Contrivance of Recognised Hand Gestures into Voice and Text Output
Communication between mute and normal people has always been very difficult. Mute people use the hand
gestures as a sign language to convey their feelings to normal people. Sign language (hand gesture) varies from person to
person and place to place. It is difficult for normal people to understand these hand gestures without learning them. To
overcome this difficulty we proposed a model which uses finger counts as the hand gestures and converts them into voice and
textual information. The model contains two techniques to perform the hand gesture recognition. In first technique the hand
gestures are captured and analysed using an algorithm written in MATLAB for image processing. The second technique is a
real-time algorithm, which is used to convert the hand gestures directly into voice and textual information without capturing
and storing any image i.e. it takes video as an input in real-time. The above mentioned techniques have been performed and
compared to design a system which can give an accurate output. From the analysis performed it was found that both the
techniques can be used to give an accurate output, if provided with good lighting conditions and even background. The
proposed model does not require any data set and hence it reduces the memory used and also reduces the complexity of the
system. The main advantage of this system is that it does not perform any comparison with the data-set hence it is not user
specific and shows less error as compared to data set based techniques.
Index Terms - Hand gesture recognition, Real-time algorithm, Data-set.