Recognition of Postural Rolling of Patients Body from Video Image for Tele-nursing System

K. Ohta, E. Yasui, T. Kawaguchi, M. Azuma, and C. Ohta (Japan)

Keywords

Pattern Recognition, Video Image,Tele-nursing, Postural Rolling

Abstract

This study has established the method of recognition of the postural rolling of a patient's body through the use of a video-mail system, in order to evaluate the priority of the video mail in the tele-nursing system. The patient's appearance is recorded by a video camera in front of patient. Next, consecutive video pictures are resolved into still and individual frame images. The face area and background are extracted by calculating the difference of the consecutive frames. The position of the pupil is extracted from the face area by using the area separation filter. To extract the postural rolling of the body, the sequential change in the position of the pupil is measured. The change in the vertical direction is almost few. We focused on the change in the horizontal direction. In addition, the behavior of the sequential change in the horizontal direction is transformed into the chaotic behavior and the Fourier power spectrum. This paper proposes that there are differences between healthy and unhealthy by examining chaotic behavior and the Fourier power spectrum, and those are useful to evaluate the priority of the video mail in the tele-nursing system.

Important Links:



Go Back