http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
PPG 신호를 활용한 고 정확도 제 1심음 및 제 2심음 자동 식별 알고리즘 개발
이수민(S. M. Lee),김부성(B. S. Kim),정동화(D. H. Jung),임홍준(H, J. Lim),박희준(H. J. Park),웨이췬(Qun Wei) 한국재활복지공학회 2021 한국재활복지공학회 학술대회논문집 Vol.2021 No.4
Heart disease is reported as a leading reason that caused human death around world. For the early prevention, using stethoscope to detect abnormalities of the heart sound is one of the most effective method so far. However, because of the noise that inside and outside of the tube of the stethoscope, and sujective judgement devided by the docter will cause the misdiagnosis. To improve the accuracy of the auscultation, in this study, a high-accuracy S1 and S2 heart sound automatic identification algorithm using PPG signal was developed. Based on the 3rd Shannon Energy algorithm, the parameters of the start and end points of S1 and S2 of the hear sound were found. Also, the P-peak value of the PPG signal were extracted and used to find the relationship with the S2 peak. The developed algorithm was conducted a performance test experiment and the rate of identification was higher than the exist algorithm.
신진과학자 세미나: 직물의 촉감을 평가할 때 일어나는 손동작의 분석
이수민 ( S M Lee ),( M Kamijo ),( T Nishimatsu ),( Y Shimizu ),권영하 ( Y R Kwon ) 한국감성과학회 2004 추계학술대회 Vol.2004 No.-
Human uses the sight and the tactilesensations to evaluate the hand of textile fabrics. Especially the tactile sensation is the important factor to decide the hand of a textile fabric. When human evaluates the hand, the physical and physiological phenomena are occurred by the finger motion(the applied force, the speed of finger movement, the touching time, the contact area and the distance of finger movement)and they use a special finger motion according to the purpose of their evaluation. Such sensory evaluation includes resilience, roughness, flexibility and softness. We observed and compared the finger motions of both high sensitive users and low sensitive users. We used a glove type measurement system with pressure sensors and accelerator sensor to investigate the characteristics of finger motion while evaluating a hand of fabric. The sensors provided data on the applied force and direction of finger motion as its patterns. This allowed us to identify and analyze the differences between high sensitive users and low sensitive users. The results show that the finger motion varied according to the user``s ability to discriminate, with high sensitive user``s finger motion being better suited to sensory evaluation.
Defect Detection of Laser Drilling Micro Holes Using Photodiode Sensor and Autoencoders
S. M. Lee(이수민),M. K. Jeong(정민기),D. W. Yeo(여대원),Y. K. Lee(이용관) Korean Society for Precision Engineering 2021 한국정밀공학회 학술발표대회 논문집 Vol.2021 No.11월
Laser micro drilling is a manufacturing method commonly used to drill holes of microscopic diameters into metals. The size and quantity of holes drilled pose a challenge in accurate, thorough, and time-efficient quality control. We observed an abnormal amount of visible light scattered when the laser failed to drill through the material and created a defective hole. Using this observation, we hypothesized there would be a difference in the pattern of light around the work area during successful and defective drilling. In this paper we propose the usage of a photodiode sensor and autoencoders for in-situ progress monitoring and defect detection of laser micro drilling. We generated simulation data of the expected power over time prior to the data collection and trained CNN and LSTM autoencoders as a proof of concept. The photodiode sensor was installed at an angle to the 1064 nm Nd:YAG Laser to collect scattered, reflected and emitted light. Data collected showed high similarity to the general pattern of simulated data.