CN112466033A - Ultraviolet pattern based bill verification detection method and detection system thereof - Google Patents
Ultraviolet pattern based bill verification detection method and detection system thereof Download PDFInfo
- Publication number
- CN112466033A CN112466033A CN202011479918.0A CN202011479918A CN112466033A CN 112466033 A CN112466033 A CN 112466033A CN 202011479918 A CN202011479918 A CN 202011479918A CN 112466033 A CN112466033 A CN 112466033A
- Authority
- CN
- China
- Prior art keywords
- bill
- feature
- pattern
- similarity
- picture
- 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
Images
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/20—Testing patterns thereon
- G07D7/2008—Testing patterns thereon using pre-processing, e.g. de-blurring, averaging, normalisation or rotation
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/20—Testing patterns thereon
- G07D7/2016—Testing patterns thereon using feature extraction, e.g. segmentation, edge detection or Hough-transformation
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/20—Testing patterns thereon
- G07D7/202—Testing patterns thereon using pattern matching
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Inspection Of Paper Currency And Valuable Securities (AREA)
Abstract
The invention discloses a bill authenticity verification detection method and a bill authenticity verification detection system based on ultraviolet patterns, and belongs to the technical field of pattern recognition and computer vision. The detection method comprises the following steps: step 1, carrying out angle correction, positioning and binarization processing on an ultraviolet image to be checked, and then intercepting a check pattern; step 2, extracting corner features, contour features and pixel point features of the picture to be detected; and 3, comparing the similarity of the feature matrix of the bill to be detected with the feature matrix of the template bill to obtain the similarity values of the two feature matrices, and finally judging whether the similarity values meet the threshold requirement or not so as to judge the authenticity of the bill. The invention judges the authenticity of the fake pattern by checking the fake pattern under the ultraviolet light, has lower requirements on the hardware equipment for image acquisition and saves the hardware debugging time; and the method for fusing and verifying the multiple characteristics of each verification pattern can give consideration to the problems of the area size of the pattern, the existence of the loss of the pattern, the existence of the change of the pattern content and the like, and improves the verification precision.
Description
Technical Field
The invention belongs to the technical field of pattern recognition and computer vision, and particularly relates to a bill verification detection method and a bill verification detection system based on ultraviolet patterns.
Background
Along with the development of national economy, the circulation of bank bills is increased, and the problem of illegal cash register by bills is often caused in the transaction process of the bills, so that serious economic loss is caused for bank customers. In order to solve the problems, the anti-counterfeiting technology of the bills is continuously improved, and the anti-counterfeiting technology comprises paper anti-counterfeiting, ink anti-counterfeiting, printing anti-counterfeiting and the like. At present, one main method for checking bills is to check the bills based on colors, and judge the authenticity by checking the color characteristics of each anti-counterfeiting point in the bills.
The applicant has found that the following disadvantages exist in the prior art after long-term practice: 1. in the traditional method, the color-based bill verification needs to collect color information of the verification point at a fixed position, when the orientation of the bill is changed, the verification point collection equipment cannot collect the color information, and the fault-tolerant rate of the whole system is low. 2. When the counterfeit checking system needs to check other types of bills, the positions of the image acquisition equipment are required to be changed according to different counterfeit checking points, and if the bills of different types are identified simultaneously, the acquisition equipment is required to be added to check all the counterfeit checking points, so that the problems of hardware cost improvement, debugging difficulty increase and the like are caused.
Disclosure of Invention
The purpose of the invention is as follows: provides a bill verification detection method based on ultraviolet patterns and a detection system thereof, which are used for solving the problems involved in the background technology.
The technical scheme is as follows: a bill verification detection method based on ultraviolet patterns comprises the following steps:
step 1, bill pretreatment: carrying out denoising treatment and binarization treatment on an ultraviolet image to be verified, and then intercepting a verification pattern;
step 2, characteristic value extraction: extracting corner feature, contour feature and pixel point feature of the picture to be detected;
step 3, threshold discrimination: and comparing the similarity of the feature matrix of the bill to be detected with the feature matrix of the template bill to obtain the similarity values of the two feature matrices, and finally judging whether the similarity values meet the threshold requirement or not so as to judge the authenticity of the bill.
