Recently, the number of disabled peoples is increasing resulting from traffic accidents and sports accidents. The aim of this study is to develop a powered orthoses to provide upper-limb motion aid for these disabled peoples. In the earlier stage, an upper-limb orthoses was developed as a prototype. In this prototype, the command input is based on the neck inclination. Such a method of command input is sometimes inconvenient for the disabled peaples. In this paper, a new orthoses is described, in which a voice recognition is introduced for the command input system to avoid the inconvenience. And also one board computer is introduced in order to realize the transportability of the orthoses system. In the system, whole control unit are installed on a wheelchair. These improvements produce the powerd orthoses, which can be controlled simply and precisely as well.

View full abstract

The purpose of the present study is to detect the gazing direction from human face images. From the preliminary experments, it was found that the image including both eyes was sufficient to detect the gazing direction. We tried to determine the gazing direction from the relative position between the irises and the eye rims automatically using the image of both eyes stored in a frame memory. The deflection angle is determined from the distance between the center of both irises and the center of symmetry, which is obtained as the interior division of the distance between the inner edges of the eyes weighted by the product of the width and the height of the eyes. The present algorithm has an inherent deflection toward one's dominant eye by about 2degrees. A 60% of the obtained deflection angles falls within2degrees of the correct value when the deflection angle is θ=0.

View full abstract