https://github.com/abdellatif-laghjaj/tomato-sorting-using-cnn
An intelligent computer vision system for automated cherry tomato sorting based on ripeness and size using YOLOv11 and Streamlit.
https://github.com/abdellatif-laghjaj/tomato-sorting-using-cnn
cnn cnn-classification object-detection realtime-classification yolo
Last synced: 8 months ago
JSON representation
An intelligent computer vision system for automated cherry tomato sorting based on ripeness and size using YOLOv11 and Streamlit.
- Host: GitHub
- URL: https://github.com/abdellatif-laghjaj/tomato-sorting-using-cnn
- Owner: abdellatif-laghjaj
- Created: 2024-12-19T12:26:15.000Z (10 months ago)
- Default Branch: main
- Last Pushed: 2024-12-29T21:50:26.000Z (9 months ago)
- Last Synced: 2025-02-22T17:45:17.785Z (8 months ago)
- Topics: cnn, cnn-classification, object-detection, realtime-classification, yolo
- Language: Jupyter Notebook
- Homepage:
- Size: 6.72 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
## 👨💻 Credits
| Contributors |
|-------------|
| Laghjaj Abdellatif |
| Reda El Bettioui |
# 🍅 Cherry Tomato Sorting System
An intelligent computer vision system for automated cherry tomato sorting based on ripeness and size using YOLOv11 and Streamlit.
## 📝 Overview
This project implements a real-time cherry tomato sorting system that:
- Detects and classifies cherry tomatoes by ripeness level (Unripe, Semi-ripe, Ripe)
- Measures tomato sizes and categorizes them (Small, Medium, Large)
- Provides real-time analytics and visualization
- Supports IP camera integration for continuous monitoring
## 🎥 Demo
Your browser does not support the video tag.
## 🔧 Features
### Real-time Detection
- Process live video feed from IP cameras
- Dynamic bounding boxes with rich labels showing:
- Ripeness classification
- Size measurements
- Confidence scores
### Multi-class Classification
- **Unripe (Green)**: Early stage tomatoes
- **Semi-ripe (Orange)**: Transitioning tomatoes
- **Ripe (Bright Red)**: Fully ripened tomatoes
### Size Categorization
- **Small**: < 20mm diameter
- **Medium**: 20-25mm diameter
- **Large**: > 25mm diameter
### Live Analytics Dashboard
- Total tomato count
- Color distribution charts
- Size distribution metrics
- Real-time performance stats
## 🛠️ Technologies Used
- **Python 3.8+**: Core programming language
- **YOLOv11**: Object detection & classification
- **Streamlit**: Interactive web interface
- **OpenCV**: Image processing & camera integration
- **Plotly**: Data visualization
- **NumPy**: Numerical operations
## 📦 Installation
```bash
# Clone repository
git https://github.com/abdellatif-laghjaj/tomato-sorting-using-CNN
cd tomato-sorting-using-CNN
# Create virtual environment
python -m venv venv
# Activate environment
source venv/bin/activate # Linux/Mac
# or
venv\Scripts\activate # Windows
# Install dependencies
pip install -r requirements.txt
```
## 💻 Usage
1. **Start the Streamlit app:**
```bash
streamlit run app.py
```
2. **Configure settings in sidebar:**
- Enter IP camera URL
- Adjust processing parameters if needed
3. **Click "Start Processing" to begin detection**
4. **View real-time results:**
- Live video feed with detections
- Dynamic analytics charts
- Running metrics
## 🤖 Model Architecture
```
Input Layer (640x640x3)
│
▼
Backbone: CSPDarknet
│ ├── Cross-Stage-Partial Connections
│ ├── Multiple Conv Blocks
│ └── Feature Pyramid Network (FPN)
│
▼
Neck: PANet (Path Aggregation Network)
│ ├── Bottom-up Path Augmentation
│ └── Adaptive Feature Pooling
│
▼
Head: Decoupled Detection Head
│ ├── Classification Branch (3 classes)
│ └── [Unripe, Semi-ripe, Ripe]
│ └── Regression Branch
│ └── [Bounding Box Coordinates]
│
▼
Output Layers
├── Class Predictions
├── Bounding Box Coordinates
└── Confidence Scores
```
### Training Configuration
```python
model.train(
data=data_yaml_path,
epochs=50,
imgsz=640,
batch=16,
name="cherry_tomato_sorting"
)
```
## 🤝 Contributing
Fork repository
Create feature branch (git checkout -b feature/AmazingFeature)
Commit changes (git commit -m 'Add some AmazingFeature')
Push to branch (git push origin feature/AmazingFeature)
Open Pull Request
## 🙏 Acknowledgments
OpenCV community for computer vision tools
Ultralytics for YOLOv11 framework
Streamlit team for the amazing dashboard framework