Tilt processing method, device and equipment for document recognition, and storage medium

Abstract

Embodiments of the present application provide a file recognition tilt processing method, device, equipment and storage medium, the method comprises: obtaining keyword information of a target recognition file, and a first character recognition result of the target recognition file; the first character recognition result is obtained based on a first image recognition of the target recognition file; compare the first character recognition result and the keyword information for the inclusion relationship, and obtain the first inclusion relationship comparison result; according to the first inclusion relationship comparison result of the first character recognition result and the keyword information, the first image is processed to obtain the tilt processing result of the target recognition file. Based on the inclusion relationship comparison result, the image is automatically processed to automatically correct the recognized file direction, without querying and tilt judgment for each character direction, without correcting based on the artificial retake mode, realizing the tilt automatic correction of the recognized character recognition result, improving the correction tilt processing efficiency of the file recognition.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of communication, in particular to a file recognition tilt processing method, a file recognition tilt processing device, a corresponding electronic device and a corresponding computer readable storage medium. BACKGROUND

[0002] OCR recognition (Optical Character Recognition) refers to a process of checking characters printed on paper by using an electronic device (such as a scanner or a digital camera), determining the shape by detecting light and dark patterns, and then translating the shape into computer text by character recognition method.

[0003] OCR recognition of the text of a paper document is a common technical means, and at this time, there is usually a need to judge and correct the tilt of the text recognized by using OCR technology. In related technologies, after the direction of all the text in the text content is counted and judged based on OCR technology, the correction is realized by manually re-shooting, but the query of judging the direction of each character takes a long time. SUMMARY

[0004] In view of the above problems, the present application is proposed to provide a file recognition tilt processing method, a file recognition tilt processing device, a corresponding electronic device and a corresponding computer readable storage medium which overcome the above problems or at least partially solve the above problems.

[0005] The present application discloses a file recognition tilt processing method, which comprises:

[0006] Obtaining keyword information of a target recognition file and a first character recognition result of the target recognition file; the first character recognition result is obtained based on a first image recognition of the target recognition file;

[0007] Comparing the first character recognition result and the keyword information to obtain a first inclusion relationship comparison result;

[0008] According to the first inclusion relationship comparison result of the first character recognition result and the keyword information, processing the first image to obtain a tilt processing result of the target recognition file.

[0009] Optionally, the keyword information is obtained based on the file type of the target recognition file; and the tilt processing result of the target recognition file is obtained according to the first inclusion relationship comparison result of the first character recognition result and the keyword information, comprising:

[0010] If the first inclusion relationship comparison result is a first inclusion relationship, determining that the first image is the tilt processing result of the target recognition file, and determining that the target recognition file meets the file requirements corresponding to the file type of the target recognition file;

[0011] And / or, if the first inclusion relationship comparison result is a second inclusion relationship, the second character recognition result of the target recognition file is obtained, and the tilt processing result for the target recognition file is obtained based on the second character recognition result and the second inclusion relationship comparison result of the keyword information.

[0012] Optionally, the number of the keyword information is at least one, the first inclusion relationship is used to indicate that the first character recognition result includes all the keyword information of the file type, and the second inclusion relationship is used to indicate that the first character recognition result partially includes or does not include the keyword information of the file type.

[0013] Optionally, obtaining the second character recognition result of the target recognition file includes:

[0014] Performing a conversion of a first image of the target recognition document at a preset angle to obtain a second image generated by the conversion at the corresponding preset angle;

[0015] The second image is recognized to obtain a second character recognition result.

[0016] Optionally, obtaining a tilt processing result for the target recognition file according to a comparison result of the second character recognition result and the second inclusion relationship of the keyword information includes:

[0017] If the second inclusion relationship comparison result is the third inclusion relationship, determining that the second image is the tilt processing result of the target recognition file, and determining that the target recognition file meets the file requirements corresponding to the file type of the target recognition file;

[0018] And / or, if the inclusion relationship comparison result is a fourth inclusion relationship, it is determined that the target identification file does not meet the file requirement corresponding to the file type of the target identification file.

[0019] Optionally, the number of keyword information is at least one, the third inclusion relationship is used to indicate that the second character recognition result contains keyword information of all the file types, and the fourth inclusion relationship is used to indicate that the second character recognition result partially contains or does not contain keyword information of the file type.

[0020] Optionally, the keyword information is obtained based on a file type of the target identification file, the tilt processing result is used to indicate a tilt-corrected image, and / or an auditing result of whether the first image meets a file requirement corresponding to the file type;

[0021] After obtaining the tilt processing result for the target identification file, the method further includes:

[0022] The tilt-corrected image and the auditing result are saved.

