Development of perceptual functions as primitive awareness for VDT work environment

Abstract

As the previous work, some perceptual functions such as sitting posture recognition or human's location point recognition have been proposed for awareness of VDT worker's activity in an effort to realize human-computer symbiosis, that is, Symbiotic Computing. However, there are some problems in their functions. For example, in case of sitting posture recognition, they can not achieve the dynamic sitting posture recognition as they confine to static sitting posture clustering. In addition, human's location point estimation has been achieved with low estimation accuracy, it is necessary to improve the accuracy. In this study, we extended the previous two perceptual functions, which are dynamic sitting posture recognition on seat and human's location detection in front of the desk. We utilizes optical fiber pressure sensor mattress and for sitting posture recognition and laser rage finder for human's location detection. Two experiments have been done to evaluate the developed functions. As the results, we show that the proposed functions are quite competent to recognize human's activities in VDT work environment rather than our previous study.