English Alphabets Recognition using Hand Gesture

Abstract

In Human-Computer Interaction (HCI) area, the sign language is needed for individuals with special needs (dumb) for the purpose of communicating with individuals with no special needs and also for communicating with each other. Thus, the sign language is determined via using the systems of hand recognition. The initial stage in utilizing manual signal recognition is deleting the background in addition to a part (arm/hand) from other body parts just as much as feasible. Concerning the presented study, we are offering a novel approach for representing the alphabet of English language as 6 pattern gestures. The hash pattern has been applied through utilizing histogram equation for gray-scale image. The process of classification has been implemented via the Euclidean distance function as well as Fisher Linear Discrimination Analysis (FLDA). Difficulties including recognizing similar gestures appear to be managed quite efficiently with the proposed technology, as the background is fixed white. The recognition rate has been 82.86 since the lighting differs between different places, in addition to the differences in resolution of the camera, that is highly important in characterizing features.