car-plate-detection-yolov26

Car License Plate Detection (YOLOv26)

This repository contains a YOLOv26 object detection model fine-tuned for Car License Plate Detection. The model is built using the Ultralytics framework.

Model Details

• Model Type: Object Detection (License Plate)
• Architecture: YOLOv26n (Nano variant)
• Training Epochs: 10
• Input Image Size: 640

Evaluation Metrics (Best Accuracy)

• Precision: 63.8%
• Recall: 73.2%
• mAP50: 74.7%
• mAP50-95: 41.3%

Dataset

The model was trained on the Car License Plate Detection dataset from Kaggle, which consists of images with bounding box annotations for license plates.

How to Use

1. Installation

To use this model, you need to install the ultralytics library:

pip install ultralytics

2. Inference Code

You can use the following snippet to run detection on an image:

from ultralytics import YOLO

# Load the fine-tuned model
model = YOLO('path/to/best.pt')

# Run inference on an image
results = model('car_image.jpg')

# Display or process results
for result in results:
    boxes = result.boxes
    for box in boxes:
        print(f"Detected License Plate: {box.xyxy}") # Bounding box coordinates
        print(f"Confidence: {box.conf}")

Repository Structure

• best.pt: The fine-tuned model weights (PyTorch format).
• args.yaml: Training configuration and hyperparameters.
• results.csv: Training logs per epoch.
• metrics.json: Final evaluation metrics.
• run_summary.json: Summary of the training run.

Credits

• Original Dataset: Andrew MVD (Kaggle)
• Framework: Ultralytics YOLO