Damaged banknote processing methods, apparatus, computer equipment, and storage media

Abstract

This application relates to a method, apparatus, computer equipment, and storage medium for processing damaged banknotes, applicable to the financial technology field or other related fields. The method includes: acquiring multiple feature images of the damaged banknote under different image types; determining the denomination and version of the damaged banknote based on the multiple feature images; and determining local feature images of multiple local regions of the damaged banknote corresponding to the regions in the feature images of each image type from the multiple feature images; for each local region, comparing the local feature image of the local region under each image type with the reference local feature image of the same region of the banknote corresponding to the denomination and version under the same image type to obtain matching information for each local region under each image type; and determining the authenticity identification result of the damaged banknote based on the matching information of each local region under each image type. This method can achieve automatic identification of the authenticity of damaged banknotes, improving the efficiency of counterfeit detection.

Description

Technical Field

[0001] This application relates to the field of financial technology or other related fields, and in particular to a method, apparatus, computer equipment, storage medium and computer program product for processing damaged currency. Background Technology

[0002] Damaged banknotes refer to banknotes that are torn, missing, or damaged due to natural wear and tear, corrosion, resulting in changes in appearance and texture, discoloration, unclear patterns, or compromised security features, rendering them unsuitable for continued circulation. To avoid losses from damaged banknotes, holders can exchange them at banks.

[0003] Currently, banks mainly rely on manual labor to identify damaged banknotes, which results in a high time consumption and low efficiency. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for processing damaged banknotes, addressing the technical problems of the excessive time and low efficiency in the aforementioned damaged banknote identification process.

[0005] Firstly, this application provides a method for processing damaged banknotes. The method includes:

[0006] Multiple feature images of the damaged banknotes are collected; these multiple feature images correspond to different image types.

[0007] Based on the multiple feature images, the denomination version corresponding to the damaged banknote is determined, and from the multiple feature images, the local feature images of the corresponding regions of multiple local areas of the damaged banknote in feature images of various image types are determined;

[0008] For each local region, the local feature image of the local region under each image type is compared with the reference local feature image of the same region of the banknote corresponding to the denomination under the same image type to obtain the matching information of each local region with the reference local feature image under each image type.

[0009] Based on the matching information of each local region under each image type, the authenticity identification result for the damaged coin is determined.

[0010] In one embodiment, the damaged banknote corresponds to multiple denominations. After determining the denomination of the damaged banknote, the method further includes:

[0011] Determine the similarity probability between the damaged coin and each denomination version;

[0012] The method further includes:

[0013] The denomination with the highest similarity probability is selected as the current denomination.

[0014] For each local region, the local feature image of the local region under each image type is compared with the reference local feature image of the same region of the banknote corresponding to the current denomination under the same image type, so as to obtain the matching information of each local region with the reference local feature image under each image type.

[0015] If the damaged banknote is determined to be counterfeit based on the matching information, the denomination with the second highest similarity probability is selected as the new denomination. The process of comparing the local feature images of the local area under each image type with the baseline local feature images of the same area of ​​the banknote corresponding to the current denomination under the same image type is repeated until the damaged banknote is determined to be genuine or all denominations have been compared, thus obtaining the authenticity identification result for the damaged banknote.

[0016] In one embodiment, determining the denomination version of the damaged banknote based on the plurality of feature images includes:

[0017] Based on the degree of correlation between each feature image and the face value version, the target feature image is determined from the multiple feature images;

[0018] Based on the target feature image, determine the denomination version corresponding to the damaged coin.

[0019] In one embodiment, determining the authenticity of the damaged banknote based on the matching information of each local region under each image type includes:

[0020] The influence factors of each image type on the authenticity identification results of the damaged banknotes are obtained; the influence factors are obtained by training the genuine and counterfeit labels of the sample damaged banknotes and the sample feature images of the sample damaged banknotes under each image type.

[0021] Using the aforementioned influence factors, the matching information of each local region under various image types is corrected to obtain the corrected matching information of each local region under various image types.

[0022] Based on the corrected matching information of each local region under each image type, the authenticity identification result for the damaged coin is determined.

[0023] In one embodiment, determining the authenticity identification result for the damaged coin based on the corrected matching information of each local region under each image type includes:

[0024] Statistical matching information is obtained by statistically processing the corrected matching information of each local region under various image types.

[0025] If the statistical matching information is greater than or equal to a preset threshold, the damaged coin is determined to be genuine; if the statistical matching information is less than the preset threshold, the damaged coin is determined to be counterfeit.

[0026] In one embodiment, the acquisition of multiple feature images of the damaged banknotes includes:

[0027] Collect multiple initial feature images of the damaged coin;

[0028] Each initial feature image is processed to obtain the multiple feature images of the damaged coin.

[0029] Secondly, this application also provides a damaged currency processing device. The device includes:

[0030] The acquisition module is used to acquire multiple feature images of the damaged banknotes; the multiple feature images correspond to different image types;

[0031] The determining module is used to determine the denomination version of the damaged banknote based on the multiple feature images, and to determine the local feature images of the corresponding regions of multiple local regions of the damaged banknote in feature images of various image types from the multiple feature images.

