• 인간식물환경학회지
• /
• 제22권2호
• /
• pp.219-229
• /
• 2019

This study was carried out to verify the accuracy of fitness tracking devices in monitoring heart rate and energy consumption and to contribute to the development of a forest exercise program that can recommend the intensity and amount of forest exercises based on personal health-related data and provide monitoring and feedback on forest exercises. Among several commercially available wearable devices, Fitbit was selected for the research, as it provides Open API and data collected by Fitbit can be utilized by third parties to develop programs. Fitbit provides users with various information collected during forest exercises including exercise time and distance, heart rate, energy consumption, as well as the altitude and slope of forests collected by GPS. However, in order to verify the usability of the heart rate and energy consumption data collected by Fitbit in forest, the accuracy of heart rate and energy consumption were verified by comparing the data collected by Fitbit and reference. In this study, 13 middle-aged women were participated, and it was found that the heart rate measured by Fitbit showed a very low error rate and high correlation with that measured by the reference. The energy consumption measured by Fitbit was not significantly different from that measured in the reference, but the error rate was slightly higher. However, there was high correlation between the results measured by Fibit and the reference, therefore, it can be concluded that Fitbit can be utilized in developing actual forest exercise programs.

• 대한임베디드공학회논문지
• /
• 제14권3호
• /
• pp.123-131
• /
• 2019

In recent years, a non-contact respiration and heart rates monitoring via IR-UWB radar has been paid much attention to in various applications - patient monitoring, occupancy detection, survivor exploring in disaster area, etc. In this paper, we address a novel approach of real time heart rate estimation using IR-UWB radar. We apply sine fitting and peak detection method for estimating respiration rate and heart rate, respectively. We also deploy two techniques to mitigate the error caused by wrong estimation of respiration rate: a moving average filter and finding the frequency of the highest occurrence. Experimental results show that the algorithm can estimate heart rate in real time when respiration rate is presumed to be estimated accurately.

• 한국정보통신학회논문지
• /
• 제25권7호
• /
• pp.971-976
• /
• 2021

본 논문은 심박수 측정을 위한 안면 얼굴 영상 데이터 수집 시스템을 설계한다. 본 논문의 설계 내용은 웹 카메라를 이용하여 사용자 얼굴 영상 정보들을 수집하고, 수집된 사용자 얼굴 영상 정보들을 이용하여 심박수를 측정하는 기능이다. 웹 카메라를 이용한 비접촉식 심박수 측정으로 인하여 오차가 발생할 가능성이 있다. 따라서 본 논문에서는 심박수 측정시 얼굴 영상 데이터 분류를 통해서 오차가 발생한 경우와 정상적인 경우를 구별하여 심박수 프로그램 오차 수정에 이용할 수 있도록 하고자 한다. 오차가 발생된 경우의 자료를 이용하여 오차를 줄이기 위한 목적으로 사용할 수 있도록 한다. 본 논문에서 제안하고 설계한 내용에 대해서 실험을 수행하였다. 실험 결과 본 논문에서 설계한 내용이 정상적으로 동작됨을 확인할 수 있었다.

• 전기학회논문지
• /
• 제56권9호
• /
• pp.1688-1693
• /
• 2007

The aim of this paper is to develop and verify the system which can detect heart rate during exercise by using conductive fabric electrode and transportable measurement module. The experiment was performed under 4 conditions(resting, walking, jogging, running) and 18 subjects data are used. By using the ECG measurement system used in cardiac stress testing as reference value in order to verify the accuracy of the developed system, the relative error and correlation coefficient was calculated for each subject at every 3 seconds. The results have shown that the high correlation between the developed system and the reference system for detecting heart rate during exercise. Relative error and correlation coefficient are 2.27% and 0.9877, respectively. 7 subjects data are omitted in these calculations because of severe noises. Therefore, it is expected that this system could be used as a health monitoring system in ubiquitous environment in the future.

• 대한임베디드공학회논문지
• /
• 제18권2호
• /
• pp.51-58
• /
• 2023

This paper proposes a deep learning method for estimating the heart rate from facial videos. Our proposed method estimates remote photoplethysmography (rPPG) signals to predict the heart rate. Although there have been proposed several methods for estimating rPPG signals, most previous methods can not be utilized in low-power single board computers due to their computational complexity. To address this problem, we construct a lightweight student model and employ a knowledge distillation technique to reduce the performance degradation of a deeper network model. The teacher model consists of 795k parameters, whereas the student model only contains 24k parameters, and therefore, the inference time was reduced with the factor of 10. By distilling the knowledge of the intermediate feature maps of the teacher model, we improved the accuracy of the student model for estimating the heart rate. Experiments were conducted on the UBFC-rPPG dataset to demonstrate the effectiveness of the proposed method. Moreover, we collected our own dataset to verify the accuracy and processing time of the proposed method on a real-world dataset. Experimental results on a NVIDIA Jetson Nano board demonstrate that our proposed method can infer the heart rate in real time with the mean absolute error of 2.5183 bpm.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 2001년도 추계학술대회 논문집
• /
• pp.91-95
• /
• 2001

In this study, we estimated the behavior of the diversity of physiological responses under varied physical environmental factors by measuring variations of heart rate variability (HRV), an index of activity of cardiac autonomic control. Seven healthy young male adults consented and participated in the study. The environmental conditions consisted of thermal, lighting, and acoustic conditions. Two components of HRV were measured. one was the low frequency (LF) component of HRV, which provided a quantitative index of the sympathetic and parasympathetic (vagal) activities controlling the heart rate (HR). The other component measured was the high frequency (HF) component, which provided an index of the vagal tone. The percent contribution of physical environmental factors to the variations in HRV indices were calculated by ANOVA. The contribution of physical environmental factors to the variations in HR was higher than the contribution of HF and LF. However, the contribution of these factors was lower than the contribution related with individual difference in all indices. This result showed that the individual diversity of physiological responses is not a negligible quantity.