In a further embodiment, the method for extracting corner features includes the following steps:
step 211, firstly, constructing a Hessian matrix, solving a Hessian matrix for each pixel point, and calculating a local maximum value by using a Hessian matrix discriminant;
step 212, using the local maximum point as a key point, then setting a threshold to filter out the key point with weak energy and the key point with wrong positioning, and screening out useful feature points;
step 213, taking a preset X1Y 1 rectangular area block around the feature point, wherein the direction of the rectangular area block is along the main direction of the feature point;
step 214, sampling each region block by preset size X2Y 2, and counting Haar wavelet characteristics of each region block in the horizontal direction and the vertical direction of T pixels of threshold value;
step 215, taking four values of the sum of the horizontal direction values, the sum of the vertical direction values, the sum of the absolute values in the horizontal direction and the sum of the absolute values in the vertical direction of the Haar wavelet features as the characteristic values of the subareas;
and step 216, obtaining a total 64-dimensional feature vector as a descriptor of the feature point, namely the feature vector of the corner point of the picture.
In a further embodiment, the contour features include changes in the outer contour curve of the image and the size of the area of the image.
In a further embodiment, the method for extracting the contour features includes the following steps: and expressing the profile characteristics of the picture to be detected by adopting the Hu moment, wherein the Hu moment is 7 invariant moments constructed by utilizing second-order and third-order normalized central moments.
In a further embodiment, the pixel point characteristics are the number of white pixel points of the binarized picture to be detected.
In a further embodiment, the method for extracting the pixel point features includes the following steps: firstly, performing morphological processing on a binarized picture to be detected to remove fine white noise points possibly existing in a background and fine black noise points in a pattern; and traversing the whole picture to be detected to obtain the number of white pixel points.
In a further embodiment, the method for determining the threshold value includes the following steps:
step 31, fusing the feature vectors after the three features are extracted to form a new feature vector matrix;
step 32, calculating the similarity of the matrix and a feature vector matrix of the template picture by using the Euclidean distance, and outputting the calculation result of the similarity of the two matrices in a one-dimensional array form;
step 33, setting members in the array which are lower than the similarity threshold value to zero, and carrying out weighting calculation on the similarity values meeting the similarity threshold value and the number of the members to obtain a value representing the similarity of the pictures;
step 34, if the value meets the authenticity detection threshold value, judging the pattern to be a real pattern, otherwise, judging the pattern to be false;
in a further embodiment, the similarity threshold and the authenticity detection threshold are obtained by testing bills under different environments.
The invention also provides a detection system based on the ultraviolet pattern bill verification detection method, which comprises the following modules:
the bill preprocessing module is used for denoising and binarizing the ultraviolet image to be verified, and then intercepting the verification pattern;
the characteristic value extraction module is used for extracting angular point characteristics, contour characteristics and pixel point characteristics of the picture to be detected;
and the threshold value judging module is used for comparing the similarity of the feature matrix of the bill to be detected with the feature matrix of the template bill to obtain the similarity value of the two feature matrices, and finally judging whether the similarity value meets the threshold value requirement or not so as to judge the authenticity of the bill.
Has the advantages that: the invention relates to a bill verification detection method and a bill verification detection system based on ultraviolet patterns, which have the following advantages compared with the prior art:
1. the invention judges the authenticity of the fake-verifying pattern under the ultraviolet light, has lower requirements on hardware equipment for image acquisition and saves hardware debugging time.
2. The method for fusing and verifying the multiple characteristics of each verification pattern can give consideration to the area size of the pattern, the existence of the loss of the pattern, the existence of the change of the pattern content and other detailed problems, and the verification precision is improved.
