Malaysian Car Plate Dataset

Project Context
This dataset was developed as part of a Final Year Project (FYP) at the Faculty of Information Science and Technology (FIST), Multimedia University (MMU). The broader project focuses on computational intelligence and smart campus infrastructure, specifically involving the conversion of standard local webcams into network-accessible IP cameras for real-time video surveillance and processing.

This dataset serves as the foundational data to train a computer vision model for Lisence Plate Recognition (LRP), aiming to automate vehicle access control and parking management within a campus or residential environment.

Content
The dataset contains a carefully curated collection of images featuring Malaysian vehicle license plates. Malaysian car plates are unique because they often feature varying font styles (including italic or cursive on older/custom plates), different sizes, and both single-row and double-row formats.

Total Images: Insert Number 500+ images
Format: .jpg and .png
Annotations: Bounding box annotations are provided in YOLO format, making it ready for plug-and-play training with modern object detection algorithms.

Target Classes:
1. car_plate
2. car

Collection Methodology
I collected data, but some of the images are using an internet dataset, which I sorted myself. The images capture vehicles from various angles (front and rear) and under different environmental/lighting conditions (daytime glare, shadows, etc.) to ensure the trained model is robust in real-world scenarios.

The images were manually reviewed, and precise bounding boxes were drawn around the license plates using Roboflow.

Inspiration & Potential Use Cases
This dataset is ideal for computer vision researchers and developers working on:
ANPR / ALPR Systems: Training object recognition models (like YOLOv8, SSD, or Faster R-CNN) to detect and crop Malaysian license plates for downstream Optical Character Recognition (OCR).
Smart Parking: Integrating real-time computer vision with IoT devices for automated boom-gate access and vehicle logging.
Feature Extraction Algorithms: Testing image processing techniques (like OpenCV morphological operations, ORB, or SIFT) to isolate plate characters against complex vehicle backgrounds.

Acknowledgements
If you utilize this dataset in your own research, algorithms, or academic projects, please consider citing this repository. Special thanks to the Faculty of Information Science and Technology (FIST) at MMU for the academic support throughout this Final Year Project.