A comprehensive automated waste sorting system that uses computer vision, machine learning, and a servo mechanism to sort waste items into recycling and garbage. The system includes a web analytics dashboard and a database for tracking sorting history.

• Computer Vision Classification: Automatically identifies cans, recyclables, and garbage
• Automatic Sorting: Mechanically sorts items into appropriate bins using a servo-controlled platform
• Analytics Dashboard: Web interface to view statistics and sorting history
• Custom Model Training: Train your own classification model for better accuracy
• Data Tracking: Store images and sorting data for analysis

• Arduino Uno or compatible board
• High-torque servo motor (at least 15-20 kg-cm torque)
• Logitech Brio 4K webcam (or similar high-resolution camera)
• Computer with USB ports (Windows/Mac/Linux)
• 5V 3A power supply for servo
• Wood and acrylic materials for the physical platform

• Python 3.8 or newer
• TensorFlow 2.x
• OpenCV
• SQLite
• Flask (for web dashboard)
• Arduino IDE (for uploading firmware)

waste-sorting-system/
├── arduino/                  # Arduino firmware
│   └── waste_sorter_arduino.ino
├── data/                     # Data storage directory
│   ├── counts.json           # Counter values
│   └── sorting_data.db       # SQLite database
├── logs/                     # Log files
├── models/                   # Trained models
│   ├── latest_model.h5       # Latest model
│   └── class_mapping.json    # Class mapping
├── static/                   # Static web files
│   ├── css/
│   ├── js/
│   └── img/
├── templates/                # HTML templates
├── training_data/            # Data for model training
│   ├── can/
│   ├── recycling/
│   └── garbage/
├── app.py                    # Web dashboard Flask app
├── database.py               # Database module
├── main.py                   # Main application
└── train_model.py            # Model training script

git clone https://github.com/BradleyNgu/cuHacking.git
cd cuHacking

pip install -r requirements.txt

Or manually install the following packages:

pip install tensorflow opencv-python numpy pillow pyserial flask matplotlib

1. Open Arduino IDE
2. Load the firmware from arduino/waste_sorter_arduino.ino
3. Connect your Arduino via USB
4. Upload the firmware to the board

mkdir -p data logs models static/css static/js static/img templates training_data/can training_data/recycling training_data/garbage

• Wood:

  • Base: 24" x 24" x 3/4" plywood
  • Sides: 2x 24" x 16" x 1/4" plywood
  • Back: 1x 24" x 16" x 1/4" plywood

• Acrylic:

  • Platform: 12" x 10" x 1/8" clear acrylic
  • Side Rails: 2x 10" x 2" x 1/8" clear acrylic

• Hardware:

  • Metal Axle: 1/4" diameter steel rod, 14" long
  • Ball Bearings: 2x flanged ball bearings for 1/4" shaft
  • Mounting Brackets: 2x metal L-brackets for bearings
  • Shaft Collars: 2x for securing axle
  • Servo Horn: Extended servo horn (6-8")
  • Screws, nuts, bolts, washers

• Bins:

  • 2x plastic bins (approximately 10" x 10" x 12")

1. Build the Frame:

   • Assemble the U-shaped frame using the plywood pieces
   • Secure with wood screws
   • Ensure corners are square

2. Install the Pivot Mechanism:

   • Mount the bearing brackets at the center height on both sides
   • Insert the axle through both bearings
   • Secure with shaft collars

3. Create the Tilting Platform:

   • Cut the acrylic platform and side rails
   • Attach side rails to the platform
   • Mount platform on the axle

4. Install the Servo:

   • Create a sturdy mount for the servo
   • Attach the extended servo horn to the underside of the platform
   • Position ~5" from the pivot point for best leverage

5. Mount the Camera:

   • Create a bracket to hold the webcam above the platform
   • Position it for a clear view of items

6. Connect Electronics:

   • Connect the Arduino to your computer via USB
   • Connect the servo to the Arduino (signal wire to pin 9)
   • Connect servo power to external 5V power supply

python main.py

• Select the Arduino port from the dropdown
• Select the camera from the dropdown
• Click "Connect"

For manual operation:

• Place item on the platform
• Click "Analyze Item"
• Click "Sort Item" once analysis is complete

For automatic operation:

• Check "Auto-Sort Mode"
• Place items on the platform one at a time

From the application menu:

• Go to Tools > Start Analytics Dashboard

Or start it separately:

python app.py

Access the dashboard at http://localhost:5000

1. Collect images for each category:

   • Place images in the appropriate folders under training_data/
   • Aim for at least 50-100 images per category

2. Launch the training dialog:

   • From the main application, go to Tools > Model Training
   • Set parameters and click "Start Training"

3. Restart the application to use the new model

• Ensure the correct port is selected
• Verify the Arduino has the correct firmware
• Check the USB cable connection
• Test with Tools > Test Arduino

• Verify the camera is connected
• Try a different USB port
• Test with Tools > Test Camera
• Check for other applications using the camera

• Calibrate the platform angles using the slider
• Ensure the servo has enough power (use external power supply)
• Make sure the platform surface is smooth
• Check if items are within the weight limit of the servo

Contributions are welcome! Please feel free to submit a Pull Request.

This project is licensed under the MIT License - see the LICENSE file for details.