AgriNex - Smart Agricultural Intelligence System 🌾

AgriNex is an end-to-end, multi-modal agricultural intelligence platform leveraging machine learning, deep neural networks, and interactive web tools to empower modern precision farming. The system brings predictive analytics directly to farmers and agricultural extension workers through intuitive dashboards and live deployed microservices.

🌾 Overview

AgriNex combines cutting-edge AI/ML technologies with user-friendly web interfaces to help farmers make data-driven decisions about crop selection, disease management, and fertilizer optimization. The platform integrates multiple machine learning models trained on agricultural datasets to provide accurate and actionable insights.

🚀 Deployed Live Streamlit Applications

AgriNex trained machine learning models are deployed live on Streamlit Cloud:

🌾 Crop Recommendation AI: smart-crop-recommendations.streamlit.app
🧪 Fertilizer Prediction AI: fertilizer-predictions.streamlit.app
🔬 Plant Disease Identification: plant-diseases-identification.streamlit.app

🏗️ System Architecture & Workflow

graph TD
    User([Farmer / Agricultural Worker]) --> WebApp[AgriSens Web Application]
    
    subgraph Frontend Layer
        WebApp --> CropPage[Crop Recommendation UI]
        WebApp --> FertPage[Fertilizer Advisor UI]
        WebApp --> DiseasePage[Disease Detection UI]
        WebApp --> Chatbot[AI Kisan Chatbot Widget]
        WebApp --> Dashboard[Farm Dashboard & Mandi Ticker]
    end
    
    subgraph ML & Intelligence Layer
        CropPage --> |Soil & Climate Data| CropML[Random Forest Classifier]
        FertPage --> |Soil Nutrient Analysis| FertML[Fertilizer XGB/RF Classifier]
        DiseasePage --> |Leaf Image Upload| DiseaseCNN[TensorFlow 38-Class CNN Model]
        Chatbot --> |Natural Language Query| KisanEngine[Knowledge Engine & Mandi Data API]
    end

    subgraph External Cloud Deployments
        CropML --> StreamlitCloud[Streamlit Cloud Services]
        FertML --> StreamlitCloud
        DiseaseCNN --> StreamlitCloud
    end

✨ Features & Capabilities

Embedded Streamlit ML Apps: Real-time interactive model predictions seamlessly embedded inside responsive web interfaces (crop-prediction.html, fertilizer.html, disease-detection.html).
Universal AI Kisan Chatbot: Multi-lingual floating conversational assistant tailored for localized agricultural advice, weather alerts, disease remedies, and APMC market prices.
Precision Crop Recommendation Engine: High-accuracy crop selection powered by Random Forest algorithms analyzing Nitrogen (N), Phosphorus (P), Potassium (K), temperature, relative humidity, pH, and annual rainfall.
Computer Vision Plant Pathology: Deep convolutional neural network (CNN) capable of identifying 38 distinct crop diseases across 14 plant species with leaf image uploaded via web interface.
Soil-Specific Fertilizer Advisor: Automated fertilizer dosage calculation matching crop requirements against current NPK depletion metrics to avoid over-fertilization.
Live APMC Mandi Ticker & Government Schemes: Real-time wholesale commodity price ticker and curated catalog of PM-KISAN, KCC, and state agricultural subsidies.
Farm Management Telemetry Dashboard: Complete farm monitoring hub with 7-day weather forecasting, spray-window advisory, yield revenue estimate tools, and direct Krishi Vigyan Kendra (KVK) contacts.

🛠️ Tech Stack

Frontend

React.js - Interactive user interface
HTML/CSS/JavaScript - Web standards

Backend & ML Services

Python - Core language for ML/AI services
Streamlit - Framework for data apps and model serving
Flask - Web framework for APIs

Machine Learning & AI

TensorFlow - Deep learning framework
Keras - Neural network API for plant disease identification
scikit-learn - Machine learning algorithms for crop and fertilizer prediction
Random Forest Classifier - Ensemble learning for crop recommendation

Data Processing & Visualization

Pandas - Data manipulation and analysis
NumPy - Numerical computing
Matplotlib - 2D plotting and visualization
Seaborn - Statistical data visualization
OpenCV - Computer vision for image processing

Datasets

PlantVillage Dataset - 38 plant disease classes
Crop Recommendation Dataset - Soil and environmental parameters
Fertilizer Recommendation Dataset - Soil nutrient analysis

