Most modern people are exposed to various musculoskeletal disorders as their sitting time increases.Persistent wrong sitting postures lead to musculoskeletal disorders as well as mental illnesses such as decreased concentration, distraction and others.The system implemented in this paper uses Kinect sensor and SDK to extract 2D image and skeleton image to determine user's movement. Thereafter, three types of skeleton information of the user are detected to measure a change amount of the skeleton coordinate value that varies according to the movement. The measured coordinate data can determination distraction degree and posture by the posture determination algorithm built in the PC program.For performance evaluation of the system, the postures determined by the implemented system were compared with the postures determined by video recording. Experiments were in which a total of 13 subjects conducted 10min. As a result of comparing the posture of the video recording and the implemented system, the concordance rate of 88.15% was confirmed. Also, distraction estimation performance was evaluated according to the posture change rate. The experimental results showed that sections with high posture change rates showed higher distraction degree than those with low posture change rates. Therefore, the experimental results confirmed that the implemented system can be used to determine posture and distraction estimate.Some errors are thought to have been caused by the photographing angle of Kinect device and the physical characteristics of user. In the future, we want to conduct comparative analytical studies using EEG to ensure the objectivity of distraction presumption indicators.
Volume 11 | 06-Special Issue
Pages: 2050-2054