A PPG Signal De-noising Method Based on The DTCWT and The Morphological
Filtering
Tong Bai
1
, Dan Li, Huiqian Wang*
1
,Yu Pang
1
, Guoquan Li
1
,Jinzhao Lin
1
, Qianneng Zhou
1
, Gwanggil Jeon
2
1: Chongqing University of Posts and Telecommunications, Chongqing, China
2: Department of Embedded Systems Engineering, College of Information and Technology, Incheon National University,
119 Academy-ro, Yeonsu-gu, Incheon 22012, Republic of Korea
*Corresponding author: wanghq@cqupt.edu.cn
Abstract—The pulse wave signal contains a variety of noises
and has strong nonlinear and non-stationary. According to
the previous wavelet transformation method, this paper
proposes a PPG signal de-noising algorithm based on dual-
tree complex wavelet transform (DTCWT) and
morphological filtering. With the advantage of simple
construction, clear mathematical implications and low
computational complexity, this method overcomes the shift
sensitive and the frequency aliasing in the discrete wavelet
transform. The simulation results show that this algorithm
could remove the power line interference and EMG
interference, and the quantitative index of SNR and mean
square error is superior to the traditional threshold de-
noising algorithm. Therefore, the DTCWT and
morphological filtering de-noising algorithm would obtain a
clear pulse wave signal.
Keywords-pulse wave signal; DTCWT; morphological
filtering
I. INTRODUCTION
Photoplethysmography
[1]
(PPG) is used to optically
obtain a volumetric measurement of an organ. Many
clinical diseases especially heart disease may cause pulse
changes, thus, the pulse wave signal of human body
contains abundant physiology and pathology information.
The changes of circulatory system cause different shapes
of the pulse wave, therefore, the changes of pulse wave
shapes could be used for diagnosing the aortic valve, the
rhythm of heart and the elasticity of artery. The frequency
range of pulse signal of a healthy normal people
[2]
is 0 to
20Hz and almost 99 percent of the energy is distributed on
0 to 10Hz. Affected by the instruments in the extracting
process, pules wave signal tend to be multiple noise
interference. The low-frequency interference caused by the
human breathing, less than 1Hz, will generate the drift
phenomenon. Therefore, the extracted pulse wave signal is
not in a horizontal line. The high-frequency interference,
power frequency interference and EMG interference, cause
the PPG signal exists thorns and slight oscillation. These
interferences generate significantly impeding the judgment
of heart function changes. Therefore, the pulse wave signal
is supposed to extract efficiently from the noise to obtain
the accurate and pure PPG signal.
In recent years, the researchers propose a variety of
methods for removing PPG signal noise. The most
common method, Adapting filtering, needs the additional
hardware to acquire the reference signal and uses the
synthesis techniques for reference signal such as Fourier
transform
[3]
, wavelet transform
[4][5]
, support vector
machine(SVM) decomposition
[6]
and independent
component analysis(ICA)
[7][8]
to eliminate the noise. Based
on the resolution characteristics, wavelet transform could
be used to the PPG signal de-noising. However, in the
traditional wavelet transform, the tiny drift of the input
signal causes a significant change of wavelet coefficient,
the missing of information. Thus, the false results generate
a strong effect on the experiment.
II. D
E-NOISING METHED BASED ON THE DTCWT AND
MORPHPLOGICAL FILTERING
TYPE STYLE AND
FONTS
[9]
A. Morphological filtering
[10][11]
Morphological filtering is a kind of nonlinear filter
technique from Mathematical morphology (MM). MM is
most commonly applied to digital images, but it can be
employed as well on graphs, signal processing and some
other fields. The basic idea in binary morphology is to
probe an image with a simple, pre-defined shape, drawing
conclusions on how this shape fits or misses the shapes in
the image. This simple "probe" is called the structuring
element, and is itself a binary image.
The MM includes four basic operators - erosion,
dilation, opening and closing. Assume the signal
sequence
1}-N,{0,1,=Ff "
, the structuring element
1}-M,{0,1,=Kk "
. The dilation of A by the
structuring element B is defined by:
^`
1)N,M,1,M(n
,k(m)-m)-f(nmaxk)(n)(f
1M,0,m
"
"
The erosion of A by the structuring element B is
defined by:
^`
M)NM0,(n
,k(m)-m)f(nmink)(n)(f
1M,0,m
"
"
Ɨ
The opening of A by B is obtained by the erosion
of A by B, followed by dilation of the resulting image
by B:
k)(n)k(fk)(n)(f ƗD
The closing of A by B is obtained by the dilation
of A by B, followed by dilation of the resulting image
by B:
k)(n)k(fk)(n)(f Ɨ
Morphological filtering correction the positive pulse
and negative pulse of the signal based on the opening
operator and closing operator.
2016 12th International Conference on Signal-Image Technology & Internet-Based Systems
978-1-5090-5698-9/16 $31.00 © 2016 IEEE
DOI 10.1109/SITIS.2016.85
503
2016 12th International Conference on Signal-Image Technology & Internet-Based Systems
978-1-5090-5698-9/16 $31.00 © 2016 IEEE
DOI 10.1109/SITIS.2016.85
503
下载后可阅读完整内容，剩余3页未读，立即下载
weixin_38671048
- 粉丝: 4
- 资源: 870
我的内容管理
展开
最新资源
- Flex垃圾回收与内存管理：防止内存泄露
- Python编程规范与最佳实践
- EJB3入门：实战教程与核心概念详解
- Python指南v2.6简体中文版——入门教程
- ANSYS单元类型详解：从Link1到Link11
- 深度解析C语言特性与实践应用
- Gentoo Linux安装与使用全面指南
- 牛津词典txt版：信息技术领域的便捷电子书
- VC++基础教程：从入门到精通
- CTO与程序员职业规划：能力提升与路径指南
- Google开放手机联盟与Android开发教程
- 探索Android触屏界面开发：从入门到设计原则
- Ajax实战：从理论到实践
- 探索Android应用开发：从入门到精通
- LM317T稳压管详解：1.5A可调输出，过载保护
- C语言实现SOCKET文件传输简单教程
资源上传下载、课程学习等过程中有任何疑问或建议，欢迎提出宝贵意见哦~我们会及时处理！
点击此处反馈
