CN203825699U - Multispectral counterfeit money detection system - Google Patents
Multispectral counterfeit money detection system Download PDFInfo
- Publication number
- CN203825699U CN203825699U CN201420093669.5U CN201420093669U CN203825699U CN 203825699 U CN203825699 U CN 203825699U CN 201420093669 U CN201420093669 U CN 201420093669U CN 203825699 U CN203825699 U CN 203825699U
- Authority
- CN
- China
- Prior art keywords
- light source
- detection system
- linear array
- multispectral
- banknote
- 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.)
- Expired - Fee Related
Links
Landscapes
- Inspection Of Paper Currency And Valuable Securities (AREA)
Abstract
The utility model discloses a multispectral counterfeit money detection system, which comprises a linear array light source system and a linear array detection system, wherein a multispectral light source is arranged on the linear array light source system; a photodetector used for detecting wide band light intensity signals is integrated in the linear array detection system; light emitted by the multispectral light source irradiates on the surface of the money, light is reflected or transmitted and reaches the linear array detection system, the photodetector inside the linear array detection system converts light intensity signals into electrical signals to be sent to an upper computer. Alternation money and joining money can be well recognized, and recognition is more accurate compared with the prior art.
Description
Technical field
The utility model relates to the interleaving techniques field of photonics and information science, particularly a kind of multispectral counterfeit money detection system.
Background technology
Currency is the important foundation stone in society life.The identification of counterfeit money is related to popular life happiness.But counterfeit money method for making of today emerges in an endless stream.Beyond the counterfeit money of " complete " is made, " technique of composite-RMB bank note " and " splicing coin " is also growing, and within a period of time " deception " ATM in bank.In order better to identify counterfeit money, people have proposed many schemes.Optical means is popular because its simple and practical feature is standby.But, simply utilize light intensity (image) information, in many cases, cannot provide the most comprehensive information.The spectral signal of object can provide more fully object essential information.Therefore, how can identify more accurately counterfeit money and become a research topic that has practical value.
Utility model content
Fundamental purpose of the present utility model is that the shortcoming that overcomes prior art is with not enough, a kind of multispectral counterfeit money detection system is provided, this system, by adopting multi-spectral linear array light source and linear array detector, can gather the information of more banknote spectrum pictures, thereby more be conducive to identify counterfeit money.
The purpose of this utility model realizes by following technical scheme: multispectral counterfeit money detection system, comprises array light source system and linear array detection system; In array light source system, multispectral light source is installed; Integrated for surveying the photodetector of broad band light intensity signal in linear array detection system; The light that multispectral light source sends is radiated at banknote surface, after banknote reflection or transmission, arrives linear array detection system, and the photodetector of linear array detection system inside sends to host computer after light intensity signal is converted to electrical signal.
Concrete, described array light source system comprises light source mount pad and the first plano-convex post lens, N is segmented out on the long limit of light source mount pad
1sub regions, installs S light source in every sub regions; The plane of the first plano-convex post lens is close to the long limit of light source mount pad.
Concrete, described linear array detection system comprises detector mount pad and the second plano-convex post lens, detector mount pad segments out N on long limit
2sub regions, installs T detector in every sub regions, on the workplace of several detectors therein, optical filter is installed; The plane of the second plano-convex post lens is close to the long limit of detector mount pad.
Preferably, if the light that multispectral light source sends is to be reflected and arrived linear array detection system by banknote, array light source system and linear array detection system are located at the same side of banknote to be measured, and 1 white light source is installed in every sub regions in described array light source system; 3 detectors are installed in every sub regions in described linear array detection system, and red, green, blue color filter are installed respectively on the workplace of 3 detectors.Adopt the detection mode of reflection, can well identify adhesive tape in " splicing coin " etc., and then identification counterfeit money.
Preferably, if the light that multispectral light source sends is to arrive linear array detection system by banknote transmission, array light source system and linear array detection system are located at not homonymy of banknote to be measured, 3 light sources are installed in every sub regions in described array light source system, and 3 light source colours are respectively red, green, blue; 1 detector is installed in every sub regions in described linear array detection system.Adopt the detection mode of transmission, can well identify and have one surperficial " technique of composite-RMB bank note " for forgery etc., and then identification counterfeit money.
