Research Scholar, Department of Information Communication Engineering, Mokwon University, 88 Doanbuk-ro, Seo-gu, Daejeon, 35349, Korea
The purpose of this study is to analyze the posture patterns of the user group with back pain and the user group without back pain using the data transmitted from the wearable sensor attached to the user's body and chair and to guide the correct posture based on this classification.
In this study, the sitting posture of the user was analyzed by attaching a tilt sensor and a pressure sensor to the wearable device and the chair. For the study, sensor data for users with and without back pain were collected via the web. The Support Vector Machine(SVM) model was used to classify the posture patterns that could cause the correct posture and pain. This posture pattern can predict future pain according to the current posture.
As a result of the study, the posture patterns of the users with and without back pain were classified by the hard margin method and it was verified through the program whether the body was inclined to the left or to the right and bent back or forth based on the posture pattern. The two groups of users were classified clearly in the sitting posture pattern and highly reliable in predicting the presence of back pain in random user data. The results of these studies are expected to be useful for predicting incorrect posture that cause back pain in the future.
This studyfocuses on back pain that leads to neck and shoulder pain. However, it is necessary to analyze the pain area in detail in future and it is also necessary to optimize the monitoring system for posture prediction that can cause pain.
Sitting Posture, Wearable Sensor, Support Vector Machine, Posture Pattern, Classification, Prediction
