Document image quality assessment method, device, computer equipment and storage medium

By preprocessing the document image and multi-dimensional clarity evaluation, the problem of low clarity of the document image is solved, and automated and accurate image quality evaluation is achieved, improving evaluation efficiency and accuracy.

CN113723392BActive Publication Date: 2024-11-12GLODON CO LTD
View PDF 12 Cites 0 Cited by

Patent Information

Application Number
CN202111149187.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-09-29
Publication Date
2024-11-12
Estimated Expiration
2041-09-29

AI Technical Summary

Technical Problem

In the prior art, the clarity of the document image is not high, which makes it difficult for computers to recognize and requires manual evaluation, and there are problems such as inconsistent evaluation standards, inaccurate results and inefficient efficiency.

Method used

By preprocessing the document images, the clarity evaluation model is used to evaluate from multiple dimensions such as reflection, printed font, handwritten font and stamp clarity, and output comprehensive definition scores to achieve automated image quality evaluation.

Benefits of technology

It realizes accurate, fast and automatic evaluation of document images, ensures the objectivity and standardization of evaluation, and has a wide range of application and flexibility.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113723392B_ABST
    Figure CN113723392B_ABST
Patent Text Reader

Abstract

The present invention provides a document image quality assessment method, device, computer equipment and computer-readable storage medium, the method comprising the following steps: obtaining a document image uploaded by a user; preprocessing the document image to obtain a target image containing only the document and having a regular shape; using a clarity assessment model to assess the clarity of the target image and output a clarity score; wherein the clarity includes any combination of illumination clarity, print font clarity, handwriting font clarity and seal clarity; and assessing the quality of the document image according to the clarity score. The present invention automatically scores the document image through a machine learning model in artificial intelligence technology, thereby achieving the purpose of accurate and rapid quantitative quality assessment.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a document image quality assessment method, device, computer equipment and storage medium. Background Art

[0002] With the development of industrial Internet, the review, statistics, and processing of various documents are completed online. For example, as an important voucher in the logistics transportation process, the transport document needs to be uploaded to the logistics management platform for unified inspection and management. Currently, users usually collect image information of various documents through mobile phones, pads, high-speed scanners, scanners and other devices, and then upload the collected image information to the management platform or server on the mobile terminal (android, ios) or PC.

[0003] During the image acquisition process, due to improper operation, equipment diversity, and complex surrounding environment, the uploaded document images are not clear and difficult for computers to identify. Therefore, the existing technology requires manual judgment to evaluate the quality of the document images uploaded by users to reduce invalid document images. However, manual evaluation often has defects such as inconsistent evaluation standards, inaccurate evaluation results, and low evaluation efficiency, resulting in the inability to achieve the expected effect of image quality evaluation. Summary of the invention

[0004] The purpose of the present invention is to provide a technical solution that can accurately and quickly automatically evaluate the quality of document images, so as to solve the above-mentioned problems existing in the prior art.

[0005] To achieve the above object, the present invention provides a document image quality assessment method, comprising the following steps:

[0006] Get the document image uploaded by the user;

[0007] Preprocessing the document image to obtain a target image containing only the document and having a regular shape;

[0008] Using a clarity evaluation model to evaluate the clarity of the target image and output a clarity score; wherein the clarity includes any combination of illumination clarity, print font clarity, handwriting font clarity and seal clarity;

[0009] The quality of the document image is evaluated according to the clarity score.

[0010] According to the document image quality assessment method provided by the present invention, the step of preprocessing the document image to obtain a target image containing only the document and having a regular shape comprises:

[0011] Segmenting the document image using an image segmentation algorithm to determine a target area in the document image;

[0012] Adjusting the vertex positions of the target area so that the target area forms a regular rectangle;

[0013] The adjusted image within the target area is output as the target image.

