Journal of the Institute of Convergence Signal Processing (융합신호처리학회논문지)

The Korea Institute of Convergence Signal Processing (한국융합신호처리학회)
권호 표지

Implementation of CNN Model for Classification of Sitting Posture Based on Multiple Pressure Distribution

다중 압력분포 기반의 착석 자세 분류를 위한 CNN 모델 구현

  • Seo, Ji-Yun (Department of Computer Engineering, Dongseo University) ;
  • Noh, Yun-Hong (Department of Computer Engineering, Busan Digital University) ;
  • Jeong, Do-Un (Department of Computer Engineering, Dongseo University)
  • 서지윤 (동서대학교 컴퓨터공학과) ;
  • 노윤홍 (부산디지털대학교 컴퓨터공학과) ;
  • 정도운 (동서대학교 컴퓨터공학과)
  • Received : 2020.04.25
  • Accepted : 2020.06.26
  • Published : 2020.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

Musculoskeletal disease is often caused by sitting down for long period's time or by bad posture habits. In order to prevent musculoskeletal disease in daily life, it is the most important to correct the bad sitting posture to the right one through real-time monitoring. In this study, to detect the sitting information of user's without any constraints, we propose posture measurement system based on multi-channel pressure sensor and CNN model for classifying sitting posture types. The proposed CNN model can analyze 5 types of sitting postures based on sitting posture information. For the performance assessment of posture classification CNN model through field test, the accuracy, recall, precision, and F1 of the classification results were checked with 10 subjects. As the experiment results, 99.84% of accuracy, 99.6% of recall, 99.6% of precision, and 99.6% of F1 were verified.

근골격 질환은 착석 자세로 업무 및 학업을 장시간 진행하거나 잘못된 자세 습관으로 발생하는 경우가 많다. 일상생활에서 근골격 질환을 예방하기 위해서는 실시간 착석자세 모니터링을 통해 잘못된 자세를 바른 자세로 유도하는 것이 가장 중요하다. 본 논문에서는 의자에 밀착된 착석 정보를 무 구속적으로 검출하기 위하여 다채널 압력센서 기반의 자세 측정 시스템과 사용자의 착석 자세 분류를 위한 CNN 모델을 제안한다. 제안된 CNN 모델은 착석 자세 정보를 기반으로 압력분포에 따른 사용자의 5가지 자세 분석이 가능하다. 필드테스트를 통한 자세 분류 신경망의 성능평가를 위하여 10명의 피실험자를 대상으로 분류결과에 대한 정확도, 재현율, 정밀도 및 조화 평균을 확인하였다. 실험 결과, 99.84%의 accuracy, 99.6%의 recall, 99.6%의 precision, 99.6%의 F1을 확인하였다.

Keywords

Acknowledgement

본 연구는 교육부의 재원으로 한국연구재단의 지원을 받아 수행된 기초연구사업(No. 2018R1D1A1B07045337)과 과학기술정보통신부 및 정보통신기획평가원의 SW 중심 대학 지원 사업(2019-0-01817) 및 사회 맞춤형 산학협력선도대학(LINC+)의 연구 결과물임.

References

  1. S. H. Kim, J. G. Kim, "The effects of stretching exercise on the musculoskeletal pain relief of workers in small workplaces," Journal of Coaching Development, vol. 21, no. 4, 129-137, 2019. https://doi.org/10.47684/jcd.2019.12.21.4.129
  2. C. S. Kim, I. C. Choi, D. C. Kim, "A study on characteristics of a weighing system based on three-point load cells," Journal of Agriculture & Life Science, vol. 56, no. 1, 127-134, 2019.
  3. M. S. Kim, J. Y. Seo, Y. H. Noh, D. U. Jeong, "Implementation of cushion type posture discrimination system using FSR sensor array," The Korea Institute of Convergence Signal Processing, vol. 20, no.2, 99-104, 2019.
  4. Y. J. Shin, H. Y. Joo, J. H. Yang, "A study on the monitoring technique for musculoskeletal safety management and implementation of the system," Journal of Korea Institute of Information, Electronics, and Communication Technology, vol. 13, no. 3, 267-276, 2020.
  5. K. H. Kim, W. S. Lee, G. H. Oh, "A research on serious game for self-rehabilitation of patients with shoulder injury," Journal of Korea Game Society, vol. 19, no. 3, 87-99, 2019. https://doi.org/10.7583/JKGS.2019.19.3.87
  6. D. W. Kim, H. J. Nam, S. Y. Lee, Y. K. Haam, O. S. Seo, H. J. Lee, "Multi-user home-training healthcare system using kinect sensor and wearable devices," The Journal of Korean Institute of Communications and Information Sciences, vol. 44, no. 4, 719-727, 2019. https://doi.org/10.7840/kics.2019.44.4.719
  7. G. J. Park, B. Binayak, "Railway facility real-time intruder monitoring system using computer vision and deep learning," The Korean Society For Railway, vol. 23, no. 1, 35-44, 2020. https://doi.org/10.7782/JKSR.2020.23.1.35
  8. J. Y. Han, J. H. Park, "Turtle neck syndrome posture correction service using CNN-based learning model," The Korea CONTENTS Society, vol. 20, no. 7, 47-55, 2020.