4 Gesture navigation
4.1 Gesture classification based on various criteria

The gesture controlled application become more and more popular. The gestures can be divided into two basic categories by user experience. Innate gestures that are based on general experience of all users such as to move an object to the right by moving hand to the right, catch an object with closed fingers, etc. Naturally, innate gesture can be effected by habits or culture. By using these gesture we can support no need to learn user to get experience for gesture control. We suppose the user can control application naturally. Second group of gestures are learned gestures. In this case it is necessary to teach the users. The gestures used for navigation of the systems and applications has to be easy, natural and has to spend minimum of human energy.

Static gestures

Static gestures do not depend on movement. Static gestures represent shapes of the gesturing limbs, which carry the meaningful information (Fig. 4.1).

They are “non-moving” type of gestures where we do not need information about the motion. Thus there is no need to investigate a sequence of frames, rather than the actual image frame. A sequence of several frames containing different gestures is only important in the higher level of comprehension.

Continual gestures

Continuous gestures serve as a base for application interaction. Continual gesture is prolonged tracking of movement where no specific pose is recognized but the movement is used to interact with the application or virtual environment (Fig. 4.1).

The typical example of continual gestures is game control using a touchless technology such as Microsoft Kinect, etc. where the system maps the changes in postures to changes in the video game, but there is no specific movement to trigger a sequence of changes.

Dynamic gestures

Dynamic gestures represent movement that allows users to directly manipulate an object or control application (Fig. 4.1).

Dynamic gestures can be defined in two ways. First approach is to algorithmically set rules and conditions that the performed gesture has to pass to be evaluated and successfully recognized. The second way is to use templates. Templates are sets of points which identify the shape of the gesture. Sophisticated algorithms are used to evaluate success rate when comparing user movements with templates.

image
Fig. 4.1 – Gesture types: a) static gesture, b) dynamic gesture, c) continual gesture