ID card OCR recognition method and system

Abstract

The present invention relates to an OCR identification method and system for ID card, the method comprising: obtaining a plurality of ID card photos that have been successfully identified by OCR, performing different degrees of blurring processing to form a training set; based on the training set data, training and optimizing the sharpening parameters through a deep learning algorithm to obtain the optimized sharpening parameters; obtaining the ID card photo uploaded by the user, performing USM sharpening processing on the ID card photo according to the optimized sharpening parameters, and sending the sharpened ID card photo to the OCR recognition system to identify the ID card information. The present invention performs pre-processing on the ID card photo through the USM sharpening algorithm, significantly enhancing the text part in the photo, so that the OCR system can more accurately identify the information on the ID card. Experimental results show that the recognition success rate of the blurred photo after pre-processing is increased from 12.1% to 56.6%, which verifies the effectiveness of the present invention.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing and optical character recognition, and in particular to an identity card OCR recognition method and system. Background Art

[0002] In consumer finance institutions, customer identity verification is a key link in the business process. In order to complete this step efficiently and accurately, it has become an industry standard for customers to upload ID card photos and then perform OCR (Optical Character Recognition) recognition. This process not only simplifies customer operations, but also greatly improves the work efficiency of financial institutions. However, in actual applications, due to various reasons, such as shooting environment, equipment quality, user operation, etc., the quality of uploaded ID card photos varies, especially blurry and unclear handwriting.

[0003] At present, most financial institutions use traditional OCR technology for OCR recognition of ID card photos. This technology has a good recognition effect on clear and high-quality photos, and can accurately extract text information in photos, such as name, ID number, address, etc. However, when faced with blurred and unclear photos, the recognition efficiency and effect of traditional OCR technology will drop significantly, or even be unable to recognize. In order to improve the accuracy of OCR recognition, some institutions try to clarify photos before recognition. For example, real-time rendering high-definition technology such as MSAA (Multisample Anti-Aliasing) is used, or super-resolution sampling technology is applied to enhance the clarity of photos. However, these technologies often require a lot of training costs and training sets, and may not work well for certain types of blur (such as inaccurate focus, reflections, etc.). Summary of the invention

[0004] In view of this, it is necessary to provide an ID card OCR recognition method and system to solve the above-mentioned defects of the prior art.

[0005] In order to solve the above problems, in a first aspect, an embodiment of the present invention provides an identity card OCR recognition method, comprising:

[0006] Obtain multiple ID card photos that have been successfully recognized by OCR, perform different degrees of blurring, and form a training set;

[0007] Based on the training set data, the sharpening parameters are trained and optimized through the deep learning algorithm to obtain the optimized sharpening parameters;

[0008] Get the ID card photo uploaded by the user, perform USM sharpening on the ID card photo according to the optimized sharpening parameters, and send the sharpened ID card photo to the OCR recognition system to recognize the ID card information.

[0009] Preferably, the method further comprises:

[0010] Obtain multiple ID card photos that failed OCR recognition in the production environment as a verification set, and use the verification set to test the optimized sharpening algorithm and OCR recognition system.

[0011] Preferably, the step of training and optimizing the sharpening parameters by a deep learning algorithm based on the training set to obtain the optimized sharpening parameters specifically includes:

[0012] Initialize the parameters of the sharpening algorithm and sharpen the blurred photos in the training set; the parameters of the sharpening algorithm include Amount, Radius and Threshold;

[0013] Send the sharpened photo to the OCR recognition system to obtain the recognition result;

[0014] By comparing the recognition results with the true results, the cross entropy loss function is calculated, and the parameter values ​​are optimized through the gradient descent algorithm;

[0015] After multiple iterations of training, when the loss function reaches a minimum or converges, the optimized sharpening parameters are obtained.

[0016] Preferably, the expression of the cross entropy loss function is:

[0017]

[0018] Where CE(x) is the cross entropy loss function value; P(xi) is the actual probability of the i-th category, Q(xi) is the predicted probability of the i-th category, and n represents the total number of categories.

[0019] Preferably, the deep learning algorithm is a feedforward neural network algorithm.