3. Because the influence of the thermal stability of the photosensitive element of the acquisition equipment is different from the structure of the sensor of the acquisition equipment, fine noise interference exists, the interference of noise points to white pixel points is eliminated through the preprocessing module on the bill picture, the bill counterfeit checking pattern is highlighted, and the extraction precision of the characteristic value is improved.
4. The invention aims at the fake-checking patterns with different characteristics, forms a new characteristic-fused matrix by coupling multiple characteristics, performs similarity calculation, and reduces the calculation complexity while ensuring higher accuracy.
5. The method extracts the corner feature, the contour feature and the pixel point feature of the bill ultraviolet picture, has higher accuracy rate by self-defining the probability of calculating the image to be true by a multi-feature matrix fusion method, and ensures the detection truth degree and high false-checking accuracy rate of the false-checking pattern.
In conclusion, the invention ensures the detection precision and improves the fault-tolerant rate of the whole system; the requirements on hardware equipment are reduced, and the hardware debugging time is saved.
Drawings
FIG. 1 is a system flow diagram of the present invention.
Fig. 2 is a schematic picture of a bill after pretreatment in the invention.
FIG. 3 is a schematic picture of a binarized ticket according to the present invention.
Fig. 4 is a schematic picture of a cut-out authentication pattern in the present invention.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the present invention. It will be apparent, however, to one skilled in the art, that the present invention may be practiced without one or more of these specific details. In other instances, well-known features have not been described in order to avoid obscuring the invention.
As shown in fig. 1, a bill verification detection system based on ultraviolet patterns comprises the following modules:
the bill preprocessing module is used for denoising and binarizing the ultraviolet image to be verified, and then intercepting the verification pattern;
the characteristic value extraction module is used for extracting angular point characteristics, contour characteristics and pixel point characteristics of the picture to be detected;
and the threshold value judging module is used for comparing the similarity of the feature matrix of the bill to be detected with the feature matrix of the template bill to obtain the similarity value of the two feature matrices, and finally judging whether the similarity value meets the threshold value requirement or not so as to judge the authenticity of the bill.
The bill verification detection method based on the ultraviolet pattern is further explained by combining the detection system, and comprises the following steps:
step 1, bill pretreatment: carrying out denoising treatment and binarization treatment on an ultraviolet image to be verified, and then intercepting a verification pattern; specifically, the ultraviolet image to be verified is subjected to angle correction and specific bill positioning, and the image to be tested is subjected to smoothing and filtering, as shown in fig. 2. And carrying out binarization processing on the picture to be detected, removing irrelevant factors in the picture and highlighting the bill verification pattern. The binarized image to be detected is shown in fig. 3. The size of the binarized picture to be detected is fixed, the positions of the counterfeit checking patterns in the bill are fixed, the fixed coordinates of the three counterfeit checking patterns in the bill are directly calculated, the counterfeit checking patterns are independently intercepted, and the drawings 2 to 4 are taken as examples and comprise tax patterns, bank logo patterns and bamboo patterns, wherein the intercepted bamboo patterns are shown in fig. 4.
Wherein, the binarization processing comprises the following steps: f for ultraviolet images to be verified0And (x, y) is expressed, wherein (x, y) is the coordinate of the pixel point. The value of the pixel point is composed of three channels of R, G and B, namely, any point (x) in the image0,y0) Has a value of F0(x0,y0)＝(R0,G0,B0). The three-channel ultraviolet image contains much unnecessary image information, so that the false-checking pattern in the image is not obviously distinguished from the blank background and is not obviousThe method obtains a complete verification pattern, thereby causing the verification accuracy to be reduced. Therefore, the graphic needs to be binarized to distinguish the verification pattern in the ultraviolet image from the background. Firstly, graying the ultraviolet image, and controlling the value of a pixel point to be between (0,255), specifically, the grayed image is G (x, y), and the calculation formula is as follows:
wherein R, G and B are respectively ultraviolet images F0And (x, y) three channel values of the pixel points. And then, carrying out binarization processing on the ultraviolet image after the gray level processing, wherein the processed image is represented by F (x, y), and the binarization processing formula is as follows:
t is a threshold value for distinguishing the fake-verifying pattern from the background pattern, and the distinguishing effect is optimal when t is 150 according to multiple test tests of the bill; after binarization, the picture is black and white, wherein F (x, y) is white when 1, and F (x, y) is black when 0.
