CN103136465A

Movatterモバイル変換

Info

Publication number: CN103136465A
Application number: CN201310070680XA
Authority: CN
Inventors: 郑刚; 李忠义
Original assignee: Tianjin University of Technology
Current assignee: Tianjin University of Technology
Priority date: 2013-03-06
Filing date: 2013-03-06
Publication date: 2013-06-05

Abstract

一种心电信号用于身份识别的方法。包括两部分，第一部分为注册，首先通过心电信号采集装置采集心电信号（30秒-60秒），通过蓝牙上传到计算机，对心电信号进行处理，包括R波检测，R点对齐，描绘颜色叠加图，确定边界，确定个人模版。第二部分为检测，首先通过心电信号采集装置采集心电信号（5秒），通过蓝牙上传到计算机，对心电信号进行处理，包括R波检测，心电图判断，输出判断结果，达到身份识别目的。本发明实现了基于心电图的身份识别，实现了心电信号的方便采集，实现了心电信号传输的无线化。

A method of using electrocardiographic signals for identification. It consists of two parts. The first part is registration. First, the ECG signal is collected by the ECG signal acquisition device (30 seconds to 60 seconds), uploaded to the computer through Bluetooth, and the ECG signal is processed, including R wave detection, R point alignment, Draw color overlays, define boundaries, define personal templates. The second part is the detection. First, collect the ECG signal (5 seconds) through the ECG signal acquisition device, upload it to the computer through Bluetooth, and process the ECG signal, including R wave detection, ECG judgment, and output judgment results to achieve identity recognition. Purpose. The invention realizes the identification based on the electrocardiogram, realizes the convenient collection of the electrocardiogram signal, and realizes the wireless transmission of the electrocardiogram signal.

Description

Electrocardiosignal is used for the method for identification

Technical field

The invention belongs to biological information recognition technology field, particularly a kind of based on Electrocardiographic personal identification method.

Background technology

In information security field, the discovery of new biological information feature will inevitably cause multidisciplinary, multi-field extensive concern, the fusion of it and other existing biological information, and extensive, the realization of using of many biological informations recognition system model, all required providing in the field for higher level of security the possibility that realizes.The ECG(electrocardiosignal) meet just the biological information of this feature.

Because ECG and heart have the correspondence of genius morbi, ECG improves constantly the reliability of its Diagnosing Cardiac disease in a century of its birth, become at present the important method of heart disease diagnosis.Over past ten years, the research of ECG is extended to biological information field, particularly biological information identification field.Has " work " property due to itself, make the password of the processing take ECG as " raw material ", possessed other biological identifying information (fingerprint, iris, palmmprint, people's face ...) not available advantage, be that ECG is at every moment changing, different people, same people's ECG signal not in the same time, different acquisition environment, people's different gestures (static, move, repose, stand ...) etc. in various degree variation all can be arranged under situation.The biological information that ECG is used for identification need to possess two key features:

1, the ECG of different people is discrepant, and this makes ECG become outstanding biological information identification material.

2, same people's ECG has very strong general character, and general character is enough to that same people is obtained under different conditions distinguishing ECG is identified as a people.

