Novel application of kinect sensor to prevent collision accidents between machine and operator

Abstract

Many collision accidents between operators and machine tools are reported. However, no effective way of preventing these accidents has been proposed to date. Therefore, the authors propose the implementation of a danger decision by measuring the human posture and face direction. By performing a danger decision, the authors aim to realize the effective prevention of collision accidents. The posture and face direction were measured by using a Kinect sensor, which has the ability to obtain information about the posture and face direction. This paper presents the results of experiments aimed at examining whether it is possible to realize a danger decision. The experimental results confirm the feasibility of using a danger decision based on the measurement of the operator's posture and face direction.