[0020] Preferably, performing USM sharpening processing on the ID card photo according to the optimized sharpening parameters includes:

[0021] Use the best sharpening parameters trained by the deep learning algorithm and pre-process the ID card photo through the USM algorithm to enhance the high-frequency part of the ID card photo and highlight the text part.

[0022] Preferably, the expression of the USM algorithm is:

[0023] y(n,m)=x(n,m)+λz(n,m)

[0024] Where x(n, m) is the input image, y(n, m) is the output image, z(n, m) is the correction signal, n and m represent the row and column indexes of the image respectively, and λ is the scaling factor used to control the enhancement effect.

[0025] In a second aspect, an embodiment of the present invention provides an identity card OCR recognition system, comprising:

[0026] The training set construction module is used to obtain multiple ID card photos that have been successfully recognized by OCR, perform different degrees of blurring, and form a training set;

[0027] The parameter training and optimization module is used to train and optimize the sharpening parameters based on the training set data through a deep learning algorithm to obtain the optimized sharpening parameters;

[0028] The OCR recognition module is used to obtain the ID card photo uploaded by the user, perform USM sharpening processing on the ID card photo according to the optimized sharpening parameters, and send the sharpened ID card photo to the OCR recognition system for identification of the ID card information.

[0029] In a third aspect, the present invention further provides an electronic device, including a memory and a processor, wherein:

[0030] The memory is used to store programs;

[0031] The processor is coupled to the memory and is used to execute the program stored in the memory to implement the steps in the ID card OCR recognition method as described in the embodiment of the first aspect of the present invention.

[0032] In a fourth aspect, the present invention further provides a computer-readable storage medium for storing computer-readable programs or instructions, which, when executed by a processor, can implement the steps in the ID card OCR recognition method as described in the embodiment of the first aspect of the present invention.

[0033] Compared with the prior art, the ID card OCR recognition method and system provided by the present invention have the following beneficial effects:

[0034] 1) The present invention uses the USM sharpening algorithm to pre-process the ID card photo, significantly enhancing the text part in the photo, so that the OCR system can more accurately recognize the information on the ID card. The USM algorithm effectively improves the visual effect of the image by enhancing the high-frequency part of the image, making the originally blurred text clearer and providing better input conditions for OCR recognition. The experimental results show that the recognition success rate of blurred photos after pre-processing is increased from 12.1% to 56.6%, verifying the effectiveness of the present invention.

[0035] 2) The present invention trains and optimizes the sharpening parameters through a deep learning algorithm. The optimized sharpening parameters can more effectively enhance the clarity of the text part of the ID card photo, so that the OCR system can more accurately recognize the information in the photo. Since the ID card photo may be blurred due to various reasons (such as lens focus, reflection, photo position, etc.) in actual application, the optimized sharpening parameters can better adapt to these different situations and improve the system's recognition ability for blurred photos. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A flow chart of the ID card OCR recognition method provided by the present invention;

[0037] Figure 2 A schematic diagram of the verification set test results provided by the present invention;

[0038] FIG. 3( a ) is a schematic diagram of an OCR recognition result of a blurred photo without pre-processing provided by an embodiment of the present invention;

[0039] FIG3( b ) is a schematic diagram of an OCR recognition result of a blurred photo after pre-processing provided by an embodiment of the present invention;

[0040] Figure 4 A structural block diagram of the ID card OCR recognition system provided by the present invention;

[0041] Figure 5 This is a structural block diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0042] The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0043] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0044] At present, most financial institutions use traditional OCR technology for OCR (Optical Character Recognition) of ID card photos. This technology has a good recognition effect on clear and high-quality photos, and can accurately extract text information in photos, such as name, ID number, address, etc. However, when faced with blurred and unclear photos, the recognition efficiency and effect of traditional OCR technology will drop significantly, or even fail to recognize. In order to improve the accuracy of OCR recognition, some institutions try to clarify photos before recognition. For example, real-time rendering high-definition technology such as MSAA (Multisample Anti-Aliasing) is used, or super-resolution sampling technology is applied to enhance the clarity of photos. However, these technologies often require a lot of training costs and training sets, and may not work well for certain types of blur (such as inaccurate focus, reflection, etc.).

