Segmentasi Karakter untuk Pengenalan Pelat Nomor Kendaraan Berbasis Clustering dan Projection Histogram

Prastomo, Mohammad Rizaldi Huzein (2021) Segmentasi Karakter untuk Pengenalan Pelat Nomor Kendaraan Berbasis Clustering dan Projection Histogram. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 05111740000024-Undergraduate_Thesis.pdf] Text
05111740000024-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 October 2023.

Download (2MB) | Request a copy

Abstract

Teknologi pengenalan pelat nomor kendaraan atau License Plate Recognition (LPR) telah diadopsi di banyak aplikasi lalu lintas modern, seperti tempat parkir, pemantauan lalu lintas, dan pengawasan keamanan jalan. Pengenalan pelat nomor biasanya terdiri dari tiga tahapan penting, yakni deteksi pelat nomor, segementasi karakter, dan pengenalan karakter. Untuk mendapatkan hasil yang baik dalam membaca pelat nomor, dibutuhkan segmentasi karakter yang akurat. Apabila segmentasi karakter gagal, maka hasil pengenalan pelat nomor kendaraan menjadi tidak maksimal.
Pada Tugas Akhir ini akan dibuat skema optimasi segmentasi karakter untuk sistem pengenalan pelat nomor kendaraan menggunakan analisa histogram proyeksi dan segmentasi citra dengan clustering. Analisa histogram proyeksi merupakan proses pemilihan kandidat daerah yang hanya terdiri atas karakter pelat nomor dengan menggunakan histogram proyeksi piksel sebagai metode perhitungannya. Selanjutnya, clustering pada citra digunakan untuk meningkatkan hasil segmentasi citra dengan melakukan pengelompokkan terhadap piksel yang memiliki kemiripan saja.
Pengujian dilakukan dalam lima skenario yakni pengujian koreksi kemiringan citra, pengujian analisa histogram proyeksi piksel, pengujian algoritma clustering, pengujian pengenalan karakter pada citra, dan pengujian pada data video yang membandingkan sistem pada penelitian sebelumnya dengan sistem yang diusulkan pada Tugas Akhir ini. Hasil optimal didapatkan dengan menggunakan koreksi kemiringan, analisa histogram, dan K-Means clustering dengan rata-rata akurasi segmentasi 90.40%. Selain itu hasil pengenalan pada majority vote dalam sistem Tugas Akhir ini memberikan lebih baik dari penelitian sebelumnya dengan rata-rata 95.34%.
======================================================================================================
License Plate Recognition (LPR) technology has been adopted in many modern traffic applications, such autonomous parking lots, traffic monitoring, and road safety surveillance. Number plate recognition usually consists of three important stages, namely license plate detection, character segmentation, and character recognition. To get the optimal results in reading license plates, accurate character segmentation is required. If the character segmentation fails, the results of the vehicle number plate recognition may not be optimal.
In this final project, a character segmentation optimization scheme will be made for the vehicle number plate recognition system using projection histogram analysis and image segmentation with clustering. Projection histogram analysis is a regional candidate selection process that only consists of number plate characters using a pixel projection histogram as the calculation method. Furthermore, image clustering is used to improve the results of image segmentation by grouping only pixels that have similarities.
The tests were carried out in five scenarios, namely image skew correction testing, pixel projection histogram analysis testing, clustering algorithm testing, character recognition testing on images, and video data testing comparing the system in previous studies with the system proposed in this research. Optimal results are obtained by using skew correction, histogram analysis, and K-Means clustering with an average on accuracy of 90.40%. In addition, the results of the introduction to the majority vote in this research are better than the previous research with an average of 95.34%.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: License Plate Recognition, Segmentasi citra, Projection Histogram, Clustering, K-Means.
Subjects: Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines.
Q Science > QA Mathematics > QA76.6 Computer programming.
R Medicine > R Medicine (General) > R858 Deep Learning
T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7882.P3 Pattern recognition systems
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User: Mohammad Rizaldi Huzein Prastomo
Date Deposited: 29 Jul 2021 09:19
Last Modified: 29 Jul 2021 09:19
URI: http://repository.its.ac.id/id/eprint/84582

Actions (login required)

View Item View Item