Step 2, characteristic value extraction: and extracting the corner feature, the contour feature and the pixel point feature of the picture to be detected. Taking the bills shown in fig. 2 to 4 as an example, since three detected patterns required for bill verification are separated, the corner features are used for verifying the whole picture to judge that the case of no corner features of the logo patterns occurs, and at this time, if false bills without logos occur, the false bills cannot be detected. The line number pattern and the bamboo pattern have more broken lines due to the fine and complicated patterns, and the error of extracting the outline characteristics is larger. Compared with the actual production bills, the corner feature has a good detection effect on the line number patterns and the bamboo patterns, and the contour feature has a good detection effect on the line emblem patterns. Therefore, aiming at the verification patterns with different characteristics, in order to ensure that the verification patterns have no problems and high verification accuracy, the verification is carried out by adopting a multi-feature extraction method combining the angular point features, the contour features and the pixel point features. Specifically, the tax pattern and the bamboo pattern of the picture are subjected to corner feature extraction, the logo pattern is subjected to contour feature extraction, and pixel point feature extraction is performed on the whole picture.
Firstly, extracting the corner feature of a picture to be detected, and extracting the feature by adopting an SURF algorithm. The specific mathematical model is as follows:
(1) firstly, constructing a Hessian matrix, wherein the matrix form is as follows: wherein F is the image F (x, y) after binarization processing, and the discriminant is
In actual use, a gaussian second-order filter is needed for calculation, and the gaussian function is as follows:
the second order gaussian differential is:
a Hessian matrix constructed by second-order Gaussian differential is
Is recorded as:
and calculating pixel points of the whole image by using the constructed Hessian matrix to obtain a scale space layer, and modifying the size of the template of the Gaussian filter and the size of the parameter of the filter to obtain space layers with different scales.
(2) Extracting characteristic points: comparing the size of each pixel point processed by the hessian matrix with the size of a point of a 3-dimensional neighborhood, detecting by using a filter with the size corresponding to the scale layer in the comparison process, recording a 3 x 3 filter template, covering 9 pixel points by the template, keeping 27 pixel points in the three-dimensional neighborhood if the central point is the maximum value or the minimum value of the 27 points, and taking the central point as a preliminarily determined feature point.
(3) In a circular neighborhood centered on the feature point, the neighborhood radius is 6s, where s is the scale value of the feature point. Counting the sum of Haar wavelet characteristics of points in a sector of 60 degrees in the x direction and the y direction, wherein the Haar wavelet characteristics are two rectangular characteristics, and the calculated sum of the characteristics is recorded as:
wherein w is the number of feature points in the statistical region. The characteristic direction is as follows:
and then, rotating the sector at certain intervals and counting the Harr wavelet characteristic values in the region again, and finally taking the direction of the sector with the maximum value as the main direction of the characteristic point.
(4) Taking a rectangular area image with the side length of 20s along the main direction by taking the feature point as the center, dividing the area image into 16 sub-block areas, counting haar wavelet features of 25 pixels in the horizontal direction and the vertical direction in each sub-block area, and counting the sum sigma dx of values in the horizontal direction, the sum sigma dy of values in the vertical direction, the sum sigma dx of absolute values in the horizontal direction and the sum sigma dy of absolute values in the vertical direction in each sub-block in 4 directions. The 4 values are taken as the feature vector of each sub-block area, the total number is 64 dimensions, the feature vector formed by all the extracted feature points is recorded as H, the vector is a matrix with n columns and 64 rows, and n is the number of the feature points.