[0045] In view of this, the present invention proposes an ID card OCR recognition method, which uses the Unsharp Mask (USM) sharpening algorithm to perform preliminary processing on the ID card photo to enhance the text part in the photo, thereby improving the recognition accuracy and efficiency of the OCR system for the ID card photo. The present invention is aimed at blurry photos caused by poor shooting conditions (such as inaccurate focus, reflection, photo position offset, etc.), and makes them clearer through sharpening processing, thereby improving the success rate of OCR recognition. The following will be explained and introduced through multiple embodiments.

[0046] Figure 1 The ID card OCR recognition method flow chart provided by the present invention is shown in FIG. Figure 1 The ID card OCR recognition method provided by the present invention at least comprises the following steps:

[0047] Step S1, obtain multiple ID card photos that have been successfully recognized by OCR, perform blurring processing to different degrees, and form a training set.

[0048] Specifically, this embodiment extracts 2000 ID card photos that have been successfully recognized by OCR, blurs them to different degrees using a Gaussian blur algorithm, and forms a training set.

[0049] Gaussian blur is an image processing technology that smoothes an image by convolving it with the kernel of a Gaussian function, and can simulate the blurring effect of an image caused by poor shooting conditions (such as inaccurate focus, hand shaking, etc.). In order to simulate various blur situations that may be encountered in actual scenes and enable the model to learn how to better perform OCR recognition under different blur levels, the embodiment of the present invention applies different degrees of blur to multiple ID card photos.

[0050] Step S2: Based on the training set data, the sharpening parameters are trained and optimized by a deep learning algorithm to obtain optimized sharpening parameters.

[0051] In a preferred embodiment of the present invention, step S2 specifically includes the following steps:

[0052] S21, initialize the parameters of the sharpening algorithm and sharpen the blurred photos in the training set.

[0053] The parameters of the sharpening algorithm include Amount, Radius, and Threshold. Amount controls the intensity of the sharpening effect. Radius determines the width of the edge area and affects the range of sharpening. Threshold sets the level of detail that needs to be sharpened to avoid sharpening noise and low-contrast areas.

[0054] In this embodiment, initial values ​​are set for these sharpening parameters, for example, the default value of Amount is 0.1, the default value of Radius is 50, and the default value of Threshold is 0. These initial values ​​will serve as the starting point for deep learning training.

[0055] S22, sending the sharpened photo to an OCR recognition system to obtain a recognition result;

[0056] The sharpening parameters are AI-learned using a feedforward neural network algorithm. The input is a blurred ID card photo, and the output is the OCR recognition result.

[0057] S23, by comparing the recognition result with the true result, the cross entropy loss function is calculated, and the parameter value is optimized by the gradient descent algorithm.

[0058] Since the OCR recognition result is a classification problem (i.e., recognition success or failure), the cross entropy loss function is selected as the optimization target. The cross entropy loss function can measure the difference between the predicted probability distribution and the true probability distribution. The cross entropy loss function is often used in deep learning to judge the accuracy of the classification algorithm. In the OCR recognition process, there is no typical linear result that can be used to calculate the error, so it is suitable to use the cross entropy loss function for OCR algorithm training and gradient descent.

[0059] In the OCR (optical character recognition) ID card information recognition task, the composition of the predicted value (Y1) and the true value (Y2) of the feedforward neural network model and how they are used in the training process are as follows:

[0060] Y1 is the result obtained by the OCR recognition system after recognizing the ID card photo. It is a vector containing multiple fields, which usually include: province, city, date of birth, ID card validity period, issuing authority and ID card number. These fields are represented by c1, c2, c3, c4, c5 and c6 respectively.

[0061] Y2 is the real information in the ID card photo uploaded by the user. It has the same structure as Y1, that is, it is also a vector containing the province, city, date of birth, ID card validity period, issuing authority, and ID card number. During the training process, Y2 is used as the real value to compare with the model prediction value (Y1) to calculate the loss and update the model parameters.