[0014] According to the document image quality assessment method provided by the present invention, the clarity assessment model includes a reflection recognition model, and the step of using the clarity assessment model to assess the clarity of the target image and output a clarity score includes:

[0015] The target image is input into the reflection recognition model to output a reflection probability score of the target image.

[0016] According to the document image quality assessment method provided by the present invention, the clarity assessment model further includes a printed font recognition model and a handwritten font recognition model, and the step of using the clarity assessment model to assess the clarity of the target image and output a clarity score includes:

[0017] Using an OCR recognition method to identify all font strings in the target image;

[0018] Using a classification algorithm to classify all the font strings into printed font strings and handwritten font strings;

[0019] Inputting the printed font string into a printed font recognition model to output a printed font clarity score of the target image;

[0020] The handwriting character string is input into a handwriting character recognition model to output a handwriting character clarity score of the target image.

[0021] According to the document image quality assessment method provided by the present invention, the printed font string and the handwritten font string each include multiple, the printed font clarity score is obtained by weighted calculation based on multiple single printed font clarity scores, and the handwritten font clarity score is obtained by weighted calculation based on multiple single handwritten font clarity scores.

[0022] According to the document image quality assessment method provided by the present invention, the clarity assessment model also includes a seal recognition model, and the step of using the clarity assessment model to assess the clarity of the target image and output a clarity score includes:

[0023] Segmenting the target image using an image segmentation algorithm to obtain a seal image containing only the seal content;

[0024] The seal image is input into the seal recognition model to output a seal clarity score of the target image.

[0025] According to the document image quality assessment method provided by the present invention, the step of using the clarity assessment model to assess the clarity of the target image and outputting a clarity score further includes:

[0026] The reflection probability score, the printed font clarity score, the handwritten font clarity score, and the seal clarity score are weighted and summed to obtain a comprehensive clarity score of the target image.

[0027] To achieve the above object, the present invention further provides a document image quality assessment device, comprising:

[0028] The document acquisition module obtains the document image uploaded by the user;

[0029] A preprocessing module, adapted to preprocess the document image to obtain a target image containing only the document and having a regular shape;

[0030] A clarity scoring module, adapted to evaluate the clarity of the target image using a clarity evaluation model and output a clarity score; wherein the clarity includes any combination of illumination clarity, print font clarity, handwriting font clarity and seal clarity;

[0031] The quality assessment module is adapted to assess the quality of the document image according to the clarity score.

[0032] To achieve the above objectives, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0033] To achieve the above object, the present invention also provides a computer-readable storage medium on which a computer program is stored, and the computer program implements the steps of the above method when executed by a processor.

[0034] The document image quality assessment method, device, computer equipment and readable storage medium provided by the present invention automatically score the document image through machine learning in artificial intelligence technology, thereby achieving the purpose of accurate and rapid quantitative quality assessment. The present invention first pre-processes the document image to remove interference factors in the document image and improve the recognition of the document image. Then, using the clarity assessment model trained by machine learning, the document image can be scored for clarity from several different dimensions such as reflection probability, print font clarity, handwritten font clarity, seal clarity, and finally the quality assessment is performed based on the clarity scores of different dimensions. Using the clarity assessment model to perform clarity scoring can ensure the objectivity, standardization and accuracy of the scoring, and the clarity scores of multiple different dimensions can be set with different weights according to different application scenarios, so that the scope of application of the present invention is wider and has higher flexibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a flow chart of a first embodiment of a method for assessing the quality of a document image according to a first embodiment of the present invention;

[0036] Figure 2 A schematic flow chart of preprocessing a document image according to an embodiment of the present invention;

[0037] Figure 3 A schematic flow chart of evaluating font clarity in an embodiment of the present invention;

[0038] Figure 4 This is a schematic flow chart of quality assessment based on clarity according to Embodiment 1 of the present invention;

[0039] Figure 5 A schematic diagram of a program module of a document image quality assessment device according to a first embodiment of the present invention;

[0040] Figure 6 Schematic diagram of the hardware structure of the document image quality assessment device according to the first embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0042] Embodiment 1

[0043] See also Figure 1This embodiment proposes a document image quality assessment method, comprising the following steps:

[0044] S100: Obtain the document image uploaded by the user. The document in this embodiment may include written certification materials in any format, such as invoices, logistics orders, bills of lading, packing lists, etc., and the present invention does not limit this. The document image may be a photo or image obtained by any image acquisition device such as a mobile phone, camera, tablet computer, etc.

