Banknote verification method of banknote counter based on edge positioning

Abstract

The present invention discloses a banknote verification method based on edge positioning. The present invention relates to the technical field of banknote detection and solves the technical problems of lacking a systematic analysis process for different detection results, being unable to comprehensively evaluate the authenticity of banknotes, and having insufficient reliability and stability in banknote verification when facing complex and diverse counterfeiting means. By sequentially performing ultraviolet fluorescence detection and infrared induction detection and combining with a systematic analysis of the detection results, the present invention can more comprehensively and accurately judge the authenticity of banknotes. In the infrared induction detection, the infrared edge features are analyzed from two aspects of stability and regularity, improving the accuracy of banknote verification, reducing the misjudgment and missed judgment rates. For abnormal comparison results, not only the relative position relationship is judged by calculating the distance between the abnormal edge features and the normal edge features, but also the texture features of the abnormal edges are deeply analyzed to effectively identify finely forged counterfeit banknotes and enhance the ability of banknote verification technology to cope with complex counterfeiting means.

Description

Technical Field

[0001] The present invention relates to the technical field of banknote detection, and specifically to a banknote verification method for a banknote detector based on edge positioning. Background Art

[0002] A banknote detector is a machine that verifies the authenticity of banknotes and counts the number of banknotes. Due to the large scale of cash circulation and the heavy workload of cash handling at bank teller counters, a banknote counting machine has become an indispensable device. A banknote detector is mainly used for counting, authenticating, and sorting cash, and is widely used in various financial industries and various enterprises and institutions with cash flow.

[0003] According to the patent application with the publication number CN118470849A, a banknote verification method and system for a banknote detector based on image processing are disclosed. Among them, the banknote verification method based on image processing includes steps such as preliminary identification, establishing a scanning area, image acquisition, image processing, mapping, feature point extraction, inclination calculation, primary correction calculation, secondary identification, verification passed, and verification failed; the banknote verification system includes an identification module, a scanning module, an acquisition module I, a processing module, a mapping module, an acquisition module II, a calculation module I, a calculation module II, an identification module II, and an output module.

[0004] Traditional banknote verification methods mainly rely on manual experience and simple physical feature recognition, such as observing watermarks, security threads, fluorescent patterns, etc. However, with the continuous upgrading of counterfeiting techniques, these methods are difficult to meet the accurate authentication requirements. Currently, image recognition technology is gradually applied in the field of banknote verification, using technologies such as edge detection and multispectral analysis to improve the accuracy of banknote verification. However, the existing technology still has deficiencies in comprehensively and accurately analyzing by integrating multiple detection means, and it is difficult to effectively identify complex counterfeiting means. Summary of the Invention

[0005] In view of the deficiencies of the prior art, the present invention provides a banknote verification method for a banknote detector based on edge positioning, which solves the problems of lacking a systematic analysis process for different detection results, being unable to comprehensively evaluate the authenticity of banknotes, and having insufficient reliability and stability in banknote verification when facing complex and diverse counterfeiting means.

[0006] To achieve the above objectives, the present invention is realized through the following technical solutions: A banknote verification method for a banknote detector based on edge positioning, which specifically includes the following steps:

[0007] Use the Canny algorithm to perform edge detection on the obtained banknote image, extract edge features, and the edge features obtained here include characters, patterns, shapes, and lines;

[0008] Perform ultraviolet fluorescence detection on the extracted edge features, compare the obtained fluorescent edge features with the comparison features of genuine banknotes, and generate abnormal or normal comparison results;

[0009] If the comparison result is normal, perform infrared induction detection on the edge of the banknote image to obtain the infrared edge features and their signal intensities. Judge the distribution of the infrared edge features through a bar chart to generate stable or unstable signals. At the same time, select a group of infrared edge features, compare their variation rules with the signal intensity of the overall image edge, and generate regular or irregular variation signals. If both stable and regular variation signals are obtained, it is determined that the banknote verification is normal, and a normal banknote verification signal is generated; otherwise, an abnormal banknote verification signal is generated.