And secondly, extracting the outline characteristics of the picture to be detected, wherein the outline characteristics comprise the change of an outline curve of the picture and the area size of the picture. Because the logo pattern is small, the situation that the logo pattern has no corner features can occur when the corner features are extracted from the whole picture. This results in the logo being undetectable when it is incorrect. Therefore, by extracting the contour features and fusing the contour features with the corner features, a feature matrix representing all the information of the fake-checking patterns is generated. And expressing the outline characteristics of the picture to be detected by using the Hu moment. The Hu moment is 7 invariant moments constructed using second and third order normalized central moments. The standard moments of the images are:
wherein p and q represent the image as p + q order moments, p, q ═ 0,1,2. M and N are the width and height of the image respectively.
The center distance is as follows:
wherein p, q is 0,1,2.,
normalized center-to-center distance of
The 7 invariant moments constructed using the second and third order normalized central moments are:
M1＝η20+η02
M2＝(η20-η02)2+4η11 2
M3＝(η30-3η12)2+(3η21-η03)2
M4＝(η30+η12)2+(η21+η03)2
M5＝(η30-3η12)(η30+η12)((η30+η12)2-3(η21+η03)2)+(3η21-η03)(η21+η03)(3(η30+η12)2-(η21+η03)2)
M6＝(η20-η02)((η30+η12)2-(η21+η03)2)+4η11(η30+η12)(η21+η03)
M7＝(3η21-η03)(η30+η12)((η30+η12)2-3(η21+η03)2)-(η30-3η12)(η21+η03)(3(η30+η12)2-(η21+η03)2)
the 7 invariant moments as feature vectors have invariance of rotation, scaling and translation, and can reflect the outline information of the image. When some verification patterns with relatively small patterns or the verification patterns are placed incorrectly, when the corner features are extracted from the whole image, the feature matrixes representing all the verification pattern information are generated by extracting the contour features and fusing the contour features with the corner features, so that the problems can be perfectly solved.
And finally, the pixel point characteristics are the number of white pixel points of the picture to be detected after binarization, and when the characteristics assist in judging that the interior of the fake-identifying pattern is less missing, the corner point characteristics and the outline characteristics are not easy to detect. Obtaining an ultraviolet image F0(x, y) since the influence of the thermal stability of the photosensitive element of the acquisition device is different from the structure of the sensor of the device, there is a small noise interference, as shown in fig. 1, in order to eliminate the interference of the noise to the white pixel point, the noise needs to be removed. And performing morphological processing on the binarized picture F (x, y) to be detected, specifically performing expansion operation on the F (x, y) and then performing corrosion operation on the F (x, y). Wherein the dilation operation is performed on the entire picture using a 3 x 3 matrix templateF (x, y) is traversed in the directions of the transverse direction x and the vertical direction y. The template is as follows:
where the corresponding pixel point value traversed is. The expansion operation is to take the minimum value in the templates as the value of the middle point of the template for each traversed template. The erosion operation takes the maximum value of the template as the value of the middle point of the template for each traversed template. The pixel point value of a fine white noise point in the image is 1, through the operation, as the background around the white noise point is black, namely the pixel point value is 0, when the template traverses the white pixel point during expansion processing, the minimum value of 0 can be obtained due to the existence of the surrounding black pixel point, the white pixel point is changed into a black pixel point, then the expanded fake-identifying pattern is restored to the original size through corrosion operation, and the processed image is marked as F1(x, y). After removing the possible fine white noise points in the picture, calculating the whole picture to be detected to obtain the number of white pixel points, wherein the calculation method comprises the following steps
Where M and N are the width and height of the image, respectively.