Obtaining by said detecting system after the spectrum picture information of banknote, at host computer, image information is processed, whether identify is counterfeit money.The recognition methods adopting in host computer is the counterfeit money detection method based on multispectral tagsort technology, comprises step:
(1) training process:
(1-1) scan genuine notes by multispectral light source, every run-down obtains a measurement vector; Use different genuine notes multiple scanning several times, obtain a canonical matrix;
(1-2) scan counterfeit money by multispectral light source, obtain the test vector of a counterfeit money, all measurement vectors in this test vector and canonical matrix are compared one by one, calculate and the each difference value of vector in normal form space of measuring, get minimum value wherein as difference value corresponding to this test vector; Use different counterfeit money multiple scanning several times, obtain a difference value vector, get minimum value in this vector as threshold value;
(2) the identification process of new banknote: scan new banknote, obtain when pre-test vector, this is worked as to pre-test vector and compare one by one with all measurement vectors in canonical matrix, get the minimum value of difference value in all normal forms space as the difference value of deserving pre-test vector correspondence; If this difference value is greater than threshold value, thinks that this new banknote is counterfeit money, otherwise think genuine notes.
Preferably, also comprise step:
(1-3) according to the vectorial comparative result of all measurements in the test vector of counterfeit money and canonical matrix, obtain the contribution weight of different spectrum to difference value, according to contribution weight, color balance parameter in light intensity or the linear array detector of light source corresponding to spectrum is adjusted, made its ratio identical with contribution weight; Then repeating step (1-1), (1-2), obtain new canonical matrix and threshold value.
Preferably, in described step (1-2), (2), between two vectors, the acquisition methods of the difference value in normal form space is defined as follows; Dev=||U × I
i-U × I
m||, wherein, Dev is the difference value in the normal form space that will ask for, I
ifor the vector of i row in canonical matrix, I
mfor vector to be compared, by principal component analytical method, obtain eigenvectors matrix U, for extracting I
iand I
mthe proper vector of vector.
Described contribution weight is obtained by following methods:
wherein Tn represents the columns in canonical matrix.
More specifically, described detection method is:
(1) training process:
(1-1) multispectral light source sends wire white light light, and light scanning is after genuine notes, and by reflection or be transmitted to linear array detection system, the result of linear array detection system single pass is defined as a dimensional vector I
e; i; Wire light continuous sweep banknote surface, total result of detection is defined as overall measurement matrix
the dimension of matrix is N × (w × M
2); Wherein, N is that in currency examine process, linear light source is at the scanning times of banknote surface, and w is detectable spectrum number in system, M
2for the number of detecting element in linear array detection system; By overall measurement matrix
reset, form a new dimensional vector
its dimension is (N × wM
2) × 1, called after is measured vector;
(1-2) select at random T to open genuine notes, the process of repeating step (1-1), obtains successively and measures vector, and all measurement Vector Groups are combined into canonical matrix: [I
1, I
2..., I
t], wherein I
irepresent to measure the measurement vector obtaining the i time;
(1-3) use counterfeit money, scan according to the mode of step (1-1), obtain test vector I
s,i; By test vector I
s,iwith canonical matrix [I
1, I
2..., I
t] in all measurement vectors compare one by one, obtain the minimum difference value Dev in normal form space
i;
(1-4) use different counterfeit money, repeating step (1-3), relatively the test vector I of counterfeit money
swith canonical matrix [I
1, I
2..., I
t], obtain the minimum difference value Dev in normal form space, minimum difference value Dev corresponding all counterfeit money is defined as to one dimension row vector L:[Dev
1, Dev
2..., Dev
s], wherein S is the number of the counterfeit money of use, Dev
jrepresent the difference value of obtaining while measuring counterfeit money the j time; Get [Dev
1, Dev
2..., Dev
s] in minimum value as threshold value L
th;
(1-5) according to the test vector I of counterfeit money in step (1-4)
swith canonical matrix [I
1, I
2..., I
t] between comparison, know the contribution weight of each spectrum to difference value Dev simultaneously, according to contribution weight, change the intensity between different colours in light source, or color balance parameter in adjustment detector, then repeating step (1-1)-step (1-4), obtains new canonical matrix [I
1, I
2..., I
t] and threshold value L
th;
(2) the identification process of new banknote: utilize multispectral light source to scan new banknote, obtain one-dimensional measurement vector I
m, to one-dimensional measurement vector I
mcanonical matrix [I with step (1-5) acquisition
1, I
2..., I
t] compare one by one, obtain minimum difference value L
m; If this difference value is greater than threshold value L
th, think that this banknote is counterfeit money, otherwise, think that this banknote is genuine notes.