[0032] The matching module is used to compare the local feature image of each local region under each image type with the reference local feature image of the same region of the banknote corresponding to the denomination under the same image type for each local region, so as to obtain the matching information of each local region with the reference local feature image under each image type.

[0033] The identification module is used to determine the authenticity of the damaged coin based on the matching information of each local region under each image type.

[0034] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0035] Multiple feature images of the damaged banknotes are collected; these multiple feature images correspond to different image types.

[0036] Based on the multiple feature images, the denomination version corresponding to the damaged banknote is determined, and from the multiple feature images, the local feature images of the corresponding regions of multiple local areas of the damaged banknote in feature images of various image types are determined;

[0037] For each local region, the local feature image of the local region under each image type is compared with the reference local feature image of the same region of the banknote corresponding to the denomination under the same image type to obtain the matching information of each local region with the reference local feature image under each image type.

[0038] Based on the matching information of each local region under each image type, the authenticity identification result for the damaged coin is determined.

[0039] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0040] Multiple feature images of the damaged banknotes are collected; these multiple feature images correspond to different image types.

[0041] Based on the multiple feature images, the denomination version corresponding to the damaged banknote is determined, and from the multiple feature images, the local feature images of the corresponding regions of multiple local areas of the damaged banknote in feature images of various image types are determined;

[0042] For each local region, the local feature image of the local region under each image type is compared with the reference local feature image of the same region of the banknote corresponding to the denomination under the same image type to obtain the matching information of each local region with the reference local feature image under each image type.

[0043] Based on the matching information of each local region under each image type, the authenticity identification result for the damaged coin is determined.

[0044] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0045] Multiple feature images of the damaged banknotes are collected; these multiple feature images correspond to different image types.

[0046] Based on the multiple feature images, the denomination version corresponding to the damaged banknote is determined, and from the multiple feature images, the local feature images of the corresponding regions of multiple local areas of the damaged banknote in feature images of various image types are determined;

[0047] For each local region, the local feature image of the local region under each image type is compared with the reference local feature image of the same region of the banknote corresponding to the denomination under the same image type to obtain the matching information of each local region with the reference local feature image under each image type.

[0048] Based on the matching information of each local region under each image type, the authenticity identification result for the damaged coin is determined.

[0049] The aforementioned method, apparatus, computer equipment, storage medium, and computer program product for processing damaged banknotes determine the corresponding denomination and version of the damaged banknote through its feature image. Then, the damaged banknote is compared with the reference local feature image of the banknote corresponding to that denomination and version. Based on the obtained matching information, the authenticity of the damaged banknote is determined. This enables automatic identification of the authenticity of damaged banknotes without human intervention, thereby improving the efficiency of counterfeit detection. Furthermore, by comparing multiple local areas of the damaged banknote with the reference local feature image, instead of comparing the complete feature image of the damaged banknote with the reference feature image, comparison time can be further saved, improving the efficiency of counterfeit detection. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the damaged coin processing system in one embodiment;

[0051] Figure 2 This is a schematic diagram of a damaged coin placement device in one embodiment;

[0052] Figure 3 This is a flowchart illustrating a damaged coin processing method in one embodiment;

[0053] Figure 4 This is a schematic diagram of the complete process of the damaged coin processing method in another embodiment;

[0054] Figure 5 This is a structural block diagram of a damaged coin processing device in one embodiment;

[0055] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0057] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0058] refer to Figure 1 This is a schematic diagram illustrating the structure of a damaged coin processing system according to an exemplary embodiment. Figure 1As shown, the system includes an interaction unit 110, an access unit 120, a scanning unit 130, a control unit 140, an evaluation unit 150, a marking unit 160, and a storage unit 170, wherein...

[0059] Interactive unit 110 is used to interact with users, guiding them to pre-sort the damaged coins, deposit the sorted coins, and providing feedback such as images, text, and receipts for users to choose whether to exchange them. (See attached image / receipt) Figure 2 The diagram shown illustrates a damaged coin placement facility. Damaged and soiled coins should be sorted and placed in advance according to requirements. Figure 2 In the diagram, 201 is the top cover, 202 is the damaged or soiled coin, and 203 is the bottom cover. The top cover 201 and the bottom cover 203 should be able to work in conjunction with the scanning unit 130 while fixing and flattening the damaged or soiled coin. Optionally, the top cover and the bottom cover can be made of a flat and transparent material such as glass or plastic.

[0060] The storage and retrieval unit 120 is used to receive the sorted damaged coins deposited by the user and return the damaged coins to the user when the user chooses not to exchange them.

[0061] The scanning unit 130 is used to scan the damaged banknotes and record their current status information, including but not limited to front and back photos, dimensions, visible light reflection images, visible light transmission images, infrared reflection images, infrared transmission images, ultraviolet reflection images, ultraviolet transmission images, fluorescent images, magnetic images, magnetic features of the security thread, optically variable printing images, optical features of the security thread (coating), finely cut-out images, electrical features, spectral absorption features, transparent window features, watermark features, and serial number, etc.