[0062] During the training process, the cross entropy loss function is used to evaluate the accuracy of the OCR recognition results. For each field (such as province, city, etc.), the cross entropy loss between its predicted probability and the true probability is calculated, and these losses are added together to get the total loss value. Then, the model parameters are updated through optimization algorithms such as gradient descent to minimize this loss value.

[0063] The expression of the cross entropy loss function is:

[0064]

[0065] Where CE(x) is the cross entropy loss function value; P(xi) is the actual probability of the i-th category, Q(xi) is the predicted probability of the i-th category, and n represents the total number of categories.

[0066] S24, after multiple iterations of training, when the loss function reaches a minimum value or converges, the optimized sharpening parameters are obtained.

[0067] Specifically, the blurred ID card photo is used as input, and after USM sharpening, it is sent to the OCR system for recognition. According to the recognition result (success or failure), the cross entropy loss is calculated, and the weight of the sharpening parameter is updated through the back propagation algorithm. This process will be iterated many times until the loss function converges to the minimum value or reaches the preset number of training rounds.

[0068] After deep learning training, a set of optimized sharpening parameters will be obtained. These parameters can make the blurred ID card photos more accurately recognized by the OCR system after sharpening.

[0069] The present invention trains and optimizes the sharpening parameters through a deep learning algorithm. The optimized sharpening parameters can more effectively enhance the clarity of the text part of the ID card photo, so that the OCR system can more accurately recognize the information in the photo. Since the ID card photo may be blurred due to various reasons (such as lens focus, reflection, photo position, etc.) in actual application, the optimized sharpening parameters can better adapt to these different situations and improve the system's recognition ability for blurred photos.

[0070] Step S3, obtaining the ID card photo uploaded by the user, performing USM sharpening processing on the ID card photo according to the optimized sharpening parameters, and sending the sharpened ID card photo to the OCR recognition system to recognize the ID card information.

[0071] Specifically, users upload their ID card photos through the client or web interface of the consumer finance institution. These photos may be blurred or unclear due to shooting environment, equipment quality or improper user operation.

[0072] The system will use the best sharpening parameters obtained through deep learning algorithm training to pre-process the ID card photo through the USM (Unsharp Mask) algorithm to enhance the high-frequency part of the ID card photo and highlight the text part. Among them, the USM algorithm is an image sharpening technology that makes the image sharper by enhancing the edge contrast of the image.

[0073] In this embodiment, the expression of the USM algorithm is:

[0074] y(n,m)=x(n,m)+λz(n,m)

[0075] Where x(n, m) is the input image, y(n, m) is the output image, z(n, m) is the correction signal, n and m represent the row and column indexes of the image respectively, and λ is the scaling factor used to control the enhancement effect.

[0076] After the sharpening process, the ID card photo will be sent to the OCR (Optical Character Recognition) recognition system. The OCR system is a software technology that can automatically recognize and convert text information in an image. In this embodiment, the OCR system recognizes various information on the ID card from the sharpened photo, such as name, gender, ethnicity, date of birth, ID card number, etc.

[0077] The present invention uses the USM sharpening algorithm to pre-process the ID card photo, significantly enhancing the text part in the photo, so that the OCR system can more accurately recognize the information on the ID card. The USM algorithm effectively improves the visual effect of the image by enhancing the high-frequency part of the image, making the originally blurred text clearer and providing better input conditions for OCR recognition. The experimental results show that the recognition success rate of blurred photos after pre-processing is increased from 12.1% to 56.6%, verifying the effectiveness of the present invention.

[0078] In a preferred embodiment of the present invention, the ID card OCR recognition method further includes:

[0079] Obtain multiple ID card photos that failed OCR recognition in the production environment as a verification set, and use the verification set to test the optimized sharpening algorithm and OCR recognition system.

