Print picture sensitive information identification method and device, equipment and storage medium

Abstract

The invention provides a printed picture sensitive information identification method and device, equipment and a storage medium, and is applied to the financial and medical fields. The method provided by the invention comprises the following steps: acquiring data of a printed picture, the data of the printed picture comprising formatted data and picture data; carrying out characterization processing on the formatted data to obtain printed basic feature information and name feature information of the picture; according to the basic feature information, the name feature information and an information identification model, primarily judging whether sensitive information exists in the printed picture or not; if it is judged for the first time that the sensitive information exists in the printed picture, character recognition is conducted on the picture data to obtain text information, the number of the sensitive data in the text information is obtained, and whether the sensitive information exists in the printed picture or not is judged again according to the number of the sensitive data. According to the method, whether the printed picture contains the sensitive information or not can be quickly identified after the picture is printed, and the performance of the terminal is not affected.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method, apparatus, device, and storage medium for identifying sensitive information in printed images. Background Technology

[0002] The financial and healthcare sectors involve large amounts of sensitive information. To prevent customer information leaks, numerous data breach prevention (DBM) products are deployed within enterprises. Current DBM products need to consider real-time performance, therefore most use regular expression matching. To prevent data leaks through printing, DBM products using regular expression matching are also typically deployed. While regular expressions are effective at handling formatted data, they cannot handle sensitive information converted to images for printing. Sensitive information in images can generally be identified using OCR (Optical Character Recognition), but DBM products are deployed on user terminals. Deploying OCR functionality on user terminals would severely impact terminal performance and the user's work experience. Summary of the Invention

[0003] The purpose of this invention is to provide a method, apparatus, device, and storage medium for identifying sensitive information in printed images, so as to solve the technical problem that identifying sensitive information in printed images can seriously affect terminal performance in the prior art.

[0004] The technical solution of the present invention is as follows: a method for identifying sensitive information in printed images is provided, comprising:

[0005] Acquire data for the printed image, wherein the data for the printed image includes formatted data and image data;

[0006] The formatted data is subjected to feature processing to obtain basic printable feature information and image name feature information;

[0007] Based on the basic feature information, the name feature information, and the information recognition model, it is initially determined whether the printed image contains sensitive information. The information recognition model is trained based on a machine learning model.

[0008] If the printed image is initially determined to contain sensitive information, then text recognition is performed on the image data to obtain text information. The number of sensitive data in the text information is then obtained, and the presence of sensitive information in the printed image is determined again based on the number of sensitive data.

[0009] Furthermore, the basic characteristic information includes the number of prints per day, the number of prints per hour, the job type, and whether the prints were made outside of working hours.

[0010] Furthermore, the name feature information includes character count information, keywords, file name similarity, and whether the last digits of the file name are consecutive.

[0011] Furthermore, the formatted data includes the printer's job information, file name, and printing time; the formatted data is then subjected to feature processing to obtain keywords, including:

[0012] The file name is segmented into words to obtain segmented data. If the segmented data exists in the preset keyword library, the segmented data is used as the keyword.

[0013] Furthermore, the training steps of the information recognition model include:

[0014] Obtain basic feature information samples and name feature information samples, and use the basic feature information samples and name feature information samples to train the machine learning model to obtain the information recognition model.

[0015] Further, text recognition is performed on the image data to obtain text information, and the amount of sensitive data in the text information is obtained, including:

[0016] The image data is used to perform text recognition using optical character recognition to obtain text information, and the number of sensitive data in the text information is obtained using regular expression matching.

[0017] Furthermore, based on the quantity of the sensitive data, it is further determined whether the printed image contains sensitive information, including:

[0018] If the number of sensitive data in the current category is greater than the corresponding preset threshold, then the printed image is determined to contain sensitive information.

[0019] Another technical solution of the present invention is as follows: a device for identifying sensitive information in printed images is provided, including a data acquisition module, a data processing module, a first identification module, and a second identification module;

[0020] The data acquisition module is used to acquire data of the printed image, which includes formatted data and image data;

[0021] The data processing module is used to perform feature processing on the formatted data to obtain basic feature information for printing and name feature information for the image;

[0022] The first identification module is used to initially determine whether the printed image contains sensitive information based on the basic feature information, the name feature information, and the information identification model, wherein the information identification model is trained based on a machine learning model.

