Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

120 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AI-Based Multi-Drone Wildlife Monitoring & Intrusion Detection System

Overview

This project presents an AI-driven, software-based multi-drone monitoring system designed to address human–wildlife conflict and wildlife conservation challenges. The system uses aerial imagery (real or simulated) to detect, track, and analyze wildlife and human activities, enabling early detection of urban wildlife intrusion and potential poaching threats.


Problem Statement

Rapid urbanization and habitat fragmentation have increased human–wildlife conflict and poaching threats. Traditional monitoring methods are manual and limited, leading to delayed detection. This project proposes an intelligent aerial monitoring system using AI and simulated multi-drone coordination to support proactive conservation and urban safety.


Features

Wildlife Detection & Tracking

Detects and tracks animals using deep learning models.

Human & Poacher Detection

Identifies suspicious human activities in restricted zones.

Movement Analysis & Prediction

Analyzes animal movement patterns and predicts high-risk zones.

Multi-Drone Coordination (Simulated)

Allocates surveillance areas and avoids redundancy across drones.

Urban Wildlife Intrusion Detection

Detects wildlife entering urban areas and triggers alerts.

Interactive Dashboard

Visualizes detections, tracking, and risk alerts in real-time.


Methodology

1. Data Collection & Preprocessing

  • Use UAV/wildlife datasets and simulated drone feeds
  • Perform cleaning, annotation, and normalization

2. Object Detection

  • YOLOv8 model for detecting animals and humans

3. Tracking

  • DeepSORT for tracking movement across frames

4. Behavior Analysis

  • Identify suspicious patterns (poaching indicators)

5. Multi-Drone Coordination

  • Simulated environment for zone allocation and coverage

6. Predictive Analysis

  • Identify high-risk zones using spatio-temporal patterns

7. Visualization

  • Dashboard for real-time monitoring and alerts

Tech Stack

  • Language: Python
  • Deep Learning: PyTorch / TensorFlow
  • Computer Vision: OpenCV, YOLOv8 (Ultralytics)
  • Tracking: DeepSORT
  • Data Handling: NumPy, Pandas
  • Visualization: Streamlit / Flask
  • Development: Jupyter Notebook / Google Colab

Evaluation Metrics

  • Accuracy
  • Precision & Recall
  • mAP (Mean Average Precision)
  • Tracking Accuracy
  • Confusion Matrix

Deliverables

  • Wildlife detection and tracking system
  • Poacher detection module
  • Urban intrusion alert system
  • Multi-drone coordination (simulation)
  • Interactive dashboard

Scope

This project focuses on software-based analysis using aerial data and simulated drones. It does not involve physical drone control and serves as a decision-support system for wildlife conservation and urban safety.


Applications

  • Forest and wildlife conservation
  • Anti-poaching surveillance
  • Urban safety monitoring
  • Smart city integration
  • Disaster and environmental monitoring

Future Work

  • Integration with real-time drone feeds
  • Thermal imaging for night detection
  • GPS-based wildlife tracking
  • Real-world multi-drone deployment

Team

  • Venkatesh Paitwar
  • Sandarbh Singh
  • Srushti Agrawal
  • Shravani Borde

Note

This project is developed as a Third Year Project (Data Science / AI domain) and focuses on real-world problem-solving using modern deep learning techniques.

About

Mini project.......

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages