🌴 SmartDate IoT — Deglet Nour Date Classification System

🌍 University Project

---

📖 Overview

SmartDate IoT is an intelligent system designed to automatically classify Deglet Nour dates using a Deep Learning model (EfficientNetB3) integrated into a simulated IoT pipeline.
The system captures images using a webcam, detects the date region with YOLOv8, classifies the date type, and publishes results in real time using MQTT toward a web dashboard.

This project demonstrates a complete pipeline:
Webcam → YOLOv8 → EfficientNetB3 → MQTT → Node.js → MongoDB → React Dashboard

🎯 Objectives

• Build an accurate deep learning model for classifying 11 types of Deglet Nour dates
• Implement an IoT-like communication system using MQTT
• Simulate a smart agriculture pipeline
• Provide a real-time monitoring dashboard for visualization and analysis

---

🧩 Project Structure

SmartDate-IoT/
│
├── Colab_Notebooks/                         → Model training, evaluation, fine-tuning
├── Date_Dataset/                            → Images for training/validation/testing
├── Docs/                                    → Documentation + architecture images
├── IOT/                                     → MQTT publisher/subscriber scripts
│   ├── .env                             
│   ├── detections_log.csv              
│   ├── inference_publish_webcam.py      
│   ├── mqtt_subscriber_save.py          
│   └── utils.py                         
│
├── Web_App/                                 → App web
│   ├── server/
│   │   ├── src/
│   │   │   ├── server.js                    → Main backend server
│   │   │   ├── db.js                    
│   │   │   ├── models/Detection.js     
│   │   │   ├── services/mqttService.js  
│   │   │   └── routes/api.js            
│   │   ├── .env
│   │   ├── package.json
│   │   └── node_modules/
│   │
│   ├── client/                              → Frontend
│   │   ├── src/
│   │   │   ├── App.jsx, main.jsx, index.css
│   │   │   ├── pages/Dashboard.jsx
│   │   │   ├── components/
│   │   │   │   ├── Header.jsx
│   │   │   │   ├── Footer.jsx
│   │   │   │   ├── KpiOverview.jsx
│   │   │   │   ├── RealtimeView.jsx
│   │   │   │   ├── HistoryList.jsx
│   │   │   │   ├── ActivityBars.jsx
│   │   │   │   ├── TypeDonut.jsx
│   │   │   │   ├── ChartPanel.jsx
│   │   │   │   └── MQTTContext.jsx
│   │   │   └── assets/
│   │   ├── public/
│   │   ├── package.json
│   │   └── node_modules/
│
├── models/
│   └── smartdate_efficientnetb3.keras       → Trained model
│
└── README.md

---

🔍 Key Features

✅ Real-time classification using webcam

Local capture → YOLOv8 detection → EfficientNetB3 classification → instant feedback.

☁️ MQTT-based IoT communication

Simulated edge device → cloud workflow using a secure TLS MQTT broker (HiveMQ Cloud).

📊 Interactive dashboard (React + Node.js + MongoDB)

• Real-time updates streamed from MQTT
• Detection history
• Daily activity timeline
• Type distribution donut chart
• Confidence score monitoring
• Intelligent recommendations

💡 Smart Recommendations

The system generates user-friendly suggestions based on:

• the detected date type
• the confidence level

---

⚙️ Workflow Phases

1️⃣ Deep Learning Model Training

• Dataset cleaning & augmentation
• EfficientNetB3 fine-tuning on Google Colab
• Evaluation + best model export (Keras format)

2️⃣ IoT Simulation with MQTT

• Webcam capture
• YOLOv8 date localization
• EfficientNet date type classification
• MQTT publishing (TLS secured)
• Subscriber saving detection logs into MongoDB + CSV

3️⃣ Web Application

• Node.js backend (REST API + MQTT listener)
• MongoDB database
• React front-end dashboard for:
  • real-time updates
  • charts
  • history
  • recommendations

---

📚 Installation & Usage

▶️ 1. Run the IoT scripts

Publisher (Webcam → YOLOv8 → EfficientNet → MQTT)

python IOT/inference_publish_webcam.py

Subscriber (MQTT → MongoDB + CSV)

python IOT/mqtt_subscriber_save.py

▶️ 2. Start the backend server

cd Web_App/server
npm install
npm start

▶️ 3. Start the frontend dashboard

cd Web_App/client
npm install
npm run dev

---

⚖️ Licence

This project is distributed under the MIT license — you are free to reuse it for educational or personal purposes, provided you cite the original author.

---

⭐ SmartDate IoT demonstrates how Deep Learning + IoT + Web technologies can work together to build a smart, connected, and real-time agricultural system.

🌾 Smart agriculture starts with smart data.