WO2012083712A1 - Paper currency identification method and device - Google Patents

Abstract

A paper currency identification method and device are provided. An anti-counterfeiting identification image of a paper currency is obtained via transmission light, and then the obtained anti-counterfeiting identification image is identified, so whether the paper currency is true or false is judged. The paper currency identification method overcomes the limitation of an algorithm treatment under reflection type imaging, and the paper currency identification method is easy to realize, low in cost, and accurate in detection, and the counterfeit money can be detected quickly.

Description

 Method and device for identifying banknotes

 The present invention relates to the field of image processing and recognition technology, and in particular to a banknote identification method and apparatus. Background technique

 Maintain financial order and the public interest, and strictly prevent counterfeit money from entering the market. The counterfeiting of banknotes is a matter of particular concern to financial management institutions. In the traditional money counters and clearing machines, the traditional methods of authenticity based on magnetism are widely used. The image detection based methods mainly focus on new and old, currency, stain, code recognition, auxiliary forgery, and auxiliary thickness detection. In addition, some of the counterfeit features, such as the 05th edition of the 100 yuan RMB on the back of the Great Hall of the People, in the infrared light, the Hall information and other feature points can be used as an auxiliary means of authentication. However, because such image security points are easily copied, their level of authentication is not high.

 At present, in the related art of image authentication, most of the counterfeiting capabilities are not reluctant, and most of them are auxiliary counterfeiting in the money counter and the sorting machine. For example, using the infrared image security of the RMB, the image of the hall in the back of the RMB version of the RMB version is half, and some counterfeit coins are available for all the information of the General Assembly Hall, or not. According to this, the counterfeiting can be performed. However, this feature is easily copied in counterfeit coins using infrared ink.

 Other image-based authentication points such as microtext or engraving printing are only suitable for static high-resolution image recognition. In high-speed banknote counters or sorters, it is difficult to capture high-definition images. It cannot be used in financial equipment for high-speed banknotes. Therefore, in the authentication, the traditional methods of magnetic authentication and fluorescence authentication are still the main methods of authentication. This type of approach is already a mature solution in the industry, but the counterfeit money of financial instruments such as counterfeit currency counterfeit detectors is also circulated in the real financial market through high-magnification information and fluorescent information.

As the counterfeit currency manufacturing technology is becoming more and more sophisticated, counterfeit currency producers are constantly updating and using the corresponding technical means to respond to the existing inspection methods. It can be said that "the road is one foot high and the magic height is one foot." In order to improve the level of counterfeiting and quickly respond to the counterfeit currency manufacturing methods that are constantly updated and developed, it is necessary to use the existing anti-counterfeiting marks or feature information on the current circulating banknotes based on the existing inspection methods, and to use new detection and authentication techniques to improve The identification performance of financial instruments, to curb the circulation of counterfeit currency, to maintain financial order and the public interest. SUMMARY OF THE INVENTION In view of the above, the present invention provides a banknote discriminating method and apparatus for obtaining banknotes by using transmitted light. An image of the security mark, and the banknote identification is performed based on the security mark. A banknote discriminating method provided by an embodiment of the present invention is provided with an image acquiring unit, a light source and a banknote passage in the banknote discriminating device for use, wherein the image acquiring unit and the light source are respectively located in the banknote passage of the banknote identifying device Side, the method includes:

 The light source illuminates the banknote to be tested during the passage of the banknote to be tested, and the image acquisition unit uses the transmitted light of the banknote to collect an anti-counterfeit logo image of a predetermined position on the banknote by using the light source. ;

 Processing the anti-counterfeit identification image collected by the collection;

 The authenticity of the banknote is determined based on the processed security image.

 The embodiment of the present invention further provides a banknote discriminating device, comprising an image acquiring unit, a light source, and a banknote passage, wherein the image acquiring unit and the light source are respectively located on two sides of the banknote passage of the banknote identification device; During the passage, the light source illuminates the banknote to be tested, and the image acquisition unit uses the transmitted light of the light source to collect an image of the anti-counterfeit mark on the banknote; the device further includes:

 An image processing unit, configured to process the collected anti-counterfeit identification image;

 And a determining unit, configured to determine the authenticity of the paper currency according to the anti-counterfeiting identification image processed by the image processing unit.

