DOI QR코드

DOI QR Code

Research on Classification of Sitting Posture with a IMU

하나의 IMU를 이용한 앉은 자세 분류 연구

  • Received : 2017.05.07
  • Accepted : 2017.08.29
  • Published : 2017.08.31

Abstract

Bad sitting postures are known to cause for a variety of diseases or physical deformation. However, it is not easy to fit right sitting posture for long periods of time. Therefore, methods of distinguishing and inducing good sitting posture have been constantly proposed. Proposed methods were image processing, using pressure sensor attached to the chair, and using the IMU (Internal Measurement Unit). The method of using IMU has advantages of simple hardware configuration and free of various constraints in measurement. In this paper, we researched on distinguishing sitting postures with a small amount of data using just one IMU. Feature extraction method was used to find data which contribution is the least for classification. Machine learning algorithms were used to find the best position to classify and we found best machine learning algorithm. Used feature extraction method was PCA(Principal Component Analysis). Used Machine learning models were five : SVM(Support Vector Machine), KNN(K Nearest Neighbor), K-means (K-means Algorithm) GMM (Gaussian Mixture Model), and HMM (Hidden Marcov Model). As a result of research, back neck is suitable position for classification because classification rate of it was highest in every model. It was confirmed that Yaw data which is one of the IMU data has the smallest contribution to classification rate using PCA and there was no changes in classification rate after removal it. SVM, KNN are suitable for classification because their classification rate are higher than the others.

바르지 못한 앉은 자세는 다양한 질병과 신체 변형을 유발한다. 하지만 오랜 시간동안 바른 앉은 자세를 유지하는 것은 쉬운 일이 아니다. 이러한 이유 때문에 그동안 자동으로 바른 앉은 자세를 유도하기 위한 다양한 시스템이 제안되어왔다. 이전에 제안되었던 앉은 자세 판별 및 바른 앉은 자세 유도 시스템은 영상 처리를 이용한 방법, 의자에 압력센서를 달아 측정하는 방법, IMU(Internal Measurement Unit)를 이용한 방법이 있었다. 이 중 IMU를 이용한 측정 방법은 하드웨어 구성이 간단하고, 공간, 광량 등의 환경적 제한이 적어 측정에 있어서 용이한 이점이 있었다. 본 논문에서는 하나의 IMU를 이용하여 적은 데이터로 효율적으로 앉은 자세를 분류하는 방법을 연구하였다. 특징추출 기법을 이용하여 데이터 분류에 기여도가 낮은 데이터를 제거하였으며, 머신러닝 기법을 이용하여 앉은 자세 분류에 적합한 센서 위치를 찾고, 여러 개의 머신러닝 모델 중 가장 분류 정확도가 높은 머신러닝 모델을 선정하였다. 특징추출 기법은 PCA(Principal Component Analysis)를 사용하였고, 머신러닝 모델은 SVM(Support Vector Machine), KNN(K Nearest Neighbor), K-means (K-means Algorithm) GMM (Gaussian Mixture Model), and HMM (Hidden Marcov Model)모델을 사용하였다. 연구결과 데이터 분류율이 높게나온 뒷목이 적합한 센서 위치가 되었으며, 센서 데이터 중 Yaw데이터는 분류 기여도가 가장 낮은 데이터임을 PCA 특징추출 기법을 이용하여 확인하고, 제거하여도 분류율에 영향이 매우 작음을 확인하였다. 적합 머신러닝 모델은 SVM, KNN 모델로 다른 모델에 비하여 분류율이 높게 나오는 것을 확인할 수 있었다.

Keywords

References

  1. Shang Hyuk Yim, Yun Geun Lee, Jung Jin Cho, Jeong Il Son, Jae Cheol Song. "Sympton Prevalence of Work-Related Musculoskeletal Disorders and Related Factors among Bank Workers by Visual Dispaly Terminal Use." ANNALS OF OCCUPATIONAL AND ENVIRONMENTAL MEDICINE, vol. 9, no. 1, pp. 85-98, 1997.
  2. Min-Seok Ko, Joon-Young Choi. "Related Factors to Visual Display Terminal Syndrome of Hospital Employees Using Lalonde Health Field Model." Health and Welfare, vol. 14, pp. 7-22, 2012. https://doi.org/10.23948/kshw.2012.12.14.7
  3. Editorial Department, "The New Danger of Modern People! VDT Syndrome." Defense & Technology, vol. 354, pp. 90-91.2008.
  4. Hee-suk Park, Yun-Geun Lee, Dong-min Oak, Dae-sung Kim, "Comparison of Several National Guidelines on Ergonomic VDT Workshops." Korean Institute of Industrial Engineers Conference, pp. 188-194. 2008.
  5. Dorn Q Kim, Soo-Hun Cho, Tai-Ryoon Han, Ho-Jang Kwon, Mina Ha, Nam-Jong Paik. "The Effect of VDT Work on Work-related Musculoskeletal Disorder." ANNALS OF OCCUPATIONAL AND ENVIRONMENTAL MEDICINE, vol. 10, no. 4, pp. 524-533, 1998.
  6. Mi Jung Sim, Young Sook Lee. "A study on the VDT syndrome and relating work factors of office workers." Korean Journal of Health Education and Promotion, vol. 25, no. 3, pp. 95-109, 2008.
  7. W. H. CHO, W. Y. LEE, H. K. Choi. "An Investigation on the Biomechanical Effects of Turtle Neck Syndrome through EMG Analysis." Inter. J. of Prec. Eng. & Man. Conf., pp. 195-196. 2008.
  8. Knave, B. G., Wibom, R. I., Voss, M., Hedstrom, L. D., and Bergqvist, U. O., Work with video display terminals among office employees : I. Subjective symptoms and discomfort, Scandinavian Journal of Work, Environment and Health, pp. 457-466, 1985.
  9. Kang, J.H., Park, R.Y., Lee, S.J., Kim, J.Y., Yoon, S.R., and Jung, K.I. The effect of the forward head posture on postural balance in long time computer based worker. Annals of Rehabilitation Medicine, vol. 36, pp. 98-104, 2012. https://doi.org/10.5535/arm.2012.36.1.98
  10. Park, J.H. et al., Body pressure distributions of the varying sitting postures in office work, Proceedings of the spring conference of Korean society of ergonomics, 2011.
  11. M.R. Kim, H.W. Kim and W.D. Cho, "posture helper using Gaussian mixture background modeling," in Proc the Korean Institute of communications and Information Sciences, Pyeongchang, Korea, pp. 25-26, 2010.
  12. S.J. Lee and S.K. Jung, "Posture symmetry based motion capture system for analysis of lower-limbs rehabilitation training," Journal of Multimedia Information System, vol.14, no. 12, pp.1517-1527, 2011.
  13. H.J. Ha and C. D. Lee, "Design of Algorithm for Guidance of Sitting Posture Correction Using Presusure Sensor and Image Processing Interpolation Technique," Journal of Korean Institute of Information Technology, vol.14, no. 1, pp. 37-44, 2016.
  14. K. R. Ko and S. B. Pan, "Feature extraction and classification of posture for four-joint based human motion data analysis," journal of the institute of electronics and information engineers, vol. 52, no. 6, pp.117-125, 2015. https://doi.org/10.5573/ieie.2015.52.6.117
  15. Shumei Zhang, Paul McCullagh, Chris Nugent, Huiru Zheng, Matthias Baumgarten, "Optimal model selection for posture recognition in home-based healthcare.," Springer-Verlag, 2010.