[0023] Embodiments of the present application also disclose a tilt processing device for file identification, and the device includes:

[0024] a keyword information obtaining module, configured to obtain keyword information of a target identification file and a first character recognition result for the target identification file, the first character recognition result being obtained based on a first image of the target identification file;

[0025] a containing relationship comparing module, configured to compare a containing relationship between the first character recognition result and the keyword information to obtain a first containing relationship comparison result;

[0026] a tilt processing module, configured to process the first image according to the first containing relationship comparison result of the first character recognition result and the keyword information to obtain a tilt processing result for the target identification file.

[0027] Optionally, the keyword information is obtained based on a file type of the target identification file; and the tilt processing module includes:

[0028] a tilt processing submodule, configured to, when the first containing relationship comparison result is a first containing relationship, determine that the first image is the tilt processing result of the target identification file, and determine that the target identification file meets a file requirement corresponding to the file type of the target identification file; and / or, when the first containing relationship comparison result is a second containing relationship, obtain a second character recognition result of the target identification file, and obtain the tilt processing result for the target identification file according to a second containing relationship comparison result of the second character recognition result and the keyword information.

[0029] Optionally, the keyword information includes at least one, the first containing relationship is used to indicate that the first character recognition result contains keyword information of all the file types, and the second containing relationship is used to indicate that the first character recognition result partially contains or does not contain keyword information of the file type.

[0030] Optionally, the tilt processing submodule includes:

[0031] The inclination processing unit is configured to convert the first image of the target recognition file by a preset angle to obtain a second image generated by conversion by the corresponding preset angle, recognize the second image to obtain a second character recognition result, and determine the second image as an inclination processing result of the target recognition file when the second containing relationship comparison result is a third containing relationship, and determine that the target recognition file meets the file requirements corresponding to the file type of the target recognition file; and / or determine that the target recognition file does not meet the file requirements corresponding to the file type of the target recognition file when the containing relationship comparison result is a fourth containing relationship.

[0032] Optionally, the number of the keyword information is at least one, the third containing relationship is used to indicate that the second character recognition result contains all the keyword information of the file type, and the fourth containing relationship is used to indicate that the second character recognition result partially contains or does not contain the keyword information of the file type.

[0033] Optionally, the keyword information is acquired based on the file type of the target recognition file, and the inclination processing result is used to indicate an image after inclination correction and / or an auditing result of whether the first image meets the file requirements corresponding to the file type.

[0034] After the inclination processing result of the target recognition file is obtained, the device further comprises:

[0035] The inclination result saving module is configured to save the image after inclination correction and the auditing result.

[0036] The embodiment of the application further discloses an electronic device, including a processor, a memory, and a computer program stored on the memory and capable of running on the processor, and the computer program is executed by the processor to implement any one of the inclination processing methods for file recognition.

[0037] The embodiment of the application further discloses a computer readable storage medium, and the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement any one of the inclination processing methods for file recognition.

[0038] The embodiment of the application has the following advantages:

[0039] In the embodiment of the present application, by performing the inclusion relationship comparison between the keyword information of the target identification file and the character recognition result obtained after recognizing the image of the target identification file, the image is automatically processed based on the inclusion relationship comparison result to automatically correct the recognized file direction, without the need to query and determine the direction of each character, and without the need to correct based on the artificial retaking method, which can reduce the unreasonable uncertainty of artificial retaking, realize automatic correction of the recognized character recognition result, and improve the correction and tilt processing efficiency of file recognition. Further, since the processing of the character recognition result is based on the inclusion relationship comparison result of the keyword information and the character recognition result, the automatic correction of the tilt can be realized based on the inclusion relationship comparison result to determine whether the image content of the target identification file is accurate, which is suitable for file auditing related work in various industries, reduces the artificial auditing cost, improves the full coverage rate of auditing, and improves the accuracy of file auditing. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 is a step flow chart of an embodiment of a file recognition tilt processing method of the present application;

[0041] Figure 2 is a step flow chart of another embodiment of a file recognition tilt processing method of the present application;

[0042] Figure 3 is an application scenario diagram of the file recognition tilt processing provided by the embodiment of the present application;

[0043] Figure 4 is a structure block diagram of an embodiment of a file recognition tilt processing device of the present application. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0045] In the related art, for the tilt determination and tilt correction of file recognition, mainly based on the OCR technology, the direction of all characters in the text content is counted and determined, and then a large number of character libraries are needed for artificial retaking, otherwise the query time is too long for determining the direction of each character, which leads to low efficiency of file recognition tilt processing.

