Aerial Vehicle Mapper 🛁

Welcome to Aerial Vehicle Mapper! This repository provides a powerful tool for detecting and mapping vehicles from aerial images using the Detectron2 deep learning framework. 🚗🛰️

---

Features ✨

• Vehicle Detection: Detects cars, trucks, buses, and motorbikes in aerial images.
• High Accuracy: Built on Detectron2's advanced object detection models.
• Custom Thresholding: Adjust detection confidence levels for greater flexibility.
• Visualization: Produces annotated images with bounding boxes for detected vehicles.

---

Applications 🌍

• Urban Planning: Analyze traffic patterns and vehicle density in cities.
• Surveillance: Monitor vehicle movements for security purposes.
• Disaster Management: Identify vehicles in affected areas during emergencies.
• Environmental Studies: Study vehicular impact on ecosystems.
• Infrastructure Development: Aid in designing better road networks and parking facilities.

---

Prerequisites 🛠️

Make sure you have the following:

• Python 3.8+
• NVIDIA GPU with CUDA support (for optimal performance)
• pip for package management

---

Examples 🗂️

Input Image

Output Image

---

How It Works ⚙️

graph TD;
    A[Load Aerial Image] -->|Preprocess Image| B(Resize & Normalize);
    B -->|Pass to Model| C[Detectron2 Model];
    C -->|Run Inference| D{Object Detection};
    D -->|Extract Vehicle Coordinates| E[Bounding Box Extraction];
    E -->|Apply Confidence Threshold| F[Filter Detections];
    F -->|Overlay Annotations| G[Visualize Results];
    G -->|Save Annotated Image| H[Store Output];
    H -->|Integrate with GIS| I[Geospatial Mapping];
    I -->|Generate Reports| J[Final Output];

Loading

---

Dependencies 🛆

• Detectron2
• PyTorch
• OpenCV
• NumPy
• PyYAML

Happy Mapping! 🗰️