Performance Analysis of License Plate Character Recognition Using Two-Stage YOLOv8n with CLAHE Gamma NLM Preprocessing

Putra, Gredy Christian Hendrawan (2026) Performance Analysis of License Plate Character Recognition Using Two-Stage YOLOv8n with CLAHE Gamma NLM Preprocessing. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) Cover.pdf Download (2MB)
	Text (Bab 1) Bab 1.pdf Download (188kB)
	Text (Bab 2) Bab 2.pdf Restricted to Repository staff only until 20 July 2029. Download (1MB) \| Request a copy
	Text (Bab 3) Bab 3.pdf Restricted to Repository staff only until 20 July 2029. Download (1MB) \| Request a copy
	Text (Bab 4) Bab 4.pdf Restricted to Repository staff only until 20 July 2029. Download (9MB) \| Request a copy
	Text (Bab 5) Bab 5.pdf Download (90kB)
	Text (Daftar Pustaka) Daftar Pustaka.pdf Download (96kB)
	Text (Lampiran) Lampiran.pdf Restricted to Repository staff only until 20 July 2029. Download (3MB) \| Request a copy

Abstract

This study discusses the development of a detection and character recognition program for Indonesian vehicle license plates using the YOLOv8n architecture in a two-stage pipeline, namely plate detection and character recognition. The research focuses on analyzing the effect of image preprocessing on model performance by comparing YOLOv8n trained and tested on preprocessed image data with YOLOv8n trained and tested on data without preprocessing. In the preprocessing scenario, the training and testing datasets are first processed with Grayscale, CLAHE, Gamma Correction, and Non-Local Means (NLM) Denoising consistently before being fed into the model. The methodology includes preparing the hardware and software, collecting Dataset Model A for plate detection and Model B for character recognition, performing data splitting, and training two separate YOLOv8n models that are evaluated using precision, recall, F1-Score, mAP@0.5, and mAP@0.5:0.95. The system is tested using data captured under morning, noon, and night conditions, and evaluated with the specific metrics Plate Success Rate per condition (PSRₖ), Character Success Rate per condition (CSRₖ), and Plate Recognition Rate per condition (PRRₖ). The same process is also applied in the scenario without preprocessing, resulting in two sets of evaluation results: with preprocessing and without preprocessing. The two results were then compared to determine which scenario is more suitable for use in object detection tasks, particularly license plate recognition.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Haromainy, Muhammad Muharrom Al	NIP199506012022031006	muhammad.muharrom.if@upnjatim.ac.id
Thesis advisor	Junaidi, Achmad	NIP197811102025211048	achmadjunaidi.if@upnjatim.ac.id

Subjects:

T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Gredy Christian Hendrawan Putra

Date Deposited:

21 Jul 2026 01:10

Last Modified:

21 Jul 2026 02:03

URI:

https://repository.upnjatim.ac.id/id/eprint/55015