[0062] Visible light reflective graphics refer to patterns and text produced during the banknote printing process that are visible to the naked eye under visible light illumination. Visible light transmission graphics refer to patterns and text produced during the banknote printing process that are visible to the naked eye when viewed against the light. Infrared reflective graphics refer to patterns and text produced during the banknote printing process that have absorption or transparency effects when illuminated from the front by an infrared light source. Infrared transmission graphics refer to patterns and text produced during the banknote printing process that have absorption or transparency effects when viewed from the back by an infrared light source. Fluorescent graphics refer to patterns and text produced during the banknote printing process that can radiate light in other wavelength ranges when excited by a light source of a specific wavelength. Magnetic graphics refer to patterns and text with magnetic characteristics produced during the banknote printing process. Magnetic characteristics of the security thread refer to the magnetic distribution characteristics of the security thread produced during the banknote printing process. Optically variable printing graphics refer to patterns and text that appear in different colors to the naked eye at different angles under natural light illumination produced during the banknote printing process. Security thread (film) optical features refer to the optical characteristics of the security thread or film under specific light sources, generated during the banknote printing process. Fine perforated graphics refer to fine patterns and text that have a perforated effect under perspective, generated during the banknote printing process. Electrical features refer to characteristics exhibiting electrical response, generated during the banknote printing process. Spectral absorption features refer to the absorption characteristics of different light spectra, generated during the banknote printing process. Transparent window features refer to features visible to the naked eye, creating a transparent window effect, generated during the banknote printing process. Watermark features refer to graphics or text with light and shadow effects visible when viewed against the light, generated during the banknote substrate manufacturing process. Serial number refers to the serial number printed on the surface of the banknote, generated during the banknote printing process.

[0063] The control unit 140 is used to control the coordinated operation of the other modules, such as the interaction unit 110, the access unit 120, the scanning unit 130, the evaluation unit 150, the marking unit 160, and the storage unit 170.

[0064] The evaluation unit 150 employs image recognition and artificial intelligence technologies to authenticate and evaluate damaged banknotes, determining whether they are genuine and whether they meet exchange standards, and estimating their redeemable value based on relevant standards. The evaluation unit 150 can be deployed within the system or remotely (e.g., on a server) and communicate with the system at the front end via a network.

[0065] The marking unit 160 is used to mark the damaged banknotes according to relevant requirements when the damaged banknotes are determined to be genuine and the user is willing to exchange them, or when the damaged banknotes are counterfeit and should be immediately recycled and not allowed to circulate.

[0066] The storage unit 170 is used to collect and store damaged banknotes in accordance with relevant requirements when the damaged banknotes are determined to be genuine and the user is willing to exchange them, or when the damaged banknotes are counterfeit and should be immediately recycled and not allowed to circulate.

[0067] It should be noted that the above description is merely illustrative and should not be construed as limiting the scope of this application. The aforementioned technical solutions may have alternatives such as remote manual assistance, network connection between the control unit and a backend server, manual or automatic distribution of the upper and lower covers before pre-sorting, manual or automatic completion of pre-sorting work, mortise and tenon joints between the upper cover 201 and the lower cover 203, hinges, or fixing devices, etc. Therefore, equivalent variations made to the solution in this application still fall within the scope of this application.

[0068] In one embodiment, such as Figure 3 As shown, a method for processing damaged banknotes is provided, which can be applied to... Figure 1 Taking evaluation unit 150 as an example, the following steps are included:

[0069] Step S310: Collect multiple feature images of the damaged coin; the multiple feature images correspond to different image types.

[0070] Each feature image corresponds to an image type. The image type may include at least one of the following: front and back photographs of damaged banknotes, external dimensions, visible light reflectance images, visible light transmission images, infrared reflectance images, infrared transmission images, ultraviolet reflectance images, ultraviolet transmission images, fluorescent images, magnetic images, magnetic features of security threads, optically variable ink images, optical features of security threads (with film), finely cut-out images, electrical features, spectral absorption features, transparent window features, watermark features, and serial number features.

[0071] In specific implementation, it can be achieved through... Figure 1 The scanning unit 130 in the middle scans the damaged banknote and acquires the feature images of the damaged banknote under various image types, thereby obtaining multiple feature images of the damaged banknote.

[0072] Step S320: Based on multiple feature images, determine the denomination version of the damaged coin, and determine the local feature images of the corresponding regions of multiple local areas of the damaged coin in the feature images of each image type from the multiple feature images.

[0073] The local area refers to the representative region identified from the damaged coins that can characterize the features of the damaged coins.

[0074] The denomination and version of damaged banknotes include both the face value and the version of that face value. The face value represents the banknote's face value. The version refers to differences in the year and design of the same denomination, including variations in pattern, design, and color. For example, a 20-yuan banknote may have versions from 1999 and 2005, among others.

