CN107203009A - A kind of mobile phone detection method extracted based on wavelet-based attribute vector - Google Patents
A kind of mobile phone detection method extracted based on wavelet-based attribute vector Download PDFInfo
- Publication number
- CN107203009A CN107203009A CN201710345505.5A CN201710345505A CN107203009A CN 107203009 A CN107203009 A CN 107203009A CN 201710345505 A CN201710345505 A CN 201710345505A CN 107203009 A CN107203009 A CN 107203009A
- Authority
- CN
- China
- Prior art keywords
- mobile phone
- signal
- wavelet
- detection
- vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V3/00—Electric or magnetic prospecting or detecting; Measuring magnetic field characteristics of the earth, e.g. declination, deviation
- G01V3/38—Processing data, e.g. for analysis, for interpretation, for correction
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V3/00—Electric or magnetic prospecting or detecting; Measuring magnetic field characteristics of the earth, e.g. declination, deviation
- G01V3/08—Electric or magnetic prospecting or detecting; Measuring magnetic field characteristics of the earth, e.g. declination, deviation operating with magnetic or electric fields produced or modified by objects or geological structures or by detecting devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Abstract
The invention discloses a kind of mobile phone detection method extracted based on wavelet-based attribute vector, including acquisition testing signal, article to be measured is passed through into detection door, detection signal is sampled, sample frequency is 30HZ, and signal duration is about 2s, and computer is sent into after A/D is changed；DB3 wavelet decompositions have been carried out to the detection signal collected using wavelet analysis, 8 component of signals has been obtained, component of signal is reconstructed and obtains different frequency range signal characteristic quantity, and then have extracted each frequency band signals characteristic quantity；Then dimensionality reduction is carried out to characteristic vector using principal component analytical method, signal characteristic quantity is transmitted to application program；Trained good BP networks output analysis result.The present invention detects magnetic array acquisition signal to mobile phone by DB3 small echos and handled, and improves collection array SNR, and carrying out wavelet decomposition for mobile phone detection magnetic array data extracts characteristic vector, carries out mobile phone and non-handset identity, substantially increases mobile phone Detection accuracy.
Description
Technical field
The present invention relates to mobile phone detection field, more particularly to a kind of mobile phone detection extracted based on wavelet-based attribute vector
Method.
Background technology
With the continuous improvement of security requirements, many occasions forbid the electronic products such as carrying mobile phone, recording pen, video recorder,
The electronic product can be detected by needing a kind of equipment badly.Current detection technique mainly has three kinds, and a kind of is traditional coil
Technology, mobile phone is detected by eddy detection technology；One kind is magnetic sensor technologies, and hand is detected by detecting the change in earth's magnetic field
The electronic products such as machine；One kind is to be based on the non-linear section detection technique of electron tube, is detected by analyzing non-linear section higher hamonic wave
Mobile phone.
Non-linear section detection technique threshold is high, costly, and current mainstream technology is coil EDDY CURRENT and magnetic field detection,
Coil eddy detection technology is for low frequency signal difference in response, and equipment volume and weight are big, and magnetic field detection technology is in sensitivity, body
There is good performance in terms of product, weight, power consumption, it is with the obvious advantage in mobile phone context of detection.
In the design of mobile phone detecting system, find mobile phone, key, wrist-watch and other items in by detection, exist wrong report or
The problem of failing to report is, it is necessary in system development, article target identification be realized using related software algorithm, so as to improve mobile phone
The accuracy rate and reliability of the detection of detector.
The main barrier of current technology is to be difficult to distinguish the ferromagnetic class article of the non-mobile phone such as mobile phone and key, exists higher
Rate of false alarm, current technology can only take a balance in rate of failing to report and rate of false alarm are accurate, be difficult to realize rate of failing to report and report by mistake forthright same
Shi Gaijin.
The content of the invention
For above-mentioned technical problem, the invention provides a kind of mobile phone detection method extracted based on wavelet-based attribute vector.