About the more existing achievements of the research of this two aspect, but be for the situation in 1 basically.Hoekema[1] the variation major part that proposed and proved the electrocardiosignal between different people comes from how much (geometrical) structures and electric physiology (physiological) information of heart.Simon[3] 1997 articles of delivering have also illustrated this point, and the feasibility of ECG as the biological information of distinguishing identity also has been described simultaneously.Biel[2] also ECG is proposed as the concrete grammar of identification in the same year, pointed out that cardiogram has " activity ", and can carry out non-active collection etc. and be different from the biological information (fingerprint, iris, palmmprint, people's face etc.) that other are used for identification.And concrete proposition utilize the shape information (time gap between key point, amplitude, slope etc.) of ECG as the information characteristics of ECG, the data message of pointing out simultaneously singly to lead can be completed identification.So far many pieces of scientific papers have been delivered in international publication, international conference from 1999.M.Ogawa[4 in 1997] distinguish two individualities by small echo and NN analysing ECG, and special article the earliest is from the Biel[2 of Sweden orebro university in 1999], this piece article was repeatedly quoted afterwards, she has proposed 30 alternative parameters that come from reference point, recycling PCA carries out dimensionality reduction, and experimental data derives from 20 individualities of testing oneself, and the identification Average Accuracy is 97.14%, simultaneously, her data that also propose singly to lead namely can be used for the ECG identification.Calendar year 2001, M. Kyoso [5] utilizes mahalanobis distance to distinguish 9 individualities.The Shen[6 of University of Washington in 2002] identification based on 7 unique points has also been proposed, by 20 individual data items that singly lead, adopt the classification of pattern-recognition and artificial neural network, reach 100% accuracy rate.The Kim[7 of Korea S Qingzhou university in 2005] by extracting the interval feature of ECG benchmark, at 10 individualities, 30 seconds I lead and analyze on the ECG data.From SAIC(Science Applications International Corporation) Israel[9] and Irvine[8] published an article on Pattern Recognition in 2005 and 2008, proposed based on 15 ECG reference points and automatically extracted their personal identification method.Another research team in this respect was from the Edward S. Rogers Sr. electronics of University of Toronto and the Dimitrios Hatzinas of Computer Engineering Dept., and this team has delivered many pieces of articles and carried out the research that ECG is used for identification from 2003.Document [10] 2006 is delivered, inherited to a great extent Israel[9] method that proposes, a kind of mode identification method of the stratification that analytic features and appearance features are combined has been proposed in order further to improve the identification accuracy.The method has been utilized interval and the amplitude parameter of 21 reference points, after analytical characteristic, classifies, recycled the fusion of two category informations, that is: utilized LDA to interval and the amplitude parameter classification of ECG, recycling PCA carries out dimensionality reduction, classify with KNN at last, reach 100% discrimination.Utilized respectively 24 features based on the ECG reference point to carry out ECG identification research from research team of Ontario, Canada Polytechnics, and reach 100% discrimination, but the individual number of identification had been only 16 people in 2008.Same team has done summary analysis in 2009 to research in the past, thinks that ECG is feasible as identification.To sum up document has provided effective solution for the identification of ECG.

Present research concentrates on two kinds of methods

1, based on the system of selection of unique point, the feature in cardiogram is selected:

1) heart rate;

2) unique point amplitude (as: R, P, Q, S, T etc.);

3) unique point interval (RR, PR, QT, ST etc.);

4) continuous information (as HRV, heart rate variability rate etc.).

2, based on cardiac electrical vector information

For example the cycle of a heartbeat bynIndividual sampled point is described, and then forms one-dimensional vector with sampled point, by intelligent algorithm (as, artificial neural network, machine learning algorithm etc.) train, reach the purpose of identification.

Shortcoming based on the system of selection of unique point:

1, positioning feature point is inaccurate.

Even 2 positioning feature point are accurate, according to feature point extraction feature (distance and amplitude), these features can not represent this waveform fully, have a lot of useful informations to be lost.

Summary of the invention

The present invention seeks to solve the problem of extracting inaccurate and INFORMATION OF INCOMPLETE at ECG identification temporal signatures point, provide a kind of electrocardiosignal to be used for the new method of identification.The present invention adopts the color addition method of image to carry out the masterplate design based on Electrocardiographic personal identification, is used for identification.

Electrocardiosignal provided by the invention is used for the method for identification, comprises following two parts:

1st, registration part

1.1st,Use electrocardiogram signal acquisition device that registrant's electrocardiosignal is gathered;

1.2nd,By the bluetooth in harvester, the electrocardiosignal that collects is uploaded to computing machine;

1.3rd,By computing machine on the electrocardiosignal that transmits process, comprising:

A) electrocardiosignal of uploading being carried out the R point detects;

B) according to the R point, signal segmentation is become complete one by one heart cycle waveform;

1.4th,Set up the template library that is used for identification, be each registrant and set up passage, simultaneously the border of Acquisition channel;

2nd, test section

2.1st,Use electrocardiogram signal acquisition device that detected person's electrocardiosignal is gathered;

2.2nd,By the bluetooth in harvester, the electrocardiosignal that collects is uploaded to computing machine;

2.3rd,By computing machine on the electrocardiosignal that transmits process, comprising:

A) electrocardiosignal of uploading being carried out the R point detects;

2.4th,The detected person is detected identification, provide recognition result.