Step 3, threshold discrimination: and comparing the similarity of the feature matrix of the bill to be detected with the feature matrix of the template bill to obtain the similarity values of the two feature matrices, and finally judging whether the similarity values meet the threshold requirement or not so as to judge the authenticity of the bill. Specifically, the method for determining the threshold includes the following steps: fusing the feature vectors after the three features are extracted to form a new feature vector matrix; specifically, seven matrix values M calculated from Hu moments1To M7And forming a 64-dimensional single-column feature vector by the white pixel point value S, fusing the feature vector and the angular point feature vector, and normalizing the matrix to form an n + 1-column and 64-row feature vector matrix T. Then, the Euclidean distance is used for calculating the similarity between the matrix and the feature vector matrix of the template picture, the feature vector matrix of the template picture is recorded as T ', the first column of T and T' is taken as an example, and the calculating method comprises the following steps:
where i is the number of columns of the eigenvector matrix. The calculation result of the similarity of the two matrixes is output in a one-dimensional array form and is marked as L, wherein the calculation result comprises the similarity of n +1 columns of characteristics, and a similarity threshold value is set as t1Setting a threshold value t smaller than the similarity in the array1Obtaining a new similarity matrix L', and carrying out custom weighting calculation on the similarity values meeting the similarity threshold and the number of members to obtain a value P representing the similarity of the picture, wherein the custom weighting calculation formula is as follows:
wherein j represents the number of columns of L'; p represents the probability that the bill is true, if P is more than or equal to 0.86, the bill is judged to be a true pattern, and if not, the bill is judged to be false.
Through carrying out actual detection on 100 bills, the detected bills comprise 5 false bills, and the actual false-checking test accuracy is 100%. Through tests, the method extracts the corner feature, the contour feature and the pixel point feature of the bill ultraviolet picture, has higher accuracy rate by self-defining the probability of the image being true through a multi-feature matrix fusion method, and simultaneously, because the similarity calculation is carried out by using the matrix after feature fusion, the calculation complexity is not obviously increased while the higher accuracy rate is ensured.
It should be noted that the various features described in the above embodiments may be combined in any suitable manner without departing from the scope of the invention. The invention is not described in detail in order to avoid unnecessary repetition.
Claims (10)
1. A bill verification detection method based on ultraviolet patterns is characterized by comprising the following steps:
step 1, bill pretreatment: carrying out denoising treatment and binarization treatment on an ultraviolet image to be verified, and then intercepting a verification pattern;
step 2, characteristic value extraction: extracting corner feature, contour feature and pixel point feature of the picture to be detected;
step 3, threshold discrimination: and comparing the similarity of the feature matrix of the bill to be detected with the feature matrix of the template bill to obtain the similarity values of the two feature matrices, and finally judging whether the similarity values meet the threshold requirement or not so as to judge the authenticity of the bill.
2. The ultraviolet pattern-based bill validation detection method according to claim 1, wherein the bill pretreatment method comprises the steps of:
step 11, firstly, graying the ultraviolet image, controlling the value of the pixel point between (0,255), and marking the grayed image as G (x, y), then:
wherein R, G and B are respectively ultraviolet images F0(x, y) three channel values of the pixel points;
and step 12, performing binarization processing on the ultraviolet image after the gray level processing, wherein the processed image is represented by F (x, y), and the binarization processing formula is as follows:
wherein t is a threshold value for distinguishing the fake-checking pattern from the background pattern, and t is 150 hours according to multiple test tests of the bill;
and step 13, after binarization, the picture is black and white, wherein if F (x, y) is 1, the picture is white, and if F (x, y) is 0, the picture is black.
3. The bill verification detection method based on the ultraviolet pattern as claimed in claim 1, wherein the corner feature extraction method comprises the following steps:
step 211, firstly, constructing a Hessian matrix, solving a Hessian matrix for each pixel point, and calculating a local maximum value by using a Hessian matrix discriminant;
step 212, using the local maximum point as a key point, then setting a threshold to filter out the key point with weak energy and the key point with wrong positioning, and screening out useful feature points;
step 213, taking a preset X1Y 1 rectangular area block around the feature point, wherein the direction of the rectangular area block is along the main direction of the feature point;
step 214, sampling each region block by preset size X2Y 2, and counting Haar wavelet characteristics of each region block in the horizontal direction and the vertical direction of T pixels of threshold value;
step 215, taking four values of the sum of the horizontal direction values, the sum of the vertical direction values, the sum of the absolute values in the horizontal direction and the sum of the absolute values in the vertical direction of the Haar wavelet features as the characteristic values of the subareas;
and step 216, obtaining a total 64-dimensional feature vector as a descriptor of the feature point, namely the feature vector of the corner point of the picture.