Compared with prior art, tool has the following advantages and beneficial effect the utility model:
1, the utility model, by adopting multispectral light source to irradiate, can make the banknote information of acquisition more comprehensive, is more conducive to truth identification below.
2, in the utility model device, according to the difference of reflection or transmission detection mode, the structure of array light source system and linear array detection system also can be adjusted, and detection mode is more various, uses more convenient.
Brief description of the drawings
Fig. 1 is the structural principle schematic diagram of embodiment 1 while adopting reflective detection mode;
Fig. 2 (a) is the side view of array light source system in embodiment 1;
Fig. 2 (b) is the front view of array light source system in embodiment 1;
Fig. 3 (a) is the side view of linear array detection system in embodiment 1;
Fig. 3 (b) is the front view of linear array detection system in embodiment 1;
Fig. 4 is the method flow diagram of embodiment 1;
Fig. 5 is the structural principle schematic diagram of embodiment 2 while adopting transmission-type detection mode;
Fig. 6 is the front view of array light source system in embodiment 2;
Fig. 7 is the front view of linear array detection system in embodiment 2.
Embodiment
Below in conjunction with embodiment and accompanying drawing, the utility model is described in further detail, but embodiment of the present utility model is not limited to this.
Embodiment 1
As shown in Figure 1, the multispectral counterfeit money detection system of the present embodiment adopts reflective detection mode, and array light source system 1 and linear array detection system 2 are located at the same side of banknote 3 to be measured.
In the present embodiment, the structure of array light source system 1 is as Fig. 2 (a), 2(b) as shown in, it comprises that the plane of light source mount pad 11 and the first plano-convex post lens 12, the first plano-convex post lens 12 is close to the long limit of light source mount pad 11.N is segmented out on the long limit of light source mount pad 11
1sub regions 14,1 white light source 13 of the interior installation of every sub regions 14.Also can adopt in actual applications the small light source element of multiple different spectrum (color).
In the present embodiment, the structure of linear array detection system 2 is as Fig. 3 (a), 3(b) as shown in, it comprises that the plane of detector mount pad 21 and the second plano-convex post lens 22, the second plano-convex post lens 22 is close to the long limit of detector mount pad 21.N is segmented out on the long limit of detector mount pad 21
2sub regions 24,3 detectors 23 of the interior installation of every sub regions 24, and the detector workplace of the top is provided with Red lightscreening plate, and lower left detector workplace is provided with green color filter, and lower right detector workplace is provided with blue color filter.
In Fig. 1, dotted arrow represents the direction of propagation of the wire light that array light source system sends, and solid arrow representative is through the direction of propagation of the light of banknote reflection.The square banknote that represents moving along lower right, the wire light illumination that has a wire region to be focused on banknote.The light that array light source system is sent is focused to wire light after the first plano-convex post lens, is radiated on banknote, after banknote reflection, is detected by linear array detection system.Banknote, in motion process, is scanned by linear light source, and its surface information is all obtained by linear array detection system.