The method that the 1.4th described foundation of step is used for the template library of identification comprises:

I. define data set

The definition data set

:

(1-1)

pBe the quantity of this data centralization heart cycle waveform,

Be data set

In

The bar heart cycle waveform,

(1-2)

And

,tDimension for heart cycle waveform.

II. set up passage

With data set

In in all data-mapping to a two-dimensional coordinate system, horizontal ordinate is the time

And

, ordinate is to exist above-mentioned cardiac cycleTValue constantly

The concrete steps of setting up passage are as follows:

1, structure

Counter matrices E, whereinmBe data set

In maximal value and data set

In the difference of minimum value add one,As formula (1-3) to (1-6)

	(1-3)
			(1-4)
	(1-5)
			(1-6)

Statistics thejIn rowThe number of times that occurs, and deposit matrix inE, wherein, 1＜j＜t, 1＜i＜p

The number of times that repeatsFreqBe stored in counter matrices

In the position

(1-7)

2, RGB andFreqMapping relations:FreqLarger, namely multiplicity is more, and RGB is more shallow, the RGB color that the below's design is a kind of 8, and the change procedure of RGB is as shown in formula 1-8

(1-8)

RGB withFreqMapping formula (1-9) as follows, then ask according to the progressive formation of RGBr,g,bValue;

(1-9)

With data setIn point according to matrixEMark in coordinate system successively after obtaining rgb value;

III. Acquisition channel border:

Above

The passage that the red area that step obtains is to locateCThe borderWH,LThe up-and-down boundary that represents respectively passage,

(1-10)

The traversal counter matricesE, find withrgbCorresponding for redness

,

The point of representative is the point in profile, and judges up-and-down boundary by the row at this place; If

Be the point in the border, itFreqThere is matrix

In, if

,

Be the coboundary

In point, be expressed as

, wherein

, andIf

,

Be lower boundaryIn point, be expressed as

, wherein

, and

,

(1-11)

The 2.4th step comprised the concrete grammar that the detected person detects identification:

In the 2.3rd divided good heart cycle waveform of step, waveform of random choose is for detection of identification;

Adopt passage to identify:

If judge certain heart cycle waveform

Whether belong to certain passageC, just need judgementBWhether on the border of this passage

Scope in,

(1-12)

As shown in formula (1-13), ifBIn have a few all and existWIn,BBelong to passageCOtherwise,, do not belong to;

(1-13)

Output display as a result with identification.

Electrocardiogram signal acquisition device of the present invention comprises microprocessor, and the signal detection module that is connected with microprocessor respectively, LCD display module, bluetooth module and key-press module; Described signal detection module comprises that the pre-amplification circuit, the low pass filtered that connect successively involve amplifying circuit and rejector circuit, the output terminal of rejector circuit (being also the output terminal of signal detection module) is connected with the I/O port of microprocessor, and microprocessor is connected with LCD display module, bluetooth module and key-press module by the I/O port respectively.

Advantage of the present invention and good effect:

The present inventionRealized based on Electrocardiographic identification

Can avoid the fraud of identity due to Electrocardiographic activity, so this invention can be avoided the generation of this type of event.

Realized that the convenient of electrocardiosignal gathers

The present invention utilizes the electrocardiogram signal acquisition device of oneself's design to complete collecting work, greatly reduces cost and the difficulty of ecg signal acquiring.Be conducive to the popularization of equipment and method.

Realized the wireless penetration of electrocardiosignal transmission

The present invention utilizes Bluetooth technology to upload electrocardiogram (ECG) data, has guaranteed that electrocardiogram (ECG) data is not disturbed by outer signals, the simultaneously wireless convenience of using that improved.

Experimental result shows, can be used for identification based on the personal identification method of electrocardiosignal, and its rate of accuracy reached to 92.7% meets the requirement of identification.

Description of drawings

Fig. 1 is the treatment scheme of registration part in the present invention.

Fig. 2 is the treatment scheme of test section in the present invention.

Fig. 3 is the individual masterplate passage of setting up in registration part of the present invention.