4. The ultraviolet pattern-based bill verification detection method according to claim 1, wherein the outline features comprise changes of outline curves of the images and area sizes of the images.
5. The ultraviolet pattern-based bill validation detection method according to claim 4, wherein the outline feature extraction method comprises the following steps: and expressing the profile characteristics of the picture to be detected by adopting the Hu moment, wherein the Hu moment is 7 invariant moments constructed by utilizing second-order and third-order normalized central moments.
6. The ultraviolet pattern-based bill validation detection method according to claim 1, wherein the pixel feature is the number of white pixels of the binarized image to be detected.
7. The ultraviolet pattern-based bill validation detection method according to claim 6, wherein the pixel point feature extraction method comprises the following steps: firstly, performing morphological processing on a binarized picture to be detected to remove fine white noise possibly existing in a background; and traversing the whole picture to be detected to obtain the number of white pixel points.
8. The ultraviolet pattern-based bill validation detection method according to claim 1, wherein the threshold discrimination method comprises the following steps:
step 31, fusing the feature vectors after the three features are extracted to form a new feature vector matrix;
step 32, calculating the similarity of the matrix and a feature vector matrix of the template picture by using the Euclidean distance, and outputting the calculation result of the similarity of the two matrices in a one-dimensional array form;
step 33, setting members in the array which are lower than the similarity threshold value to zero, and carrying out weighting calculation on the similarity values meeting the similarity threshold value and the number of the members to obtain a value representing the similarity of the pictures;
and step 34, if the picture similarity value meets the authenticity detection threshold value, judging the picture to be a real pattern, otherwise, judging the picture to be false.
9. The ultraviolet pattern-based bill verification detection method according to claim 1, wherein the similarity threshold value and the authenticity detection threshold value are obtained by testing bills under different environments.
10. A detection system based on the ultraviolet pattern bill verification detection method according to any one of claims 1 to 9, characterized by comprising the following modules:
the bill preprocessing module is used for denoising and binarizing the ultraviolet image to be verified, and then intercepting the verification pattern;
the characteristic value extraction module is used for extracting angular point characteristics, contour characteristics and pixel point characteristics of the picture to be detected;
and the threshold value judging module is used for comparing the similarity of the feature matrix of the bill to be detected with the feature matrix of the template bill to obtain the similarity value of the two feature matrices, and finally judging whether the similarity value meets the threshold value requirement or not so as to judge the authenticity of the bill.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011479918.0A CN112466033B (en) | 2020-12-15 | 2020-12-15 | Ultraviolet pattern based bill verification detection method and detection system thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011479918.0A CN112466033B (en) | 2020-12-15 | 2020-12-15 | Ultraviolet pattern based bill verification detection method and detection system thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112466033A true CN112466033A (en) | 2021-03-09 |
| CN112466033B CN112466033B (en) | 2022-09-02 |
Family
ID=74804386
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011479918.0A Active CN112466033B (en) | 2020-12-15 | 2020-12-15 | Ultraviolet pattern based bill verification detection method and detection system thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112466033B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103617673A (en) * | 2013-12-10 | 2014-03-05 | 江苏国光信息产业股份有限公司 | Ultraviolet image characteristic-based check true and false identification system and method |
