Challenges and opportunities in finger spelling recognition in air

Finger spelling in air refers to writing in the air using finger movements. In the noisy environment it is difficult to communicate with the person at several distances which causes the problem. In such condition writing in air can be used as a communication mode. This paper studies the challenges occur during communication and presents a real-time video based method which allows writing alphabets and numerals in free air in front of system camera. Proposed method tracks the color of fingertip and extracts the color using Euclidean filter which gives better results in complex and noisy background. Then system applies OCR to recognize the written character. The proposed system is software based and relevant. This provides a human-machine interaction which does not require sensors or additional hardware. Only the system camera and a bright color tape are used. This system can be used efficiently at any noisy places like railway station, airport, and traffic signal for effective communication.