---

Smart Agri Project

A comprehensive IoT-powered agriculture monitoring and plant health prediction system. This project integrates ESP32-based hardware sensors, a Python Flask web server, and a modern web interface to provide real-time environmental data and AI-driven plant disease classification.

---

Table of Contents

• Overview
• Features
• Hardware Components
• Software Components
• Installation & Setup
• Usage
• Directory Structure
• API Endpoints
• License
• Mobile App Availability

---

Overview

This project enables smart agriculture by:

• Collecting real-time data (temperature, humidity, soil moisture, NPK values) from the field using ESP32 microcontrollers and sensors.
• Sending this data to a Flask-based backend for visualization and further processing.
• Providing a web interface for live monitoring and plant disease prediction using deep learning (ResNet-50).

Features

• Real-Time Sensor Data: View live temperature, humidity, soil moisture, and NPK (Nitrogen, Phosphorus, Potassium) values from the field.
• AI Plant Disease Detection: Upload plant images for instant health classification (Healthy, Potassium/Nitrogen deficiencies, etc.) using a ResNet-50 model.
• Remote Hardware Control: Toggle field devices (e.g., LEDs/lights) via the web interface.
• SocketIO Live Updates: Instant frontend updates when new sensor data arrives.
• ESP32 Integration: Ready-to-flash Arduino code for ESP32 boards to connect sensors and communicate with the server.

Hardware Components

• ESP32 Boards: For WiFi connectivity and sensor interfacing.
• Sensors:
  • DHT11/DHT22 (Temperature & Humidity)
  • NPK Soil Sensor (for Nitrogen, Phosphorus, Potassium)
• Relay/LED Modules: For remote control demonstration.

Software Components

• Flask Web Server: Handles API requests, serves the frontend, and manages SocketIO events.
• PyTorch/Transformers: For running the ResNet-50 plant disease classifier.
• Frontend: HTML/CSS/JS (with SocketIO) for real-time dashboard and image upload.
• ESP32 Arduino Code: For sensor reading and WiFi communication (see hardware/ folder).

Installation & Setup

1. Clone the Repository

git clone https://github.com/R3tr0gh057/Agri-project
cd Agri-project

2. Install Python Dependencies

You can use the provided script or manually install:

# Using the script (Linux)
bash install_packages.sh
# Or manually
pip install -r requirements.txt

3. Flash ESP32 Boards

• Use the code in hardware/esp-32/, hardware/npk-code/, and hardware/light-code/ for respective ESP32 modules.
• Install required Arduino libraries (see hardware/esp-32/lib-requirements.txt).

4. Run the Server

# Using the helper script (Linux)
bash run_server.sh
# Or manually
python server.py

5. Access the Web Interface

• Open your browser and go to: http://localhost:5000
• For remote access, use the ngrok tunnel provided by run_server.sh.

Usage

• Live Dashboard: View real-time sensor data as it updates.
• Image Prediction: Upload plant images for instant health/disease classification.
• Test Prediction Locally:

  python Test/image.py

• Hardware Control: Use the web interface to toggle LEDs/lights connected to ESP32.

Directory Structure

Agri-project/
├── hardware/
│   ├── esp-32/         # ESP32 main board code & requirements
│   ├── light-code/     # ESP32 code for light/relay control
│   └── npk-code/       # ESP32 code for NPK sensor
├── lib/                # Python utility and model code
├── static/             # Frontend JS and CSS
├── templates/          # HTML templates
├── Test/               # Image prediction test scripts
├── server.py           # Main Flask server
├── requirements.txt    # Python dependencies
├── install_packages.sh # Setup script
├── run_server.sh       # Server + ngrok launcher
└── README.md           # This file

API Endpoints

• GET / — Main dashboard page
• POST /predict — Accepts base64-encoded plant images, returns class & description
• POST /update-data — ESP32 posts sensor data (JSON)
• GET /fetch-data — Returns latest sensor data (JSON)
• GET /toggleLED — Toggles field LED/relay via ESP32

License

This project is licensed under the MIT License. See LICENSE for details.

Mobile App Availability

A companion mobile app is available and works seamlessly with this system once the server and hardware are set up and running. Contact the project maintainer for access or more details.

---