A Review on Vision-Based Hand Gesture Recognition and Applications

Introduction

Presently, vision based hand gesture recognition has become a highly developing research field for the purpose of human computer interaction. Such recognition systems are deployed to serve as a replacement for the commonly used human-machine interactive (HCI) devices such as keyboard, mouse, joystick etc. in real world situations. Recently, gestures have become an important segment of such HCI devices. A gesture may be defined as a physical movement of body parts such as hands, arms, head, face etc. to express some information or feelings (Murthy & Jadon, 2009). Gestures play an important role in our day to day communication. The ability of a computer or any processing system to understand the meaning of these gestures is referred to as gesture recognition. Among the various types of gestures, hand gestures are the most commonly used, as they are natural, easy to use and more convenient for communication. Hand gestures are basically of two types- static and dynamic (as shown in Figure 1). Static hand gestures do not involve any kind of hand movement in comparison to dynamic hand gestures, where either the entire hand moves (global motion) or only the fingers move (local motion) (Ahmeda, Alexanderb, & Anagnostopoulos, 2008).

Figure 1.

Types of hand gestures

There are mainly two approaches of hand gesture recognition: glove-based and vision-based. In glove-based method, the user needs to wear a sensor glove or a colored glove, which serves as an interface to communicate with the computer. Although, this approach gives accurate results, it affects the ease and naturalness with which the user interacts with the computer. Vision-based approach overcomes the drawback of glove based approach as it serves as a natural means of interaction. This method adopts computer vision and machine learning algorithms for recognizing the hand gestures. However, obtaining highly accurate results is a challenging task for vision based approaches (Rautaray & Agarwal, 2012). In this chapter we have surveyed the literature on various vision-based hand gesture recognition systems and its application in different domains.

There are several applications of hand gesture recognition systems such as sign language recognition, human-robot interaction, controller less video gaming, smart TV, video surveillance etc.. Among these, sign language recognition has become an active topic of research as it provides an opportunity for the hearing impaired to communicate with the normal people without the need of an interpreter.

This chapter provides an insight into the basic process of hand gesture recognition, the detailed description of each of the sections and the related works which have been done in this field.

Background

Gesture is a natural mode of communication which enables human beings to interact with each other from distance without the need of physical devices. The necessity of computing devices has increased tremendously in recent times. There are many physical devices like keyboards, mouses, joysticks, remote control etc. which can be used as interface for communicating with computers. However, the ease of this interaction can be greatly enhanced by the use of hand gestures. The ability of hand gestures to communicate with computing devices led to the existence of hand gesture recognition systems.

Hand gloves or sensor gloves were primitively used in 1990 as interface technique for communication between humans and computers. Although they provide very accurate recognition results, yet they are very cumbersome and unnatural. So, this led to the development of vision-based techniques for human-computer interaction (HCI).