Virtual Keyboard

Project Description:

Technologies: Python, OpenCV, NumPy, PyAutoGUI

Human-computer interaction inspired me to design a project where a person could type without touching a keyboard. Using a webcam feed, I developed a gesture recognition system with OpenCV that detected hand contours and mapped them to keyboard events.

The pipeline included image preprocessing, contour detection, and finger-tracking logic. Once gestures were recognized, I used PyAutoGUI to simulate keystrokes on the system. This created a functioning virtual keyboard controlled entirely by hand movements.

While experimental, the project showcases how computer vision techniques can be applied to accessibility solutions and innovative interfaces.

Skills Showcase:

• Real-time computer vision with OpenCV

• Gesture recognition using contour and feature detection

• Mapping gestures to simulated keystrokes with PyAutoGUI

•  Python modules:

  Pickle: To serialize the data from the image objects.

  Argparse: To check and use command line arguments.

  Pyautogui: To draw and show the virtual keyboard on the screen.

• Image preprocessing and mask creation for robustness

• Applied HCI concepts to accessibility design

  
  

Key Insights:

• Webcam input can reliably substitute for hardware input in controlled settings

• Hand contour recognition proved more robust than color thresholding

• Prototype demonstrated feasibility of accessibility-oriented HCI solutions