Air Canvas

EasyChair Preprint no. 12863

Abstract

In recent years, air based writing has emerged as a compelling area in pattern recognition and image processing, crucial for advancing automation and enhancing manmachine interaction. With a focus on reducing processing time and enhancing recognition accuracy, researchers have explored novel techniques. Object tracking, an integral part of Computer Vision, has gained momentum with faster computers and affordable, high-quality video cameras. This involves three key steps: object detection, frame-to-frame tracking, and behavioral analysis. Object tracking entails addressing issues such as suitable representation, feature selection, detection, and tracking. Its applications span automatic surveillance, video indexing, and navigation. Leveraging this, a project aims to create a motion-totext converter for intelligent wearable devices, allowing writing in air for various purposes, aiding communication for the deaf and reducing reliance on traditional devices.

Keyphrases: gestures, grabbing, MediaPipe, ndarray, OpenCV

@Booklet{EasyChair:12863,
  author = {Jagadeesh Vavili and Sai Mohan Reddy Vurabavi and Lokesh Chitturi and Lokesh Palaparthi and Vaibhavee Jani},
  title = {Air Canvas},
  howpublished = {EasyChair Preprint no. 12863},

  year = {EasyChair, 2024}}