C242-PS377

Waste Classification with MobileNetV2

This project aims to classify waste types into several categories (cardboard, glass, metal, paper, plastic, trash, and organic) using a pre-trained MobileNetV2 model. The model can be used to help in automatic waste sorting, which can be beneficial for recycling and more efficient waste management.

Contributors

• M239b4ky4526 - Yeheskiel Yunus Tame - Universitas Kristen Duta Wacana - Machine Learning
• M239b4ky1557 - Frederik Samra Sarongallo - Universitas Kristen Duta Wacana - Machine Learning
• M314b4ky2563 - Mohammad Baharudin Yusuf - Universitas Singaperbangsa Karawang - Machine Learning

Features

• Multi-class Classification: The model can classify waste into 8 different categories.
• High Accuracy: The model is trained on a large dataset and achieves high accuracy on validation data.
• Efficiency: Using MobileNetV2, this model is relatively lightweight and fast, making it suitable for implementation on resource-limited devices.
• Data Augmentation: Employs data augmentation techniques to increase training data variation and prevent overfitting.
• Transfer Learning: Leverages transfer learning from MobileNetV2 pre-trained on the ImageNet dataset, allowing the model to learn faster and more effectively.

How to Use

1. Prepare Dataset

• Location: Place your waste images in the Capstone-Project/Dataset folder
• Folder Structure: Each waste category should have its own folder
• Supported Categories:
  • Cardboard
  • Glass
  • Metal
  • Paper
  • Plastic
  • Trash
  • Organic

2. Run the Notebook

• Platform: Use Google Colab or Jupyter Notebook
• File: Open model_waste_classification.ipynb
• Process:
  1. Load required libraries
  2. Prepare dataset
  3. Train the model
  4. Evaluate the model
• Run all cells sequentially

3. Model Testing

• Prediction Function: Use predict_image(image_path)
• Parameter: Input the path of the image to be tested
• Usage Example:

  # Example of calling the prediction function
  result = predict_image('path/to/your/image.jpg')
  print(result)