A Streamlit-based AI web app to promote a clean and sustainable campus at VIT Chennai through intelligent waste segregation and automated maintenance alerts.
This project uses Machine Learning + Image Processing to automatically classify waste as:
- 🟢 Biodegradable
- 🔵 Non-Biodegradable
Once classified, the system:
- Suggests the nearest disposal point on campus.
- Sends automatic maintenance email notifications to the collection team.
- Tracks basic statistics like waste reports, users, and requests processed.
- 📸 Upload waste images for AI-based classification
- 🧮 Real-time confidence score with image feature extraction
- 📍 Location-based disposal guidance
- ✉️ Automatic email alerts to maintenance team
- 🔐 Secure Gmail App Password integration
- 🧠 Built with Random Forest model trained on custom waste dataset
| Category | Tools / Libraries |
|---|---|
| Frontend | Streamlit |
| Backend / ML | scikit-learn, joblib, OpenCV, NumPy, Pandas |
| Image Processing | PIL (Pillow), NumPy |
| Email Notifications | smtplib, email.mime |
| Deployment | Streamlit Cloud |
| Version Control | Git + GitHub |
- Algorithm: Random Forest Classifier
- Input: Extracted visual + color + shape features
- Output: Waste type prediction (biodegradable / non-biodegradable)
- Scaling: StandardScaler used for normalization
Model files:
random_forest_model_fixed.pklscaler_fixed.pkl
- Clone the repository:
git clone https://github.com/Akhila707/Waste-Management.git cd Waste-Management