 The invention provides a banknote identification method and device, which uses the transmitted light to obtain an image of the anti-counterfeit mark on the banknote, and determines the authenticity of the banknote by identifying the acquired image of the anti-counterfeit mark. The limitations of algorithmic processing under reflective imaging can be overcome. The banknote detecting solution provided by the invention is easy to implement, has low realization cost, and the detection is fast and accurate, so that the counterfeit currency can be detected quickly, and the effect is more obvious. Instruction sheet

 1 is a flow chart of a method for authenticating a banknote according to an embodiment of the present invention;

 2 is a schematic structural view of a banknote discriminating device provided in an embodiment of the present invention;

 3 is a flowchart of processing in a method for discriminating bills according to an embodiment of the present invention; FIG. 4Aa is an image of a complementary counterprint pattern of genuine and counterfeit coins collected in an embodiment of the present invention; FIG. 4Bb is an embodiment of the present invention; The grayscale image of the complementary print pattern of the true and false coins provided in the example; FIG. 4Cc is an image of the true and false coin complementary to the ancient coins pattern after the denoising and binarization processing provided in the embodiment of the present invention;

FIG. 4Dd is a representation of the true and false coin complementary to the edge of the Indian coin pattern respectively provided in the embodiment of the present invention. Binary map

 4Ee are respectively a gray-scale binary image of a true and a counterfeit coin complementary to an ancient coin pattern provided in an embodiment of the present invention;

 Figure 5ab is a binary projection curve of the true and false coin complementary to the ancient coins pattern provided in the embodiment of the present invention. detailed description

 In view of the deficiencies in the prior art, the present invention provides a banknote discriminating method and apparatus, which uses transmitted light to acquire an image of an anti-counterfeit mark on a banknote, and determines the banknote by recognizing the acquired image of the anti-counterfeit mark. The authenticity. The limitations of algorithmic processing under reflective imaging can be overcome. The banknote detection solution provided by the invention is easy to implement, the realization cost is low, the detection is fast and accurate, and the counterfeit currency can be detected quickly, and the effect is more obvious.

 Since the genuine coin complementary printing feature is a front-back one-time printing technique, the printed pattern can be accurately overlapped. Generally, the ^_coin is printed twice on the front side and the back side, so that it is impossible to accurately overlap the complementary print pattern, and often misalignment, overlap, etc. can be detected only in the case of transmission.

 In the banknote identification device provided by the present invention, an image acquisition unit, a light source and a banknote passage are disposed in the banknote identification device, and the image acquisition unit and the light source are respectively located on both sides of the banknote passage of the banknote identification device, 1. The method includes the following steps:

 S01, using the light source to transmit the image of the anti-counterfeit mark on the banknote through the transmitted light of the banknote; in the process of the banknote to be tested passing through the banknote passage, the light source illuminates the banknote to be tested, and the image obtaining unit uses the light source to collect the banknote through the transmitted light of the banknote An image of the anti-counterfeit logo;

 Among them, the image of the anti-counterfeit mark on the banknote can be realized by the following means:

 La) determining, according to the detected banknote specification, a range of regions in which the anti-counterfeiting mark may appear; lb) controlling the light source to illuminate the range of the area on the banknote to be tested, and controlling the image collecting unit to collect an image of the range of the area.

 Or you can collect images of the anti-counterfeit logo on the banknotes, or by:

 IA) collecting a full-size image of the banknote to be tested, and positioning the anti-counterfeiting mark area according to the detected banknote specification;

 IB) Obtain an image of the anti-counterfeit identification area.

The image acquisition unit may use a contact image sensor CIS or a CMOS or a CCD. The anti-counterfeiting mark is a front-back complementary print pattern, and may also be other marks such as a watermark (image or number), a holographic magnetic security line, and the like. 502, processing the anti-counterfeit identification image collected by the collection;

 Since the image collected by the 釆 has noise influence or the background has color, it is necessary to perform filtering processing, morphological denoising, etc. on the anti-counterfeit logo image; and/or

 Perform image binarization and the like on the anti-counterfeit logo image.

 503. Determine the authenticity of the banknote according to the processed anti-counterfeiting logo image.

 Specifically, for some anti-counterfeiting marks, such as the front and back complementary printing patterns, the overlapping effect cannot be seen under reflection, and the recognition and judgment cannot be performed, but in the transmission, the images formed by the overlapping of the front and back sides can be easily collected. The authenticity of the banknote can be judged based on the image formed by the coincidence. For example, for the ancient coins in the 05 version of the coin, the ancient coins can be printed on the back and back, and the ancient coins can be formed according to the collected and processed images.