[0010] If the comparison result is abnormal, obtain the corresponding abnormal edge features, select a group of normal edge features as the target object, find the two closest points between the abnormal and normal edge features as feature points, calculate the distance between the two points and compare it with the standard distance to generate a secondary analysis signal.

[0011] Obtain the secondary analysis signal, obtain all the pixel points of the image corresponding to the abnormal edge features, and obtain the corresponding pixel points of the pixel points. At the same time, obtain the gray level after image quantization, construct a gray-level co-occurrence matrix, and calculate the texture correlation value corresponding to the abnormal edge features. Then compare it with the preset value to generate a normal or abnormal banknote verification signal.

[0012] As a further solution of the present invention, the specific method for performing ultraviolet fluorescence detection on the extracted edge features is as follows:

[0013] Label the obtained edge features as i, and i = 1, 2,..., j. Here, i represents the type label of the edge features, and j represents the specific type corresponding to the edge features. Then perform ultraviolet fluorescence detection on the obtained edge features. Specifically, irradiate the banknote with ultraviolet light to obtain the fluorescent edge features.

[0014] As a further solution of the present invention, the specific method for generating abnormal or normal comparison results is as follows:

[0015] Obtain the fluorescent edge corresponding to the genuine banknote as the comparison feature, and compare it with the fluorescent edge features. If the fluorescent edge features are the same as the comparison feature, a normal comparison result is generated; otherwise, an abnormal comparison result is generated.

[0016] As a further solution of the present invention, the specific method for judging the distribution of infrared edge features through a bar chart to generate stable or unstable signals is as follows:

[0017] Perform infrared induction detection on the edge features to obtain the infrared edge features. Obtain all the infrared edge features and label them as n, and n = 1, 2,..., m, where m is the type of infrared edge features.

[0018] Obtain the edge signal intensity corresponding to the infrared edge feature n and denote it as Kn. Based on the height of the bar chart, judge the distribution of the infrared edge feature, obtain the height difference of the edge signal intensity corresponding to the genuine banknote, and compare the two. If the two are the same, generate a stable signal; otherwise, if the two are different, generate an unstable signal. Specifically, the same here means that the difference between the two is within the preset range. Otherwise, if it is not within the preset range, it means they are different, and the value of the preset range is set by the operator based on a large amount of data.

[0019] As a further solution of the present invention, the specific method for generating a regular or irregular change signal is as follows:

[0020] Obtain the edge signal intensity Kn corresponding to the infrared edge feature n. At the same time, take one group of infrared edge features as the analysis object, obtain the edge signal intensity Kn of the analysis object and the change rule of the analysis object, and compare the two. If the change rules of the two are the same, it means that the edge signal intensity changes regularly, and a regular change signal is generated; otherwise, if the change rules of the two are different, it means that the edge signal intensity changes irregularly, and an irregular change signal is generated.

[0021] As a further solution of the present invention, the specific method for generating a secondary analysis signal is as follows:

[0022] Obtain any group of normal edge features and denote them as the target object. Establish a corresponding rectangular coordinate system with the center point of the current image, and represent the abnormal edge features and the target object in the form of coordinates. At the same time, determine the feature points corresponding to the abnormal edge features, and here the feature points are the points closest to the target object, and obtain the coordinate positions corresponding to the feature points. Similarly, obtain the coordinate positions of the feature points corresponding to the target object. Then, according to the coordinate distance formula Calculate the distance between the two feature points, and at the same time compare the distance with the standard distance value. If the distance is the same as the standard distance value, it means that the relative position of the abnormal edge feature meets the requirements, and a secondary analysis signal is generated; otherwise, if the distance is different from the standard distance value, it means that the relative position of the abnormal edge feature does not meet the requirements, and a banknote verification anomaly signal is generated.