The features such as present invention considers mobile phone and other items detection signal unpredictability, non-stationary and instantaneity, it is known that wavelet transformation
It is especially suitable for handling this burst, it is different with traditional processing method, signal characteristic can be showed simultaneously in the domain of time-frequency two.Examine
Consider the pattern recognition problem of mobile phone and other items, mobile phone, key and other items by when detection signal should in time domain and frequency domain
Feature with differentiation is, it is necessary to which the Feature Extraction Technology of the signal using wavelet transformation, carries out the extraction of characteristic vector.Extract
During, the characteristic vector that sampling is extracted there may be the problem of dimension is excessive, be unfavorable for subsequent data processing and analysis, can
Using PCA) dimension-reduction treatment is carried out to the characteristic vector extracted；Finally it assign treated characteristic vector as god
An input vector through network realizes the identification of mobile phone article pattern.
The present invention is achieved by the following technical solutions：
A kind of mobile phone detection method extracted based on wavelet-based attribute vector, is comprised the following steps：
First by article to be measured by detection door, by array of magnetic sensors gather article to be measured by when signal, it is right
Detection signal is sampled, and setting sample frequency is 30HZ, and the duration of signal is about 2s, and meter is sent into after A/D is changed
Calculation machine；
DB3 wavelet decompositions have been carried out to the detection signal collected using wavelet analysis, 8 component of signals have been obtained, to each
Individual signal, which is reconstructed, obtains different frequency range signal characteristic quantity, and then extracts each frequency band signals characteristic quantity；
Then using principal component analytical method to characteristic vector carry out dimension-reduction treatment, then signal characteristic quantity transmit to should
With in program；
The characteristic vector extracted is inputted into BP networks, the principal character of BP Web-based Self-regulated Learning mobile phone signals, through excessive
The mobile phone characteristic signal training study of amount, exports analysis result, it is mobile phone or non-mobile phone article to recognize article to be measured.
Further, concretely comprising the following steps for each frequency band signals characteristic quantity is extracted：
(1) N layers of orthogonal wavelet decomposition are carried out to the detection signal changed through A/D, obtains the 1st layer to the common N number of high frequency of n-th layer
Coefficient of wavelet decomposition sequence：
{ d1, d2 ..., dN }
(2) energy of each floor height frequency coefficient of wavelet decomposition sequence is sought, Ej is set as jth floor height frequency coefficient of wavelet decomposition sequence
Dj energy, then have
N is middle component dk number in formula, and j, k are the integer more than or equal to 1.
(3) composition of characteristic vector, yardstick order with the energy normalized of each floor height frequency coefficient of wavelet decomposition sequence, group
Into characteristic vector group, i.e.,
F=(E1,E2,…,EN)。
The present invention compared with prior art, with following technique effect：
The present invention can be brought up to mobile phone detection success rate from 80% using wavelet analysis characteristic vector pickup technology
95%, while rate of false alarm does not increase, greatly improve the key performance of mobile phone detecting system.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the accompanying drawing used required in technology description to be briefly described.It is clear that drawings in the following description are only this hair
Some bright embodiments, for those of ordinary skill in the art, on the premise of not paying creative work, can be with root
Other accompanying drawings are obtained according to these accompanying drawings.
The schematic diagram that Fig. 1 recognizes for the mobile phone detection pattern of the present invention；
Primary signal and corresponding wavelet decomposition signal schematic representation that Fig. 2 gathers for the mobile phone and key of the present invention.
Embodiment
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention
In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is
A part of embodiment of the present invention, rather than whole embodiments.
The schematic diagram that Fig. 1 recognizes for the mobile phone detection pattern of the present invention；As shown in figure 1, one kind of the present invention is based on small echo
The mobile phone detection method of characteristic vector pickup, comprises the following steps：
First by article to be measured by detection door, by array of magnetic sensors gather article to be measured by when signal, it is right
Detection signal is sampled, and setting sample frequency is 30HZ, and the duration of signal is about 2s, and meter is sent into after A/D is changed
Calculation machine；Wherein, the data for having 2k bytes are got off by instrument record, but from the point of view of electric capacity discharge waveform, detection adds with Acquisition Circuit
The actual process of load is not essentially exceeding 500ms, and the signal that sensor is collected can continue 1s, so the only data to preceding 1k
Analyzed.
