基于视频图像和信号加权的皮肤光学特性测量方法研究

发布时间：2018-03-25 05:16

本文选题：光学检测方法　切入点：非接触测量　出处：《天津大学》2016年博士论文

【摘要】：皮肤光学特性与人体生理状态密切相关,在无创非接触人体生理信息测量领域中起着非常重要的作用。本文利用视频成像和信号加权方法,分析和讨论皮肤视频图像序列信号的光强变化特性,进而快速精确地测量皮肤光学特性;基于这种非接触光学测量方法,在心率测量的基础上,研究人体皮肤的光学特异性,探索经络穴位处皮肤的光学特性。本课题研究的用普通摄像头测量皮肤表面光学特性的方法,不仅为人体生理参数快速测量提供一种简便安全的方法,也为经络穴位研究提供一种新的科学研究和技术手段。首先,在综述和分析心率和经络穴位测量技术的基础上,讨论了用皮肤光学特性测量心率和提取穴位信息的物理基础,以及可见光在测量人体生理参数和穴位信息的适用性。针对人体生理/穴位信息测量要求,分析了皮肤特性测量条件及其对测量结果的影响,并采取照明光源设计、图像匹配和光照补偿等方法减小这种影响,有效地提高了结果的测量精度。其次,通过对视频图像进行三通道分离、信号生成、盲源分离、时频变换等分析,快速准确地测量心率等人体关键生理参数,为进一步研究人体穴位奇异性奠定了测量基础。最后,系统地研究通过皮肤表面图像RGB信号的加权合成方法,提取人体穴位信息。针对从视频信号提取穴位信息的方法,采用信号加权方法减小了脉搏波对穴位信息的影响程度,提取了人体生理测量信号,并采用峰度系数参量进行分析,研究人体皮肤光学特异性,得到人体穴位信息;进而采用基于AR的双谱技术分析和探讨人体穴位区域的皮肤光学特性。结果表明,在可见光条件下,初步观察到了与穴位重合的皮肤表面光学特异性区域。本文的主要创新点如下:1.提出了一种基于皮肤表面光学图像确定人体穴位区域的方法,为定量地研究穴位开拓了新的途径;2.研究通过对皮肤表面图像的RGB信号分离和盲源分析,提高图像信号质量的方法,成功提取了心率信号;3.研究通过皮肤表面图像RGB信号的加权合成方法,初步观察到了与穴位重合的皮肤表面光学特异性区域。
[Abstract]:The optical properties of skin are closely related to the physiological state of human body and play a very important role in the field of non-invasive non-contact human physiological information measurement. This paper analyzes and discusses the characteristics of light intensity variation of skin video image sequence signals, and then measures the optical characteristics of skin quickly and accurately. Based on this non-contact optical measurement method, the optical specificity of human skin is studied on the basis of heart rate measurement. To explore the optical properties of skin at meridian acupoints. The method of measuring the optical properties of skin surface by ordinary camera in this subject not only provides a simple and safe method for the rapid measurement of physiological parameters of human body. It also provides a new scientific research and technical means for the study of meridian acupoints. Firstly, on the basis of summarizing and analyzing the measuring techniques of heart rate and meridian acupoints, the physical basis of measuring heart rate and extracting acupoint information by optical characteristics of skin is discussed. And the applicability of visible light in measuring physiological parameters of human body and acupoint information. According to the requirement of measuring physiological and acupoint information of human body, the measurement conditions of skin characteristics and their influence on the measurement results are analyzed, and the design of lighting source is adopted. The method of image matching and illumination compensation can reduce the influence and improve the accuracy of the result. Secondly, through the analysis of three-channel separation, signal generation, blind source separation, time-frequency transform, etc. The rapid and accurate measurement of key physiological parameters of human body, such as heart rate, lays a foundation for further study on the singularity of human acupoints. Finally, the weighted synthesis method of RGB signal from skin surface image is studied systematically. According to the method of extracting acupoint information from video signal, the influence of pulse wave on acupoint information is reduced by signal weighting method, and the physiological measurement signal of human body is extracted, and the kurtosis coefficient parameter is used to analyze it. By studying the optical specificity of human skin and obtaining the information of human acupoints, the optical properties of human skin in acupoints are analyzed and discussed by using the bispectral technique based on AR. The results show that, under visible light, The optical specific area of skin surface that overlaps with acupoint is observed preliminarily. The main innovation of this paper is as follows: 1. A method to determine the acupoint area of human body based on the optical image of skin surface is proposed. In order to quantitatively study acupoints, a new approach has been developed. 2. The method of improving image quality by separating RGB signals from skin surface images and analyzing blind sources is studied. Heart rate signal was extracted successfully. By using the weighted synthesis method of RGB signal of skin surface image, the optical specific area of skin surface was preliminarily observed to coincide with acupoint.
【学位授予单位】：天津大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TP391.41

【参考文献】