 Further, a detection module may be provided at the entrance of the banknote passage of the banknote discriminating device, and when it is detected that the banknote passes through the entrance of the banknote passage, the light source is turned on, and the image acquisition unit is triggered to start collecting images on the banknote. This controls the operating state of the relevant functional unit of the bill discriminating device.

 Referring to Fig. 2, the present invention further provides a banknote discriminating device 200 having an image capturing unit 10, a light source 20, and a banknote passage 30. The banknote detecting device further includes:

 The image acquisition unit 10 and the light source 20 are respectively located on both sides of the banknote passage 30 of the banknote identification device. During the process of the banknotes to be tested passing through the banknote passage, the light source 20 illuminates the banknote to be tested, and the image acquisition unit 10 transmits through the light source 20. The transmitted light of the banknote collects an image of the anti-counterfeit mark on the banknote; the image processing unit 40 is configured to process the collected anti-counterfeit mark image;

 Since the image collected by the 釆 has noise influence or the background has color, it is necessary to perform filtering processing, morphological denoising, etc. on the anti-counterfeit logo image; and/or

 Perform image binarization processing on the anti-counterfeit logo image.

 The determining unit 50 is configured to determine the authenticity of the banknote according to the anti-counterfeiting logo image processed by the image processing unit.

 The image acquisition unit 10 uses a contact image sensor CIS or CMOS or CCD.

 The anti-counterfeiting logo is a complementary printed image of the offset back, specifically, the fifth set of coins used in the renminbi. The determining unit forms a complete graphic based on whether the printed coins are coincident in the processed image.

 The bill discriminating device 200 further includes:

The sensing unit 60a is configured to detect the banknote specification to determine a range of areas in which the banknote security mark may appear; In a specific embodiment, an infrared pair tube (or other device) can be used as the sensing unit.

 The control unit 70 is configured to control the light source to illuminate the area of the area on the banknote to be tested, and control the image collection unit to collect an image of the area range.

 The image collection unit collects a full-size image of the banknote to be tested; the device further includes:

 a positioning unit 80, configured to locate an anti-counterfeit identification area based on the full-size image;

 The image acquisition unit 10 acquires an image of the security feature based on the area determined by the positioning unit 80.

 The image processing unit 40 specifically includes:

 The first processing module 40a performs filtering processing and morphological denoising processing on the image of the area where the anti-counterfeiting mark is located; and/or

 The second processing module 40b performs image binarization on the image of the area where the anti-counterfeiting mark is located; and/or

 The third processing module 40c is configured to remove fixed noise.

 The banknote discriminating device provided by the invention further comprises:

 The detecting unit 60b is disposed in the banknote passage for detecting the banknote entering the banknote passage; when detecting that the banknote passes through the banknote passage entrance, the detecting unit 60b issues an indication signal, turns on the light source, and triggers the image acquiring unit 10 to start 釆An image on the banknote is collected.

 In a specific embodiment of the present invention, the image of the anti-counterfeit mark on the banknote can be realized in different ways. For example, the full-size image of the banknote to be tested can be collected first, and then the area of the pseudo-marker is located in the full-size image. , then intercept the image of the anti-counterfeit logo. Alternatively, it is also possible to directly determine the range of areas in which the anti-counterfeiting mark may appear based on the detected banknote specifications, and then control the image of the range of the anti-counterfeit mark area.

 In a specific embodiment of the present invention, an image obtained by intercepting an anti-counterfeit mark from a full-size image is used. Referring to FIG. 3 and FIG. 4, an example of a coin-complementing coin-type coin pattern is taken as an example. The specific processing flow is as follows:

5301, the authentication device is activated, and the banknote is started;

 5302, collecting a full image of the banknote to be tested;

 The image acquisition unit (such as CIS, CMOS or CCD) and the light source are respectively located on both sides of the banknote passage of the banknote identification device, and the light source transmits the full image of the banknote through the transmitted light of the banknote.

 5303, detecting the collected image, positioning the position of the crown number to determine the banknote facing;

The position of the banknote can be determined based on the front and back of the crown number, thereby determining the position of the offset print pattern (i.e., the ancient coin pattern shown in Figs. 4A and 4a).

5304, the crown number coordinate is used as a reference point for the offset printing on the printed pattern, and the complementary offset printing is intercepted. Like

 Obtaining a complementary offset image based on the determined position of the offset printing pattern;

 5305, converting the complementary offset image into a grayscale image, extracting the edge of the complementary offset image, and performing morphological denoising;

 The grayscale images of the true and false coins complementary to the ancient coins pattern are shown in Figures 4B and 4b, respectively.