DB3 wavelet decompositions have been carried out to the detection signal collected using wavelet analysis, 8 component of signals have been obtained, to each
Individual signal, which is reconstructed, obtains different frequency range signal characteristic quantity, and then extracts each frequency band signals characteristic quantity；
Then dimensionality reduction is carried out to characteristic vector using principal component analytical method, then signal characteristic quantity is transmitted to applying journey
In sequence；
The characteristic vector extracted is inputted into BP networks, the principal character of BP Web-based Self-regulated Learning mobile phone signals, through excessive
The mobile phone characteristic signal training study of amount, exports analysis result, it is mobile phone or non-mobile phone to recognize article to be measured.
Wherein, Daubechie3 small echos, abbreviation DB3 is near symmetrical, biorthogonal, and its time domain supportive is strong, and frequency range is drawn
Divide effect and real-time preferable.And precision of prediction can not be significantly improved very much greatly to what load sequence decomposition scale was selected, on the contrary also
The efficiency of calculating can be reduced.Selected DB3 is morther wavelet, uses the Mallat algorithms of multiresolution analysis to carry out yardstick to load for 3
Decomposition.
From the perspective of signal filtering, signal to be decomposed is passed through a high-pass filter and one by orthogonal wavelet decomposition
Low pass filter is filtered, and obtains one group of low frequency signal and one group of high-frequency signal, and decompose N always to low frequency signal
Layer, decompose obtained low frequency signal every time and high-frequency signal length be all original signal length half, both are equal to length sum
The length of original signal, can be regarded as having carried out dot interlace sampling after the filtering, and decomposition result neither redundancy does not also lose original signal
Any information.Therefore, the present invention is arranged with the higher frequency signal energy of each metric space after orthogonal wavelet decomposition by yardstick order
Into vector as characteristic vector, specifically, extracting concretely comprising the following steps for each frequency band signals characteristic quantity：
(1) N layers of orthogonal wavelet decomposition are carried out to the detection signal changed through A/D, obtains the 1st layer to the common N number of high frequency of n-th layer
Coefficient of wavelet decomposition sequence：
{ d1, d2 ..., dN }
(2) energy of each floor height frequency coefficient of wavelet decomposition sequence is sought, Ej is set as jth floor height frequency coefficient of wavelet decomposition sequence
Dj energy, then have
N is middle component dk number in formula.
(3) composition of characteristic vector, yardstick order with the energy normalized of each floor height frequency coefficient of wavelet decomposition sequence, group
Into characteristic vector group, i.e.,
F=(E1,E2,…,EN)。
Primary signal and corresponding wavelet decomposition signal schematic representation that Fig. 2 gathers for the mobile phone and key of the present invention.
Find after tested, shown in the following Tables 1 and 2 of characteristic vector of mobile phone and key signal, Tables 1 and 2 is respectively hand
8 component of signals of machine and key.
Table 1
Table 2
The present invention detects magnetic array acquisition signal to mobile phone by DB3 wavelet basis and handled, and improves collection array SNR,
Wavelet decomposition is carried out for mobile phone detection magnetic array data and extracts characteristic vector, carries out mobile phone and non-handset identity.
Appeal characteristic vector is subjected to dimensionality reduction and neural network learning, mobile phone Detection accuracy is substantially increased.
Based on the embodiment in the present invention, those of ordinary skill in the art are obtained under the premise of creative work is not made
The all other embodiment obtained, belongs to the scope of protection of the invention.Although the present invention is illustrated with regard to preferred embodiment
And description, can be with it is understood by those skilled in the art that without departing from scope defined by the claims of the present invention
Variations and modifications are carried out to the present invention.
