Raharja, Adnan Erlangga (2021) Koreksi Kemiringan Citra Menggunakan Planar Homography Untuk Pengenalan Pelat Nomor Kendaraan. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Text
05111740000100-Adnan-Erlangga-Raharja-Buku_TA.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (8MB) | Request a copy |
Abstract
Pengenalan pelat nomor kendaraan atau dikenal dengan istilah License Plate Recognition (LPR) merupakan sebuah bidang permasalahan yang berfokus pada pendeteksian dan pengenalan pelat nomor kendaraan. Dalam proses pengenalan pelat nomor akan terdapat banyak faktor eksternal yang dapat mempengaruhi proses kerja sistem, salah satunya adalah posisi dan orientasi kendaraan relatif dengan posisi kamera ketika proses pengambilan gambar. Kemiringan ini akan menimbulkan masalah seperti karakter yang bersentuhan ataupun karakter pada pelat nomor menjadi rusak. Akibatnya, ini akan memiliki efek semakin sulitnya untuk melakukan segmentasi dan pengenalan karakter pada pelat nomor. Untuk mengatasi masalah tersebut, Tugas Akhir ini membuat sistem koreksi kemiringan pelat nomor menggunakan planar homography yang kemudian akan di implementasikan pada sistem pengenalan pelat nomor kendaraan otomatis.
Koreksi kemiringan dilakukan dengan mendeteksi tepi pelat nomor menggunakan algoritma Canny Edge Detection, kemudian mencari titik sudut pelat nomor, dan melakukan koreksi kemiringan dengan menggunakan planar homography. Setelah citra pelat nomor diperbaiki kemiringannya maka proses segmentasi dan prediksi dapat dijalankan dengan baik dan tepat.
Pengujian dilakukan dalam tiga skenario yaitu pengujian deteksi sudut manual, pengujian algoritma edge detection, dan pengujian pada data video. Hasil segmentasi optimal pada video yang optimal di dapatkan dengan menggunakan algoritma Canny edge detection yang memiliki nilai rata-rata akurasi 75.88%, presisi 87.26%, dan recall 76.24% dan hasil prediksi pada video optimal dengan nilai rata-rata akurasi 77.45%, presisi 82.50%, dan recall 77.72%.
=====================================================================================================
License Plate Recognition (LPR) is a problem area that focuses on detecting and recognizing vehicle number plates. In the number plate recognition process, there will be many external factors that can affect the working process of the system, one of which is the position and orientation of the vehicle relative to the camera position during the shooting process. This slope will cause problems such as characters that touch or characters on the number plate become damaged. As a result, this will have the effect of making it more difficult to segment and recognize characters on license plates. To overcome this problem, this Final Project makes a system for correcting the slope of the license plate using planar homography which will then be implemented in an automatic vehicle number plate recognition system.
Skew correction is done by detecting the edge of the license plate using the Canny Edge Detection algorithm, then looking for the corner point of the license plate and correcting the skew using planar homography. After correcting the skew of the license plate image, the segmentation and prediction process can be carried out properly and precisely.
The tests were carried out in three scenarios, namely manual angle detection testing, edge detection algorithm testing, and video data testing. Optimal character segmentation results on optimal video are obtained using the Canny edge detection algorithm which has an average value of 75.88% accuracy, 87.26% precision, and 76.24% recall and character prediction results on optimal video with an average accuracy value of 77.45%, precision 82.50 %, and recall 77.72%.
Item Type: | Thesis (Undergraduate) |
---|---|
Uncontrolled Keywords: | License Plate Recognition, Planar Homography, Kemiringan, Canny edge detection, Slope |
Subjects: | T Technology > T Technology (General) > T385 Visualization--Technique T Technology > T Technology (General) > T57.5 Data Processing |
Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis |
Depositing User: | Adnan Erlangga Raharja |
Date Deposited: | 23 Aug 2021 05:45 |
Last Modified: | 23 Aug 2021 05:45 |
URI: | http://repository.its.ac.id/id/eprint/88898 |
Actions (login required)
View Item |