[0075] In practice, since the denomination and version represent the differences in both the face value and the printing method of a banknote, the face value and version can be distinguished based on the characteristics of the banknote. Therefore, the denomination and version of a damaged banknote can be determined by pre-scanning multiple feature images containing characteristic information of the damaged banknote. More specifically, since not all of the scanned feature images can be used to determine the denomination and version, and some feature images may not reflect the denomination and version information of the damaged banknote, after obtaining multiple feature images of the damaged banknote, it is necessary to filter the multiple feature images to obtain target feature images associated with the denomination and version, and determine the corresponding denomination and version of the damaged banknote based on the target feature images.

[0076] In addition, after obtaining multiple feature images of the damaged banknote, it is necessary to select multiple local regions from different positions on the damaged banknote. For each local region, the local feature image of the corresponding region in each feature image is determined. That is, the local feature image of each local region will be determined in each feature image.

[0077] Step S330: For each local region, compare the local feature image of the local region under each image type with the reference local feature image of the same region of the banknote corresponding to the denomination under the same image type to obtain the matching information of each local region with the reference local feature image under each image type.

[0078] In the specific implementation, suppose there are N feature images of the scanned damaged banknote (denoted as 1, 2, ..., N), and M local regions are identified from the damaged banknote (denoted as 1, 2, ..., M). Then each local region will correspond to N local feature images, that is, local feature images of N image types. Each feature image will identify M local feature images. For each local region, the local feature images of the local region under each image type are compared with the reference local feature images of the same region of the banknote corresponding to the denomination under the same image type. This will give the matching information between each local region and the reference local feature images under each image type.

[0079] For example, for a local region M, let the local feature images of this local region under various image types be: M1, M2, ..., M N Suppose the damaged banknotes are of the 20 yuan (2020 edition). Obtain the baseline local feature images B1, B2, ..., B of the same region for the corresponding banknotes of the 20 yuan (2020 edition) under the same image type. NThe local feature images of region M under various image types are compared with the baseline local feature images of the corresponding denominations of the 20-yuan banknote (2020 edition) under various image types. Specifically, local feature image M1 of the same image type is compared with the baseline local feature image B1, local feature image M2 of the same image type is compared with the baseline local feature image B2, and so on. N Compared with the baseline local feature image B N By comparing the results, we can obtain the matching information of each local region with the corresponding baseline local feature image under each image type.

[0080] Step S340: Determine the authenticity recognition result for the damaged coin based on the matching information of each local region under each image type.

[0081] In practice, after obtaining the matching information of each local region with the corresponding reference local feature image under each image type, the matching information can be statistically processed to obtain the statistical matching information of the damaged banknote and the banknote corresponding to the denomination of the damaged banknote. The statistical matching information is compared with a preset threshold. If the statistical matching information is greater than or equal to the preset threshold, the damaged banknote is determined to be genuine; if the statistical matching information is less than the preset threshold, the damaged banknote is determined to be counterfeit.

[0082] Furthermore, if the damaged banknote is determined to be genuine, computer image analysis is used to determine the remaining width and location of the damage, and the exchangeable value of the damaged banknote is calculated according to the exchange rules.

[0083] In the above-mentioned damaged banknote processing method, firstly, multiple feature images of the damaged banknote under different image types are acquired. Based on these feature images, the denomination and version of the damaged banknote are determined. Then, from the multiple feature images, local feature images of multiple local regions of the damaged banknote corresponding to the regions in the feature images of each image type are identified. For each local region, the local feature image of the local region under each image type is compared with the baseline local feature image of the same region of the banknote corresponding to the denomination and version under the same image type to obtain matching information between each local region and the baseline local feature image under each image type. Finally, based on the matching information of each local region under each image type, the authenticity identification result of the damaged banknote is determined. This method determines the denomination and version of a damaged banknote by analyzing its feature images. Then, it compares the damaged banknote with the reference local feature images of the banknote corresponding to that denomination and version. Based on the matching information, it determines the authenticity of the damaged banknote. This method enables automatic identification of the authenticity of damaged banknotes without human intervention, thereby improving the efficiency of counterfeit detection. Furthermore, by comparing multiple local areas of the damaged banknote with the reference local feature images, instead of comparing the complete feature images of the damaged banknote with the reference feature images, it can further save comparison time and improve the efficiency of counterfeit detection.

[0084] In an exemplary embodiment, there are multiple denomination versions corresponding to the above-mentioned damaged banknotes. After determining the denomination version corresponding to the damaged banknotes in step S320, the method further includes: determining the similarity probability between the damaged banknotes and each denomination version.

[0085] The method further includes:

[0086] Step S321: Select the denomination version with the highest similarity probability as the current denomination version;

[0087] Step S322: For each local region, compare the local feature image of the local region under each image type with the reference local feature image of the same region of the banknote corresponding to the current denomination under the same image type to obtain the matching information of each local region with the reference local feature image under each image type.

[0088] Step S323: If the damaged banknote is determined to be counterfeit based on the matching information, the denomination version with the second highest similarity probability is selected as the new denomination version. The process of comparing the local feature images of the local area under each image type with the benchmark local feature images of the same area of ​​the banknote corresponding to the current denomination version under the same image type is repeated until the damaged banknote is determined to be genuine or all denomination versions have been compared to obtain the authenticity identification result for the damaged banknote.

