MFCC와 DTW를 사용한 노인 낙상사고 인지 방안 연구
Senior citizens fall accident Situation Recognition using Mel Frequency Cepstral Coefficients (MFCC) and Dynamic Time Warping (DTW) Techniques
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
댁내 환경 상에서 노인들의 낙상사고가 많이 발생하고 있으며, 특히 낙상사고의 대부분은 화장실에서 발생한다. 비록 노인 분들의 사고를 방지하기 위한 방식은 많으나, 노인 분들이 119에 신고하기에는 어렵다는 문제가 있다. 사물인터넷 환경에서는 IoT Broker가 많은 센서들로 부터 정보를 수집하며, 지능적 정보 분석을 통해, 사용자에게 서비스를 제공한다. 본 논문에서는 MFCC와 DTW를 사용하여 소리인식 기반의 노인 낙상사고 상황 인지를 개발 및 제안하고 있으며, 상황 인지를 위해, 4개의 상황 카테고리 별 소리 수집 및 매트랩을 통한 분석을 진행하였다. 분석 결과 약 87.5%의 정확도를 가지며, 향후 노인 분들의 화장실에서의 낙상사고 시 대응에 도움이 될 것으로 보인다.
Abstract
There many senior citizens accident in home, especially, falling accidents at bathroom hold a large majority of senior citizen. Although, there are many preventive measures against senior citizen accident, it is difficult for senior citizen to report to 911. In Internet of Things environment, IoT broker collects information from a lot of sensors and operate intelligent data analysis for providing services to users. We develop seniors falling accident Situation Recognition based on sound recognition using Mel Frequency Cepstral Coefficients (MFCC) and Dynamic Time Warping(DTW)Techniques. For situation recognition, we made 4 situation categories at bathroom, recorded WAV files and analyzed sound based on Matlab, its correct rate is 87.5%. This seniors falling accident situation recognition will be helpful to react quickly on senior falling accident situation at bathroom.
참고문헌 (12)
- Yu, Xinguo, "Approaches and principles of fall detection for elderly and patient", e-health Networking, Applications and Services, 2008. HealthCom 2008. 10th International Conference on. IEEE (2008)
- Muda, Lindasalwa, "Voice recognition algorithms using mel frequency cepstral coefficient (mfcc) and dynamic time warping (dtw) techniques", arXiv preprint arXiv:1003.4083, ~ (2010)
- Meinard Muller, "Information Retrieval for Music and Motion", pp.69~84 (2007), Springer-Verlag
- Cowling, Michael, "Comparison of techniques for environmental sound recognition", Pattern Recognition Letters, 24(15), pp.2895~2907 (2003)
- Hayashida, K., "Multi-stage identification for abnormal", warning sounds detection based on maximum likelihood classification, 2013, pp.19(1), 055020~[DOI]
- Speech recognition with dynamic time warping using MATLAB, "Lama,Palden", ~, pp.Project Report, CS 525 (2010)
- Valero,Xavier, "Gamma tone cepstral coefficients: Biologically inspired features for non-speech audio Categoryification", Multimedia, IEEE Transactions, 14(6), pp.1684~1689 (2012)
- Ittichaichareon, Chadawan, "Speech recognition using MFCC", International Conference on Computer Graphics, Simulation and Modeling (ICGSM'2012) (2012)
- Kopparapu, Sunil Kumar, "Choice of Mel filter bankin computing MFCC of are sampled speech", Information Sciences Signal Processing and their Applications (ISSPA), 2010 10th International Conference on. IEEE (2010)
- Sim, S. Y., "Fall detection algorithm for the elderly using acceleration sensors on the shoes", Engineering in Medicine and Biology Society, EMBC, 2011 Annual International Conference of the IEEE, pp.4935~4938 (2011), IEEE
- "Samsung Galaxy S4"[URL]
- "Inside the Samsung Galaxy S4" (2013)[URL]