[0023] The second recognition module is used to perform text recognition on the image data when it is initially determined that the printed image contains sensitive information, obtain text information, acquire the number of sensitive data in the text information, and determine again whether the printed image contains sensitive information based on the number of sensitive data.

[0024] Another technical solution of the present invention is as follows: an electronic device is provided, including a memory and a processor. The memory stores a computer program that can be executed by the processor. When the processor executes the computer program, it implements the method for identifying sensitive information in printed images as described in any of the above technical solutions.

[0025] Another technical solution of the present invention is as follows: a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and the computer program, when executed by a processor, implements the method for identifying sensitive information in printed images as described in any of the above technical solutions.

[0026] The beneficial effects of this invention are as follows: It acquires data of a printed image, including formatted data and image data; it performs feature processing on the formatted data to obtain basic print feature information and image name feature information; based on the basic feature information, the name feature information, and an information recognition model (trained using a machine learning model), it initially determines whether the printed image contains sensitive information; if it initially determines that the printed image contains sensitive information, it performs text recognition on the image data to obtain text information, obtains the number of sensitive data points in the text information, and further determines whether the printed image contains sensitive information based on the number of sensitive data points; through the above technical solution, it is possible to quickly identify whether a printed image contains sensitive information after it has been printed, without affecting terminal performance. Attached Figure Description

[0027] Figure 1 A flowchart illustrating the method for identifying sensitive information in printed images provided in an embodiment of the present invention;

[0028] Figure 2 This is a schematic diagram of the structure of the printed image sensitive information recognition device provided in an embodiment of the present invention;

[0029] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0031] In the description of this application, the terms "first," "second," etc., are used only for distinguishing purposes and should not be construed as indicating or implying relative importance or order. In this specification, the terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0032] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0033] Figure 1 This is a flowchart illustrating the method for identifying sensitive information in printed images according to an embodiment of the present invention. It should be noted that if substantially the same result is obtained, the method for identifying sensitive information in printed images of the present invention does not necessarily reflect the intended meaning. Figure 1 The illustrated process sequence is limited. For example... Figure 1 As shown, this method for identifying sensitive information in printed images mainly includes the following steps:

[0034] S101, Obtain the data for the printed image, wherein the data for the printed image includes formatted data and image data;

[0035] In some embodiments, the data breach prevention product can be set to a print image audit mode, which obtains print image data from the print module of the data breach prevention product. The print image data includes formatted data and image data.

[0036] S102, Perform feature processing on the formatted data to obtain basic feature information for printing and name feature information for the image;

[0037] In one optional implementation, the basic characteristic information includes the number of prints per day, the number of prints per hour, the job type, and whether the prints were made outside of working hours.

[0038] In some embodiments, images containing sensitive information are typically printed in batches within a short period of time. The formatted data can be segmented by hour and / or day, and the segmented data can be characterized to obtain basic feature information such as the number of prints per day, the number of prints per hour, the job type, and whether the prints were made outside of working hours.

[0039] In an optional implementation, the name feature information includes character count information, keywords, file name similarity, and whether the last digits of the file name are consecutive.

[0040] In some embodiments, the file names of images containing sensitive information generally lack explicit semantics and mainly consist of numbers and symbols, such as 245123644-4564.jpg, 245123644-4565.jpg, and 245123644-4566.jpg. Normal file names, on the other hand, contain some special keywords. Therefore, name feature information such as character count and keywords is associated with images containing sensitive information. Among them, character count information can include the number of Chinese characters and its ratio to the total number of characters, the number of numbers and its ratio to the total number of characters, and the number of symbols and its ratio to the total number of characters.

[0041] In some embodiments, the Damerau-Levenshtein algorithm (best string matching algorithm) can be used to compare the similarity of file names at adjacent time points one by one, obtain a set of similarity values, and calculate the peak, minimum, average, and variance of this set of similarities. For example, a set of file names includes "a_5421.jpg", "a_5422.jpg", "a_5423.jpg", and "a_5424.jpg", and a set of approximations of 90%, 90%, and 90% is calculated. Then the peak of this set of approximations is 90%, the minimum is 90%, the average is 90%, and the variance is 0. At least one of the peak, minimum, average, and variance of the similarity is used as the feature value of the file name similarity.

