pada penelitian ini dikembangkan suatu sistem pendeteksian plat nomor kendaraan secara otomatis menggunakan Algoritma YOLO V3 dengan framework darknet. Langkah awal yang dilakukan dalam pembuatan sistem deteksi adalah pendeteksian objek menggunakan Algoritma YOLO V3 dan dilanjutkan dengan pendeteksian karakter menggunakan Tesseract OCR. Hasil pengujian menunjukan bahwa sistem pendeteksian plat nomor otomatis mencapai tingkat akurasi 98,52% dengan rata-rata pembacaan pada 3,03 detik dan untuk hasil OCR menggunakan Software Tesseract di dapatkan hasil deteksi 20% dimana sistem berhasil untuk mengenali seluruh karakter pada plat mobil yang berupa karakter Alphanumerik sebanyak 6-7 karakter
==================================================================================================================
Automation of the system in transportation surveillance has an important role for the detection of violations that occur on the highway. Surveillance of vehicles that violate one of them can be done by identifying the license plates of vehicles with SNI standards. Stages in the detection of vehicle license plates, starting from localization of vehicle license plates and then followed by reading the characters on the license plate. In this study a vehicle license plate detection system was developed automatically using the YOLO V3 algorithm with the darknet framework. The first step taken in making the detection system is object detection using the YOLO V3 algorithm and continued with character detection using the Tesseract OCR. The test results show that the automatic number plate detection system reaches an accuracy level of 98.52% with an average reading of 3.03 secon and for the OCR results using the Tesseract Software, a 20% detection result was obtained where the system succeeded in recognizing all characters on the car plate in the form of 6-7 alphanumeric character.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | CNN, Deep learning, Object detection, Tesseract, YOLO. |
| Subjects: | T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing |
| Divisions: | Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | Farah Qoonita Syuhaila |
| Date Deposited: | 28 Aug 2020 05:26 |
| Last Modified: | 23 May 2023 05:09 |
| URI: | http://repository.its.ac.id/id/eprint/81598 |
Actions (login required)
 |
View Item |