Fig. 4 is the visualize figure of the waveform that can be identified in test section of the present invention.

Fig. 5 is the visualize figure of the waveform that is not identified in test section of the present invention.

Fig. 6 is electrocardiogram signal acquisition device block scheme of the present invention.

Fig. 7 is the circuit diagram of microprocessor portion.

Fig. 8 is the circuit diagram that low pass filtered involves amplification circuits.

Fig. 9 is 50HZ rejector circuit circuit diagram partly.

Figure 10 is the circuit diagram of LCD display module part.

Figure 11 is the circuit diagram of bluetooth module part.

The electrocardiosignal figure that Figure 12 collects for oneself's design electrocardiogram signal acquisition device.

Figure 13 is that 20 of electrocardiosignal are cut apart good heart cycle waveform array of figure.

Figure 14 is No. 100 constructed passages of people's electrocardiosignal.

Figure 15 is the border of the constructed passage of No. 100 people's electrocardiosignals.

Whether Figure 16 is for belonging to the audio-visual picture of No. 100 people's passages detected person's cardiac cycle.

Embodiment

Embodiment 1, electrocardiogram signal acquisition device

Figure 6 shows that electrocardiogram signal acquisition device, this device comprises microprocessor, and the signal detection module that is connected with microprocessor respectively, LCD display module, bluetooth module and key-press module, the output terminal of signal detection module is connected to the I/O port of microprocessor, and microprocessor is connected with key-press module with LCD display module, bluetooth module (seeing Figure 11) by the I/O port respectively.

Described signal detection module comprises that the pre-amplification circuit, the low pass filtered that connect successively involve amplifying circuit (see figure 8) and rejector circuit's (see figure 9), and the output terminal of rejector circuit is connected with the I/O port of microprocessor.

Described rejector circuit is 50HZ rejector circuit.

Described button can comprise up and down key, mode key, acknowledgement key, cancel key and power knob.

What described microprocessor adopted is AT91SAM7S64 chip (see figure 7), and what described LCD display module adopted is HTM12864 module (see figure 10).

Embodiment 2

The process that the electrocardiosignal that the present invention proposes is used for the method for identification comprises two parts, as shown in Figure 1, 2.

First is registration, at first gathers electrocardiosignal (30 seconds-60 seconds) byembodiment 1 oneself's design electrocardiogram signal acquisition device, uploads to computing machine by bluetooth, electrocardiosignal is processed, comprised that the R ripple detects, the alignment of R point, describe color addition figure, determine the border, determine individual masterplate.

Second portion is for detecting, and at first the electrocardiogram signal acquisition device byembodiment 1 oneself's design gathers electrocardiosignal (5 seconds), uploads to computing machine by bluetooth, electrocardiosignal is processed, comprised that the R ripple detects, the cardiogram judgement, the output judged result reaches the identification purpose.

Specific implementation process of the present invention is as follows:

1st, registration part

1.1st,Use oneself's design electrocardiogram signal acquisition device that registrant's electrocardiosignal is gathered, as Figure 12;

A) electrocardiosignal of uploading being carried out the R point detects;

B) according to the R point, signal segmentation is become complete one by one heart cycle waveform, cut apart good heart cycle waveform such as Figure 13;

The method of setting up the template library that is used for identification comprises:

I. define data set:

Describe with No. 100 artificial examples in Massachusetts Institute of Technology's arrhythmia cordis ecg database (MIT-BIH).

The data set of No. 100 people's electrocardiosignals of definition

, what p=1119 represented is the cardiac cycle quantity that this data centralization comprises, what t=180 represented is the holocentric dimension in moving cycle, so data set

Be defined as follows:

II. set up passage:

Data set with No. 100 people's electrocardiosignals

In in all data-mapping to a two-dimensional coordinate system, horizontal ordinate is

, ordinate is to exist cardiac cycleTValue constantly is

The concrete steps of setting up passage are as follows:

1. construct

Counter matrices E, whereinmBe data set

In maximal value and data set

In the difference of minimum value add one,

The number of times that repeatsFreqBe stored in count matrixIn the position.According to this statistical, well imagine, maximum numerals is 0 in matrix E, because matrix is larger, so can only simply present.

