Informatics Publishing Limited

Abstract

Sign language based communication is basically essential by consistently changing distinctive hand shapes and development by a signer. The proposed framework going to recognize the sign of gesture-based communication from a video stream of the signer and deciphers into relating words and sentences in Hindi. The proposed framework utilizes the image preparing strategies and synthetic intelligence procedures to get the target. To do the proposed task execution it utilizes picture handling strategies, for example, frame differencing based tracking, edge recognition, wavelet change, picture combination systems to area shapes in our videos. It moreover utilizes descriptors of elliptical Fourier for feature extraction of shape and also uses analysis for the purpose of feature set reduction as well as optimization. Extracted features are stored into the database and compared with given input video of the signer by a trained unclear inference system

Keywords

Full Text