 5306, binarizing the complementary offset image, and detecting the offset pattern according to the edge information; FIG. 4C and FIG. 4c are images of the true and false coin complementary to the ancient coin pattern after denoising and binarization respectively, and comparing Out, the obvious difference between true and false coins.

 5307, Is the detection of the complementary offset printing pattern qualified?

 It can be seen from the comparison that the true and false coins have obvious differences. The complementary offset ancient coin pattern of the genuine coin shown in Fig. 4D is complete and clear, and the offset printing of the genuine coin shown in Fig. 4d is vague and incomplete.

 5308, Is the registration information based on grayscale binary image and grayscale image normal?

 According to the binarization process, whether the ancient coin graphics in the image coincide to form a complete ancient coin graphic. As can be seen from Figures 4E and 4e, the difference between true and false coins.

 Based on the detected offset complementary printed pattern (ie, ancient coin pattern) outline, image registration recognition based on gray scale and binary map based can be used.

 Based on the binary image, the projected image is projected according to the symmetry direction of the pattern, and the registration detection is performed according to the symmetry of the projected image. For example, for the ancient coin map of the renminbi, after obtaining the gradation-based binary map, it can be projected simultaneously in the horizontal direction and the vertical direction, since the ancient coins are symmetrical in both directions. Therefore, the projection map is symmetrical based on the center position of the pattern, as shown in Fig. 5a, so that the left and right data are subtracted correspondingly at the center point for such features, and the absolute values of all subtraction results are summed. The residuals that can be obtained are much smaller in the real currency than the counterfeit currency, because the real coins are symmetrical, and the counterfeit projections are asymmetrical based on the center of the pattern, as shown in Figure 5b.

Based on the gray scale similarity calculation, the center point of the pattern can already be positioned from above. From the direction of the symmetry axis, the real coins will be basically symmetrical, while the 4 renminbi cannot be overlapped due to printing misalignment and the like, and is not completely symmetrical. Therefore, after positioning the symmetry axis, the left and right halves are cropped, and the mature gradation-based template matching technique can be used for matching. Such as image difference sum (SAD), image difference sum of squares (SSD), normalized cross-correlation coefficient (NCC) and other methods. Symmetrical images for SAD and SSD will be much smaller than the values for asymmetric images. The more symmetric the NCC, the closer the image value is to 1. Therefore, based on the calculation results, the experience threshold is set. You can distinguish between genuine and counterfeit currency. In another embodiment of the present invention, the image of the anti-counterfeit mark (e.g., the complementary pair of coins of the renminbi) is collected directly according to the range of areas in which the detected banknote security mark may appear. The specific processing flow is as follows:

 The image position of the area where the anti-counterfeiting mark is located (vertical direction, the position in the horizontal direction is determined according to the denomination and the face of the banknote), and the processing can be speeded up.

 In the process of image collection in the area where the anti-counterfeiting mark is located, the position of the set can be set by software without changing the hardware structure. According to the information-oriented, the area where the anti-counterfeiting mark is located is used as the coarse positioning position. . Then, the image is collected, and the image of the image is binarized and morphologically filtered. The printed pattern is accurately positioned according to the processed image. The symmetry of the pattern is then used to register in the binary image and the grayscale image, thereby identifying whether the banknote is a genuine coin or a counterfeit currency.

 The invention provides a banknote identification method and device, which uses the transmitted light to obtain an image of the anti-counterfeit mark on the banknote, and determines the authenticity of the banknote by identifying the acquired image of the anti-counterfeit mark. The limitations of algorithmic processing under reflective imaging can be overcome. The banknote detecting solution provided by the invention is easy to implement, has low realization cost, and the detection is fast and accurate, so that the counterfeit currency can be detected quickly, and the effect is more obvious.

 The transmission method used in the present invention performs counterfeiting, and the hardware cost is reduced relative to the reflective image analysis, making it easier to productize. The invention provides a new anti-counterfeiting means, provides a new means of authentication, increases the strength of the authentication, and combines the existing authentication means to further prevent the circulation of counterfeit currency from the root. Because complementary printing technology is a high-end technology in printing technology, and the requirements for printing presses are very high. It is difficult for counterfeit currency to achieve the same effect as the true coin in the degree of coincidence of the chroma and the complementary image. The counterfeit currency is difficult to achieve complete registration, and generally there will be misalignment and overlap.