[0023] As a further solution of the present invention, the specific method for constructing the gray-level co-occurrence matrix is as follows:

[0024] Obtain the abnormal edge features corresponding to the secondary analysis signal, and at the same time obtain the corresponding image. Then, obtain the pixel points in the image and denote them as a, where a = 1, 2,..., b, and b represents the number of pixel points. Represent the pixel point a as (i a , j a ). Then, obtain the preset direction and distance, and obtain the pixel point corresponding to the pixel point a, denoted as the point to be analyzed ( , ), where represents the coordinate offset from the current pixel (i a , j a ) to another relevant pixel in the vertical direction (row direction), represents the coordinate offset from the current pixel (i a , j a ) to another relevant pixel in the horizontal direction (column direction), and and are determined by the direction and distance, and the number of occurrences of the gray value combination of these two pixels is counted;

[0025] Obtain the gray level L after image quantization, and construct a gray level co-occurrence matrix P of L×L. The element P(i a , j a ) in the matrix represents the frequency of occurrence of the pixel pair a with gray values i and j under a given direction and distance.

[0026] As a further solution of the present invention, the specific method for generating the banknote verification normal or abnormal signal is:

[0027] According to the formula calculate the correlation value COR of the image, where and are the means of the gray values i and j respectively, and are the standard deviations of the gray values i and j respectively. At the same time, compare the calculated correlation value COR with a preset value;

[0028] If the correlation value COR is greater than the preset value, it means that the texture feature of the current image has a high correlation with the texture feature of the genuine banknote image, and a banknote verification normal signal is generated. On the contrary, if the correlation value COR is less than the preset value, it means that the texture feature of the current image has a low correlation with the texture feature of the genuine banknote image, and a banknote verification abnormal signal is generated.

[0029] The present invention provides a banknote verification method for a banknote detector based on edge positioning. Compared with the prior art, it has the following beneficial effects:

[0030] Through ultraviolet fluorescence detection and infrared induction detection in sequence, and combined with systematic analysis of the detection results, the present invention can more comprehensively and accurately judge the authenticity of banknotes. The edge features are carefully labeled and compared with the fluorescence edge features of genuine banknotes. In infrared induction detection, the infrared edge features are analyzed from two aspects of stability and regularity, greatly improving the accuracy of banknote verification, reducing the misjudgment and missed judgment rates. For abnormal comparison results, not only the relative position relationship is judged by calculating the distance between abnormal edge features and normal edge features, but also the texture features of abnormal edges are deeply analyzed. For example, the correlation value is calculated using the gray-level co-occurrence matrix and compared with the preset value to effectively identify counterfeit banknotes with sophisticated forgery methods, enhancing the ability of banknote verification technology to cope with complex forgery means. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a flowchart of the method steps of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0032] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0033] Example 1, please refer to Figure 1 , the present application provides a banknote verification method for a banknote verification machine based on edge positioning, and the method specifically includes the following steps:

[0034] Step S1: The banknote verification machine obtains an image of the banknote through an optical sensor, and at the same time uses an image processing algorithm to perform edge detection on the obtained image, extracts the edge features corresponding to the image, and the edge features obtained here include characters, patterns, shapes, and lines.

[0035] Specifically, when using an image processing algorithm (such as the Canny algorithm) for processing, first convert the color image into a grayscale image, use a Gaussian filter to smooth the grayscale image to reduce the noise in the image, calculate the gradients of the image in the horizontal and vertical directions through the Sobel operator, and then obtain the amplitude and direction of the gradients. The amplitude represents the intensity of the edge, and the direction represents the orientation of the edge. In the gradient amplitude image, check the neighboring pixels in the gradient direction of each pixel point. If the gradient amplitude of this pixel point is not the local maximum, set it to 0 to refine the edge. Set two thresholds (a low threshold and a high threshold), determine the pixel points with a gradient amplitude greater than the high threshold as strong edges, eliminate the pixel points with a gradient amplitude less than the low threshold, and retain the pixel points between the two if they are connected to the strong edges, otherwise eliminate them, and further obtain the edge features of the image, which are represented by the following code:

[0036] import cv2 import numpy as np

[0037] import matplotlib.pyplot as plt

[0038] # Read the image

[0039] image = cv2.imread('your_image.jpg')