As shown in Figure 4, in the present embodiment, detection method is as follows:
One, the training stage
Step 1: send wire white light light in array light source, light is irradiated to after genuine notes, will reflex in corresponding linear array detection system.Linear array detection system can be obtained the intensity signal of three kinds of colors of red, green, blue simultaneously, is designated as single measurement matrix [I
eR; 1, I
eG; 1, I
eB; 1], this matrix is divided into three parts, respectively the reflective light intensity information of three kinds of colors of corresponding red, green, blue.This matrix column vector dimension is N
2, i.e. the number of subregion in linear array detection system.The all surfaces of light scanning banknote, obtains overall measurement matrix [I
eR; 1, I
eG; 1... I
eB; N], wherein N is the scanning times of light on banknote, and this matrix is rearranged into new column vector a: I, its dimension is (N × 3N
2) × 1, is defined as and measures vector.
Step 2: adopt different genuine notes to repeat said process, once the number of times of multiple scanning reaches predetermined number of times T time, stop scanning, obtain and measure vectorial combination, be defined as canonical matrix: [I
1, I
2..., I
t].
Step 3: use conterfeit note, test according to the mode of the step 1 in the training stage.Obtain test vector I
s,i.By test vector I
s,iwith canonical matrix [I
1, I
2..., I
t] in institute's directed quantity compare one by one, obtain the difference value Dev in minimum normal form space
i.
Step 4: use a large amount of different conterfeit notes, repeating step 3, relatively test vector and the canonical matrix of counterfeit money, obtains the difference value Dev:Dev=[Dev in normal form space
1, Dev
2..., Dev
s], wherein S is the number of the counterfeit money of use, Dev
jrepresent the difference value of obtaining while measuring counterfeit money the j time.Get [Dev
1, Dev
2..., Dev
s] in minimum value as threshold value L
th.
Step 5: according to the test vector I of counterfeit money in step 4
swith canonical matrix [I
1, I
2..., I
t] between comparison, know simultaneously difference value contributed to maximum spectral components, be defined as optimal spectrum component.If the contribution maximum of C component now to difference value, can improve the C component intensity factor in linear array detector, then repeating step 1-step 4, obtains new canonical matrix [I
1, I
2..., I
t] and threshold value L
th.Described C component is the one in three kinds of spectral components of red, green, blue.
Two, the cognitive phase of new banknote
Step 6: utilize multispectral Line of light line sweep banknote to be detected, according to the method for step 1, use Line of light line sweep banknote, obtain one-dimensional measurement vector I
m, to I
mcanonical matrix [the I obtaining with step 5
1, I
2..., I
t] compare one by one, obtain minimum difference value L
m; If this difference value is greater than threshold value L
th, think that this banknote is counterfeit money, otherwise, think that this banknote is genuine notes.
The acquisition methods of two vectorial normal form spatial diversity values in described step 3 and step 6 is defined as follows; Dev=||U × I
i-U × I
m||, wherein I
ifor the vector of i row in canonical matrix, I
mfor vector to be compared, by principal component analytical method, can obtain eigenvectors matrix U, for extracting I
iand I
mthe proper vector of vector.
Embodiment 2
The present embodiment except following characteristics other structures with embodiment 1:
As shown in Figure 5, the multispectral counterfeit money detection system of the present embodiment adopts transmission-type detection mode, and array light source system 1 and linear array detection system 2 are located at the not homonymy of banknote 3 to be measured.
In the present embodiment, the structure of array light source system 1 as shown in Figure 6,3 LED light sources of the interior installation of its every sub regions 14.Wherein the light source of the top is red LED, and lower left light source is green LED, and lower right light source is blue led.The structure of linear array detection system 2 as shown in Figure 7,1 detector 23 of the interior installation of its every sub regions 24.N is segmented out on the long limit of light source mount pad
1sub regions 14, N is segmented out on the long limit of detector mount pad
2sub regions 24,
In Fig. 5, dotted arrow represents the direction of propagation of the wire light that array light source system sends, and solid arrow representative is through the direction of propagation of the light of banknote transmission.The square banknote that represents moving along lower right, the wire light illumination that has a wire region to be focused on banknote.The light that array light source system is sent is focused to wire light after the first plano-convex post lens, is radiated on banknote, after banknote transmission, is detected by linear array detection system.