 The processing algorithm provided by the invention has universality, and the complementary printing technology is an anti-counterfeiting technology commonly used in various countries, and the implementation characteristics are similar. Therefore, the algorithm is similar in implementation, and it is easier to modularize when upgrading, and the upgrade is more convenient. It is easier to reach the requirement of identifying multiple currencies in one machine. The present invention can be implemented or used by those skilled in the art in light of the disclosed embodiments. Various modifications to these embodiments are obvious to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the scope and spirit of the invention. The above-mentioned embodiments are merely preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalents, improvements, etc., which are included in the spirit and principles of the present invention, should be included in Within the scope of protection of the present invention.

Claims

Rights request A method for discriminating a banknote, characterized in that an image acquisition unit, a light source and a banknote passage are provided in the banknote discriminating device for use, wherein the image acquisition unit and the light source are respectively located in the banknote passage of the banknote identification device Side, the method includes:  The light source illuminates the banknote to be tested during the passage of the banknote to be tested, and the image acquisition unit uses the transmitted light of the banknote to collect an anti-counterfeit logo image of a predetermined position on the banknote by using the light source. ;  Processing the anti-counterfeit identification image collected by the collection;  The authenticity of the banknote is determined based on the processed security image.  The method for discriminating the banknote according to claim 1, wherein the image of the anti-counterfeit mark on the banknote comprises:  Determining, according to the detected banknote specification, a range of regions in which the anti-counterfeiting identifier may appear; controlling the image collection unit to collect an image of the region range.  The method for discriminating the banknote according to claim 1, wherein the image of the anti-counterfeit mark on the banknote comprises:  Collecting a full-size image of the banknote to be tested, positioning the anti-counterfeit identification area according to the detected banknote specification; and acquiring an image of the anti-counterfeit identification area.  The method for discriminating the banknote according to claim 1, wherein the processing of the image of the area in which the anti-counterfeiting mark is performed comprises:  Perform filtering processing and morphological denoising on the image of the area where the anti-counterfeiting mark is located; and/or perform image binarization on the image of the area where the anti-counterfeiting mark is located.  The method for identifying a banknote according to claim 1, further comprising:  The banknote is detected to pass through the banknote passage, the light source is turned on, and the image acquisition unit is triggered to start collecting images on the banknote.  The method for discriminating a banknote according to any one of claims 1 to 5, wherein the anti-counterfeiting mark is a front-back complementary print pattern, and the true color of the banknote is determined according to the processed anti-counterfeit mark image Pseudo, specifically including:  A determination is made based on whether or not the printed patterns are coincident in the processed image to form a complete pattern. 7. A banknote discriminating device, comprising: an image acquisition unit, a light source, and a banknote passage, wherein the image acquisition unit and the light source are respectively located on two sides of the banknote passage of the banknote identification device; During the process of measuring the banknote passing through the banknote passage, the light source illuminates the banknote to be tested, and the image acquisition unit uses the transmitted light of the banknote to collect an image of the anti-counterfeit mark on the banknote;  The device also includes:  An image processing unit, configured to process the collected anti-counterfeit identification image;  And a determining unit, configured to determine the authenticity of the banknote according to the anti-counterfeiting identification image processed by the image processing unit.  8. The bill discriminating device according to claim 7, further comprising: a sensing unit configured to detect the banknote specifications to determine a range of regions in which the banknote anti-counterfeiting label may appear;  And a control unit, configured to control the light source to illuminate the area of the area on the banknote to be tested, and control the image collection unit to collect an image of the area range.  9. The banknote discriminating device according to claim 7, wherein the image collecting unit collects a full-size image of the banknote to be tested; the device further comprises:  a positioning module, configured to locate an anti-counterfeit identification area based on the full-size image;  An image acquisition module is configured to acquire an image of the anti-counterfeit identification area.  10. The banknote detecting device according to claim 7, wherein the image processing unit comprises:  a first processing module, performing filtering processing and morphological denoising processing on the image of the area where the anti-counterfeiting mark is located; and/or  The second processing module performs image binarization processing on the image of the area where the anti-counterfeiting mark is located; and the third processing module is configured to remove fixed noise.  11. The banknote discriminating device according to claim 7, further comprising: a detecting unit disposed at the entrance of the banknote passage for detecting the banknote entering the banknote passage; when detecting that the banknote passes through the banknote passage entrance, The detecting unit issues an indication signal to turn on the light source and trigger the image acquisition unit to start collecting images on the banknote.  The method for discriminating a banknote according to any one of claims 7 to 10, wherein the anti-counterfeiting mark is a front-back complementary pair printing pattern, and the determining unit determines whether the printed image is in accordance with the processed image. Coincidence forms a complete graph for decision.