Internship project: AICTE || SHELL || EDUNET FOUNDATION
.gif){kind=link}
.gif){kind=link}
An intuitive web-based application to classify waste into Organic or Inorganic categories, promoting sustainability and eco-friendly practices. The app leverages the power of TensorFlow, OpenCV, and Streamlit to deliver real-time predictions with an elegant user interface.
.gif)
- 📤 Upload Image: Upload an image of waste and get it classified.
- 📸 Capture Image: Use your device's camera to classify waste on the spot.
- 🎨 Interactive UI: Minimalistic and user-friendly design for smooth interaction.
- 📊 Real-time Predictions: Classifies waste with confidence scores.
- 📦 Powered by AI: Built using a custom-trained CNN model for accurate results.
| Technology | Description |
|---|---|
| TensorFlow | Deep Learning framework for training and inference of the waste classifier. |
| OpenCV | Used for image preprocessing and overlaying bounding boxes. |
| Streamlit | Simplified the creation of an interactive web app. |
| PIL | Image manipulation library. |
Directory structure:
└── hiteshydv001-waste-classification-model-cnn/
├── README.md
├── app.py
├── requirements.txt
└── waste_classification_model.h5
- Upload or Capture an image of the waste.
- The image is processed and resized to match the input requirements of the model.
- A Convolutional Neural Network (CNN) predicts whether the waste is:
- Organic (e.g., food scraps, plants).
- Inorganic (e.g., plastics, metals).
- The result is displayed along with a bounding box around the waste object and a confidence score.
- Model Type: Convolutional Neural Network (CNN)
- Layers: Multiple convolutional and pooling layers for feature extraction, followed by fully connected layers for classification.
- Image Dimensions: 150x150 pixels.
- Organic Waste (O)
- Inorganic Waste (R)
- Normalization: Pixel values are normalized to a range of [0, 1].
- Data Augmentation: Techniques like rotation, zoom, and horizontal flip applied to increase dataset variability and model robustness.
- Sustainability-Oriented: Encourages proper waste segregation for a cleaner environment.
- Real-World Application: Designed for use in homes, industries, and municipalities.
Hitesh Kumar
📧 [email protected]
🌐 LinkedIn | GitHub
This project is licensed under the MIT License. See the LICENSE file for details.
- TensorFlow Community for extensive documentation and support in building and training the model.
- Streamlit Team for providing a fantastic framework to develop and deploy web applications.
- OpenCV for offering efficient image processing tools to enhance data handling and manipulation.