In the present embodiment, detection method is as follows:
One, the training stage
Step 1: in array light source, the LED of three kinds of different colours (spectrum) is alternately luminous fast, linear array detection system can be complied with hyposynchronous acquisition transmitted spectrum, is designated as single measurement matrix [I
eR; 1, I
eG; 1, I
eB; 1].This matrix column vector dimension is N
2, i.e. the number of subregion in linear array detection system.The all surfaces of light scanning banknote, obtains overall measurement matrix [I
eR; 1, I
eG; 1... I
eB; N], wherein N is the scanning times of light on banknote, and this matrix is rearranged into new column vector a: I, its dimension is (N × 3 × N
2) × 1, is defined as and measures vector.
Step 2: said process repeats T time, obtains canonical matrix: [I
1, I
2..., I
t].
Step 3: use conterfeit note, test according to the mode of the step 1 in the training stage.Obtain test vector I
s,i.By test vector I
s,iwith canonical matrix [I
1, I
2..., I
t] in institute's directed quantity compare one by one, obtain the difference value Dev in normal form space
i.
Step 4: use a large amount of different conterfeit notes, repeating step 3, relatively test vector and the canonical matrix of counterfeit money, obtains the difference value Dev:Dev=[Dev in normal form space
1, Dev
2..., Dev
s], wherein S is the number of the counterfeit money of use, Dev
jrepresent the difference value of obtaining while measuring counterfeit money the j time.Get [Dev
1, Dev
2..., Dev
s] in minimum value as threshold value L
th.
Step 5: according to the comparison between the testing result of counterfeit money in step 4 and canonical matrix, simultaneously know that the LED light source of Ct different colours is to the contribution weight [w of difference value
1, w
2, w
3], according to contribution weight, adjust the light intensity of the LED light source of 3 different colours, make its ratio identical with contribution weight, then repeating step 1-step 4, obtains new canonical matrix [I
1, I
2..., I
t] and threshold value L
th.
Two, the cognitive phase of new banknote
Step 6: utilize multispectral Line of light line sweep banknote to be detected, according to the method for step 1, use Line of light line sweep banknote, obtain one-dimensional measurement vector I
m, compare I by algorithm
mthe measurement vector obtaining with step 5 combines [I
1, I
2..., I
t] compare one by one, obtain minimum difference value L
m; If this difference value is greater than threshold value L
th, think that this banknote is counterfeit money, otherwise, think that this banknote is genuine notes.
Above-described embodiment is preferably embodiment of the utility model; but embodiment of the present utility model is not restricted to the described embodiments; other any do not deviate from change, the modification done under Spirit Essence of the present utility model and principle, substitutes, combination, simplify; all should be equivalent substitute mode, within being included in protection domain of the present utility model.
Claims (5)
1. multispectral counterfeit money detection system, is characterized in that, comprises array light source system and linear array detection system; In array light source system, multispectral light source is installed; Integrated for surveying the photodetector of broad band light intensity signal in linear array detection system; The light that multispectral light source sends is radiated at banknote surface, after banknote reflection or transmission, arrives linear array detection system, and the photodetector of linear array detection system inside sends to host computer after light intensity signal is converted to electrical signal.
2. multispectral counterfeit money detection system according to claim 1, is characterized in that, described array light source system comprises light source mount pad and the first plano-convex post lens, and N is segmented out on the long limit of light source mount pad
1sub regions, installs S light source in every sub regions; The plane of the first plano-convex post lens is close to the long limit of light source mount pad.
3. multispectral counterfeit money detection system according to claim 2, is characterized in that, described linear array detection system comprises detector mount pad and the second plano-convex post lens, and detector mount pad segments out N on long limit
2sub regions, installs T detector in every sub regions, on the workplace of several detectors therein, optical filter is installed; The plane of the second plano-convex post lens is close to the long limit of detector mount pad.
