Hand Gesture Recognition using Deep Learning Models

Abstract

Hand Gesture Recognition (HGR) is most widely used in many applications to detect the hand movements of humans in front of the camera. HGR is the most innovative system that provides a user-friendly nature that can interact with a system that is known to humans. Applications such as machine interaction with humans, language signs, and analyzing the original hand gestures by the system is a very difficult task. HGR detection can be done based on the shapes of the four fingers and one thumb to recognize the hand and their respective location in the image. In this paper, a new HGR technique is introduced based on the shape metrics with the text-to-image conversion technique. For implementation, the Deep Convolution Neural Networks (DCNN) have been used with inception V3. The model considers the skin color and texture because the features of the image are specifically different conditions that are influenced. A webcam is used in the proposed system which is working on 20 fps with 7-megapixel intensity.