Nirikshon: Real-Time Object Detection System

Author: Imtiaz Khandoker Email: czar.imtiaz@gmail.com GitHub: @1mt142

Figure 1: System Architecture Overview

This system utilizes an HD camera and AI-powered object detection to identify harmful or unexpected objects in different environments using trained data models. It operates in two modes:

1. Household or Ground-Level Security: The system can detect threats such as intruders or even stray animals like dogs entering a restricted area (e.g., your home, backyard, or office premises). When a threat is identified through the camera, it captures an image and sends a real-time alert — including the image — to your WhatsApp or other preferred communication channels.

2. Sky Protection: For aerial threats, such as missiles or flying objects, the system uses the HD camera to perform real-time detection. Upon identifying a potential airborne threat, it captures the image and sends an alert. Following detection, a radar module is activated to continuously track the object’s movement in the sky, enabling enhanced surveillance and threat monitoring.

By combining AI with visual and radar-based tracking, the system ensures flexible protection for both ground-level and sky-based security threats.

How It Works

1. Detection: The system continuously analyzes video feeds for objects using predefined detection models.
2. Capture: Once a object is detected, the system captures an image of the object.
3. Notification: The captured image is sent instantly to your WhatsApp via a pre-configured session.

Setup

Ensure your detection model includes harmful objects and is properly configured. Set up WhatsApp Web integration as per the installation instructions.

📌 Table of Contents

• Features
• Installation
• Configuration
• Usage
• Technical Details
• Troubleshooting

🌟 Features

Core Functionality

• Real-time object detection using YOLOv8
• Multi-threaded processing for smooth performance
• Configurable target object detection
• Confidence threshold filtering

Alert Systems

• WhatsApp instant notifications
• Local sound alerts (.mp3)
• Visual on-screen counters

Camera Support

• Built-in webcams
• IP cameras (RTSP/HTTP)
• Auto-fallback to available sources

Data Management

• Automatic snapshot saving
• Organized detection archive
• Timestamped evidence logging

🛠 Installation

Prerequisites

• Python 3.8+
• OpenCV-compatible camera
• WhatsApp Web session

Setup Steps

# Clone repository
# Nirikshon: Real-Time Object Detection System
1mtgit clone https://github.com/1mt142/nirikshon_ai.git
cd nirikshon_ai

# Create virtual environment
python -m venv venv

# Activate environment
# Windows:
.\venv\Scripts\activate
# Linux/Mac:
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

🤝 Contributing

Pull requests are welcome! For major changes, please open an issue first.