[0089] In practice, because damaged banknotes are incomplete or soiled, the feature images collected may be incomplete, making it difficult to accurately determine the denomination and version of the banknote. Therefore, multiple denominations and versions may be identified for the damaged banknote. In this case, it is necessary to determine the similarity probability between the damaged banknote and each denomination and version, i.e., the probability that the damaged banknote belongs to each denomination and version. Then, when determining the authenticity of the damaged banknote, it is compared with the baseline feature images of the banknotes corresponding to each denomination and version in descending order of similarity probability to determine the authenticity identification result.

[0090] More specifically, the denomination with the highest similarity probability is first selected as the current denomination. Then, for each local area, the local feature image of that local area under each image type is compared with the baseline local feature image of the same area of ​​the banknote corresponding to the current denomination under the same image type. This yields matching information between each local area and the baseline local feature image under each image type. If the damaged banknote is determined to be counterfeit based on the matching information, the denomination with the second highest similarity probability is selected as the new denomination, and the process returns to step S322. This process continues until the damaged banknote is determined to be genuine or all denominations have been compared, resulting in a genuine / counterfeit identification result for the damaged banknote. If all denominations have been compared and the damaged banknote is still determined to be counterfeit, then the genuine / counterfeit identification result for the damaged banknote is confirmed as counterfeit.

[0091] In practical applications, after obtaining the matching information of each local region with the reference local feature image under various image types, statistical processing (e.g., summation) can be performed on the matching information of each local region with the reference local feature image under various image types to obtain statistical matching information. If the statistical matching information is greater than or equal to a threshold, the damaged coin is determined to be genuine; otherwise, it is determined to be counterfeit. Let R be the matching information of each local region. ij Threshold is the threshold value, then if If the damaged coin is identified as genuine, it is determined to be genuine; otherwise, it is identified as counterfeit. Here, M represents the number of local regions identified from the damaged coin, and N represents the number of feature images of the damaged coin collected.

[0092] In this embodiment, considering the case where there are multiple denominations of the damaged banknote, a method is provided to determine the authenticity of the damaged banknote in this situation. The authenticity of the damaged banknote is determined in descending order of the similarity probability between each denomination and the damaged banknote. When the damaged banknote is determined to be counterfeit based on one denomination, the authenticity is determined by the next denomination. By combining the identification results of multiple denominations, the authenticity of the damaged banknote can be determined, thereby improving the accuracy of the determined authenticity of the damaged banknote.

[0093] In an exemplary embodiment, the determination of the denomination version of the damaged coin based on multiple feature images in step S320 can be achieved through the following steps:

[0094] Step S3201: Based on the correlation between each feature image and the face value version, determine the target feature image from multiple feature images;

[0095] Step S3202: Determine the denomination version of the damaged coin based on the target feature image.

[0096] There are multiple target feature images.

[0097] In practice, not all of the multiple feature images obtained from scanning can be used to determine the denomination version. Some feature images may not reflect the denomination version information of the damaged banknote. Therefore, it is necessary to first determine the degree of correlation between each feature image and the denomination version, that is, the degree of influence of each feature image on the determination result of the denomination version. Then, based on the degree of correlation, the target feature image is determined from the multiple feature images. Based on the target feature image, the denomination version corresponding to the damaged banknote is determined.

[0098] More specifically, for example, features that affect the recognition result of the face value version include color, size, text, numbers, printed faces, and the position of anti-counterfeiting mark points. The target feature image can be determined from multiple feature images based on whether these features are included.

[0099] In this embodiment, a target feature image is determined from multiple feature images by analyzing the correlation between each feature image and its denomination. Then, based on the target feature image, the denomination corresponding to the damaged banknote is determined, so as to facilitate accurate identification of the authenticity of the damaged banknote based on its denomination.

[0100] In an exemplary embodiment, in step S340 above, the determination of the authenticity identification result for the damaged coin based on the matching information of each local region under each image type can be specifically implemented in the following way:

[0101] Step S3401: Obtain the influence factors of each image type on the recognition results of the authenticity of damaged banknotes; the influence factors are obtained by training the authenticity labels of the sample damaged banknotes and the sample feature images of the sample damaged banknotes under each image type.

[0102] Step S3402: By using the influence factor, the matching information of each local region under each image type is corrected to obtain the corrected matching information of each local region under each image type.

[0103] Step S3403: Determine the authenticity recognition result for the damaged coin based on the corrected matching information of each local region under each image type.

[0104] Among them, the influence factor can represent the degree of influence of image type on the results of identifying the authenticity of damaged banknotes.

[0105] In practice, multiple damaged banknotes can be pre-acquired as sample damaged banknotes, and sample feature images of these banknotes under various image types can be obtained. Based on the authenticity labels of the sample damaged banknotes and the sample feature images under various image types, training is performed to obtain the influence factors of the sample feature images under each image type on the authenticity recognition results of the damaged banknotes. For each image type, the influence factor of that image type is multiplied by the matching information of each local region under that image type to correct the matching information of each local region under that image type, resulting in corrected matching information for each local region under each image type. Statistical processing is then performed on the corrected matching information of each local region under each image type, and the authenticity recognition result for the damaged banknotes is determined based on the obtained statistical matching information.

