Skip to content

Latest commit

 

History

History
117 lines (80 loc) · 3.34 KB

README.md

File metadata and controls

117 lines (80 loc) · 3.34 KB

HandSense 360 V3.1.1

This program is a Python-based hand recognition system designed to track hand gestures, calculate metrics like completion rate and hand stability, and interact with external applications using sockets and mouse simulations. It leverages MediaPipe Hands, OpenCV, and PyAutoGUI to provide real-time feedback and control.

Features

  • Hand Gesture Recognition: Detects and tracks hand gestures using MediaPipe.
  • Real-Time Metrics: Calculates completion rate, hand stability, and hand flip direction.
  • External Integration: Sends data to external programs via socket communication.
  • Mouse Simulation: Controls mouse cursor based on hand position.
  • High Frame Rate Support: Configured for 120 FPS for smooth tracking.

Prerequisites

Before running the program, ensure you have the following installed:

  • Python 3.8+
  • OpenCV
  • MediaPipe
  • PyAutoGUI
  • NumPy
  • Keyboard

Install the dependencies with:

pip install opencv-python mediapipe pyautogui numpy

Usage

1. Clone the Repository

git clone https://github.com/YourUsername/Parkinson-Hand-Recognizing.git
cd Parkinson-Hand-Recognizing

2. Run the Program

Start the program with:

python hand_recognition.py

3. Control and Metrics

  • Place your palm in view of the camera for the program to start recognizing gestures.
  • Metrics such as Completion Rate, Flip Direction, and FPS will be displayed on the screen.
  • Press 'Q' to quit the program.

Customization

Configure External Socket Communication

  • Update the HOST and PORT variables in the code to match your external application:
HOST = 'localhost'  # Replace with the external program's IP address
PORT = 10888  # Replace with the desired port number

Adjust Camera Settings

You can modify the camera resolution and frame rate:

cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 1920)
cap.set(cv2.CAP_PROP_FRAME_WIDTH, 1080)
cap.set(cv2.CAP_PROP_FPS, 120)

Debugging

To debug connection issues or errors, monitor the terminal for exceptions or print statements.

Key Features in Detail

Hand Gesture Detection

  • Uses MediaPipe Hands to extract key hand landmarks in real time.
  • Tracks critical points like wrist, thumb tip, and index finger MCP for calculating distances and stability.

Mouse Simulation

  • Calculates the center position of the hand and moves the mouse cursor accordingly.
  • Smooth and responsive control using threading.

Completion Rate and Stability

  • Dynamically calculates a "completion rate" based on the relative positions of hand landmarks.
  • Stabilizes the rate to avoid false spikes or noise.

Socket Communication

  • Sends real-time metrics to external applications using TCP sockets.

Troubleshooting

  • Camera Not Detected: Ensure your camera is connected and accessible. Check the cv2.VideoCapture settings.
  • Low FPS: Reduce resolution or adjust cap.set(cv2.CAP_PROP_FPS) for better performance on lower-end hardware.
  • Dependencies Missing: Use pip install -r requirements.txt to install all dependencies at once.

License

This project is licensed under the MIT License. See the LICENSE file for more details.

Contact

For questions or support, contact Steven-ZN or open an issue on GitHub.