| CN103729656A (en) * | 2014-01-22 | 2014-04-16 | 河北远东通信***工程有限公司 | Seal imprint identification method and identification system |
| CN110598699A (en) * | 2019-09-16 | 2019-12-20 | 华中科技大学 | Anti-counterfeiting bill authenticity distinguishing system and method based on multispectral image |
| CN111310628A (en) * | 2020-02-10 | 2020-06-19 | 武汉科技大学 | Paper currency forming mode inspection and identification method based on paper currency printing pattern characteristics |
| CN111986378A (en) * | 2020-07-30 | 2020-11-24 | 湖南长城信息金融设备有限责任公司 | Bill color fiber yarn detection method and system |
-
2020
- 2020-12-15 CN CN202011479918.0A patent/CN112466033B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103617673A (en) * | 2013-12-10 | 2014-03-05 | 江苏国光信息产业股份有限公司 | Ultraviolet image characteristic-based check true and false identification system and method |
| CN103729656A (en) * | 2014-01-22 | 2014-04-16 | 河北远东通信***工程有限公司 | Seal imprint identification method and identification system |
| CN110598699A (en) * | 2019-09-16 | 2019-12-20 | 华中科技大学 | Anti-counterfeiting bill authenticity distinguishing system and method based on multispectral image |
| CN111310628A (en) * | 2020-02-10 | 2020-06-19 | 武汉科技大学 | Paper currency forming mode inspection and identification method based on paper currency printing pattern characteristics |
| CN111986378A (en) * | 2020-07-30 | 2020-11-24 | 湖南长城信息金融设备有限责任公司 | Bill color fiber yarn detection method and system |
Non-Patent Citations (4)
| Title |
|---|
| 宁柏锋: "基于图像识别的移动台账核查应用研究", 《自动化与仪器仪表》 * |
| 宁柏锋: "基于图像识别的移动台账核查应用研究", 《自动化与仪器仪表》, no. 10, 25 October 2018 (2018-10-25), pages 213 * |
| 聂生显: "基于形状特征的图像检索研究和应用", 《硕士学位论文》 * |
| 聂生显: "基于形状特征的图像检索研究和应用", 《硕士学位论文》, 1 September 2016 (2016-09-01), pages 43 - 50 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112466033B (en) | 2022-09-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10885644B2 (en) | Detecting specified image identifiers on objects | |
| CN106778586B (en) | Off-line handwritten signature identification method and system | |
| CN109658584B (en) | Bill information identification method and device | |
| US7590275B2 (en) | Method and system for recognizing a candidate character in a captured image | |
| CN108596166A (en) | A kind of container number identification method based on convolutional neural networks classification | |
| CN108694393A (en) | A kind of certificate image text area extraction method based on depth convolution | |
| CN104680161A (en) | Digit recognition method for identification cards | |
| CN103914676A (en) | Method and apparatus for use in face recognition | |
| CN102789658A (en) | Ultraviolet anti-counterfeiting check authenticity verification method | |
| CN104217221A (en) | Method for detecting calligraphy and paintings based on textural features | |
| CN107067000B (en) | A kind of recognition methods of bank note degree and device | |
| CN104680130A (en) | Chinese character recognition method for identification cards | |
| CN103034848A (en) | Identification method of form type | |
| CN109409158B (en) | Anti-counterfeiting method based on two-dimensional code edge roughness | |
| CN110738216A (en) | Medicine identification method based on improved SURF algorithm | |
| CN110378351B (en) | Seal identification method and device | |
| CN107195069A (en) | A kind of RMB crown word number automatic identifying method | |
| CN107180422A (en) | A kind of labeling damage testing method based on bag of words feature | |
| CN107784263A (en) | Based on the method for improving the Plane Rotation Face datection for accelerating robust features | |
| CN113011426A (en) | Method and device for identifying certificate | |
| CN106599923B (en) | Method and device for detecting seal anti-counterfeiting features | |
| CN107742357A (en) | A kind of recognition methods of paper money number and device | |
| KR101781351B1 (en) | Forgery passport inspection method and recording medium thereof | |
| Pan et al. | An efficient method for skew correction of license plate | |
| CN110826551B (en) | Intelligent screening method for VIN code topology printing mode image of motor vehicle |
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 |