[0040] # 1. Grayscale the image

[0041] gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

[0042] # 2. Gaussian smoothing

[0043] blurred = cv2.GaussianBlur(gray, (5, 5), 0)

[0044] # 3. Edge detection using the Canny algorithm

[0045] # Here, the low threshold is set to 50 and the high threshold is set to 150. You can adjust according to the actual situation

[0046] edges = cv2.Canny(blurred, 50, 150)

[0047] # Display the original image and the edge detection result

[0048] plt.subplot(121),plt.imshow(cv2.cvtColor(image,cv2.COLOR_BGR2RGB))

[0049] plt.title('Original Image'), plt.xticks([]), plt.yticks([])

[0050] plt.subplot(122), plt.imshow(edges, cmap='gray')

[0051] plt.title('Edge Image'), plt.xticks([]), plt.yticks([])

[0052] plt.show()

[0053] Step S2: Detect the obtained edge features through ultraviolet fluorescence detection. By comparing with the edge features of genuine banknotes, normal comparison results and abnormal comparison results are generated.

[0054] Specifically, label the obtained edge features as i, where i = 1, 2, …, j. Here, i represents the type label of the edge features, and j represents the specific type corresponding to the edge features. Then, perform ultraviolet fluorescence detection on the obtained edge features. Specifically, irradiate the banknote with ultraviolet light. The fluorescent anti-counterfeiting marks on the banknote will emit fluorescence of specific colors and intensities under the excitation of ultraviolet light. For example, in the case of the Chinese yuan, fluorescent patterns at certain specific positions will clearly appear under ultraviolet light, and their edges are clear and regular, obtaining fluorescent edge features. At the same time, obtain the corresponding fluorescent edges of genuine banknotes as comparison features. The fluorescent edges of genuine banknotes are stored in the banknote verification system of the banknote checker from the beginning. Here, the comparison features are of the same type as the obtained image features. Then, compare the two.

[0055] If the fluorescent edge features are the same as the comparison features, it indicates that the fluorescent edge features are initially shown to be normal, and a normal comparison result is generated. If the fluorescent edges are different from the comparison features, it indicates that the fluorescent edge features are initially shown to be abnormal, and an abnormal comparison result is generated.

[0056] Step S3: Analyze the obtained normal comparison results, and combine with the infrared edges obtained through infrared induction detection. The specific analysis method is as follows:

[0057] While performing ultraviolet fluorescence detection on the edge features, perform infrared induction detection on the edge features to obtain infrared edge features. Specifically, irradiate the banknote with infrared light. Due to the different absorption and reflection characteristics of different materials of the banknote to infrared light, the edge of the banknote will present a unique infrared image under infrared light. Then, analyze the stability and regularity of the infrared edge features. First, obtain all the infrared edge features and label them as n, where n = 1, 2, …, m, and m is the type of infrared edge features.

[0058] The specific method for analyzing the stability of infrared edge features is as follows: Obtain the edge signal intensity Kn corresponding to the infrared edge feature n, and at the same time use a bar chart to represent the edge signal intensity Kn. Specifically, use the number n of the infrared edge feature as the abscissa and the corresponding edge signal intensity Kn as the ordinate to draw a bar chart, and judge the distribution of the infrared edge feature based on the height of the bar chart. Obtain the height difference of the edge signal intensity corresponding to the genuine banknote, and compare it with the height difference corresponding to the current infrared edge feature. If the two are the same, it means that the distribution of the infrared edge feature is stable and a stable signal is generated. On the contrary, if the two are different, it means that the distribution of the infrared edge feature is unstable and an unstable signal is generated. Specifically, the same here means that the difference between the two is within a preset range. On the contrary, if it is not within the preset range, it means they are different, and the value of the preset range is set by the operator based on a large amount of data.

