License Plate Detection using YOLOv10

This project demonstrates real-time license plate detection using the YOLOv10 model. It leverages a custom dataset to detect number plates in live video streams.

Table of Contents

• Overview
• Installation
• Dataset
• Model
• Usage
• Results
• License

Overview

This project aims to detect and localize license plates in real-time using the YOLOv10 architecture. The model is trained on a custom dataset to accurately identify number plates in various conditions.

Installation

1. Clone the repository:

   git clone https://github.com/alihassanml/License-Plate-Detection-using-YOLOv10.git
   cd License-Plate-Detection-using-YOLOv10

2. Install the required dependencies:

   pip install -r requirements.txt

3. Download the dataset from Kaggle:

   !kaggle datasets download -d alihassanml/yolo-number-plates

4. Extract the dataset:

   unzip yolo-number-plates.zip -d dataset/

Dataset

The dataset used for this project is a custom set containing labeled images of number plates. You can download it from Kaggle.

Model

YOLOv10 is used for real-time object detection, known for its speed and accuracy. The model is fine-tuned to detect number plates in various lighting and environmental conditions.

Usage

1. Train the model using the dataset:

   python train.py --data dataset/data.yaml --cfg cfg/yolov10.yaml --weights yolov10.pt

2. Run the live number plate detection:

   python detect.py --source 0 --weights yolov10.pt --conf 0.5

   Replace 0 with a video file path if needed.

Results

The model is capable of detecting number plates in real-time with high accuracy. Sample results are shown below:

• Detected license plates in various live video streams
• Accurate bounding boxes even in low-light conditions

License

This project is licensed under the MIT License. See the LICENSE file for details.