[0046] The embodiment of the present application compares the keyword information of the target recognition file with the character recognition result obtained by recognizing the image of the target recognition file based on the inclusion relationship, automatically processes the image based on the inclusion relationship comparison result to automatically correct the recognized file direction, does not need to query and judge the inclination of each character direction, is time-saving, and does not need to correct the inclination based on the artificial retaking mode, changes from artificial correction to machine automatic correction, can reduce the unreasonable uncertainty of artificial retaking, realizes automatic correction of the inclination of the recognized character recognition result, and improves the correction inclination processing efficiency of file recognition. Further, since the processing of the character recognition result is based on the inclusion relationship comparison result of the keyword information and the character recognition result, the automatic correction of the inclination can be realized based on the inclusion relationship comparison result to determine whether the image content of the target recognition file is accurate, other tools are not needed, is suitable for file auditing related work in various industries, reduces the artificial auditing cost, improves the auditing coverage rate, and improves the accuracy of file auditing. In addition, the simple dictionary of the keyword information of the target recognition file can be used for targeted inclination verification and text content verification, and the corrected image can be directly saved based on the processed image, without human intervention.

[0047] Reference Figure 1 , a step flow chart of a file recognition inclination processing method embodiment of the present application is shown, which can specifically include the following steps:

[0048] Step 101, obtaining keyword information of a target recognition file and a first character recognition result of the target recognition file;

[0049] The target recognition file can refer to a recognized and / or to-be-recognized file to be audited. The recognized file has a corresponding file type and / or file category. The file content of different file types / file categories has different keywords. The number of keyword information can be at least one. For a certain file type, the keywords contained in the file content can be fixed, or the keywords of the type can be reduced based on the update of the file content.

[0050] In actual application, the keyword information can be stored in a dictionary library. When the file is audited, the keyword information of the target recognition file can be obtained from the dictionary library based on the file type of the target recognition file, so that the simple dictionary of the keyword information of the target recognition file can be used for targeted inclination verification and text content verification.

[0051] Specifically, the auditing on the file is performed on the character recognition result of the target recognition file. In a specific implementation, the first character recognition result of the target recognition file is mainly obtained by performing OCR recognition based on a first image captured from the target recognition file. The first image can specifically refer to an image to be input into the auditing system and needs to be audited.

[0052] In step 102, a first inclusion relationship comparison result is obtained by comparing the first character recognition result and the keyword information.

[0053] The keyword information is a simple dictionary, and the first character recognition result of the target recognition file can be subjected to targeted skew verification and text content verification. The specific verification process can be an inclusion relationship comparison of the first character recognition result and the keyword information.

[0054] The inclusion relationship comparison is performed on whether the character recognition result contains the keyword information obtained in advance, so as to determine whether the file content meets the file requirements corresponding to the file type of the target recognition file. In this way, the first image can be processed to obtain a skew processing result of the target recognition file in the case where the file content meets the file requirements, so as to automatically correct the file direction of the recognition image and reduce the unreasonable uncertainty of manual photographing again in the related skew correction technology.

[0055] In step 103, the first image is processed according to the first inclusion relationship comparison result of the first character recognition result and the keyword information to obtain a skew processing result of the target recognition file.

[0056] After the inclusion relationship comparison based on the first character recognition result and the keyword information obtains the first inclusion relationship comparison result, the first image can be processed based on the first inclusion relationship comparison result to obtain a skew processing result of the target recognition file.

[0057] In an embodiment of the present application, whether the file content meets the file requirements corresponding to the file type of the target recognition file can be determined based on whether the character recognition content contains all the keywords. The skew processing result can be used to indicate an image after skew correction and / or an auditing result of whether the first image meets the file requirements corresponding to the file type, wherein the image after skew correction mainly contains all the keyword information.

[0058] In the embodiment of the present application, the containing relationship comparison result can include a case where the character recognition result contains all the keyword information of the file type, and also include a case where the character recognition result partially contains or does not contain the keyword information of the file type, so as to perform different processing on the image based on different containing relationship comparison results, and obtain the tilt processing result for the target identification file.

[0059] It should be noted that, since the image after tilt correction mainly contains all the keyword information, the processing performed on the first image can be automatic correction processing, and the tilt correction operation performed on the image can refer to angle adjustment of the image, which can be specifically implemented by converting the image at different angles. For example, the image can be regenerated after being converted at angles of 45°, 90°, 135°, 180°, 225°, 270°, and 315°, respectively. If the regenerated image contains all the keyword information, the regenerated image can be directly saved as the image after tilt correction, so as to realize the conversion of the paper single into an electronic single (i.e., a file picture) and store the electronic single in the terminal device.