[0106] For example, if we denote the matching information corresponding to each local region as R ij The impact factor is K. N Then, the corrected matching information obtained by correcting the matching information of each local region under various image types through the influence factor can be expressed as: R ij K N .

[0107] In this embodiment, the influence factors of each image type on the authenticity recognition result of the damaged banknote are obtained through sample damaged banknote training. Then, the matching information of each local region under each image type is corrected according to the influence factors. Based on the corrected matching information of each local region under each image type, the authenticity recognition result of the damaged banknote is determined, thereby improving the accuracy of the determined authenticity recognition result of the damaged banknote.

[0108] In an exemplary embodiment, step S3403 above, determining the authenticity identification result for damaged currency based on the corrected matching information of each local region under each image type, can be achieved through the following steps:

[0109] Step S3403a: Perform statistical processing on the corrected matching information of each local region under each image type to obtain statistical matching information;

[0110] In step S3403b, if the statistical matching information is greater than or equal to a preset threshold, the damaged coin is determined to be genuine; if the statistical matching information is less than the preset threshold, the damaged coin is determined to be counterfeit.

[0111] In practice, the threshold can also be obtained by training the genuine and counterfeit labels of the damaged banknotes and the sample feature images of the damaged banknotes under various image types. Statistical processing is performed on the corrected matching information of each local region under various image types; that is, the corrected matching information of each local region under various image types is summed, and the sum of the corrected matching information is used as statistical matching information. Then, based on the comparison between the statistical matching information and the preset threshold, the genuine and counterfeit identification result of the damaged banknote is determined.

[0112] For example, if we denote the matching information corresponding to each local region as R ij The impact factor is K. N Let Threshold be the threshold value, then the relationship for determining the authenticity of damaged coins can be expressed as:

[0113]

[0114] In this embodiment, statistical matching information is obtained by statistically processing the corrected matching information of each local region under various image types. Based on the comparison results of the statistical matching information and the preset threshold, the authenticity of the damaged banknote is determined. This method performs matching through local regions without matching the entire image of the damaged banknote. On the other hand, it uses the corrected matching information of multiple image types to jointly determine the authenticity of the damaged banknote. This ensures both the efficiency and accuracy of the authenticity identification results.

[0115] In an exemplary embodiment, the acquisition of multiple feature images of the damaged coin in step S310 above can be achieved through the following steps:

[0116] Step S3101: Collect multiple initial feature images of the damaged coin;

[0117] Step S3102: Perform a reorganization process on each initial feature image to obtain multiple feature images of the damaged coin.

[0118] In practice, because damaged banknotes often have wrinkles and curls after use, even if they can be placed stably and relatively flat... Figure 2 The upper cover 201 and the lower cover 203 shown are not completely flat and can not be perfectly aligned. Therefore, the initial feature image obtained from the initial scan may be an image with wrinkles, curling, etc. So after collecting multiple initial feature images of the damaged banknote, it is necessary to process each initial feature image through computer image processing to obtain an image when the damaged banknote is completely aligned and unfolded, which can be used as multiple feature images for subsequent identification of the authenticity of the damaged banknote.

[0119] In this embodiment, by regularizing the damaged banknotes, the resulting multiple feature images of the damaged banknotes are images of the banknotes when they are completely flat and unfolded. This avoids the problem that some features of the damaged banknotes cannot be identified due to wrinkles, curling, etc., thereby improving the accuracy of subsequent face value identification and authenticity identification results of the damaged banknotes.

[0120] In one embodiment, to facilitate understanding of the embodiments of this application by those skilled in the art, the following will be described in conjunction with the appendix. Figure 4 Specific examples will be provided. (Refer to...) Figure 4 This diagram illustrates a complete process flow for a damaged coin processing method. It includes the following steps:

[0121] Step S401: Guide the user through the interaction unit 110 to sort out the damaged and soiled coins in advance, so that the single damaged and soiled coin is placed stably and relatively flat between the upper cover 201 and the lower cover 203.

[0122] Step S402: Guide the user through the interaction unit 110 to put the pre-organized top cover 201, damaged and soiled coins and bottom cover 203 into the storage unit 120.

[0123] Step S403: Under the control of the control unit 140, the damaged and soiled banknotes are scanned by the scanning unit 130 to record their current status information, including but not limited to front and back photos, dimensions, visible light reflection images, visible light transmission images, infrared reflection images, infrared transmission images, ultraviolet reflection images, ultraviolet transmission images, fluorescent images, magnetic images, magnetic features of the security thread, optically variable printing images, optical features of the security thread (coating), finely cut-out images, electrical features, spectral absorption features, transparent window features, watermark features, serial number, etc.

[0124] Step S404: The evaluation unit 150 performs authentication and evaluation on the current status information of the damaged and defaced banknotes to determine whether they are genuine and whether they meet the exchange standards.

[0125] Step S405: Using image recognition and artificial intelligence technologies, the counterfeit and damaged banknotes are identified and evaluated to determine whether they are genuine.

