A STUDY OF DIFFERENT METHODOLOGIES, APPROACHES & ALGORITHMS EFFICIENCIES FOR HAND GESTURE RECOGNITION

Abstract

Hand gesture recognition has been most sought to be research field of artificial intelligence and image processing sub domain as well. This chapter has been structured with introduction and related work in this domain .Our work basically focuses on the different methodologies and drawbacks associated with them. The domain has been discussed right from hand gesture recognition robot control to haar transform. The other dimension for tracking and classifications along with different pose and convexity defect has also been discussed in detail and efficient concept has been proposed. In the next section of chapter convolution neural network for hand gesture recognition has been discussed in detail with advantages and disadvantages. Computer vision technique and image is used to utilize hand motions through implementing monitoring and controlling of computer by “monitoring the computer and control with hand movementsâ€. In proposed study, we use YCbCr model to detect the region of hand and Harr classifier is used to detect the user face. Means Blob 10 technique is used to trace the hand and Convexity Defect-Hull is used to extract the feature. at last according to hand region feature the mouse position is controlled.