Skip to content

BuiQuangChien_6_readme.md

Latest commit

 

History

History
71 lines (52 loc) · 2.44 KB

BuiQuangChien_6_readme.md

File metadata and controls

71 lines (52 loc) · 2.44 KB
Raw

PROJECT INFORMATION: WASTE CLASSIFICATION ASSISTANT

Course: Computer Vision Group: Team 6

1. Team members and responsibilities

No. Name Student ID Role & Responsibilities
1 Bui Quang Chien (Team Lead) 23001837 Literature review; image preprocessing (Otsu thresholding, morphological operations); train and evaluate classical ML models (SVM, Random Forest, KNN); compile report (LaTeX).
2 Nguyen Anh Duc 23001865 Develop the Laravel web application; implement frontend UI and backend-to-Python inference integration; write deployment and experiment scripts.
3 Nguyen Le Ngoc Bao 23001832 Research and train deep learning models (MobileNetV2, ResNet50, EfficientNetB0); apply transfer learning and fine-tuning; prepare presentation slides.

2. Dataset and download links

  • Dataset used: TrashNet (Gary Thung & Mindy Yang, 2016)
  • Description: ~2,390 images divided into 5 classes: Glass, Paper, Cardboard, Plastic, Metal.
  • Original dataset: https://github.com/garythung/trashnet

3. Project structure and experiment workflow

Example project layout:

/
├── BuiQuangChien_6_report.pdf
├── BuiQuangChien_6_presentation.pdf
├── BuiQuangChien_6_readme.md   # this file
├── app/
│   └── Waste-classification-main/
│       └── rac-ai/
│           ├── app/
│           ├── resources/views/
│           ├── routes/
│           ├── scripts/
│           │   └── predict_waste.py
│           └── storage/app/model/
└── figures/

3.1 Experiment workflow (run the web app)

The web app calls Python inference scripts from the PHP backend. Steps to run locally:

  1. Install Python dependencies:
pip install numpy opencv-python scikit-image scikit-learn
  1. Prepare model files

Place trained model artifacts into app/Waste-classification-main/rac-ai/storage/app/model/ (e.g. best_ml_model.pkl, best_ml_scaler.pkl).

  1. Configure and run the Laravel backend
cd app/Waste-classification-main/rac-ai/
composer install
cp .env.example .env   # Windows: copy .env.example .env
php artisan key:generate
php artisan migrate
php artisan storage:link
php artisan serve
  1. Use the application
  • Open your browser at http://localhost:8000.
  • Upload a test image and the backend will run preprocessing + inference and return the predicted label with confidence.