• 전자공학회논문지
• /
• 제52권8호
• /
• pp.18-26
• /
• 2015

올바른 운동량 조절을 위해선 운동중의 심박수 측정이 중요하다. 최근 스마트 디바이스가 활발하게 사용됨에 따라, 운동중의 실시간 심박수 측정에 대한 관심이 급격하게 증가하고 있다. 고강도 운동 중에는 동잡음으로 인하여 손목 밴드 유형의 광혈류 (PPG : photoplethysmography) 측정기 신호로부터 정확한 심박수를 추정하는 것이 매우 어렵다. 본 논문에서는 손목밴드 유형의 광혈류 측정기 신호로부터 정확한 심박수 추정을 위한 효율적인 알고리즘을 제안하였다. 12개의 데이터 세트에 대하여 제안하는 알고리즘을 적용한 결과, 1.38의 분당심박수(BPM) 평균 절대 오차를 기록하였고, 0.9922의 추정 심박수와 실제 심박수간의 Pearson 상관계수를 얻었다. 제안하는 알고리즘은 웨어러블 디바이스에 적합한 빠른 연산속도와 정확한 추정을 가능케 한다.

• 한국멀티미디어학회논문지
• /
• 제24권11호
• /
• pp.1481-1491
• /
• 2021

Since most biosignals rely on contact-based measurement, there is still a problem in that it is hard to provide convenience to users by applying them to daily life. In this paper, we present a mobile application for estimating heart rate based on a deep learning model. The proposed application measures heart rate by capturing real-time face images in a non-contact manner. We trained a three-dimensional convolutional neural network to predict photoplethysmography (PPG) from face images. The face images used for training were taken in various movements and situations. To evaluate the performance of the proposed system, we used a pulse oximeter to measure a ground truth PPG. As a result, the deviation of the calculated root means square error between the heart rate from remote PPG measured by the proposed system and the heart rate from the ground truth was about 1.14, showing no significant difference. Our findings suggest that heart rate measurement by mobile applications is accurate enough to help manage health during daily life.

• 한국멀티미디어학회논문지
• /
• 제20권2호
• /
• pp.234-243
• /
• 2017

With the development of smartphone technologies enable photoplethysmogram (PPG) acquisition by camera and heart rate (HR) measurement. This papers presents improved algorithm to extract HR from PPG signal recorded by smartphone camera and to develop real-time PPG signal processing Android application. 400 video samples recorded by Samsung smartphone camera are imported as input data for further processing and evaluating algorithm on MATLAB. An optimized algorithm is developed and tested on Android platform with different kind of Samsung smartphones. To assess algorithm's performance, medical device Beurer BC08 is used as reference. According to related works, accuracy parameters includes 90% number of samples that has relative errors less than 5%, Person correlation (r) more than 0.9, and standard estimated error (SEE) less than 5 beats-per-minutes (bpm).

• Korean Journal of Radiology
• /
• 제20권1호
• /
• pp.94-101
• /
• 2019

Objective: To investigate the efficacy of motion-correction algorithm (MCA) in improving coronary artery image quality and measurement accuracy using an anthropomorphic dynamic heart phantom and 256-detector row computed tomography (CT) scanner. Materials and Methods: An anthropomorphic dynamic heart phantom was scanned under a static condition and under heart rate (HR) simulation of 50-120 beats per minute (bpm), and the obtained images were reconstructed using conventional algorithm (CA) and MCA. We compared the subjective image quality of coronary arteries using a four-point scale (1, excellent; 2, good; 3, fair; 4, poor) and measurement accuracy using measurement errors of the minimal luminal diameter (MLD) and minimal luminal area (MLA). Results: Compared with CA, MCA significantly improved the subjective image quality at HRs of 110 bpm (1.3 ± 0.3 vs. 1.9 ± 0.8, p = 0.003) and 120 bpm (1.7 ± 0.7 vs. 2.3 ± 0.6, p = 0.006). The measurement error of MLD significantly decreased on using MCA at 110 bpm (11.7 ± 5.9% vs. 18.4 ± 9.4%, p = 0.013) and 120 bpm (10.0 ± 7.3% vs. 25.0 ± 16.5%, p = 0.013). The measurement error of the MLA was also reduced using MCA at 110 bpm (19.2 ± 28.1% vs. 26.4 ± 21.6%, p = 0.028) and 120 bpm (17.9 ± 17.7% vs. 34.8 ± 19.6%, p = 0.018). Conclusion: Motion-correction algorithm can improve the coronary artery image quality and measurement accuracy at a high HR using an anthropomorphic dynamic heart phantom and 256-detector row CT scanner.