[0045] S200: Preprocessing the document image to obtain a target image containing only the document and having a regular shape.

[0046] The purpose of preprocessing in this step is to remove invalid information in the document image and improve the recognition of the document image. The document image can be preprocessed using any image preprocessing technology in the prior art, such as image noise reduction, image filtering, image segmentation, etc. This step does not limit this.

[0047] S300: Using a clarity evaluation model to evaluate the clarity of the target image and output a clarity score; wherein the clarity includes any combination of illumination clarity, print font clarity, handwriting font clarity and seal clarity.

[0048] The clarity combination model in this embodiment is an artificial neural network model obtained through machine learning training based on big data technology. Specifically, different neural network structures such as BP neural network, deep neural network, convolutional neural network, recursive neural network, etc. can be used, and this embodiment does not limit this.

[0049] The clarity evaluation model in this embodiment may include multiple ones, each used to evaluate the clarity of the target image from different dimensions. For example, a lighting clarity model for evaluating whether there is a reflection, a printed font clarity model for evaluating the clarity of printed fonts, a handwritten font clarity model for evaluating the clarity of handwritten fonts, and a seal clarity model for evaluating the clarity of seals. Each model outputs scores of different dimensions for the input target image.

[0050] The training processes of the above-mentioned different clarity models are relatively similar, and the only difference lies in the selection of training samples. Taking the illumination clarity model as an example, its training samples include multiple training images and the reflection probability score corresponding to each training image, wherein the reflection probability score can be pre-assessed manually based on specific standards. The above-mentioned training images are used as input data, and the corresponding reflection probability scores are used as output data to train the weight values ​​of the latent variables in the illumination clarity model until the illumination clarity model converges and the training process ends. At this point, the weight values ​​of the latent variables in the illumination clarity model have been fixed and will no longer be adjusted. For any image input in the subsequent stage, the illumination clarity model can output the reflection probability score corresponding to the arbitrary image. The training process of other clarity models is similar to the above process and will not be repeated in this embodiment.

[0051] S400: Evaluate the quality of the document image according to the clarity score.

[0052] As described above, in the case where the clarity assessment model includes multiple ones, multiple clarity scores can be obtained accordingly. This embodiment can set different weights for different clarity scores based on different actual needs, and finally obtain a comprehensive clarity score by weighted summation. In this way, the quality of the document image can be evaluated according to the comprehensive clarity score. For example, the comprehensive clarity score can be divided into different data intervals, and each data interval corresponds to a quality level, such as excellent, good, medium, poor, etc., so as to achieve quantitative to qualitative conversion, which is convenient for subsequent unified processing according to different quality levels, such as deleting document images with poor quality levels.

[0053] Figure 2 FIG. 2 shows a schematic flow chart of preprocessing a document image in Embodiment 1 of the present invention. Figure 2 As shown, step S200 includes:

[0054] S210: Segment the document image using an image segmentation algorithm to determine a target area in the document image.

[0055] In one example, the image segmentation algorithm can be implemented by a UNET model. The UNET model is an improvement based on FCNs. The trained UNET model can segment the target object to be identified from the input image. In this embodiment, the UNET model is used to segment the target area containing only the document content from the document image. For example, the document image often contains some background information, such as the desktop where the document is placed. This step can determine the boundary between the document content and the background information, thereby obtaining the target area.

[0056] S220: Adjust the vertex positions of the target area so that the target area forms a regular rectangle.

