A Review Paper on Air Canvas Application and Text-to-Speech System using Deep Learning

Abstract

Air Handwriting is one of the growing technologies in day by day life. Air handwriting enables user to write in air by using finger movements. The web camera detects the finger movements and converts it into readable format. This will be the natural way of communication with computer system. This will remove the need of physical input devices like keyboard, touchscreen and digital pen. This paper aims to provide an effective platform for both communication and practicing. The existing system has some disadvantages that are overcome in this paper. The existing system requires multiple fingers for writing. But using multiple fingers for different tasks like writing, changing color and erasing is complicated thing to remember. A strong, robust and efficient algorithm is proposed that will extract all the air writing trajectories or curves that are collected using a single web camera. The algorithm avoids restrictions on user’s writing without using a delimiter and an imaginary box. The deep learning CNN algorithms are also used for converting hand written text into user readable text. Additionally, Optimizing algorithms for efficiency can help ensure smooth and responsive performance