[0060] In an optional embodiment, after obtaining the tilt processing result for the target identification file, the audit result of whether the photographed first image meets the corresponding file requirements of the file type can also be saved, so as to obtain the result of whether the file identification is correct.

[0061] In the embodiment of the present application, the keyword information of the target identification file is compared with the character recognition result obtained by recognizing the image of the target identification file, the image is automatically processed based on the comparison result of the containing relationship, and the direction of the identified file is automatically corrected. The query and tilt judgment of each character direction are not required, and the correction is not required to be implemented based on the manual re-shooting manner. The tilt automatic correction of the identified character recognition result can be realized, the correction tilt processing efficiency for the file identification is improved, the artificial re-shooting uncertainty is reduced, and the tilt automatic correction of the identified character recognition result is realized. Further, since the processing of the character recognition result is based on the comparison result of the containing relationship between the keyword information and the character recognition result, the tilt automatic correction can be realized based on the comparison result of the containing relationship to determine whether the image content of the target identification file is accurate, which is suitable for file audit related work in various industries, reduces the artificial audit cost, improves the audit coverage rate, and improves the accuracy of the file audit.

[0062] Reference Figure 2 FIG. 6 shows a step flowchart of another embodiment of the tilt processing method for file identification of the present application, which can specifically include the following steps:

[0063] In step 201, if the first character recognition result of the target recognition file and the first inclusion relationship comparison result of the keyword information are in the first inclusion relationship, it is determined that the first image is the tilt processing result of the target recognition file.

[0064] In the embodiment of the present application, the keyword information of the target recognition file can be obtained, and the first character recognition result of the target recognition file can be obtained, then the inclusion relationship comparison between the first character recognition result and the keyword information can be performed to obtain the first inclusion relationship comparison result, and the first image can be processed according to the first inclusion relationship comparison result between the first character recognition result and the keyword information to obtain the tilt processing result of the target recognition file. The image is automatically processed based on the inclusion relationship comparison result to automatically correct the recognized file direction.

[0065] Specifically, the target recognition file can refer to the recognized and / or to-be-recognized file to be audited. The recognized file has a corresponding file type and / or file category. The file content of different file types / file categories has different keywords. The number of keyword information can be at least one. In actual application, the keyword information can be stored in a dictionary library. When auditing the file, the keyword information of the target recognition file can be obtained from the dictionary library based on the file type of the target recognition file, so as to perform targeted tilt checking and text content checking based on the keyword information of the target recognition file as a simple dictionary.

[0066] The inclusion relationship comparison is performed to determine whether the character recognition result contains the keyword information obtained in the foregoing, so as to determine whether the file content meets the file requirements corresponding to the file type of the target recognition file, and facilitate processing of the first image in the case that the file content meets the file requirements.

[0067] In the embodiment of the present application, the inclusion relationship comparison result can include the case that the character recognition result contains all the keyword information of the file type, and can also include the case that the character recognition result partially contains or does not contain the keyword information of the file type, so as to perform different processing on the image based on different inclusion relationship comparison results to obtain the tilt processing result of the target recognition file.

[0068] In one case, the first inclusion relationship comparison result can be the first inclusion relationship, the first inclusion relationship can be used to indicate that the first character recognition result contains all the keyword information of the file type, at this time it can be determined that the file content recognition is accurate, and the file content of the first image meets the corresponding file requirements of the file type of the target recognition file, and at this time the first character recognition result used for inclusion relationship comparison does not exist. Inclination, no need to correct the inclination of the first image corresponding to the first character recognition result, that is, the first image can be directly determined as the inclination processing result of the target recognition file, the first image is used as the image after inclination correction, and the file content recognition accurate and meeting the file requirements are saved as the audit result.

[0069] Step 202, if the first inclusion relationship comparison result of the first character recognition result of the target recognition file and the keyword information is the second inclusion relationship, the second character recognition result of the target recognition file and the second inclusion relationship comparison result of the keyword information are obtained, and if the second inclusion relationship comparison result is the third inclusion relationship, the second image is determined as the inclination processing result of the target recognition file.

