基于改进色度模型的非接触式心率检测与估计

发布时间：2018-05-05 06:38

本文选题：心率估计 + 非接触式测量　；参考：《中国生物医学工程学报》2017年06期

【摘要】：成像式光电容积描记技术能够实现低成本、非接触的远程生理参数检测与估计。其中,基于色度模型的视频图像心率估计算法计算简单、效率高,且具有较强的抗运动干扰能力,但并未考虑实验中光照变化产生的信号干扰。针对该问题,提出一种基于改进色度模型的非接触心率估计算法,通过定义感兴趣(ROI)区域和背景区域的反射光模型,利用两种区域颜色通道信息的比值来消除光照干扰,从而得到消除光照干扰的色度模型,并进行脉搏波的提取与心率估计。对10名在校大学生采集到的人脸图像进行分析,发现该算法估计心率的相对误差为3.52%±2.53%,原始色度模型算法的相对误差为7.23%±2.82%,两者对比有显著性差异(P0.01)。该算法能有效抑制光照变化所引起的干扰,使所提取脉搏波信号的信噪比平均提高1.52 d B,这对更精确地进行非接触式心率估计与脉搏波信号提取具有重要意义。
[Abstract]:Imaging photoelectric plethysmography can realize low cost, contactless remote physiological parameter detection and estimation. The algorithm based on chrominance model is simple in calculation, high in efficiency and strong in resisting motion interference, but it does not take into account the signal interference caused by the illumination change in the experiment. To solve this problem, a non-contact heart rate estimation algorithm based on the improved chrominance model is proposed. By defining the reflected light models in the region of interest (ROI) and the background region, the ratio of color channel information of the two regions is used to eliminate the illumination interference. A chrominance model for eliminating illumination interference was obtained, and pulse wave extraction and heart rate estimation were carried out. Based on the analysis of face images collected by 10 college students, it is found that the relative error of the estimated heart rate is 3.52% 卤2.53, and the relative error of the original chrominance model algorithm is 7.23% 卤2.82. There is a significant difference between the two methods. The algorithm can effectively suppress the interference caused by light change and increase the SNR of the extracted pulse wave signal by an average of 1.52 dB, which is of great significance for more accurate non-contact heart rate estimation and pulse wave signal extraction.
【作者单位】：大连理工大学电子信息与电气工程学部;
【基金】：国家自然科学基金(61671105,81241059) 国家科技支撑计划项目(2012BAJ18B06-04)
【分类号】：R318;TP391.41

【相似文献】