Measurement and Evaluation System for Self-Training System of Bed-Making Activity

This paper aims to construct a nursing self-training system of bed-making in which nurses must perform some skills for handling equipment (a bed, a bed pad, and a sheet) and for avoiding bodily injury. Eight evaluation points of the bed-making task, related to trainees’ posture and states of equipment, were identified through nursing textbooks and discussed with nursing teachers. To recognize and evaluate the points by image processing, we developed a system using three RGB-D (RGB color and depth) sensors. To increase the recognition rate, we clustered the color information by the K-means clustering method and then divided the bed-making procedure into three segments using color and depth information of the whole images. The average accuracy achieved by the proposed system in the evaluation experiment where 15 trainees participated was 80 %, an average consistent with that achieved by a human nursing teacher.