[0080] Specifically, 200 ID card photos that cannot be recognized by OCR in the production environment but can be recognized by the naked eye are selected as the verification set. In addition, in order to avoid overfitting, the present invention uses a part of blurred photos that have not participated in the training before as the verification set. The effect of sharpening pre-processing on improving the OCR recognition success rate is evaluated on the verification set. Figure 2 The schematic diagram of the verification set test results provided by the present invention is as follows: Figure 2 As shown in the figure, after 10 gradient descent trainings, the OCR recognition success rate of the ID card photos in the training set has been significantly improved after pre-processing (i.e., USM sharpening). This shows that the parameters obtained through training can effectively improve the recognition rate of blurred ID card photos.

[0081] Furthermore, in actual business scenarios, the recognition success rate of blurred photos without pre-processing is compared with the recognition success rate after pre-processing to verify the actual effect of the present invention. Figure 3(a) is a schematic diagram of the OCR recognition results of blurred photos without pre-processing provided by an embodiment of the present invention, and Figure 3(b) is a schematic diagram of the OCR recognition results of blurred photos after pre-processing provided by an embodiment of the present invention. Referring to Figures 3(a) and 3(b), the recognition success rate of blurred photos without pre-processing is 12.1%, indicating that the success rate of directly using the original blurred photos for OCR recognition is relatively low. The recognition success rate of blurred photos after pre-processing (i.e., using the trained USM sharpening algorithm) is increased to 56.6%. It can be seen that after pre-processing using the trained USM sharpening algorithm of the present invention, the recognition success rate of blurred photos is significantly improved.

[0082] The ID card OCR recognition method provided by the present invention is integrated into the relevant processes of the consumer finance system, such as customer identity verification, loan application, etc. Feedback data from actual applications is continuously collected to further optimize the algorithm and improve the recognition effect.

[0083] Figure 4 The structure diagram of the ID card OCR recognition system provided by the present invention is shown in FIG. Figure 4 , the ID card OCR recognition system 400 includes:

[0084] The training set construction module 401 is used to obtain a plurality of ID card photos that have been successfully recognized by OCR, and perform different degrees of blurring to form a training set;

[0085] The parameter training and optimization module 402 is used to train and optimize the sharpening parameters through a deep learning algorithm based on the training set data to obtain optimized sharpening parameters;

[0086] The OCR recognition module 403 is used to obtain the ID card photo uploaded by the user, perform USM sharpening processing on the ID card photo according to the optimized sharpening parameters, and send the sharpened ID card photo to the OCR recognition system to recognize the ID card information.

[0087] The ID card OCR recognition system provided by the present invention executes the ID card OCR recognition method provided by the aforementioned embodiments through the aforementioned modules. The ID card OCR recognition method has been described in detail in the aforementioned embodiments, and will not be repeated in this embodiment.

[0088] Figure 5 The structural block diagram of the electronic device provided by the present invention is as follows: Figure 5 As shown, the present invention also provides an electronic device, the electronic device 500 can be a computing device such as a mobile terminal, a desktop computer, a notebook, a palm computer, and a server. The electronic device 500 includes a processor 501 and a memory 502, wherein the memory 502 stores an ID card OCR recognition program 503.

[0089] In some embodiments, the memory 502 may be an internal storage unit of a computer device, such as a hard disk or memory of the computer device. In other embodiments, the memory 502 may also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the computer device. Furthermore, the memory 502 may also include both an internal storage unit of the computer device and an external storage device. The memory 502 is used to store application software and various types of data installed in the computer device, such as program codes installed in the computer device. The memory 502 may also be used to temporarily store data that has been output or is to be output. In one embodiment, when the ID card OCR recognition program 503 is executed by the processor 501, the following steps are implemented:

[0090] Obtain multiple ID card photos that have been successfully recognized by OCR, perform different degrees of blurring, and form a training set;

[0091] Based on the training set data, the sharpening parameters are trained and optimized through the deep learning algorithm to obtain the optimized sharpening parameters;

[0092] Get the ID card photo uploaded by the user, perform USM sharpening on the ID card photo according to the optimized sharpening parameters, and send the sharpened ID card photo to the OCR recognition system to recognize the ID card information.

[0093] In some embodiments, the processor 501 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run program codes or process data stored in the memory 502, such as executing an ID card OCR recognition program.