📁 Project Structure

AgriNex/
├── CROP PREDICT/                          # React frontend + ML model training
│   ├── frontend/                          # React application
│   │   ├── public/
│   │   │   ├── index.html
│   │   │   └── manifest.json
│   │   └── src/
│   │       ├── App.js
│   │       ├── home.js
│   │       ├── index.js
│   │       └── index.css
│   ├── saved_model/                       # Pre-trained crop prediction models
│   │   ├── model_v1.h5
│   │   ├── model_v2.h5
│   │   └── model_v3.h5
│   └── training/
│       ├── model.ipynb                    # Model training notebook
│       └── PlantVillage/                  # Training dataset
│
├── CROP-RECOMMENDATION/                   # Streamlit crop recommendation app
│   ├── webapp.py                          # Main Streamlit application
│   ├── Crop_recommendation.csv            # Dataset
│   ├── requirements.txt
│   └── fgd.html
│
├── PLANT-DISEASE-IDENTIFICATION/          # Streamlit disease detection app
│   ├── main.py                            # Main Streamlit application
│   ├── trained_plant_disease_model.keras  # Trained Keras model
│   ├── Train_plant_disease.ipynb          # Training notebook
│   ├── Test_plant_disease.ipynb           # Testing notebook
│   ├── training_hist.json                 # Training history
│   ├── settings.json                      # Configuration
│   ├── requirements.txt
│   └── test/                              # Test images
│
├── fertilizer_prediction/                 # Fertilizer prediction module
│   └── fertilizer_prediction/
│       ├── fert.py                        # Fertilizer prediction script
│       ├── p1.ipynb                       # Analysis notebook
│       ├── data_core.csv                  # Dataset
│       └── requirements.txt
│
├── Datasets/                              # Consolidated datasets
│   ├── Crop_recommendation.csv
│   ├── Fertilizer_recommendation.csv
│   ├── README.md
│   └── PlantVillage/                      # Plant disease images
│       ├── Potato___healthy/
│       ├── Potato___Late_blight/
│       └── [38+ disease classes]
│
└── AgriSens-web-app/                      # Integrated web application

🚀 Getting Started

Prerequisites

Python 3.8+
Node.js 14+ (for frontend web apps)
Docker & Docker Compose (optional for containerized setup)
pip / venv or conda for environment isolation
Git

Quick Setup with Virtual Environment

1. Clone the Repository

git clone https://github.com/SQUADRON-LEADER/AgriNex.git
cd AgriNex

2. Environment Setup (Recommended)

# Create and activate virtual environment
python -m venv venv

# On Windows
.\venv\Scripts\activate

# On Linux/macOS
source venv/bin/activate

3. Setup & Run Crop Recommendation Microservice

cd CROP-RECOMMENDATION
pip install -r requirements.txt
streamlit run webapp.py

Access interface at http://localhost:8501

4. Setup & Run Plant Disease Detection Microservice

cd ../PLANT-DISEASE-IDENTIFICATION
pip install -r requirements.txt
streamlit run main.py --server.port 8502

Access interface at http://localhost:8502

5. Setup & Run Fertilizer Advisor Microservice

cd ../fertilizer_prediction/fertilizer_prediction
pip install -r requirements.txt
python fert.py

6. Launch AgriSens Integrated Web Portal

cd ../AgriSens-web-app
# Open index.html directly in browser or serve via npx http-server
npx http-server -p 8080

🐳 Containerized Deployment (Docker)

# Build and run containers via Docker Compose
docker-compose up --build -d

📊 Model Details

Crop Recommendation Model

Algorithm: Random Forest Classifier
Features: N, P, K (soil nutrients), Temperature, Humidity, pH, Rainfall
Output Classes: 22 different crops (Rice, Maize, Chickpea, Kidney Beans, Pigeon Peas, Moth Beans, Mung Bean, Black Gram, Lentil, Pomegranate, Banana, Mango, Grapes, Watermelon, Muskmelon, Apple, Orange, Papaya, Coconut, Cotton, Jute, Coffee)
Accuracy: ~99% cross-validation accuracy
Training Data: 2,200+ soil & climate sample observations

🧪 Input Parameter Specifications Table

Feature Parameter	Symbol	Measurement Unit	Typical Value Range	Description
Nitrogen	`N`	ratio / mg/kg	0 - 140	Soil ratio of Nitrogen content
Phosphorus	`P`	ratio / mg/kg	5 - 145	Soil ratio of Phosphorus content
Potassium	`K`	ratio / mg/kg	5 - 205	Soil ratio of Potassium content
Temperature	`temp`	°C	8.8 - 43.7	Ambient atmospheric temperature
Humidity	`humidity`	%	14.3 - 99.9	Relative air humidity percentage
pH Level	`ph`	pH scale	3.5 - 9.9	Acidity / alkalinity index of soil
Rainfall	`rainfall`	mm	20.2 - 298.6	Cumulative seasonal rainfall

