INTERNATIONAL RESEARCH JOURNAL OF HUMANITIES,
ENGINEERING & PHARMACEUTICAL SCIENCES

Smart wheelchair is a significant solution for assisting the individuals which are unable to perform their daily activities by normal powered wheelchairs. For this purpose this paper proposed an application based on communication modes of human mainly by vision to develop a man-machine interface. It would be very intuitive if modern wheel chair has greater use of vision and speech recognition related control. The disabled user can control a wheel chair from the distance and without any physical contact with the computer. Instead of audio commands we have emphasized on visual system because they are more preferable in noisy environments or in places where other sounds can cause a disturbance in audio control signals. Authors have designed an algorithm using single color camera on the computer. The user will wear a colored wrist band for gathering shape and position information of the hand using detection of skin and wrist band color. Authors have applied image processing techniques for refining shape information for getting a robust comparison with a set of predefined real data to recognize the sign indicated by hand gesture. Our technique has given its contribution in developing hand gesture reorganization with a higher speed in comparison process that can be utilized to recognize 5 different hand gestures in same backgrounds for five movements of wheelchair.