Claims (2)
1. a kind of mobile phone detection method extracted based on wavelet-based attribute vector, it is characterised in that comprise the following steps：
First by article to be measured by detection door, by array of magnetic sensors gather article to be measured by when signal, to detection
Signal is sampled, and setting sample frequency is 30HZ, and the duration of signal is about 2s, and computer is sent into after A/D is changed；
DB3 wavelet decompositions have been carried out to the detection signal collected using wavelet analysis, 8 component of signals are obtained, to each letter
Number it is reconstructed and obtains different frequency range signal characteristic quantity, and then extracts each frequency band signals characteristic quantity；
Then dimension-reduction treatment is carried out to characteristic vector using principal component analytical method, then signal characteristic quantity is transmitted to applying journey
In sequence；
The characteristic vector extracted is inputted into BP networks, the principal character of BP Web-based Self-regulated Learning mobile phone signals, by substantial amounts of
The training study of mobile phone characteristic signal, exports analysis result, and it is mobile phone or non-mobile phone article to recognize article to be measured.
2. a kind of mobile phone detection method extracted based on wavelet-based attribute vector according to claim 1, it is characterised in that：
Extract concretely comprising the following steps for each frequency band signals characteristic quantity：
(1) N layers of orthogonal wavelet decomposition are carried out to the detection signal changed through A/D, obtains the 1st layer to the common N number of high frequency wavelet of n-th layer
Decomposition coefficient sequence：
{ d1, d2 ..., dN }
(2) energy of each floor height frequency coefficient of wavelet decomposition sequence is sought, E is setjFor jth floor height frequency coefficient of wavelet decomposition sequence dj's
Energy, then have
<mrow>
<msub>
<mi>E</mi>
<mi>j</mi>
</msub>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msup>
<mrow>
<mo>|</mo>
<msubsup>
<mi>d</mi>
<mi>k</mi>
<mi>j</mi>
</msubsup>
<mo>|</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
N is middle component d in formulakNumber, j, k are integer more than or equal to 1；
(3) composition of characteristic vector, yardstick order with the energy normalized of each floor height frequency coefficient of wavelet decomposition sequence, composition is special
Vector Groups are levied, i.e.,
F=(E1,E2,…,EN)。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710345505.5A CN107203009B (en) | 2017-05-17 | 2017-05-17 | A kind of mobile phone detection method extracted based on wavelet-based attribute vector |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710345505.5A CN107203009B (en) | 2017-05-17 | 2017-05-17 | A kind of mobile phone detection method extracted based on wavelet-based attribute vector |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107203009A true CN107203009A (en) | 2017-09-26 |
| CN107203009B CN107203009B (en) | 2019-05-14 |
Family
ID=59905284
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710345505.5A Active CN107203009B (en) | 2017-05-17 | 2017-05-17 | A kind of mobile phone detection method extracted based on wavelet-based attribute vector |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107203009B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110110796A (en) * | 2019-05-13 | 2019-08-09 | 哈尔滨工程大学 | A kind of analysis method of the marine ships time series data based on deep learning |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050207664A1 (en) * | 2004-03-10 | 2005-09-22 | Jayaram Ramasastry | Methods and apparatuses for compressing digital image data |
| DE102009017436A1 (en) * | 2009-04-15 | 2010-11-04 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | Detecting a change between images or in a sequence of images |
| US20130004091A1 (en) * | 2011-06-28 | 2013-01-03 | Nokia Corporation | Methods, apparatuses and computer program products for utilizing wireless links for communication of compressed data |
| CN105654445A (en) * | 2016-01-28 | 2016-06-08 | 东南大学 | Mobile phone image denoising method based on wavelet transform edge detection |
| CN205484844U (en) * | 2016-03-08 | 2016-08-17 | 北京智信天一科技有限公司 | Multimode mobile phone test alarm device |
-
2017
- 2017-05-17 CN CN201710345505.5A patent/CN107203009B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050207664A1 (en) * | 2004-03-10 | 2005-09-22 | Jayaram Ramasastry | Methods and apparatuses for compressing digital image data |