[0070] In another case, the first inclusion relationship comparison result can be the second inclusion relationship, the second inclusion relationship can be used to indicate that the first character recognition result partially contains or does not contain the keyword information of the file type, at this time it can be determined that the file content may not be accurately recognized, or the file content of the first image may not meet the corresponding file requirements of the file type of the target recognition file, or there may be an inclination, at this time the second character recognition result of the target recognition file is obtained, and the second character recognition result and the second inclusion relationship comparison result of the keyword information are compared again, in order to obtain the inclination processing result for the target recognition file based on the second inclusion relationship comparison result.

[0071] Among them, the second character recognition result of the target recognition file is mainly reflected in that the first character recognition result partially contains or does not contain the keyword information of the file type, that is, the image obtained by performing OCR recognition on the image processed after the first image exists inclination.

[0072] Specifically, it can be manifested as converting the first image of the target recognition file by a preset angle to obtain a second image generated by converting by a corresponding preset angle, and then recognizing the second image to obtain a second character recognition result. For example, after the first image is converted by 45°, 90°, 135°, 180°, 225°, 270°, 315° respectively, the second image is regenerated.

[0073] At this time, the containing relationship comparison operation can be performed again, specifically, the second character recognition result and the keyword information are compared in containing relationship to obtain a second containing relationship comparison result, so as to facilitate automatic correction of the tilt based on the second containing relationship comparison result in the case of determining whether the image content of the target recognition file is accurate based on the containing relationship comparison result, without occupying other tools.

[0074] In some embodiments of the application, the second containing relationship comparison result can be a third containing relationship, the third containing relationship can be used to indicate that the second character recognition result contains all the keyword information of the file type, then it can be determined that the file content recognition is accurate, and the file content of the second image meets the file requirements corresponding to the file type of the target recognition file, and at this time, the second character recognition result used for containing relationship comparison does not exist in the tilt situation, at this time, it can be determined that the second image is the tilt processing result of the target recognition file, and it is determined that the target recognition file meets the file requirements corresponding to the file type of the target recognition file, that is, the regenerated second image can be directly saved as the image after tilt correction, realizing the conversion of paper single into electronic single (i.e. file picture) stored in the terminal device, directly saving the processed image to obtain the corrected picture, without manual intervention.

[0075] In step 203, if the second containing relationship comparison result of the first character recognition result of the target recognition file and the keyword information is a fourth containing relationship, it is determined that the target recognition file does not meet the file requirements corresponding to the file type of the target recognition file.

[0076] In some embodiments of the application, the second containing relationship comparison result can be a fourth containing relationship, the fourth containing relationship can be used to indicate that the second character recognition result partially contains or does not contain the keyword information of the file type, at this time, it can be determined that the target recognition file does not meet the file requirements corresponding to the file type of the target recognition file, and the non-compliance requirement is saved as an audit result.

[0077] In the embodiment of the present application, by comparing the keyword information of the target recognition file with the character recognition result obtained after recognizing the image of the target recognition file, the image is automatically processed based on the comparison result to automatically correct the recognized file direction, without the need to query and determine the inclination of each character direction, and without the need to correct based on the artificial retake method, which can reduce the unreasonable uncertainty of artificial retaking, automatically correct the inclination of the recognized character recognition result, and improve the correction inclination processing efficiency of file recognition. Further, since the processing of the character recognition result is based on the comparison result of the keyword information and the character recognition result, the automatic correction of the inclination can be realized based on the comparison result to determine whether the image content of the target recognition file is accurate, which is suitable for file auditing related work in various industries, reduces the cost of artificial auditing, improves the full coverage rate of auditing, and improves the accuracy of file auditing.

[0078] Referring to Figure 3 , a schematic diagram of an application scenario of the inclination processing of file recognition provided by the embodiment of the present application is shown, which can be applied to a file auditing system 310 and is suitable for file auditing related work in various industries, and specifically relates to inclination correction of a recognized file in a business requirement file auditing scenario, for example, only the required file needs to be confirmed to have been taken, and the auditing work does not need to accurately obtain all the contents of the text in the file.

[0079] In actual application, the auditing program can be triggered and started based on receiving an auditing instruction for a certain file.

[0080] The specific implementation process of the auditing program can be as follows:

[0081] First, the keyword information of the target recognition file and the first character recognition result of the target recognition file can be obtained. In actual application, the keyword information can be obtained based on the determination of the file type / file category of the target recognition file, and the corresponding keyword information of the file type / file category can be obtained from a dictionary library. Specifically, the keyword information of the confirmed file category can be obtained from the dictionary library. The first character recognition result of the target recognition file can be obtained based on the first image recognition of the target recognition file. The first image can refer to the image to be input into the auditing system for auditing, i.e., the image input can be subjected to OCR recognition to obtain the first character recognition result.