[0059] The specific method for analyzing the regularity of infrared edge features is as follows: Obtain the edge signal intensity Kn corresponding to the infrared edge feature n, and at the same time take one group of infrared edge features as the analysis object to obtain the change rule of the edge signal intensity Kn of the analysis object. At the same time, obtain the change rule of the corresponding genuine banknote of the analysis object. Generally speaking, for parts related to the anti-counterfeiting design of banknotes, such as specific ink printing areas, the edge intensity of their infrared signals will show specific change rules, such as gradually weakening from the center to the edge, etc., and compare the two. If the change rules of the two are the same, it means that the edge signal intensity changes regularly and a regular change signal is generated. On the contrary, if the change rules of the two are different, it means that the edge signal intensity changes irregularly and an irregular change signal is generated.

[0060] For example, measure the signal intensity Kn of the infrared edge feature of the security thread at different scanning angles (scanning once every 10° from 0° to 360°). After data analysis, it is found that the signal intensity Kn of the security thread of the genuine banknote shows a periodic change rule: when the scanning angle is near 30°, 120°, 210°, and 300°, the signal intensity reaches the peak value, and the peak intensity is between 80 - 90; when the scanning angle is near 60°, 150°, 240°, and 330°, the signal intensity reaches the valley value, and the valley intensity is between 20 - 30, and the change period is 90°. Record this rule in the genuine banknote change rule database.

[0061] Analyze the stability and regularity of the infrared edge features comprehensively. If both are satisfied, and here it means generating a stable signal and a regular change signal, it means that the result of the current banknote verification is normal and a banknote verification normal signal is generated. On the contrary, if any one of the two is not satisfied, it means that the result of the current banknote verification is abnormal and a banknote verification abnormal signal is generated.

[0062] Step S4: Analyze the obtained abnormal comparison results, determine the abnormal edge features through comparison, obtain any set of normal edge features as the target object, and analyze the correlation between the target object and the abnormal edge features. The specific analysis method is as follows:

[0063] Then, establish a corresponding rectangular coordinate system with the center point of the current image, represent the abnormal edge features and the target object in the form of coordinates, and determine the feature points corresponding to the abnormal edge features. Here, the feature points are the points closest to the target object, and obtain the coordinate positions corresponding to the feature points. Similarly, obtain the coordinate positions of the feature points corresponding to the target object. Then, according to the coordinate distance formula Calculate the distance between the two feature points, and compare the distance with the standard distance value. The standard distance value is calculated and determined from the distances of multiple genuine banknotes. If the distance is the same as the standard distance value, it means that the relative position of the abnormal edge features meets the requirements, and a secondary analysis signal is generated. On the contrary, if the distance is different from the standard distance value, it means that the relative position of the abnormal edge features does not meet the requirements, and a banknote verification abnormal signal is generated. Specifically, various edge features of genuine banknotes have relatively fixed positional relationships in space. For example, the edge of the serial number on the RMB should maintain a relatively fixed spacing and positional relationship with the edge of the surrounding decorative pattern. If the edge of the serial number appears abnormal and the spacing from the edge of the decorative pattern suddenly becomes larger or smaller, further analysis is required.

[0064] Then, process the generated secondary analysis signal by judging the texture features of the abnormal edge features. When judging the texture features, analyze by establishing a corresponding gray-level co-occurrence matrix. The specific analysis method is as follows:

[0065] Obtain the abnormal edge features corresponding to the secondary analysis signal, and at the same time obtain the corresponding image. Then, obtain the pixel points in the image and denote them as a, where a = 1, 2,..., b, and b represents the number of pixel points. Represent the pixel point a as (i a , j a ). Then, obtain the preset direction and distance, and obtain the pixel point corresponding to the pixel point a, denoted as the point to be analyzed ( , ), where represents the coordinate offset from the current pixel (i a , j a ) to another related pixel in the vertical direction (row direction), represents the coordinate offset from the current pixel (i a , j a ) to another related pixel in the horizontal direction (column direction), and and Determined by direction and distance, and count the number of occurrences of the gray value combinations of these two pixels;

