Computer Vision-Based Signal Processing For Gesture- Controlled Air Canvas

Abstract

Human–Computer Interaction (HCI) is rapidly evolving towards touchless and intuitive systems. This project presents 
a Computer Vision-Based Signal Processing approach for a Gesture-Controlled Air Canvas, enabling users to interact 
with a computer using hand gestures without physical input devices. 
The system uses a webcam to capture real-time video input and processes it using computer vision techniques. 
MediaPipe is employed for accurate hand detection and tracking of 21 hand landmarks, while OpenCV is used for 
image processing and visualization. Signal processing techniques such as noise reduction and smoothing are applied 
to improve the stability and accuracy of hand movement tracking. 
Based on detected gestures, the system allows users to draw, erase, and select colors on a virtual canvas. The fingertip 
position is tracked continuously to create smooth and natural drawing strokes in real time. 
This project provides a low-cost, efficient, and user-friendly solution for touchless interaction. It has potential 
applications in smart classrooms, virtual presentations, accessibility systems, and human-computer interaction 
interfaces.