Finally, the result of count matrix E presents as follows:

RGB withFreqMapping relations:FreqLarger, namely multiplicity is more, and RGB is more shallow, and the RGB color that the below's design is a kind of 8 is shown in the following formula of the change procedure of RGB

RGB withFreqThe mapping formula as follows, then ask according to the progressive formation of RGBr,g,bValue;

For No. 100 people, the counter matrices that it is correspondingEIn maximal value be 1000, namely

And data set

In certain the point

, thisrgbFor

, namely

With data set

In point according to matrixEMark in coordinate system successively after obtaining rgb value, Figure 14 is exactly the passage that No. 100 people's electrocardiosignal is set up.

III. Acquisition channel border:

Red area in Figure 14 is the border W of No. 100 constructed channel C of people.H, L represent respectively the up-and-down boundary of passage, according to the algorithm on Acquisition channel border, extract No. 100 people's channel boundaries as shown in figure 15.

2nd, test section

2.1st, makeDesigning electrocardiogram signal acquisition device with the oneself gathers detected person's electrocardiosignal;

2.3rd, byComputing machine on the electrocardiosignal that transmits process, comprising:

A) electrocardiosignal of uploading being carried out the R point detects;

The concrete grammar that the detected person is detected identification comprises:

In divided good heart cycle waveform, waveform of random choose is for detection of identification.

Adopt passage to identify:

If judge certain heart cycle waveform

The passage that whether belongs to No. 100 peopleC, just need judgementBWhether on the border of this passage

Scope in, according to above-mentioned recognizer,BDifferentiation visual result figure as shown in figure 16.

By identification, recognition result is cardiac cycleBThe passage that belongs to No. 100 peopleCAccording to passageCInformation so that identify corresponding people.

List of references

[1] R.Hoekema，G.J.H.Uijen，A.van Oosterom．Geometrical aspect of the interindividual variability of multilead ECG recordings[J]．IEEE Transactions on Biomedical Engineering，2001，48（5）：551~559．

[2] L.Biel，O.Patterson，L. Philipson，et al．ECG analysis: a new approach in human identification[J]．IEEE Transactions on Instrumentation and Measurement,，2001， 50（3）：808~812．

[3] B.P.Simon，C.Eswaran．An EGG classifier designed using modified decision based neural network[J]．Computer and Biomedical Research，1997，30（4）：257~272．

[4] M.Ogawa，T.Tamura，et al．Fully Automated Biosignal Acquisition System for Home Health Monitoring[C]．In：Proceeding of the 19th IEEE EMBS International Conference.Chicago，USA，1997．

[5] M.Kyoso，A.Uchiyama．Developrnent of an ECG Identification System[C]．In：Proceedings of the 23th Annual EMBS International Conference，Istanbul，Turekey，2001，4：3721~3723．

[6] T.W.Shen，W.J.Tompkinsl，Y.H.Hu．ONE-LEAD ECG FOR IDENTITY VERIFICATION[C]．In：Proceedings of the 2nd Joint Engineering in Medicine and Biology，24th Annual Conference and the Annual Fall Meeting of the Biomedical Engineering Society（EMBS/BMES’02），Houston，October 2002，1：62~63．

[7] KS Kim，TH Yoon，JW Lee，et al．A Robust Human Identification by Normalized Time-Domain Features of Electrocardiogram[C]．In：Proceeding of the 2005 IEEE，Engineering in Medicine and Biology 27th Annual Conference．Santa Barbara：2005，2：1114~1117．

[8] John M. Irvine，Steven A. Israel，W. Todd Scruggs，William J. Worek．eigenPulse: Robust Human Identification from Cardiovascular Function，Pattern Recognition, Vol. 41, 2008, pp. 3427-3435．

[9] S.A.Israel，J.M.Irvine，A. Cheng，et al．ECG to Identify Individuals[J]．Pattern Recognition，2005，38（1）：133~142．

[10] Yongjin Wang，Konstantinos N.Plataniotis，Dimitrios Hatzinakos．INTEGRATING ANALYTIC AND APPEARANCE ATTRIBUTES FOR HUMAN IDENTIFICATION FROM ECG SIGNALS[C]．In：Proceedings of Biometrics Symposiums，Baltimore，USA，2006：1~6．