[0126] Step S406: If it is genuine currency, the evaluation conclusion of the evaluation unit 150 is displayed through the interaction unit 110, such as the specific terms of the exchange method and the redeemable value, and the user can choose whether to exchange it.

[0127] Step S407: If it is counterfeit money, under the control of the control unit 140, the marking unit 160 marks it, it is recycled to the storage unit 170, and the interactive unit 110 provides feedback information such as pictures, text and receipts to the user.

[0128] Step S408: The user chooses whether to redeem.

[0129] Step S409: If the user chooses not to exchange, under the control of the control unit 140, the upper cover 201, the damaged coin and the lower cover 203 are returned to the user through the storage and retrieval unit 120, and information such as pictures, text and receipts are fed back to the user through the interaction unit 110.

[0130] Step S410: If the user chooses to exchange, under the control of the control unit 140, the marking unit 160 marks the coin and stores the upper cover 201, the damaged coin, and the lower cover 203 together in the storage unit 170, and provides the user with information such as pictures, text, and receipts.

[0131] Furthermore, step S404 can be further refined into the following steps:

[0132] Step S4041: Assume that a total of N image MAP sets (N greater than 1) were collected in step S403; although a single damaged banknote is placed stably and relatively flat between the upper cover 201 and the lower cover 203, it cannot be completely flat against the upper cover 201 and the lower cover 203. The N images can be regularized by computer image processing to obtain N image MAP sets 2 when the damaged banknote is completely flat and stretched.

[0133] Step S4042: Identify key information (such as color, size, text, numbers, printed faces, and the location of anti-counterfeiting marks) from N images of the damaged or defaced banknotes using computer image recognition.

[0134] Step S4043: Determine the set of face value versions corresponding to the damaged or defaced coin in descending order of probability using key information.

[0135] Step S4044: Select M local regions on the damaged or defaced coin (preferably, for example, the damaged or defaced coin can be divided into M regions by computer image processing, and the geometric center region of each region can be selected). According to the coordinates of the M local regions, obtain the image information of the corresponding regions in the N-image MAP set 2 of the damaged or defaced coin, and compare it with the image information of the corresponding regions in the N-image MAP set 2 of the complete standard coin of the highest probability in step S4043. Each region obtains a matching degree.

[0136] Step S4045: If If the damaged or soiled coin is determined to be genuine, then the information of the complete standard coin of the next possible denomination determined in step S4043 is taken, and step S4044 is repeated until all tests are completed.

[0137] Step S4046: If the damaged or soiled banknote is determined to be genuine, the remaining width and location of the damage are determined by computer image analysis, and the redeemable amount is calculated according to the exchange rules.

[0138] The damaged currency processing method provided in this application automatically or semi-automatically completes the exchange of damaged and defaced currency through pre-sorting, receiving, scanning, evaluation, display, payment, and storage. It can achieve efficient, secure, and standardized damaged and defaced currency exchange services.

[0139] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0140] Based on the same inventive concept, this application also provides a damaged currency processing device for implementing the damaged currency processing method described above. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more damaged currency processing device embodiments provided below can be found in the limitations of the damaged currency processing method above, and will not be repeated here.

[0141] In one embodiment, such as Figure 5 As shown, a damaged banknote processing device is provided, comprising: a collection module 510, a determination module 520, a matching module 530, and an identification module 540, wherein:

[0142] The acquisition module 510 is used to acquire multiple feature images of the damaged banknotes; the multiple feature images correspond to different image types;

[0143] The determining module 520 is used to determine the denomination version of the damaged banknote based on the multiple feature images, and to determine the local feature images of the corresponding regions of multiple local regions of the damaged banknote in feature images of various image types from the multiple feature images.

[0144] The matching module 530 is used to compare the local feature image of each local region under each image type with the reference local feature image of the same region of the banknote corresponding to the denomination under the same image type for each local region, so as to obtain the matching information of each local region with the reference local feature image under each image type.

[0145] The identification module 540 is used to determine the authenticity identification result of the damaged coin based on the matching information of each local region under each image type.

[0146] In one embodiment, the damaged banknote may correspond to multiple denominations, and the above-mentioned device further includes a probability determination module for determining the similarity probability between the damaged banknote and each denomination.

[0147] The matching module 530 is also used to select the denomination version with the highest similarity probability as the current denomination version; for each local area, the local feature image of the local area under each image type is compared with the reference local feature image of the same area of ​​the banknote corresponding to the current denomination version under the same image type, so as to obtain the matching information of each local area with the reference local feature image under each image type.

[0148] The identification module 540 is further configured to, if the damaged banknote is determined to be counterfeit based on the matching information, select the denomination version with the second highest similarity probability as the new denomination version, and return the local feature images of the local area under each image type to compare with the reference local feature images of the same area of ​​the banknote corresponding to the current denomination version under the same image type, until the damaged banknote is determined to be genuine or all denomination versions have been compared to obtain the authenticity identification result for the damaged banknote.

[0149] In one embodiment, the determining module 520 is further configured to determine a target feature image from multiple feature images based on the degree of correlation between each feature image and the denomination version; and to determine the denomination version corresponding to the damaged coin based on the target feature image.