| DE102009017436A1 (en) * | 2009-04-15 | 2010-11-04 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | Detecting a change between images or in a sequence of images |
| US20130004091A1 (en) * | 2011-06-28 | 2013-01-03 | Nokia Corporation | Methods, apparatuses and computer program products for utilizing wireless links for communication of compressed data |
| CN105654445A (en) * | 2016-01-28 | 2016-06-08 | 东南大学 | Mobile phone image denoising method based on wavelet transform edge detection |
| CN205484844U (en) * | 2016-03-08 | 2016-08-17 | 北京智信天一科技有限公司 | Multimode mobile phone test alarm device |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110110796A (en) * | 2019-05-13 | 2019-08-09 | 哈尔滨工程大学 | A kind of analysis method of the marine ships time series data based on deep learning |
| CN110110796B (en) * | 2019-05-13 | 2020-12-18 | 哈尔滨工程大学 | Deep learning-based marine vessel time sequence data analysis method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107203009B (en) | 2019-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105844216B (en) | Detection and matching mechanism for recognizing handwritten letters by WiFi signals | |
| CN106909784A (en) | Epileptic electroencephalogram (eeg) recognition methods based on two-dimentional time-frequency image depth convolutional neural networks | |
| CN105701470A (en) | Analog circuit fault characteristic extraction method based on optimal wavelet packet decomposition | |
| CN110610717B (en) | Separation method of mixed signals in complex frequency spectrum environment | |
| CN107392123B (en) | Radio frequency fingerprint feature extraction and identification method based on coherent accumulation noise elimination | |
| CN106653062A (en) | Spectrum-entropy improvement based speech endpoint detection method in low signal-to-noise ratio environment | |
| CN108680796A (en) | Electromagnetic information leakage detecting system and method for computer display | |
| CN105678273A (en) | Initial point detection algorithm of transient signal in radio frequency fingerprint identification technology | |
| CN108535636A (en) | A kind of analog circuit is distributed the neighbouring embedded fault signature extracting method that the victor is a king based on stochastic parameter | |
| CN104535905A (en) | Partial discharge diagnosis method based on naive bayesian classification | |
| CN107942214A (en) | A kind of feature extracting method of transformer partial discharge signal, device | |
| CN107590455A (en) | A kind of high frequency partial discharge adaptive-filtering clustering method and device based on wavelet decomposition | |
| CN113673359A (en) | Method for improving accuracy of eddy current testing metal defects based on deep learning | |
| CN114048773A (en) | Behavior identification method and system based on transfer learning and WiFi | |
| CN111027488A (en) | Signal classification method and device | |
| Chen et al. | Automatic modulation classification of radar signals utilizing X-net | |
| CN102988041A (en) | Selective cardiac-magnetic signal averaging method in signal noise suppression | |
| CN107203009B (en) | A kind of mobile phone detection method extracted based on wavelet-based attribute vector | |
| JP2009244138A (en) | Apparatus and method for measuring/separating/analyzing signal | |
| CN111481203B (en) | Indoor static passive human body detection method based on channel state information | |
| CN204330901U (en) | A kind of innovative noise sensor | |
| CN111652132B (en) | Non-line-of-sight identity recognition method and device based on deep learning and storage medium | |
| CN112883895A (en) | Illegal electromagnetic signal detection method based on self-adaptive weighted PCA and realization system thereof | |
| Huang et al. | FM broadcast monitoring using artificial intelligence | |
| CN111181574A (en) | End point detection method, device and equipment based on multi-layer feature fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Mobile phone detection method based on wavelet feature vector extraction Effective date of registration: 20191107 Granted publication date: 20190514 Pledgee: China Co truction Bank Corp Guangzhou Tianhe subbranch Pledgor: Jiechuang Intelligent Technology Co., Ltd. Registration number: Y2019440000161 |
|
| PC01 | Cancellation of the registration of the contract for pledge of patent right | ||
| PC01 | Cancellation of the registration of the contract for pledge of patent right |
Date of cancellation: 20210521 Granted publication date: 20190514 Pledgee: China Co. truction Bank Corp Guangzhou Tianhe subbranch Pledgor: NEXWISE INTELLIGENCE CHINA Ltd. Registration number: Y2019440000161 |