Claims

Translated fromChinese

1.一种心电信号用于身份识别的方法，其特征在于该方法包括：1. A method for electrocardiographic signal identification, characterized in that the method comprises:

第1、注册部分Part 1. Registration

第1.1、使用心电信号采集装置对注册者心电信号进行采集；1.1. Use the ECG signal acquisition device to collect the registrant's ECG signal;

第1.2、通过采集装置中的蓝牙，将采集到的心电信号上传到计算机；1.2. Upload the collected ECG signal to the computer through the Bluetooth in the collection device;

第1.3、由计算机对上传来的心电信号进行处理，包括：1.3. The computer processes the uploaded ECG signal, including:

A）对上传的心电信号进行R点检测；A) R point detection is performed on the uploaded ECG signal;

B）根据R点，将信号分割成一个个完整的心动周期波形；B) According to the R point, the signal is divided into complete cardiac cycle waveforms;

第1.4、建立用于身份识别的模版库，即为每个注册者建立通道，同时获取通道的边界；1.4. Establish a template library for identification, that is, establish a channel for each registrant, and obtain the boundary of the channel at the same time;

第2、检测部分2. Detection part

第2.1、使用心电信号采集装置对被检测者的心电信号进行采集；2.1, use the ECG signal acquisition device to collect the ECG signal of the detected person;

第2.2、通过采集装置中的蓝牙，将采集到的心电信号上传到计算机；2.2. Upload the collected ECG signal to the computer through the Bluetooth in the collection device;

第2.3、由计算机对上传来的心电信号进行处理，包括：2.3. The computer processes the ECG signal uploaded, including:

第2.4、对被检测者进行检测识别，给出识别结果。Step 2.4. Perform detection and recognition on the detected person, and give the recognition result.

2.根据权利要求1所述的方法，其特征在于第1.4步所述的建立用于身份识别的模版库的方法包括：2. The method according to claim 1, characterized in that the method for establishing a template library for identification in step 1.4 includes:

I.定义数据集I. Defining Datasets

定义数据集A*：Define the dataset A*:

{A A}^{* *} = = (\begin{matrix} {A A}^{11} \\ {A A}^{22} \\ \cdot \cdot \\ \cdot &Center Dot; \\ \cdot &Center Dot; \\ {A A}^{P P} \end{matrix}) - - - - - - ((11 - - 11))

p为该数据集中心动周期波形的数量，Aⁱ为数据集A*中的第i条心动周期波形数据，p is the number of cardiac cycle waveforms in the data set, Aⁱ is the ith cardiac cycle waveform data in the data set A*,

{A A}^{i i} = = (({a a}_{11}^{i i},, {a a}_{22}^{i i},, . . . . . . . . . . . .,, {a a}_{t t}^{i i})) - - - - - - ((11 - - 22))

且1≤i≤p，t为心动周期波形的维度；And 1≤i≤p, t is the dimension of cardiac cycle waveform;

II.建立通道II. Establish a channel

将数据集A*中所有的数据映射到一个二维坐标系中，横坐标为时间T=j且1≤j≤t，纵坐标为上述心动周期在T时刻的取值为

Map all the data in the data set A* to a two-dimensional coordinate system, the abscissa is time T=j and 1≤j≤t, and the ordinate is the value of the above cardiac cycle at time T

建立通道的具体步骤如下：The specific steps to establish a channel are as follows:

1.构造m*n的计数器矩阵E，其中m为数据集A*中的最大值与数据集A*中的最小值的差值加一，n=t；如公式（1-3）至（1-6）1. Construct an m*n counter matrix E, where m is the difference between the maximum value in the data set A* and the minimum value in the data set A* plus one, n=t; such as formula (1-3) to ( 1-6)