[0094] This embodiment also provides a computer-readable storage medium on which an ID card OCR recognition program is stored. When the ID card OCR recognition program is executed by a processor, the following steps are implemented:

[0095] Obtain multiple ID card photos that have been successfully recognized by OCR, perform different degrees of blurring, and form a training set;

[0096] Based on the training set data, the sharpening parameters are trained and optimized through the deep learning algorithm to obtain the optimized sharpening parameters;

[0097] Get the ID card photo uploaded by the user, perform USM sharpening on the ID card photo according to the optimized sharpening parameters, and send the sharpened ID card photo to the OCR recognition system to recognize the ID card information.

[0098] The above-described embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An ID card OCR recognition method, characterized in that: include: Obtain multiple ID card photos that have been successfully recognized by OCR, perform different degrees of blurring, and form a training set; Based on the training set data, the sharpening parameters are trained and optimized through the deep learning algorithm to obtain the optimized sharpening parameters; Get the ID card photo uploaded by the user, perform USM sharpening on the ID card photo according to the optimized sharpening parameters, and send the sharpened ID card photo to the OCR recognition system to recognize the ID card information.

2. The ID card OCR recognition method according to claim 1, characterized in that: Also includes: Obtain multiple ID card photos that failed OCR recognition in the production environment as a verification set, and use the verification set to test the optimized sharpening algorithm and OCR recognition system.

3. The ID card OCR recognition method according to claim 1, characterized in that: The training and optimization of the sharpening parameters by a deep learning algorithm based on the training set to obtain the optimized sharpening parameters specifically includes: Initialize the parameters of the sharpening algorithm and sharpen the blurred photos in the training set; the parameters of the sharpening algorithm include Amount, Radius and Threshold; Send the sharpened photo to the OCR recognition system to obtain the recognition result; By comparing the recognition results with the true results, the cross entropy loss function is calculated, and the parameter values ​​are optimized through the gradient descent algorithm; After multiple iterations of training, when the loss function reaches a minimum or converges, the optimized sharpening parameters are obtained.

4. The ID card OCR recognition method according to claim 3, characterized in that: The expression of the cross entropy loss function is: Where CE(x) is the cross entropy loss function value; P(x i ) is the actual probability of the i-th category, Q(x i ) is the predicted probability of the i-th category, and n represents the total number of categories.

5. The ID card OCR recognition method according to claim 1, characterized in that: The deep learning algorithm is a feedforward neural network algorithm.

6. The ID card OCR recognition method according to claim 1, characterized in that: The USM sharpening process is performed on the ID card photo according to the optimized sharpening parameters, including: Use the best sharpening parameters trained by the deep learning algorithm and pre-process the ID card photo through the USM algorithm to enhance the high-frequency part of the ID card photo and highlight the text part.

7. The ID card OCR recognition method according to claim 6, characterized in that: The expression of the USM algorithm is: y(n,m)=x(n,m)+λz(n,m) Where x(n, m) is the input image, y(n, m) is the output image, z(n, m) is the correction signal, n and m represent the row and column indexes of the image respectively, and λ is the scaling factor used to control the enhancement effect.

8. An ID card OCR recognition system applied to the ID card OCR recognition method according to any one of claims 1 to 7, characterized in that: include: The training set construction module is used to obtain multiple ID card photos that have been successfully recognized by OCR, perform different degrees of blurring, and form a training set; The parameter training and optimization module is used to train and optimize the sharpening parameters based on the training set data through a deep learning algorithm to obtain the optimized sharpening parameters; The OCR recognition module is used to obtain the ID card photo uploaded by the user, perform USM sharpening processing on the ID card photo according to the optimized sharpening parameters, and send the sharpened ID card photo to the OCR recognition system for identification of the ID card information.

9. An electronic device, characterized in that: comprising a memory and a processor, wherein: The memory is used to store programs; The processor is coupled to the memory and is used to execute the program stored in the memory to implement the steps in the ID card OCR recognition method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store computer-readable programs or instructions, which, when executed by a processor, can implement the steps of the ID card OCR recognition method described in any one of claims 1 to 7.

Images