[0082] Then, the first character recognition result and the keyword information can be compared for inclusion relationship, i.e., the keyword information and the first character recognition result obtained by OCR recognition can be compared for inclusion relationship to determine whether the text content obtained by OCR recognition contains all the keyword information.

[0083] In one case, if the text content recognized by the OCR does not contain all the keyword information, the first image can be processed at this time, which can be specifically manifested as converting the text image at different angles to regenerate a plurality of pictures at corresponding angles, for example, converting the file image at 7 different orientations such as 45°, 90°, 135°, 180°, 225°, 270°, 315°, etc. to regenerate the image, at this time, the generated second image can be asynchronously subjected to OCR recognition to obtain a second character recognition result.

[0084] Then, the keyword information and the second character recognition result recognized by the OCR can be compared in a containing relationship to determine whether the second character recognition result recognized by the OCR contains all the keyword information, if the second character recognition result does not contain all the keyword information, it can be determined that the second image of the target recognition file does not exist tilt, but the paper file shot does not meet the requirements. For example, the content shot is not the corresponding file at all, or the text in the picture is upright, but the content does not belong to the required must-shoot file, etc. The common paper file requirements of the required must-shoot file can include but are not limited to network risk notice, card number receipt single, business handling entrustment letter, network fraud risk prompt notice, etc., which are not limited by the embodiments of the present application.

[0085] In another case, if the text content recognized by the OCR contains all the keyword information, that is, whether the first character recognition result contains all the keyword information or the second character recognition result contains all the keyword information, the picture containing the keyword can be saved at this time and the confirmation result of the AI audit is saved to the database, to complete the judgment of whether the file recognition of the file to be audited is accurate or not and the tilt processing process.

[0086] In the embodiment of the present application, by comparing the keyword information of the target recognition file with the character recognition result obtained after recognizing the image of the target recognition file in a containing relationship, the image is automatically processed based on the containing relationship comparison result to automatically correct the recognized file direction, without querying and judging the inclination of each character, which is time-consuming and short, and does not need to correct based on the artificial retake method, changes from artificial correction to machine automatic correction, can reduce the unreasonable uncertainty of artificial retaking, realize the automatic correction of the inclination of the recognized character recognition result, and improve the correction inclination processing efficiency of file recognition. Further, since the processing of the character recognition result is based on the containing relationship comparison result, the automatic correction of the inclination can be realized based on the containing relationship comparison result to determine whether the image content of the target recognition file is accurate, without occupying other tools, suitable for file auditing related work in various industries, reducing the artificial auditing cost, improving the auditing coverage rate and improving the accuracy of file auditing. In addition, the target recognition file keyword information can be used for targeted inclination verification and text content verification, and the processed image can be directly saved to obtain the corrected picture without manual intervention.

[0087] It should be noted that for the method embodiment, in order to simply describe, it is expressed as a series of action combinations, but those skilled in the art should know that the embodiments of the present application are not limited by the order of the described actions, because according to the embodiments of the present application, certain steps can be performed in other order or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily necessary for the embodiments of the present application.

[0088] Referring to Figure 4 , a structural block diagram of an embodiment of a file recognition inclination processing device of the present application is shown, which can specifically include the following modules:

[0089] The keyword information acquisition module 401 is configured to acquire keyword information of a target recognition file and a first character recognition result of the target recognition file, and the first character recognition result is obtained based on a first image recognition of the target recognition file.

[0090] The containing relationship comparison module 402 is configured to compare the first character recognition result and the keyword information in a containing relationship to obtain a first containing relationship comparison result.

[0091] The inclination processing module 403 is configured to process the first image according to the first containing relationship comparison result of the first character recognition result and the keyword information to obtain an inclination processing result of the target recognition file.

[0092] In an embodiment of the present application, the keyword information is obtained based on a file type of the target recognition file; the tilt processing module 403 can comprise the following sub-modules:

[0093] a tilt processing sub-module, configured to: when the first inclusion relationship comparison result is the first inclusion relationship, determine the first image as a tilt processing result of the target recognition file, and determine that the target recognition file meets a file requirement corresponding to the file type of the target recognition file; and / or when the first inclusion relationship comparison result is the second inclusion relationship, obtain a second character recognition result of the target recognition file, and obtain the tilt processing result of the target recognition file according to the second character recognition result and the second inclusion relationship comparison result of the keyword information.

