A Comprehensive Study of Real-Time Hand GestureDetectionand Recognition

Abstract

Hand gesture is an important form of communication used to convey instructions, give commands, and is an importantmeans for mute people to express themselves. The vast majority of the world population do not know about sign language thatthemutepeopleusetocommunicatewitheachotherandhencefinditdifficulttocomprehendwhatthedisabledpersonistryingto convey. With the growth in the field of human-computer interaction, a significant amount of research work has been devotedto bridging this communication gap.Breakthroughs in the field of computer vision, deep learning, Internet-of-Things, andcomputational power or capability has motivated a significant amount of work towards real-time hand gesture detection andrecognition. The primary reason for the popularity of this field of research is the challenge of generalization of the system tomakeitaccessibletoallthepeoplewhoneeditandtheproblems associatedwithhanddetection.Inthispaper,weexploresomenovel approaches used to solve this problem and compare the results of these systems to determine their compatibility underdifferentscenariosandtheirimpactonsociety.