4. multispectral counterfeit money detection system according to claim 3, it is characterized in that, if the light that multispectral light source sends is to be reflected and arrived linear array detection system by banknote, array light source system and linear array detection system are located at the same side of banknote to be measured, and 1 white light source is installed in every sub regions in described array light source system; 3 detectors are installed in every sub regions in described linear array detection system, and red, green, blue color filter are installed respectively on the workplace of 3 detectors.
5. multispectral counterfeit money detection system according to claim 3, it is characterized in that, if the light that multispectral light source sends is to arrive linear array detection system by banknote transmission, array light source system and linear array detection system are located at not homonymy of banknote to be measured, 3 light sources are installed in every sub regions in described array light source system, and 3 light source colours are respectively red, green, blue; 1 detector is installed in every sub regions in described linear array detection system.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201420093669.5U CN203825699U (en) | 2014-03-03 | 2014-03-03 | Multispectral counterfeit money detection system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201420093669.5U CN203825699U (en) | 2014-03-03 | 2014-03-03 | Multispectral counterfeit money detection system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN203825699U true CN203825699U (en) | 2014-09-10 |
Family
ID=51481174
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201420093669.5U Expired - Fee Related CN203825699U (en) | 2014-03-03 | 2014-03-03 | Multispectral counterfeit money detection system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN203825699U (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103903326A (en) * | 2014-03-03 | 2014-07-02 | 广州科珥光电科技有限公司 | Multi-spectral counterfeit bank note detection system and method |
-
2014
- 2014-03-03 CN CN201420093669.5U patent/CN203825699U/en not_active Expired - Fee Related
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103903326A (en) * | 2014-03-03 | 2014-07-02 | 广州科珥光电科技有限公司 | Multi-spectral counterfeit bank note detection system and method |
CN103903326B (en) * | 2014-03-03 | 2016-07-06 | 广州科珥光电科技有限公司 | Multispectral counterfeit money detection system and detection method thereof |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103903326A (en) | Multi-spectral counterfeit bank note detection system and method | |
US11861889B2 (en) | Analysis device | |
CN107192456B (en) | A kind of color measuring multi-optical spectrum imaging system based on LED illumination | |
CN100457031C (en) | Method and device for recognition of natural skin | |
CN105842173B (en) | A kind of EO-1 hyperion material discrimination method | |
CN105787508B (en) | Textile color identification method and system | |
US20160162907A1 (en) | System and method for identifying and authenticating a tag | |
CN104502917B (en) | Method and system for enhancing detection sensitivity of photon counting laser radar by utilizing photon regulation and control | |
US8263948B2 (en) | Authentication apparatus for moving value documents | |
AU2011257540B2 (en) | Device for testing the authenticity of valuable documents | |
CN102610023B (en) | Optical detection device for identifying counterfeiting of security and detection method thereof | |
CN109613556A (en) | Photon counting laser three-D detection imaging method based on sparse representation | |
US20240044712A1 (en) | Polarimetry | |
CN106680186B (en) | A kind of flow cytometer polymorphic type scattering optical detection system | |
CN203825699U (en) | Multispectral counterfeit money detection system | |
CN103438980B (en) | Method and device for liquid surface wave detection based on linear array CCD and linear infrared laser | |
CN102387631B (en) | Automatic judgment and adjustment method of light intensity in immunochromatography interpretation recorder | |
CN208076382U (en) | Water body multi-wavelength optical attenuation coefficient measuring device | |
CN108414464A (en) | Water body multi-wavelength optical attenuation coefficient measuring device and method | |
KR20170052297A (en) | Device and method for inspecting external appearance of display panel | |
CN104897580A (en) | Optical system and optical method for detecting atmospheric components through non-intuitive imaging | |
CN107293037A (en) | Bill false distinguishing method and device based on anti-fake material spectral characteristic | |
Bo et al. | A multi-wavelength canopy LiDAR for vegetation monitoring: system implementation and laboratory-based tests | |
CN208187381U (en) | Laser leveler automatic checkout system | |
CN203444122U (en) | Underwater hyperspectral imaging system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140910 Termination date: 20200303 |