[0094] In an embodiment of the present application, the number of keyword information is at least one, the first inclusion relationship is used to indicate that the first character recognition result contains keyword information of all file types, and the second inclusion relationship is used to indicate that the first character recognition result partially contains or does not contain keyword information of the file type.

[0095] In an embodiment of the present application, the tilt processing sub-module can comprise the following units:

[0096] a tilt processing unit, configured to: perform a preset angle conversion on the first image of the target recognition file to obtain a second image generated by conversion via a corresponding preset angle, perform recognition on the second image to obtain a second character recognition result; and when the second inclusion relationship comparison result is a third inclusion relationship, determine the second image as the tilt processing result of the target recognition file, and determine that the target recognition file meets a file requirement corresponding to the file type of the target recognition file; and / or when the inclusion relationship comparison result is a fourth inclusion relationship, determine that the target recognition file does not meet the file requirement corresponding to the file type of the target recognition file.

[0097] In an embodiment of the present application, the number of keyword information is at least one, the third inclusion relationship is used to indicate that the second character recognition result contains keyword information of all file types, and the fourth inclusion relationship is used to indicate that the second character recognition result partially contains or does not contain keyword information of the file type.

[0098] In an embodiment of the present application, the keyword information is obtained based on a file type of the target recognition file, and the tilt processing result is used to indicate a tilt-corrected image and / or an auditing result of whether the first image meets a file requirement corresponding to the file type.

[0099] After obtaining the tilt processing result for the target identification file, the device provided by the embodiment of the application further comprises the following modules:

[0100] The tilt result saving module is configured to save the image after tilt correction and the audit result.

[0101] In the embodiment of the application, the tilt processing device for file identification provided by the embodiment of the application automatically corrects the direction of the identified file by performing a containment relationship comparison between the keyword information of the target identification file and the character recognition result obtained by recognizing the image of the target identification file, and automatically processing the image based on the containment relationship comparison result. This method does not need to query and determine the tilt of each character, and does not need to correct the tilt based on the artificial retaking method. This method can reduce the unreasonable uncertainty of artificial retaking, automatically correct the tilt of the character recognition result, and improve the tilt correction efficiency of file identification. Further, since the processing of the character recognition result is based on the containment relationship comparison result, the tilt can be automatically corrected based on the containment relationship comparison result to determine whether the image content of the target identification file is accurate. This method is suitable for file audit related work in various industries, reduces the cost of artificial audit, improves the full coverage rate of audit, and improves the accuracy of file audit.

[0102] For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the related parts are described in the part of the method embodiment.

[0103] The embodiment of the application further provides an electronic device, comprising:

[0104] The electronic device comprises a processor, a memory, and a computer program stored on the memory and capable of running on the processor. When the computer program is executed by the processor, it implements each process of the above-mentioned file identification tilt processing method embodiment and achieves the same technical effect. To avoid repetition, this will not be repeated here.

[0105] The embodiment of the application further provides a computer readable storage medium, which stores a computer program. When the computer program is executed by the processor, it implements each process of the above-mentioned file identification tilt processing method embodiment and achieves the same technical effect. To avoid repetition, this will not be repeated here.

[0106] Each embodiment in the specification is described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same and similar parts of each embodiment can be referred to each other.

[0107] Those skilled in the art will appreciate that embodiments of the present application can be readily used as a method, apparatus, or computer program product. Accordingly, embodiments of the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, and the like) embodying computer program instructions.

[0108] Embodiments of the present application are described herein with reference to the Figure 1 one or more functions specified in a flow or multiple flows and / or blocks. Figure 1 means for performing one or more functions specified in a flow or multiple flows and / or blocks.

[0109] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the Figure 1 one or more functions specified in a flow or multiple flows and / or blocks. Figure 1 means for performing one or more functions specified in a flow or multiple flows and / or blocks.

[0110] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the Figure 1 one or more functions specified in a flow or multiple flows and / or blocks. Figure 1 means for performing one or more functions specified in a flow or multiple flows and / or blocks.

[0111] While preferred embodiments of the present application have been described, additional variations and modifications can be made to these embodiments by those skilled in the art once they have the benefit of the foregoing description. Therefore, the appended claims are intended to encompass within their scope all such variations and modifications as are included within the scope of the present application.

