Hand Gesture Detection using Deep Learning

1S.Malli Babu


Gesture Recognition is a form of user experience with contextual technology which makes devices to recognize and recognize physical actions as directions. The general idea of  fingers recognition is a device's ability to identify movements and to perform instructions based on such gestures. Gesture Recognition is actually the new subject in the area of computational science and technology with just the aim of decoding physical gestures through a variety of mathematical models. We are currently concentrating on understanding hand movements in our design. Management awareness could be seen as a path for machines to start learn the language of the human body. Therefore creating a stronger interface between computers and users than simple text user interface design or even GUIs that still reduce most inputs to the wireless controller. In this project, we are using the Gaussian Mixture, Backgroundsegmentation , Foreground Segmentation Algorithms to represent an application to real-time hand gesture identification. We introduce a method for extracting of the functionality including measures on the fingers of the bodies removed. The current algorithm should create a model for subtracting context to get the object in the foreground. For boolean pictures we apply Gaussian blur algorithm to the foreground picture to be detected and threshold algorithm is included in this functionThe convex hull and convexity can be used to construct a hand gesture observation of 3D object. We may create a data set, and we will describe the direction of the finger such as up, down, left, right and stop. By certain expectations we train them by gesture classifications so that they can be easily understand and gives the output properly for the given output. Experimental findings will show the feasibility of our strategy and its efficiency


Deep learning, Gaussian algorithm, Hang gestures

Paper Details
IssueIssue 3