{a a}_{min min} = = min min (\begin{matrix} {a a}_{11}^{11} & {a a}_{22}^{11} & \cdot &Center Dot; \cdot &Center Dot; \cdot &Center Dot; & {a a}_{t t}^{11} \\ {a a}_{11}^{22} & {a a}_{22}^{22} & \cdot &Center Dot; \cdot &Center Dot; \cdot &Center Dot; & {a a}_{t t}^{22} \\ \cdot &Center Dot; & \cdot &Center Dot; & \cdot &Center Dot; & \cdot &Center Dot; \\ {a a}_{11}^{p p} & {a a}_{22}^{p p} & \cdot &Center Dot; \cdot &Center Dot; \cdot &Center Dot; & {a a}_{t t}^{p p} \end{matrix}) - - - - - - ((11 - - 33))

{a a}_{max max} = = max max (\begin{matrix} {a a}_{11}^{11} & {a a}_{22}^{11} & \cdot &Center Dot; \cdot &Center Dot; \cdot &Center Dot; & {a a}_{t t}^{11} \\ {a a}_{11}^{22} & {a a}_{22}^{22} & \cdot &Center Dot; \cdot \cdot \cdot &Center Dot; & {a a}_{t t}^{22} \\ \cdot \cdot & \cdot \cdot & \cdot \cdot & \cdot &Center Dot; \\ {a a}_{11}^{p p} & {a a}_{22}^{p p} & \cdot &Center Dot; \cdot \cdot \cdot \cdot & {a a}_{t t}^{p p} \end{matrix}) - - - - - - ((11 - - 44))

m=a_max-a_min+1 (1-5)m=a_max -a_min +1 (1-5)

n=t (1-6)n=t (1-6)

统计第j列中

出现的次数，并存入矩阵E，其中，1<j<t，1<i<p；

重复出现的次数freq存储在计数器矩阵

位置中Statistics in column j

The number of occurrences, and stored in the matrix E, where, 1<j<t, 1<i<p;

The number of repeated occurrences of freq is stored in the counter matrix

in position

freq freq__{a a}^{i i}_{j j} = = [[{a a}_{max max} - - {a a}^{i i}_{j j},, j j]] - - - - - - ((11 - - 77))

2.RGB与freq的映射关系：freq越大，即重复次数越多，RGB越浅，下面设计一种8位的RGB颜色，RGB的变化过程如公式1-8所示2. The mapping relationship between RGB and freq: the larger the freq, the more repetitions, and the lighter the RGB, an 8-bit RGB color is designed below, and the change process of RGB is shown in formula 1-8

(0,0,0)→(0,0,255)→(0,255,255)→(255,255,255) (1-8)(0,0,0)→(0,0,255)→(0,255,255)→(255,255,255) (1-8)

RGB与freq的映射公式（1-9）如下，再根据RGB的渐变过程求r，g，b的值；The mapping formula (1-9) between RGB and freq is as follows, and then calculate the values of r, g, and b according to the gradient process of RGB;

\frac{f f \times \times ((33 \times \times 22^{88}))}{max max ((E E.))} &RightArrow; &Right Arrow; ((r r,, g g,, b b)) - - - - - - ((11 - - 99))

将数据集A*中的点根据矩阵E获取RGB值后依次在坐标系中标出；Obtain the RGB values of the points in the data set A* according to the matrix E and mark them in the coordinate system in turn;

III.获取通道边界：III. Get the channel boundary:

以上II步得到的红色区域是要找到的通道C的边界W；H、L分别表示通道的上下边界，The red area obtained in the above II step is the boundary W of the channel C to be found; H and L represent the upper and lower boundaries of the channel respectively,

W=[L,H] (1-10)W=[L,H] (1-10)

遍历计数器矩阵E，找到与rgb为红色的相对应的E[i,j]，E[i,j]代表的点即为边界中的点，并通过该点所在的列来判断上下边界；若

为边界中的点，它的freq存在矩阵E[i,j]中，若E[i-1,j]<E[i,j]<E[i+1,j]，则为上边界H中的点，表示为h_i，其中h_i∈H，且H={h₁,h₂,...,h_n}；若E[i-1,j]>E[i,j]>E[i+1,j]，则

为下边界L中的点，表示为l_i，其中l_i∈L，且L={l₁,l₂,...,l_n}，Traverse the counter matrix E, find the E[i,j] corresponding to the red rgb, the point represented by E[i,j] is the point in the boundary, and judge the upper and lower boundaries by the column where the point is located; if

is a point in the boundary, its freq is stored in the matrix E[i,j], if E[i-1,j]<E[i,j]<E[i+1,j], then is a point in the upper boundary H, expressed as h_i , where h_i ∈ H, and H={h₁ ,h₂ ,...,h_n }; if E[i-1,j]>E[i ,j]>E[i+1,j], then

is a point in the lower boundary L, expressed as l_i , where l_i ∈ L, and L={l₁ ,l₂ ,...,l_n },

