Gesture Recognition Using Modified HSV Segmentation

Gesture recognition system is an important link between humans and machines, in this paper we are going to present a new gesture system that used to control the human made machine by teleoperations, we have applied in this system the HSV color model that used for segmentation operation with some modification to overcome the problem of incomplete segmentation, the input gesture is divided into blocks and the features are extracted using our suggested brightness feature vector, we have made a study by considering all the block sizes starting from single pixel block size which is 1x1 and going onward passing by all the odd blocks numbers up to 23x23 block size in order to reveal the performance of gesture system through different block sizes, we have achieved an average recognition time of 1.676 seconds per gesture, in this paper we have focused on hand gestures instead of other organ because the hand can express much meaning and this is why it is used commonly by deaf people and we have applied static gesture system.