[0042] In some embodiments, when printing batches of images, the last digits will typically be consecutive. The last digit of each filename can be obtained in chronological order to form a set of numbers. If the last character of a filename is not a number, no operation is performed. After obtaining the set of numbers, the difference between them in chronological order is calculated. If the difference is +1 or -1, the last digits are considered consecutive. For example, if a set of images has filenames "a_5421.jpg", "a_5422.jpg", "a_5423.jpg", and "a_5424.jpg", and the last digits are 1, 2, 3, and 4 respectively, then the difference is 1, 1, and 1, indicating that the last digits of the filenames are consecutive.

[0043] In an optional embodiment, the formatted data includes the position information of the printer, the file name, and the printing time; the formatted data is characterized to obtain keywords, including:

[0044] The file name is segmented to obtain the segmented data. If the segmented data exists in the preset keyword library, the segmented data is used as the keyword.

[0045] In some embodiments, the formatted data may further include the name of the printer and the interception code. In order to obtain keywords from the formatted data, a keyword library (preset keyword library) can be established in advance. The process of establishing the keyword library may include obtaining a sample set of file names, which may include file names such as tender documents and business information; screening specific positions, such as information printed by positions such as legal positions, and obtaining the file name information therein; removing meaningless words from the sample set of file names, such as "of", "is", "about", removing symbols and only retaining the text, obtaining the filtered file name set, segmenting words, obtaining the word set, calculating the relevance weight of the word set. Specifically, the TF-IDF (term frequency-inverse document frequency) algorithm can be used to calculate the weight of the words, and the word set with a weight higher than the preset weight threshold is screened to obtain the preset keyword library.

[0046] As an example, a sample set of file names includes a tender document for a certain project, a litigation request for a certain person, and a tender document for a certain project. After removing meaningless words, we get "tender document for a certain project, litigation request for a certain person, tender document for a certain project", then segment the words to get "project, tender document, litigation, request", and then obtain the weight of the word vector through the TF-IDF algorithm. For example, the weight of "tender document" is -0.73, and the weight of "litigation" is -0.79. If the preset weight threshold is 0.7, the word set with a weight higher than 0.7 [tender document, litigation] can be added to the keyword library.

[0047] In some embodiments, the file name is segmented to obtain the segmented data. If the segmented data exists in the preset keyword library, the segmented data is used as the keyword. For example, the file name is "a certain legal litigation document", and after segmentation, we get "law, litigation, a certain person, document". Then check whether this group of words is in the preset keyword library. If "litigation" is in the preset keyword library, then "litigation" is used as the keyword.

[0048] S103, based on the basic feature information, the name feature information, and the information recognition model, initially determine whether there is sensitive information in the printed picture. The information recognition model is obtained by training a machine learning model;

[0049] In an optional embodiment, the training steps of the information recognition model include:

[0050] Obtain basic feature information samples and name feature information samples, and use the basic feature information samples and name feature information samples to train the machine learning model to obtain the information recognition model.

[0051] In some embodiments, basic feature information samples and name feature information samples of printed images containing sensitive information are selected as white samples, while basic feature information samples and name feature information samples of printed images not containing sensitive information are selected as black text. An artificial intelligence model is selected, such as a machine learning model like logistic regression or SVM, to train the model, obtain feature weights, and perform logical judgments through the model. The model outputs 1, indicating that the printed image contains sensitive information, and outputs 0, indicating that the printed image does not contain sensitive information.

[0052] S104, if it is initially determined that the printed image contains sensitive information, then the image data is subjected to text recognition to obtain text information, the number of sensitive data in the text information is obtained, and the presence of sensitive information in the printed image is determined again based on the number of sensitive data.

[0053] In an optional implementation, text recognition is performed on the image data to obtain text information, and the number of sensitive data in the text information is obtained, including:

[0054] The image data is used to perform text recognition using optical character recognition to obtain text information, and the number of sensitive data in the text information is obtained using regular expression matching.