[0057] This step is used to deal with the situation where the target area is irregular due to the tilted shooting angle. For example, the camera plane and the document plane are not parallel, resulting in the target area in the document image to appear as a trapezoid, rhombus or irregular quadrilateral, which is easy to interfere with the subsequent recognition process. In this step, the shape of the target area can be adjusted by adjusting the vertices of the target area. For example, the initial first position coordinates of the four vertices in the target area are obtained respectively, and the second position coordinates of the four vertices in the target area are calculated by a geometric function, so that the target area forms a regular rectangle.

[0058] S230: Outputting the adjusted image in the target area as the target image, which is more consistent with the actual geometric shape of the document and is conducive to improving the accuracy of quality assessment.

[0059] Those skilled in the art understand that font clarity has a significant impact on the clarity of a document image. In one example, the present application evaluates font clarity from two aspects, namely, printed fonts and handwritten fonts. Figure 3 is a schematic flow chart of evaluating the clarity of a font in this embodiment, as shown in FIG. Figure 3 As shown, step S300 includes:

[0060] S310: Using an OCR recognition method to identify all font strings in the target image.

[0061] Optical character recognition (OCR) technology recognizes font strings in a target image line by line, where each font string refers to a continuous font recognized in a rectangular box. Those skilled in the art understand that when other non-font elements are contained between two fonts, such as images, blank spaces, etc., the two fonts are recognized as two font strings. Therefore, multiple font strings can usually be obtained from a target image through the OCR recognition method.

[0062] S320: Using a classification algorithm to classify all the font strings into printed font strings and handwritten font strings.

[0063] Multiple font strings can be classified by any existing classification algorithm, so that different font strings can be divided into printed font strings or handwritten font strings. Existing classification algorithms include NBC algorithm, LR algorithm, decision tree algorithm, SVM algorithm, ANN algorithm, etc., which are not limited in this embodiment.

[0064] S330: Input the printed font string into a printed font recognition model to output a printed font clarity score of the target image.

[0065] S340: Input the handwriting font string into a handwriting font recognition model to output a handwriting font clarity score of the target image.

[0066] Steps S330 and S340 obtain the print font clarity score and the handwritten font clarity score respectively through the print font recognition model and the handwritten font recognition model. It should be noted that, since each font string contains multiple, the font clarity score output by the above font recognition model is the score after weighted averaging. For example, ten print font strings are input into the print font recognition model, and the print font recognition model will first calculate the individual score of each print font string, and then perform weighted averaging calculation on the ten individual scores according to the preset weight values, and finally output the average print font clarity score corresponding to the ten print font strings. The weight value can be determined according to the number of fonts contained in the font string, for example, the more fonts there are, the larger the corresponding weight value is, etc. This embodiment also performs the same weighted average calculation on the handwritten font clarity score output by the handwritten font recognition model, and no examples are given here one by one.

[0067] By classifying printed fonts and handwritten fonts and outputting clarity scores respectively, different scoring standards can be adopted for different fonts accordingly, thereby improving the accuracy and reliability of font clarity scores.

[0068] In one example, the clarity evaluation model also includes a seal recognition model, so step S300 also includes:

[0069] S350: Segment the target image using an image segmentation algorithm to obtain a seal image containing only the seal content. The seal image can be segmented from the target image using any existing image segmentation algorithm, such as a UNET model. The UNET model can be obtained through big data training, and for any input image, the seal therein can be identified and the seal image can be segmented out.

[0070] S360: Input the seal image into the seal recognition model to output the seal clarity score of the target image. The seal recognition model is similar to the illumination clarity model, the printed font clarity model, and the handwritten font clarity model described above, and evaluates the clarity of the target image from the seal clarity dimension to obtain the seal clarity score.

