OpenCV-based Squat Posture Correction Model
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The pandemic caused by COVID-19 has affected many fields at home and abroad, and many changes have occurred. In particular, the home training market has expanded through e-commerce, and the population training at home has rapidly spread. However, if you exercise alone at home without the help of experts, you are inevitably exposed to the risk of injury, and watching videos online has many limitations. Squat, one of the types of training, has the best exercise effect for lower body exercise and is programmed and operated for therapeutic purposes such as performance improvement and injury prevention. However, due to an unstable posture, damage to the waist and knees may occur and may be a factor threatening health. In this paper, we propose a model that can analyze unstable postures based on OpenCV and MediaPipe to provide comprehensive information on problem-solving and posture correction for injury prevention and feedback on squat posture. The proposed model is based on the standard squat and wide squat, which are routinely applied by beginners to experts. As a result of measuring the accuracy of the squat pose of the proposed model, it showed an accuracy of 97.85%, and as a result of a survey of users, 83.3% of them were satisfied. The proposed model is expected to contribute to health by maintaining the correct posture of squat, one of the exercise types of home training.
Abstract
The pandemic caused by COVID-19 has affected many fields at home and abroad, and many changes have occurred. In particular, the home training market has expanded through e-commerce, and the population training at home has rapidly spread. However, if you exercise alone at home without the help of experts, you are inevitably exposed to the risk of injury, and watching videos online has many limitations. Squat, one of the types of training, has the best exercise effect for lower body exercise and is programmed and operated for therapeutic purposes such as performance improvement and injury prevention. However, due to an unstable posture, damage to the waist and knees may occur and may be a factor threatening health. In this paper, we propose a model that can analyze unstable postures based on OpenCV and MediaPipe to provide comprehensive information on problem-solving and posture correction for injury prevention and feedback on squat posture. The proposed model is based on the standard squat and wide squat, which are routinely applied by beginners to experts. As a result of measuring the accuracy of the squat pose of the proposed model, it showed an accuracy of 97.85%, and as a result of a survey of users, 83.3% of them were satisfied. The proposed model is expected to contribute to health by maintaining the correct posture of squat, one of the exercise types of home training.
참고문헌 (15)
- 이용준, "딥러닝 기반 포즈 인식 및 교정을 통한 효율적인 홈 트레이닝 시스템 개", 한국차세대컴퓨팅학회 논문지, 17(6), pp.89~99 (2021), 한국차세대컴퓨팅학회[DOI]
- 권재윤, "빅데이터를 활용한 홈트레이닝 인식 분석 : COVID-19 확산시점을 중심으로", 한국체육과학회지, 30(2), pp.447~459 (2021), 한국체육과학회[DOI]
- 김기홍, "스탠스 너비와 무릎 각도가 스쿼트 운동시 반복횟수, 총운동량및 근전도 반응에 미치는 영향", 아시아 운동학 학술지, 19(3), pp.27~34 (2017), 대한운동학회[DOI]
- 권재윤, "코로나19 펜데믹 전후로 홈트레이닝 트렌드는 어떻게 변화했는가?", 한국사회체육학회지, (90), pp.279~293 (2022), 한국사회체육학회[DOI]
- 박만석, "데드리프트 운동 유형에 따른 생체역학적 신체 변화에 미치는 영향", 한국체육과학회지, 24(6), pp.981~989 (2015), 한국체육과학회
- K. K. Chul, "Effect of type-specific squat exercise on lower body muscle activity", Department of Physical Education, pp.2018
- I. Culjak, "A brief introduction to OpenCV", Proceedings of the 2012 35th International Convention MIPRO, pp.1725~1730 (2012)
- 김진영, "청각장애인의 수어 교육을 위한 MediaPipe 활용 수어 학습 보조 시스템 개발", 한국전자통신학회 논문지, 16(6), pp.1355~1362 (2021), 한국전자통신학회
- "IEEE Robotics & Automation Magazine", 2008, 100~
- D. Kavitha, "Multiple object recognition using opencv", REVISTA GEINTEC-GESTAO INOVACAO E TECNOLOGIAS, 11(2), pp.1736~1747 (2021)
- H. Adusumalli, "Face mask detection using opencv", 2021 Third International Conference on Intelligent Communication Technologies and Virtual Mobile Networks (ICICV), pp.1304~1309 (2021)
- C. Lugaresi, "Mediapipe : A framework for building perception pipelines", arXiv:1906.08172, 1~9 (2019)
- 권예진, "Real-Time Workout Posture Correction using OpenCV and MediaPipe", 한국정보기술학회논문지, 20(1), pp.199~208 (2022), 한국정보기술학회[DOI]
- V. Bazarevsky, "Blazepose: On-device real-time body pose tracking", arXiv preprint arXiv:2006.10204, 1~4 (2020)
- Y. Kartynnik, "Real-time facial surface geometry from monocular video on mobile GPUs", arXiv preprint arXiv : 1907, 1~4 (2019)