[0055] In an optional implementation, the presence of sensitive information in the printed image is determined again based on the amount of sensitive data, including:

[0056] If the number of sensitive data in the current category is greater than the corresponding preset threshold, then the printed image is determined to contain sensitive information.

[0057] In some embodiments, optical character recognition (OCR) is used to perform text recognition on image data to obtain text information. The number of sensitive data points within the text information is then determined using regular expression matching. If the number of sensitive data points in the current category exceeds a corresponding preset threshold, the printed image is determined to contain sensitive information; otherwise, the printed image is determined not to actually contain sensitive information. For example, the text information obtained through OCR might include: "Name: Li Ming, Gender: Male, Mobile Number: 15700000001, Name: Wang Ming, Gender: Male, Mobile Number: 15700000002, Name: Liu Ming, Gender: Female, Mobile Number: 15700000003, Name: Sun Ming, Gender: Male, Mobile Number: 15700000004". Using mobile phone numbers as a category of sensitive data, with a preset quantity threshold of 2, a set of mobile phone numbers is obtained, including 15700000001, 15700000002, 15700000003, and 15700000004. Since the preset quantity threshold is 2, there are 4 mobile phone numbers at this time, which is considered to indicate that the printed image contains sensitive information.

[0058] The method for identifying sensitive information in printed images provided in this invention involves acquiring printed image data, including formatted data and image data; performing feature processing on the formatted data to obtain basic print feature information and image name feature information; initially determining whether the printed image contains sensitive information based on the basic feature information, the name feature information, and an information recognition model trained using a machine learning model; if the printed image is initially determined to contain sensitive information, then performing text recognition on the image data to obtain text information, obtaining the number of sensitive data in the text information, and further determining whether the printed image contains sensitive information based on the number of sensitive data; this method can quickly identify whether a printed image contains sensitive information after it has been printed, without affecting terminal performance or daily user operations, without relying on manual inspection, reducing labor costs, preventing data leakage due to employee misconduct, effectively protecting corporate assets, and effectively maintaining the security of trade secrets.

[0059] The method for identifying sensitive information in printed images provided in this invention can be built on artificial intelligence. It acquires and processes relevant data using AI technology, achieving unattended, AI-powered identification of sensitive information in printed images. Artificial intelligence (AI) is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0060] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0061] Figure 2 This is a schematic diagram of the structure of the printed image sensitive information recognition device according to an embodiment of the present invention, as shown below. Figure 2 As shown, the printed image sensitive information identification device 20 includes a data acquisition module 21, a data processing module 22, a first identification module 23, and a second identification module 24;

[0062] The data acquisition module 21 is used to acquire data of the printed image, the data of the printed image including formatted data and image data;

[0063] The data processing module 22 is used to perform feature processing on the formatted data to obtain basic feature information for printing and name feature information for the image.

[0064] The first identification module 23 is used to initially determine whether the printed image contains sensitive information based on the basic feature information, the name feature information, and the information identification model, wherein the information identification model is trained based on a machine learning model.

[0065] The second recognition module 24 is used to perform text recognition on the image data when it is initially determined that the printed image contains sensitive information, obtain text information, acquire the number of sensitive data in the text information, and determine again whether the printed image contains sensitive information based on the number of sensitive data.

[0066] In one optional implementation, the basic characteristic information includes the number of prints per day, the number of prints per hour, the job type, and whether the prints were made outside of working hours.

[0067] In some embodiments, images containing sensitive information are typically printed in batches within a short period of time. The formatted data can be segmented by hour and / or day, and the segmented data can be characterized to obtain basic feature information such as the number of prints per day, the number of prints per hour, the job type, and whether the prints were made outside of working hours.

[0068] In an optional implementation, the name feature information includes character count information, keywords, file name similarity, and whether the last digits of the file name are consecutive.

