Cursor Controller using Hand Gesture Recognition

This paper discusses mouse cursor controller using hand gesture recognition which is an application in the field of Human Computer Interaction. The author has developed an application using the MediaPipe framework and also compared its accuracy and performance with approaches available in a variety of literature. The goal of this paper is to discuss methods to get rid of the extra hardware we need for cursor control, improve user interaction with comfortable hand gestures, reduce the computing time and increase the hit rate (accuracy/performance), as well as to use the discussed methodologies to develop software consisting of all features that a regular mouse provides. Our software must be able to perform almost all features like left Click, right-click, Brightness and volume increase/decrease, scrolling, selection, etc. By tracking the hand gestures of our user our system can perform all the above-mentioned operations with the use of a low-resolution camera only.