OmniSense Edge Security System

OmniSense is an enterprise-grade IoT edge computing platform designed for smart classroom monitoring, automated attendance, and real-time security. It leverages distributed ESP32 microcontrollers, Time-of-Flight (ToF) laser sensors, and a centralized Python gateway powered by AI facial recognition.

System Architecture

The OmniSense infrastructure is built on three core components:

1. OmniSense Edge Gateway (gateway.py)

A centralized management console built with Python and Tkinter. It acts as the brain of the operation, processing incoming telemetry and video streams.

• AI Face Recognition: Uses insightface and faiss to perform real-time facial recognition on incoming camera streams to automate student attendance and detect unauthorized strangers (spoofing/liveness detection included).
• Live CCTV Dashboard: Features a dynamic UI with a live CCTV feed viewer, allowing security personnel to monitor any classroom over the local network.
• Hardware Integration: Constantly polls connected hardware nodes to update the dashboard with live room occupancy, connection status, and telemetry.
• Auto-Discovery: Automatically discovers connected fleet nodes on the local subnet.

2. Main Hardware Node (espMain.ino)

An ESP32-based microcontroller placed at the entrance of a room.

• ToF Occupancy Tracking: Uses a SparkFun VL53L1X Time-of-Flight sensor to accurately track people entering and exiting the room via bi-directional laser zones, maintaining a highly accurate occupancy count.
• Automated Relay Control: Automatically toggles a relay (e.g., room lights or HVAC) based on the occupancy count (turns off when empty).
• Access Acknowledgment: Features a buzzer that is triggered remotely by the gateway to provide a short audible beep when an authorized (known) person is recognized. Unauthorized (unknown) persons are logged silently without triggering an alarm.
• RESTful API: Hosts local endpoints like /status (returns live occupancy JSON) and /ring (triggers the buzzer).

3. Camera Node (server_ip.ino)

An ESP32-CAM module positioned for facial capture.

• MJPEG Streaming: Serves a continuous, high-framerate MJPEG video stream over HTTP.
• MQTT Integration: Connects to an MQTT broker for lightweight telemetry and remote status broadcasting.
• Optimized for Edge: Operates efficiently to provide low-latency frames directly to the AI Gateway.

Installation & Setup

Prerequisites

• Python 3.10+
• pip install -r requirements.txt (Required libraries: opencv-python, numpy, pillow, insightface, faiss-cpu, requests)
• Arduino IDE for flashing ESP32 modules.

Flashing the Hardware

1. ESP32 Main Node: Open espMain/espMain.ino. Update the ssid and password variables with your local Wi-Fi credentials. Flash to a standard ESP32.
2. ESP32-CAM Node: Open server_ip/server_ip.ino. Update the ssid and password variables. Flash to an AI Thinker ESP32-CAM module.

Running the Gateway

1. Ensure all ESP32 modules are powered on and connected to the same Wi-Fi network as the host machine.
2. Run the gateway:

   python gateway.py

3. Use the Dynamic Fleet Setup panel to add your rooms. Enter the Room Name, the IP of the ESP32-CAM, and the URL of the ESP32 Main Node.
4. Click Initialize Infrastructure to start the AI processing pipeline and live monitoring.

Directory Structure

• /espMain/ - Source code for the ESP32 Main Node (ToF, LCD, Buzzer).
• /server_ip/ - Source code for the ESP32-CAM Node (MJPEG Streamer).
• gateway.py - Main Python application and AI pipeline.
• omnisense_config.json - Automatically generated configuration file storing fleet IPs and settings.
• /dataset/ - Directory for storing known face reference images.
• /Attendance_Logs/ - Auto-generated CSV logs and audit images for tracking events.