{a a}^{i i}_{j j} &RightArrow; &Right Arrow; \{\begin{matrix} {h h}_{i i} &Element; &Element; H h,, H h = = (({h h}_{11},, {h h}_{22},, \cdot &Center Dot; \cdot &Center Dot; \cdot &Center Dot; {,, h h}_{n no})) \\ whenE whenE [[i i - - 11,, j j]] < < E E. [[i i,, j j]] < < E E. [[i i + + 11,, j j]] \\ {l l}_{i i} &Element; &Element; L L,, L L = = (({l l}_{11},, {l l}_{22},, \cdot &Center Dot; \cdot \cdot \cdot \cdot,, {l l}_{n no})) \\ whenE whenE [[i i - - 11,, j j]] > > E E. [[i i,, j j]] > > E E. [[i i + + 11,, j j]] \end{matrix} - - - - - - ((11 - - 1111))

。.

3.根据权利要求1所述的方法，其特征在于第2.4步对被检测者进行检测识别的具体方法包括：3. The method according to claim 1, characterized in that the specific method for detecting and identifying the subject in step 2.4 comprises:

I.在第2.3步被分割好的心动周期波形中，随机挑选一个波形用于检测识别；I. In the segmented cardiac cycle waveform in step 2.3, randomly select a waveform for detection and identification;

II.采用通道进行识别：II. Using channels for identification:

若判断某个心动周期波形B=(b₁,b₂,......b_t)是否属于某通道C，就需要判断B是否在该通道的边界W=[L,H]的范围以内，To determine whether a certain cardiac cycle waveform B=(b₁ ,b₂ ,...b_t ) belongs to a certain channel C, it is necessary to determine whether B is within the range of W=[L,H] at the boundary of the channel within,

W=[L,H]W=[L,H]

L=(l₁,l₂,…,l_t) (1-12)L=(l₁ ,l₂ ,…,l_t ) (1-12)

H=(h₁,h₂,…,h_t)H=(h₁ ,h₂ ,…,h_t )

如公式（1-13）所示，若B中的所有点均在W内，则B属于通道C，反之，则不属于；As shown in formula (1-13), if all points in B are within W, then B belongs to channel C, otherwise, it does not;

W W ((B B)) \{\begin{matrix} True True \\ when when {∩ ∩}_{i i = = 11}^{t t} {l l}_{i i} \leq \leq {b b}_{i i} \leq \leq {h h}_{i i},, ((B B &Element; &Element; W W)) \\ False False \\ when when {∩ ∩}_{i i = = 11}^{t t} {l l}_{i i} \leq \leq {b b}_{i i} \leq \leq {h h}_{i i},, ((B B &NotElement; &NotElement; W W)) \end{matrix} - - - - - - ((11 - - 1313))

III.将识别的结果输出显示。III. Output and display the recognition result.

4.根据权利要求1所述的方法，其特征在于所述的心电信号采集装置包括微处理器，以及分别与微处理器连接的信号检测模块、LCD显示模块、蓝牙模块和按键模块；所述的信号检测模块包括依次连接的电极、前置放大电路、低通滤波及放大电路和带阻滤波电路，带阻滤波电路的输出端与微处理器的I/O端口相连接，微处理器分别通过I/O端口与LCD显示模块、蓝牙模块和按键模块相连接。4. method according to claim 1, it is characterized in that described electrocardiogram acquisition device comprises microprocessor, and the signal detection module that is connected with microprocessor respectively, LCD display module, bluetooth module and button module; The signal detection module described includes successively connected electrodes, a preamplifier circuit, a low-pass filter and an amplifying circuit, and a band-stop filter circuit. The output end of the band-stop filter circuit is connected to the I/O port of the microprocessor, and the microprocessor The LCD display module, the bluetooth module and the button module are respectively connected through the I/O port.