[0150] In one embodiment, the identification module 540 is further configured to obtain the influence factors of each image type on the authenticity identification result of the damaged banknote; the influence factors are obtained by training the authenticity labels of the sample damaged banknotes and the sample feature images of the sample damaged banknotes under each image type; the matching information of each local region under each image type is corrected by the influence factors to obtain the corrected matching information of each local region under each image type; and the authenticity identification result of the damaged banknote is determined based on the corrected matching information of each local region under each image type.

[0151] In one embodiment, the identification module 540 is further configured to perform statistical processing on the corrected matching information of each local region under each image type to obtain statistical matching information; if the statistical matching information is greater than or equal to a preset threshold, the damaged coin is determined to be genuine; if the statistical matching information is less than the preset threshold, the damaged coin is determined to be counterfeit.

[0152] In one embodiment, the acquisition module 510 is further configured to acquire multiple initial feature images of the damaged coin; and to perform a reorganization process on each initial feature image to obtain multiple feature images of the damaged coin.

[0153] Each module in the aforementioned damaged currency processing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0154] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When executed by the processor, the computer program implements a method for handling damaged currency. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0155] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0156] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0157] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0158] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0159] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0160] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0161] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for processing damaged banknotes, characterized in that, The method includes: Multiple feature images of the damaged banknotes are collected; these multiple feature images correspond to different image types. Based on the multiple feature images, the denomination version corresponding to the damaged banknote is determined, and from the multiple feature images, local feature images of multiple local regions of the damaged banknote corresponding to the regions in feature images of various image types are determined, wherein the local regions are representative regions that can characterize the features of the damaged banknote. For each local region, the local feature image of the local region under each image type is compared with the reference local feature image of the same region of the banknote corresponding to the denomination under the same image type to obtain the matching information of each local region with the reference local feature image under each image type. Based on the matching information of each local region under each image type, the authenticity identification result for the damaged coin is determined.

2. The method according to claim 1, characterized in that, The damaged banknotes may correspond to multiple denominations. After determining the denomination of the damaged banknotes, the following steps are also included: Determine the similarity probability between the damaged coin and each denomination version; The method further includes: The denomination with the highest similarity probability is selected as the current denomination. For each local region, the local feature image of the local region under each image type is compared with the reference local feature image of the same region of the banknote corresponding to the current denomination under the same image type, so as to obtain the matching information of each local region with the reference local feature image under each image type. If the damaged banknote is determined to be counterfeit based on the matching information, the denomination with the second highest similarity probability is selected as the new denomination. The process of comparing the local feature images of the local area under each image type with the baseline local feature images of the same area of ​​the banknote corresponding to the current denomination under the same image type is repeated until the damaged banknote is determined to be genuine or all denominations have been compared, thus obtaining the authenticity identification result for the damaged banknote.

3. The method according to claim 1, characterized in that, The step of determining the denomination and version of the damaged banknote based on the multiple feature images includes: Based on the degree of correlation between each feature image and the face value version, the target feature image is determined from the multiple feature images; Based on the target feature image, determine the denomination version corresponding to the damaged coin.

4. The method according to claim 1, characterized in that, The step of determining the authenticity of the damaged banknote based on the matching information of each local region under each image type includes: The influence factors of each image type on the authenticity identification results of the damaged banknotes are obtained; the influence factors are obtained by training the genuine and counterfeit labels of the sample damaged banknotes and the sample feature images of the sample damaged banknotes under each image type. Using the aforementioned influence factors, the matching information of each local region under various image types is corrected to obtain the corrected matching information of each local region under various image types. Based on the corrected matching information of each local region under each image type, the authenticity identification result for the damaged coin is determined.

5. The method according to claim 4, characterized in that, The step of determining the authenticity identification result for the damaged coin based on the corrected matching information of each local region under each image type includes: Statistical matching information is obtained by statistically processing the corrected matching information of each local region under various image types. If the statistical matching information is greater than or equal to a preset threshold, the damaged coin is determined to be genuine; if the statistical matching information is less than the preset threshold, the damaged coin is determined to be counterfeit.

6. The method according to claim 1, characterized in that, The collection of multiple feature images of the damaged banknotes includes: Collect multiple initial feature images of the damaged coin; Each initial feature image is processed to obtain the multiple feature images of the damaged coin.

7. A damaged coin processing device, characterized in that, The device includes: The acquisition module is used to acquire multiple feature images of the damaged banknotes; the multiple feature images correspond to different image types; The determining module is used to determine the denomination version corresponding to the damaged banknote based on the multiple feature images, and to determine the local feature images of the corresponding regions of multiple local regions of the damaged banknote in feature images of various image types from the multiple feature images, wherein the local regions are representative regions that can characterize the features of the damaged banknote determined from the damaged banknote. The matching module is used to compare the local feature image of each local region under each image type with the reference local feature image of the same region of the banknote corresponding to the denomination under the same image type for each local region, so as to obtain the matching information of each local region with the reference local feature image under each image type. The identification module is used to determine the authenticity of the damaged coin based on the matching information of each local region under each image type.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images