[0112] Finally, it is to be understood that the phraseology or terminology such as "first" and "second" etc. used herein is merely intended to differentiate one entity or operation from another entity or operation, without necessarily requiring or implying any actual such relationship or order between such entities or operations. Moreover, the terms "comprising", "including", or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0113] The above describes in detail the file identification inclination processing method, the file identification inclination processing device, the corresponding electronic device and the corresponding computer readable storage medium provided by the present application. The principles and implementation manners of the present application are described by using specific examples. The above description of the embodiments is only used to help understand the method of the present application and its core idea. Meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation manners and application ranges will be changed. In summary, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A skew processing method of file recognition, characterized by, The method comprises: obtaining keyword information of a target recognition file and a first character recognition result of the target recognition file; the first character recognition result is obtained based on a first image recognition of the target recognition file; comparing a containing relationship of the first character recognition result and the keyword information to obtain a first containing relationship comparison result; processing the first image according to the first containing relationship comparison result of the first character recognition result and the keyword information to obtain a tilt processing result of the target recognition file; wherein the keyword information is obtained based on a file type of the target recognition file; the tilt processing result of the target recognition file is obtained according to the first containing relationship comparison result of the first character recognition result and the keyword information, comprising: if the first containing relationship comparison result is a first containing relationship, it is determined that the first image is the tilt processing result of the target recognition file, and it is determined that the target recognition file meets the file requirements corresponding to the file type of the target recognition file; and / or, if the first containing relationship comparison result is a second containing relationship, a second character recognition result of the target recognition file is obtained, and the tilt processing result of the target recognition file is obtained according to a second containing relationship comparison result of the second character recognition result and the keyword information; the number of the keyword information is at least one, the first containing relationship is used to indicate that the first character recognition result contains all keyword information of the file type, and the second containing relationship is used to indicate that the first character recognition result partially contains or does not contain keyword information of the file type.

2. The method of claim 1, wherein, The second character recognition result of the target recognition file is obtained, comprising: performing a preset angle conversion on the first image of the target recognition file to obtain a second image generated by converting through a corresponding preset angle; performing recognition on the second image to obtain a second character recognition result.

3. The method of claim 2, wherein, The tilt processing result of the target recognition file is obtained according to the second containing relationship comparison result of the second character recognition result and the keyword information, comprising: if the second containing relationship comparison result is a third containing relationship, it is determined that the second image is the tilt processing result of the target recognition file, and it is determined that the target recognition file meets the file requirements corresponding to the file type of the target recognition file; and / or, if the containing relationship comparison result is a fourth containing relationship, it is determined that the target recognition file does not meet the file requirements corresponding to the file type of the target recognition file.

4. The method of claim 3, wherein, The number of the keyword information is at least one, the third containing relationship is used to indicate that the second character recognition result contains all keyword information of the file type, and the fourth containing relationship is used to indicate that the second character recognition result partially contains or does not contain keyword information of the file type.

5. The method of claim 1, wherein, The keyword information is obtained based on a file type of the target identification file, the tilt processing result is used for indicating a tilt-corrected image, and / or an auditing result of whether the first image meets a file requirement corresponding to the file type; After obtaining the tilt processing result of the target identification file, the method further includes: The tilt-corrected image and the auditing result are saved.

6. A tilt processing apparatus for file recognition, characterized by comprising: The apparatus includes: A keyword information obtaining module is configured to obtain keyword information of a target identification file and a first character recognition result of the target identification file, wherein the first character recognition result is obtained based on a first image of the target identification file; A containing relationship comparing module is configured to compare a containing relationship between the first character recognition result and the keyword information to obtain a first containing relationship comparison result; A tilt processing module is configured to process the first image according to the first containing relationship comparison result of the first character recognition result and the keyword information to obtain a tilt processing result of the target identification file; The keyword information is obtained based on a file type of the target identification file; and the tilt processing module includes: A tilt processing submodule is configured to, when the first containing relationship comparison result is a first containing relationship, determine that the first image is the tilt processing result of the target identification file, and determine that the target identification file meets a file requirement corresponding to the file type of the target identification file; and / or, when the first containing relationship comparison result is a second containing relationship, obtain a second character recognition result of the target identification file, and obtain the tilt processing result of the target identification file according to a second containing relationship comparison result of the second character recognition result and the keyword information; The keyword information includes at least one, the first containing relationship indicates that the first character recognition result contains keyword information of all the file types, and the second containing relationship indicates that the first character recognition result partially contains or does not contain keyword information of the file type.

7. An electronic device, comprising: The apparatus includes: A processor, a memory, and a computer program stored on the memory and capable of running on the processor, wherein the computer program is executed by the processor to implement the tilt processing method for file recognition according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, A computer program is stored on the computer readable storage medium, and the computer program is executed by the processor to implement the tilt processing method for file recognition according to any one of claims 1 to 5.

Images