[0066] Obtain the gray level L after image quantization, and construct a gray level co-occurrence matrix P of L×L. The element P(i a , j a ) represents the frequency of occurrence of pixel pairs a with gray values i and j at a given direction and distance. For example, if the pixel pair with gray values 5 and 7 in the image appears 10 times at the specified direction and distance, then P(5, 7) = 10. After traversing all pixels, a complete gray level co-occurrence matrix is obtained. Then, according to the formula Calculate the correlation value COR of the image, where and are the means of gray values i and j respectively, and are the standard deviations of gray values i and j respectively. At the same time, compare the calculated correlation value COR with a preset value, and the specific value of the preset value is calculated according to the gray values corresponding to the genuine banknote image;

[0067] If the correlation value COR is greater than the preset value, it means that the texture feature of the current image has a high correlation with the texture feature of the genuine banknote image, and a normal banknote verification signal is generated. On the contrary, if the correlation value COR is less than the preset value, it means that the texture feature of the current image has a low correlation with the texture feature of the genuine banknote image, and an abnormal banknote verification signal is generated.

[0068] Example 2. This example is implemented on the basis of Example 1, and the difference from Example 1 is as follows:

[0069] When analyzing the abnormal edge texture feature in this example, by calculating the contrast of the abnormal edge image and analyzing and judging through the contrast, a normal banknote verification signal and an abnormal banknote verification signal are generated, and the specific calculation formula is .

[0070] Example 3. This example is implemented on the basis of Example 1, and the difference from Example 1 is as follows:

[0071] When analyzing the abnormal edge texture feature in this example, by calculating the entropy of the abnormal edge image and analyzing and judging through the calculated entropy, a normal banknote verification signal and an abnormal banknote verification signal are generated, and the specific calculation formula is .

[0072] For some data in the above formulas, only their numerical values are taken for calculation, and parameter units are not substituted for calculation. At the same time, the content not described in detail in this specification belongs to the prior art well-known to those skilled in the art.

[0073] The above embodiments are only used to illustrate the technical method of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A banknote verification method based on edge positioning, characterized in that, The method specifically includes the following steps: Use the Canny algorithm to perform edge detection on the obtained banknote image and extract edge features; Perform ultraviolet fluorescence detection on the extracted edge features, compare the obtained fluorescent edge features with the comparison features of genuine banknotes, and generate abnormal or normal comparison results; If the comparison result is normal, perform infrared induction detection on the edge of the banknote image, obtain the infrared edge features and their signal intensities, judge the distribution of the infrared edge features through a bar chart, generate a first signal of stable or unstable signal, and at the same time select a group of infrared edge features, compare its change rule with the signal intensity change rule of the overall image edge, generate a second signal of regular or irregular change signal. If both a stable first signal and a regularly changing second signal are obtained, it is determined that the banknote verification is normal and a banknote verification normal signal is generated, otherwise a banknote verification abnormal signal is generated; If the comparison result is abnormal, obtain the corresponding abnormal edge features, select a group of normal edge features as the target object, find the two closest points between the abnormal and normal edge features as feature points, calculate the distance between the two points and compare it with the standard distance to generate a secondary analysis signal; Obtain the secondary analysis signal, obtain all pixel points of the image corresponding to the abnormal edge features, and obtain the corresponding pixel points of the pixel points. Specifically, when constructing the gray-level co-occurrence matrix, for the current pixel point, according to the preset direction and distance parameters, obtain the target pixel points in its neighborhood, and at the same time obtain the gray levels after image quantization, construct the gray-level co-occurrence matrix, and calculate the texture-related value corresponding to the abnormal edge features, and compare it with the preset value to generate a banknote verification normal or abnormal signal.

2. The banknote verification method of the banknote counter based on edge positioning according to claim 1, wherein The specific method for performing ultraviolet fluorescence detection on the extracted edge features is as follows: Label the obtained edge features as i, and i = 1, 2,..., j. Here, i represents the type label of the edge features, and j represents the specific type corresponding to the edge features. Then perform ultraviolet fluorescence detection on the obtained edge features. Specifically, irradiate the banknote with ultraviolet light to obtain fluorescent edge features.

3. The banknote verification method of a banknote counter based on edge positioning according to claim 1, wherein The specific method for generating abnormal or normal comparison results is as follows: Obtain the fluorescent edge corresponding to the genuine banknote as the comparison feature, and compare it with the fluorescent edge features. If the fluorescent edge features are the same as the comparison features, a normal comparison result is generated, otherwise, an abnormal comparison result is generated.

