A State-of-Art in Mid-Air Handwriting Recognition techniques

A dynamic gesture is one in which the hand is moved while attempting to transmit a letter, symbol, or phrase. These motions are distinguished by projected hand movement trajectories. Due to their intrinsic complexity, automated identification of these air trajectories is still a work in progress. From single character trajectory recognition through continuous word trajectory identification, research has been done in this field. In an effort to replicate the natural pen-and-paper writing process, in-air writing devices have been developed. On the other hand, algorithms that assume the presence of undesired, ligature strokes inside the genuine strokes have been recorded. The main difficulties range from acquiring trajectories to identifying acceptable gesture bounds. This document provides a thorough review of the many methods used to create Mid-Air Writing (MAW) systems. The majority of the systems in question, according to a thorough reading of the literature, are variants of the Hidden Markov Model (HMM), Dynamic Time Warping (DTW), and Longest Common Subsequence (LCS) strategies. However, this study also discusses a few unique MAW system-targeting strategies.