[0069] In some embodiments, the file names of images containing sensitive information generally lack explicit semantics and mainly consist of numbers and symbols, such as 245123644-4564.jpg, 245123644-4565.jpg, and 245123644-4566.jpg. Normal file names, on the other hand, contain some special keywords. Therefore, name feature information such as character count and keywords is associated with images containing sensitive information. Among them, character count information can include the number of Chinese characters and its ratio to the total number of characters, the number of numbers and its ratio to the total number of characters, and the number of symbols and its ratio to the total number of characters.

[0070] In some embodiments, an optimal string matching algorithm can be used to compare the similarity of file names at adjacent time points one by one to obtain a set of similarity values. The peak, minimum, average, and variance of this set of similarities are then calculated. For example, if a set of file names includes "a_5421.jpg", "a_5422.jpg", "a_5423.jpg", and "a_5424.jpg", and a set of approximations of 90%, 90%, and 90% is calculated, then the peak of this set of approximations is 90%, the minimum is 90%, the average is 90%, and the variance is 0. At least one of the peak, minimum, average, and variance of the similarity is used as a feature value of the file name similarity.

[0071] In some embodiments, when printing batches of images, the last digits will typically be consecutive. The last digit of each filename can be obtained in chronological order to form a set of numbers. If the last character of a filename is not a number, no operation is performed. After obtaining the set of numbers, the difference between them in chronological order is calculated. If the difference is +1 or -1, the last digits are considered consecutive. For example, if a set of images has filenames "a_5421.jpg", "a_5422.jpg", "a_5423.jpg", and "a_5424.jpg", and the last digits are 1, 2, 3, and 4 respectively, then the difference is 1, 1, and 1, indicating that the last digits of the filenames are consecutive.

[0072] In an optional implementation, the formatted data includes the printer's job information, file name, and printing time; the data processing module 22 performs feature processing on the formatted data to obtain keywords, including: performing word segmentation on the file name to obtain segmented data, and if the segmented data exists in a preset keyword library, then the segmented data is used as the keyword.

[0073] In some embodiments, the formatted data may further include the name of the printer and the interception code. In order to obtain keywords from the formatted data, a keyword library can be established in advance. The process of establishing the keyword library may include obtaining a sample set of file names, which may include file names such as printed tender documents, business information, etc.; screening for information printed by specific positions, such as legal positions, and obtaining the file name information therein; removing meaningless words from the sample set of file names, such as "of", "is", "about", removing symbols and only retaining the text, obtaining a filtered set of file names, performing word segmentation to obtain a set of words, calculating the correlation weights of the set of words. Specifically, the TF-IDF algorithm can be used to calculate the weights of the words, and the set of words with weights higher than the preset weight threshold can be screened to obtain the preset keyword library.

[0074] In an optional implementation manner, the printed picture sensitive information recognition device 20 further includes a model training module, which is used to obtain a basic feature information sample and a name feature information sample, and use the basic feature information sample and the name feature information sample to train the machine learning model to obtain the information recognition model.

[0075] In some embodiments, select the basic feature information sample and the name feature information sample of the printed picture containing sensitive information as the white sample, and select the basic feature information sample and the name feature information sample of the printed picture not containing sensitive information as the black text. Select an artificial intelligence model, such as a machine learning model such as logistic regression, SVM, etc., train the model to obtain feature weights, and make a logical judgment through the model. If the model outputs 1, it means that the printed picture contains sensitive information. If the model outputs 0, it means that the printed picture does not contain sensitive information.

[0076] In an optional implementation manner, the second recognition module 24 performs character recognition on the picture data to obtain text information, and obtains the number of sensitive data in the text information, including: performing character recognition on the picture data using optical character recognition to obtain text information, and using regular matching to obtain the number of sensitive data in the text information.

[0077] In an optional implementation manner, the second recognition module 24 determines again whether the printed picture has sensitive information according to the number of sensitive data, including: if the number of sensitive data in the current category is greater than the corresponding preset quantity threshold, it is determined that the printed picture has sensitive information.

[0078] In some embodiments, optical character recognition (OCR) is used to perform text recognition on the image data to obtain text information. The number of sensitive data points within the text information is then determined using regular expression matching. If the number of sensitive data points in the current category exceeds a corresponding preset threshold, the printed image is determined to contain sensitive information; otherwise, the printed image is determined not to actually contain sensitive information.