[0071] Furthermore, on the basis of obtaining the clarity scores of the target image in different dimensions, the reflection probability score, the print font clarity score, the handwritten font clarity score, and the seal clarity score are weighted and summed to obtain a comprehensive clarity score of the target image. The weight values ​​of the above clarity scores can be set according to different document characteristics. For example, for documents with more print fonts, the weight of the print font clarity score is set higher than the weight of the handwritten font clarity score; for documents with more handwritten fonts, the weight of the handwritten font clarity score is set higher than the weight of the print font clarity score, etc. In this way, the comprehensive clarity score can be more in line with user needs and the accuracy of image quality assessment can be improved.

[0072] Figure 4 FIG. 1 is a schematic flow chart of quality assessment based on clarity in Embodiment 1 of the present invention. Figure 4 As shown, this embodiment first inputs an image, detects and crops the input image (equivalent to preprocessing the document image), and further performs clarity recognition from multiple dimensions, including reflection detection (i.e., illumination clarity detection), print font clarity recognition, handwritten font clarity recognition, and seal clarity recognition. Based on the above dimensions, a comprehensive clarity score is obtained, and the image quality is finally evaluated based on the relationship between the comprehensive clarity score and the preset threshold.

[0073] Please continue reading Figure 5 , shows a document image quality assessment device. In this embodiment, the document image quality assessment device 50 may include or be divided into one or more program modules. One or more program modules are stored in a storage medium and executed by one or more processors to complete the present invention and implement the document image quality assessment method described above. The program module referred to in the present invention refers to a series of computer program instruction segments that can complete specific functions, which is more suitable for describing the execution process of the document image quality assessment device 50 in the storage medium than the program itself. The following description will specifically introduce the functions of each program module in this embodiment:

[0074] The document acquisition module 51 acquires the document image uploaded by the user;

[0075] A preprocessing module 52, adapted to preprocess the document image to obtain a target image containing only the document and having a regular shape;

[0076] A clarity scoring module 53 is adapted to evaluate the clarity of the target image using a clarity evaluation model and output a clarity score; wherein the clarity includes any combination of illumination clarity, print font clarity, handwriting font clarity and seal clarity;

[0077] The quality assessment module 54 is adapted to assess the quality of the document image according to the clarity score.

[0078] The present embodiment also provides a computer device, such as a smart phone, tablet computer, laptop computer, desktop computer, rack server, blade server, tower server or cabinet server (including an independent server or a server cluster composed of multiple servers) that can execute programs. The computer device 60 of the present embodiment includes at least but is not limited to: a memory 61 and a processor 62 that can communicate with each other via a system bus, such as Figure 6 It should be pointed out that Figure 6 Only a computer device 60 having components 61 - 62 is shown, but it should be understood that implementing all of the components shown is not a requirement, and more or fewer components may alternatively be implemented.

[0079] In this embodiment, the memory 61 (i.e., readable storage medium) includes flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 61 may be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. In other embodiments, the memory 61 may also be an external storage device of the computer device 60, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card, etc. equipped on the computer device 60. Of course, the memory 61 may also include both the internal storage unit of the computer device 60 and its external storage device. In this embodiment, the memory 61 is generally used to store the operating system and various application software installed on the computer device 60, such as the program code of the document image quality assessment device 50 of the first embodiment. In addition, the memory 61 can also be used to temporarily store various types of data that have been output or are to be output.

[0080] In some embodiments, the processor 62 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips. The processor 62 is generally used to control the overall operation of the computer device 60. In this embodiment, the processor 62 is used to run the program code stored in the memory 61 or process data, such as running the document image quality assessment device 50 to implement the document image quality assessment method of the first embodiment.

[0081] This embodiment also provides a computer-readable storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a disk, an optical disk, a server, an App application store, etc., on which a computer program is stored, and when the program is executed by a processor, a corresponding function is realized. The computer-readable storage medium of this embodiment is used to store the document image quality assessment device 50, and when executed by the processor, the document image quality assessment method of the first embodiment is realized.

