MarkHmnv / MineGuard

The repository contains software and a neural network model specifically developed for the MineGuard project

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

MineGuard

The repository contains software and a neural network model specifically developed for the MineGuard project

Overview

Mineguard is a comprehensive kit designed for drone-based remote landmine detection. It utilizes three primary sensors: an RGB camera, a thermal imager, and a metal detector. These sensors gather data from the drone's surroundings, which is then processed by independent neural networks for the detection and classification of landmines. The project leverages advanced technologies such as PyTorch, YOLOv8, deep learning, and data engineering to achieve accurate and efficient landmine detection.

Features

  • Multi-sensor Integration: Mineguard combines data from RGB camera, thermal imager, and metal detector for comprehensive landmine detection.
  • Neural Network Processing: Data from each sensor is processed by independent neural networks for precise classification of detected landmines.
  • Real-time Monitoring: The kit provides real-time monitoring of landmine detection activities, enabling swift response and action.
  • Interactive Map: Detected landmines are plotted on an interactive map for visualization and analysis.
  • Scalability: Mineguard is scalable and can be deployed on various platforms for different applications and terrains.

Technologies Used

  • PyTorch: Deep learning framework for implementing neural network models.
  • YOLOv8: Object detection model used for identifying objects in images and video streams.
  • Folium: Python library for creating interactive maps.
  • OpenCV: Library for computer vision tasks such as image processing and object detection.
  • TorchAudio: PyTorch extension for audio processing tasks.

Getting Started

Our software is originally used on the Nvidia Jetson Nano which is then attached to any drone

  1. Clone the Mineguard repository to your local machine.
  2. Install the required dependencies using pip install -r requirements.txt.
  3. Download pre-trained models for YOLOv8 and the landmine classification model.
  4. Connect the RGB camera, thermal imager, and metal detector to the device.
  5. Run the main.py script to start the Mineguard system.
  6. Monitor the drone's flight and view detected landmines on the interactive map.

Additional Information

Check out our PDF presentation for additional information.

About

The repository contains software and a neural network model specifically developed for the MineGuard project

License:GNU General Public License v3.0


Languages

Language:Python 98.0%Language:HTML 2.0%