An industrial AI prototype focused on waste steel image recognition, intelligent grading, and digital management.
This project explores how AI can support the recycling industry by helping classify and analyze waste steel images more efficiently.
In waste steel recycling scenarios, image-based inspection and grading can improve efficiency, reduce manual workload, and support more standardized quality evaluation.
This project was developed as part of my industrial AI practice and later became the foundation for my software copyright registration and innovation competition projects.
- Waste steel image data processing
- Basic visual analysis workflow
- AI-assisted classification prototype (K-means++ clustering + Mahalanobis distance)
- Data visualization for clustering / distribution analysis
- Lightweight demo interface (Streamlit)
- Python
- Streamlit / Web UI
- Pandas
- NumPy
- Matplotlib
- scikit-learn
- Pillow
- Image processing workflow
I worked on:
- Requirement analysis
- System framework design
- AI-assisted coding
- Project documentation
- PPT and pitch material preparation
- Market research
- Software copyright material preparation
This project helped me understand how AI can move from abstract models into real industrial workflows.
It also became an important foundation for my later innovation competition and AI application projects.
- Improve image classification accuracy
- Add clearer model evaluation
- Build a better demo interface
- Add more documentation and screenshots
- Explore Agent-based industrial inspection workflow