Plant Disease Detection Model

Framework: TensorFlow / Keras 2.x
Architecture: Convolutional Neural Network (CNN) with Softmax classification layer
Input Dimensions: 128x128 pixel RGB tensor
Output Classes: 38 distinct plant pathology classes
Supported Crops: Apple, Blueberry, Cherry, Corn, Grape, Orange, Peach, Pepper, Potato, Raspberry, Soybean, Squash, Strawberry, Tomato

Fertilizer Prediction Model

Algorithm: Machine Learning Classifier (Random Forest / XGBoost hybrid)
Features: Soil NPK baseline, soil type, crop type, humidity, temperature
Output: Optimized target fertilizer blend (e.g., Urea, DAP, 14-35-14, 28-28, 17-17-17, 20-20, 10-26-26) with calculated application rate

💻 Usage Examples

Crop Recommendation

Input: N=90, P=42, K=43, Temperature=20.87, Humidity=82.00, pH=6.5, Rainfall=202.9
Output: Recommendation for "Maize" crop

Disease Detection

Input: Plant leaf image
Output: Disease identified with confidence score
Example: "Potato___Late_blight - 95% confidence"

Fertilizer Recommendation

Input: Soil parameters and crop type
Output: Specific fertilizer recommendation with dosage

📈 Dataset Information

PlantVillage Dataset: 61,486 images across 38 disease classes
Crop Dataset: Features covering 22 different crops with environmental parameters
Fertilizer Dataset: Comprehensive soil and crop parameters with recommendations

🤝 Contributing

Contributions are welcome! Please follow these steps:

Fork the repository
Create your feature branch (git checkout -b feature/AmazingFeature)
Commit your changes (git commit -m 'Add some AmazingFeature')
Push to the branch (git push origin feature/AmazingFeature)
Open a Pull Request

📝 Model Training & Evaluation

To retrain or evaluate models:

Crop Prediction

cd CROP\ PREDICT/training
jupyter notebook model.ipynb

Plant Disease Detection

cd PLANT-DISEASE-IDENTIFICATION
jupyter notebook Train_plant_disease.ipynb

🔧 Configuration

Each service has a settings.json or configuration file for model parameters and settings. Modify these files to adjust:

Model paths
Input parameters
Output formats
Service ports

📦 Dependencies

See individual requirements.txt files in each module directory for specific dependencies:

Core ML dependencies: TensorFlow, Keras, scikit-learn
Data processing: Pandas, NumPy
Visualization: Matplotlib, Seaborn
Web frameworks: Streamlit, Flask
Image processing: OpenCV, Pillow

🐛 Troubleshooting

Model Loading Issues

Ensure model files (.keras, .h5) are in the correct directories
Verify TensorFlow and Keras versions match training environment

Streamlit Port Already in Use

streamlit run app.py --server.port 8503

Missing Dependencies

pip install --upgrade -r requirements.txt

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

👥 Team & Credits

AgriNex is developed as an agricultural AI solution combining expertise in:

Machine Learning & Deep Learning
Agricultural Science
Full-Stack Web Development

📞 Support

For issues, questions, or suggestions:

Open an issue on GitHub
Check existing documentation
Review dataset README files for data-specific questions

🌱 Future Enhancements & Strategic Roadmap

Multi-modal Web Dashboard Integration: Unified interface for crop recommendation, disease detection, and fertilizer calculation.
Live Streamlit Cloud Deployment: Microservices published live for testing and remote inference.
Hyper-local Weather & Microclimate API: Integrate OpenWeatherMap API for live rainfall and moisture forecast alerts.
Real-time Commodity Price Forecasting: Time-series predictive models for APMC market trends.
Native Mobile Application (React Native): Offline-first mobile app for field operation with localized camera scan.
Regional Language Localization: Support for Hindi, Marathi, Punjabi, Tamil, Telugu, and Kannada.
IoT Sensor Grid Integration: Direct Bluetooth / LoRaWAN telemetry ingestion from soil NPK hardware probes.

AgriNex • Empowering Sustainable Farming with Artificial Intelligence
Developed with ❤️ for farmers, agronomists, and precision agriculture worldwide.

Name		Name	Last commit message	Last commit date
.agents		.agents
AgriSens-web-app		AgriSens-web-app
CROP PREDICT		CROP PREDICT
CROP-RECOMMENDATION		CROP-RECOMMENDATION
Datasets		Datasets
PLANT-DISEASE-IDENTIFICATION		PLANT-DISEASE-IDENTIFICATION
fertilizer_prediction/fertilizer_prediction		fertilizer_prediction/fertilizer_prediction
graphify-out		graphify-out
.gitattributes		.gitattributes
.gitignore		.gitignore
README.md		README.md

Latest commit

History

Folders and files

Repository files navigation