Setiawan, Muh. Shahid (2020) Deteksi dan Identifikasi Nomor Registrasi Plat Kendaraan Indoneisa Berbasis Convolutional Neural Network (CNN). Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
07211540000044-Undergraduate_Thesis.pdf Download (45MB) | Preview |
Abstract
Salah satu penerapan pengolahan citra ialah pendeteksian dan identifikasi nomor registrasi plat kendaraan atau biasa dikenal License Plate Recognition (LPR). Diluar negeri, LPR telah diterapkan ke berbagai sistem. Di Indonesia LPR sudah dikembangkan, namun pengembangannya tidak dilakukan secara komperhensif sehingga belum diterapkan ke berbagai sistem. Oleh karena itu dengan memanfaatkan Convolutional Neural Network, dibuatlah sebuah sistem yang dapat mendeteksi dan mengidentifikasi nomor registrasi plat kendaraan Indonesia yang akan diterapkan di pintu gerbang kompleks perumahan. Proses training dilakukan menggunakan YOLOv3 dengan menggunakan 1200 dataset. Data tersebut dibagi menjadi 888 image train dan 312 image test. Hasil training tertinggi yang diperoleh sebesar 55.2% mAP. Kesalahan dalam deteksi dan identifikasi nomor registrasi plat kendaraan Indonesia yaitu paling rendah hanya 1,43%. Proses komputasi dalam deteksi dan identifikasi untuk memproses 3668 frame rata-rata membutuhkan 0,585145763 tiap frame serta membutuhkan waktu 464,177 detik.
===============================================================================================================================
One application of image processing is the detection and identifica-tion of vehicle license plate numbers or commonly known as License Plate Recognition (LPR). Overseas, LPR has been applied to vari- ous systems. In Indonesia LPR has been developed, but its develo-pment is not carried out in a comprehensive manner so it has not been applied to various systems. Therefore by utilizing Convolutio-nal Neural Network , a system can be made to detect and identify Indonesian vehicle license plates number that will be implemented at the gate of the housing complex. The training process is done using YOLOv3 with 1200 datasets. The data is divided into 888 image train and 312 image test. The highest training results obtained we-re 55.2% mAP. The lowest Error in detection and identification of Indonesian vehicle plates number is only 1.43 %. The computation process in detection and identification to process 3668 frames requi-res an average of 0.585145763 per frame and takes 464,177 seconds.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSKom 006.32 Set d-1 2020 |
| Uncontrolled Keywords: | Deteksi, Identifikasi, License Plate Recognition (LPR), You Only Look Once (YOLO), Convolutional Neural Network (CNN), |
| Subjects: | Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis |
| Depositing User: | Muhammad Shahid Setiawan |
| Date Deposited: | 18 Apr 2023 03:05 |
| Last Modified: | 18 Apr 2023 03:05 |
| URI: | http://repository.its.ac.id/id/eprint/74678 |
Actions (login required)
 |
View Item |