[0079] Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Figure 3 As shown, the electronic device 30 includes a processor 31 and a memory 32 communicatively connected to the processor 31.

[0080] The memory 32 stores program instructions for implementing the printed image sensitive information identification method of any of the above embodiments.

[0081] The processor 31 is used to execute program instructions stored in the memory 32 to identify sensitive information in the printed image.

[0082] The processor 31 can also be referred to as a CPU (Central Processing Unit). The processor 31 may be an integrated circuit chip with signal processing capabilities. The processor 31 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor can be a microprocessor or any conventional processor.

[0083] This invention provides a storage medium that stores program instructions capable of implementing all the methods described above. The storage medium can be non-volatile or volatile. These program instructions can be stored in the storage medium as a software product, including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, or terminal devices such as computers, servers, mobile phones, and tablets.

[0084] In the several embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0085] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. The above are merely embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

[0086] The above description is merely an embodiment of the present invention. It should be noted that those skilled in the art can make improvements without departing from the inventive concept of the present invention, but these improvements all fall within the protection scope of the present invention.

Claims

1. A method for identifying sensitive information in printed images, characterized in that, include: Acquire data for the printed image, wherein the data for the printed image includes formatted data and image data; The formatted data is subjected to feature processing to obtain basic printable feature information and image name feature information; Based on the basic feature information, the name feature information, and the information recognition model, an initial determination is made as to whether the printed image contains sensitive information. The information recognition model is trained based on a machine learning model. If the printed image is initially determined to contain sensitive information, then text recognition is performed on the image data to obtain text information. The number of sensitive data in the text information is then obtained, and the presence of sensitive information in the printed image is determined again based on the number of sensitive data.

2. The method for identifying sensitive information in printed images according to claim 1, characterized in that, The basic characteristic information includes the number of prints per day, the number of prints per hour, the job type, and whether the prints were made outside of working hours.

3. The method for identifying sensitive information in printed images according to claim 1, characterized in that, The name feature information includes character count, keywords, file name similarity, and whether the last digits of the file name are consecutive.

4. The method for identifying sensitive information in printed images according to claim 3, characterized in that, The formatted data includes the printer's job information, the file name, and the printing time; The formatted data is characterized to obtain keywords, including: The file name is segmented into words to obtain segmented data. If the segmented data exists in the preset keyword library, the segmented data is used as the keyword.

5. The method for identifying sensitive information in printed images according to claim 1, characterized in that, The training steps of the information recognition model include: Obtain basic feature information samples and name feature information samples, and use the basic feature information samples and name feature information samples to train the machine learning model to obtain the information recognition model.

6. The method for identifying sensitive information in printed images according to claim 1, characterized in that, Perform text recognition on the image data to obtain text information, and obtain the number of sensitive data in the text information, including: The image data is used to perform text recognition using optical character recognition to obtain text information, and the number of sensitive data in the text information is obtained using regular expression matching.

7. The method for identifying sensitive information in printed images according to claim 1, characterized in that, Based on the quantity of the sensitive data, determine again whether the printed image contains sensitive information, including: If the number of sensitive data in the current category is greater than the corresponding preset threshold, then the printed image is determined to contain sensitive information.

8. A device for identifying sensitive information in printed images, characterized in that, It includes a data acquisition module, a data processing module, a first identification module, and a second identification module; The data acquisition module is used to acquire data of the printed image, which includes formatted data and image data; The data processing module is used to perform feature processing on the formatted data to obtain basic feature information for printing and name feature information for the image; The first identification module is used to initially determine whether the printed image contains sensitive information based on the basic feature information, the name feature information, and the information identification model, wherein the information identification model is trained based on a machine learning model. The second recognition module is used to perform text recognition on the image data when it is initially determined that the printed image contains sensitive information, obtain text information, acquire the number of sensitive data in the text information, and determine again whether the printed image contains sensitive information based on the number of sensitive data.

9. An electronic device, comprising a memory and a processor, wherein the memory stores a computer program executable by the processor, characterized in that, When the processor executes the computer program, it implements the method for identifying sensitive information in printed images as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for identifying sensitive information in printed images as described in any one of claims 1 to 7.

Images