[0082] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0083] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0084] A person of ordinary skill in the art can understand that all or part of the steps carried out in the above-mentioned embodiment method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable medium. When the program is executed, it includes one of the steps of the method embodiment or a combination thereof.

[0085] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0086] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method.

[0087] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A document image quality assessment method, characterized in that: The following steps are involved: Get the document image uploaded by the user; Preprocessing the document image to obtain a target image containing only the document and having a regular shape; The clarity of the target image is evaluated by using a clarity evaluation model and a clarity score is output; wherein the clarity includes illumination clarity, print clarity, handwriting clarity and seal clarity, and the clarity score includes reflection probability score, print clarity score, handwriting clarity score and seal clarity score; Taking a weighted sum of the reflection probability score, the print font clarity score, the handwritten font clarity score, and the seal clarity score to obtain a comprehensive clarity score of the target image; The quality of the document image is evaluated according to the comprehensive clarity score.

2. The document image quality assessment method according to claim 1, characterized in that: The step of preprocessing the document image to obtain a target image containing only the document and having a regular shape comprises: Segmenting the document image using an image segmentation algorithm to determine a target area in the document image; Adjusting the vertex positions of the target area so that the target area forms a regular rectangle; The adjusted image within the target area is output as the target image.

3. The document image quality assessment method according to claim 1 or 2, characterized in that: The clarity evaluation model includes a reflection recognition model, and the step of using the clarity evaluation model to evaluate the clarity of the target image and outputting a clarity score includes: The target image is input into the reflection recognition model to output a reflection probability score of the target image.

4. The document image quality assessment method according to claim 3, characterized in that: The clarity evaluation model further includes a printed font recognition model and a handwritten font recognition model. The step of using the clarity evaluation model to evaluate the clarity of the target image and outputting a clarity score includes: Using an OCR recognition method to identify all font strings in the target image; Using a classification algorithm to classify all the font strings into printed font strings and handwritten font strings; Inputting the printed font string into a printed font recognition model to output a printed font clarity score of the target image; The handwriting character string is input into a handwriting character recognition model to output a handwriting character clarity score of the target image.

5. The document image quality assessment method according to claim 4, characterized in that: The printed font string and the handwritten font string both include multiple characters, the printed font clarity score is obtained by weighted calculation based on multiple single printed font clarity scores, and the handwritten font clarity score is obtained by weighted calculation based on multiple single handwritten font clarity scores.

6. The document image quality assessment method according to claim 5, characterized in that: The clarity evaluation model also includes a seal recognition model. The step of using the clarity evaluation model to evaluate the clarity of the target image and outputting a clarity score includes: Segmenting the target image using an image segmentation algorithm to obtain a seal image containing only the seal content; The seal image is input into the seal recognition model to output a seal clarity score of the target image.

7. A document image quality assessment device, characterized in that: include: The document acquisition module obtains the document image uploaded by the user; A preprocessing module, adapted to preprocess the document image to obtain a target image containing only the document and having a regular shape; A clarity scoring module, adapted to evaluate the clarity of the target image using a clarity evaluation model and output a clarity score; wherein the clarity includes illumination clarity, print font clarity, handwriting font clarity and seal clarity, and the clarity score includes a reflection probability score, a print font clarity score, a handwriting font clarity score and a seal clarity score; Taking a weighted sum of the reflection probability score, the print font clarity score, the handwritten font clarity score, and the seal clarity score to obtain a comprehensive clarity score of the target image; The quality assessment module is adapted to assess the quality of the document image according to the comprehensive clarity score.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

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

Citation Information

Patent Citations

  • Method and device for collecting and indentifying digital information of bill stamp pattern

    CN102236927A

  • Image definition evaluating method, device, computer device and storage medium

    CN107958455A

  • A document image instance segmentation method and device

    CN109948510A

  • Bill processing method, bill processing device and computer readable storage medium

    CN111476109A

  • Method and device for testing image text recognition performance, test equipment and medium

    CN111598099A