4. The banknote verification method of a banknote counter based on edge positioning according to claim 1, wherein The specific method for generating a first signal of stable or unstable signal by judging the distribution of infrared edge features through a bar chart is as follows: Perform infrared induction detection on the edge features to obtain infrared edge features. Obtain all the infrared edge features and label them as n, and n = 1, 2,..., m, where m is the type of infrared edge features; Obtain the edge signal intensity corresponding to the infrared edge feature n as Kn, and judge the distribution of the infrared edge features based on the height of the bar chart. Obtain the height difference of the edge signal intensity corresponding to the genuine banknote, and compare the two. If the two are the same, a stable signal is generated, otherwise, if the two are different, an unstable signal is generated.

5. The banknote verification method of a banknote counter based on edge positioning according to claim 1, characterized in that, The specific method for generating a second signal of regular or irregular change signal is as follows: Obtain the edge signal intensity Kn corresponding to the infrared edge feature n. At the same time, take one group of infrared edge features as the analysis object, obtain the edge signal intensity Kn of the analysis object and the variation law of the analysis object, and compare the two. If the variation laws of the two are the same, generate a regular variation signal; otherwise, if the variation laws of the two are different, generate an irregular variation signal.

6. The banknote verification method of a banknote counter based on edge positioning according to claim 1, characterized in that The specific method for generating the secondary analysis signal is as follows: Obtain any set of normal edge features as the target object, establish a corresponding rectangular coordinate system with the center point of the current image, and represent the abnormal edge features and the target object in the form of coordinates. At the same time, determine the feature points corresponding to the abnormal edge features, and obtain the coordinate positions corresponding to the feature points. Similarly, obtain the coordinate positions of the feature points corresponding to the target object. Then, according to the coordinate distance formula Calculate the distance between two feature points. The feature point coordinates (x1, y1) represent the points on the abnormal edge feature, specifically the horizontal and vertical coordinates of the abnormal edge feature points, and (x2, y2) represent the nearest points on the normal edge feature, specifically the horizontal and vertical coordinates of the normal edge feature points. At the same time, compare the distance with the standard distance value. If the distance is the same as the standard distance value, generate a secondary analysis signal. Otherwise, if the distance is different from the standard distance value, generate a counterfeit detection anomaly signal.

7. The banknote verification method of a banknote counter based on edge positioning according to claim 1, characterized in that, The specific method for constructing the gray-level co-occurrence matrix is as follows: Obtain the abnormal edge features corresponding to the secondary analysis signal, and at the same time obtain the corresponding image. Then, obtain the pixel points in the image and denote them as a, where a = 1, 2, …, b, and b represents the number of pixel points. The pixel point a is represented as (i a , j a ). Then, obtain the preset direction and distance, and obtain the pixel point corresponding to the pixel point a, denoted as the point to be analyzed ( , ). represents the coordinate offset in the vertical direction, represents the coordinate offset in the horizontal direction, and count the number of times the gray value combination of these two pixels appears; Obtain the gray level L after image quantization, and construct a gray-level co-occurrence matrix P of L×L. The element P(i a , j a ) in the matrix represents the frequency of occurrence of pixel pairs with gray values i and j at a given direction and distance.

8. The banknote verification method of a banknote counter based on edge positioning according to claim 1, characterized in that, The specific method for generating the banknote verification normal or abnormal signal is as follows: According to the formula calculate the correlation value COR of the image, where and are the mean values of the gray values i and j respectively, and are the standard deviations of the gray values i and j respectively, L is the gray level after image quantization, P(i, j) represents the joint probability of the pixel pairs where the pixel with gray value i and its corresponding pixel with gray value j appear. At the same time, compare the calculated correlation value COR with the preset value; If the correlation value COR is greater than the preset value, generate a banknote verification normal signal; otherwise, if the correlation value COR is less than the preset value, generate a banknote verification abnormal signal.

Images