Material compliance examination method and device, electronic equipment and storage medium

Abstract

The invention is suitable for the technical field of compliance inspection, and provides a material compliance inspection method and device, electronic equipment and a storage medium. The material compliance examination method comprises the steps of obtaining a to-be-examined picture in a feasibility research report, wherein the to-be-examined picture comprises at least one target object; based on the to-be-examined picture and the segmentation model, an optical character recognition technology is adopted, an object knowledge base is constructed, and the object knowledge base comprises a mapping relation among a target object, the area of the target object and the number of the to-be-examined picture; on the basis of the object knowledge base, constructing cue words; and performing compliance review on the target object based on the feasibility research report, the object knowledge base, the cue word and the big language generation model. According to the technical scheme provided by the invention, automatic compliance review can be performed on the materials in the picture of the feasibility research report through the big language generation model, the segmentation model and the optical character recognition technology, and the efficiency and accuracy of the compliance review of the materials are improved.

Description

Technical Field

[0001] The present application relates to the field of compliance review technology, and in particular to a material compliance review method, device, electronic device and storage medium. Background Art

[0002] When conducting a compliance review of a feasibility study report, it is necessary not only to review any inconsistencies between text and images and any obvious illogicalities in the report, but also to determine whether the size and quantity of objects in the images are compliant.

[0003] With the widespread use of artificial intelligence (AI) generation tools, the text and images in the feasibility study report that do not match and the obvious illogical places can be reviewed by using the inductive summary and comprehension capabilities of the large language model. When encountering pictures in the article, although there are tools such as optical character recognition (OCR) text extraction to achieve text extraction, it is still impossible to judge whether the size and quantity of the objects in the picture are appropriate for compliance. Currently, it can only be reviewed manually, resulting in low efficiency and accuracy of the review.

[0004] Therefore, how to achieve automated material compliance review has become an urgent issue to be addressed. Summary of the invention

[0005] The embodiments of the present application provide a material compliance review method, device, electronic device and storage medium, which realize automated compliance review of materials in pictures of feasibility study reports, thereby improving the efficiency and accuracy of material compliance review.

[0006] In a first aspect, an embodiment of the present application provides a material compliance review method, the method comprising: obtaining a picture to be reviewed in a feasibility study report, the picture to be reviewed including at least one target object; based on the picture to be reviewed and a segmentation model, using optical character recognition technology, constructing an object knowledge base, the object knowledge base including a mapping relationship between the target object, the area of ​​the target object and the number of the picture to be reviewed; based on the object knowledge base, constructing prompt words; based on the feasibility study report, the object knowledge base, the prompt words and a large language generation model, performing a compliance review on the target object.

[0007] In a possible implementation, the object knowledge base is constructed based on the picture to be reviewed and the segmentation model using optical character recognition technology, including: extracting text information from the picture to be reviewed based on optical character recognition technology, the text information including the object name of the target object and the number of the picture to be reviewed; segmenting the target object from the picture to be reviewed based on the segmentation model, and determining the total number of pixels of the target object; matching the text information with the target object based on the picture to be reviewed; determining the area of ​​the target object based on the total number of pixels of the picture to be reviewed and the target object; and constructing an object knowledge base based on the target object, the text information and the area of ​​the target object.

[0008] In a possible implementation, matching the text information with the target object based on the image to be reviewed includes: determining the positions of text detection boxes corresponding to multiple text information based on the image to be reviewed; calculating the distance between each text detection box and the target object; and matching the text information with the target object based on the distance between each text detection box and the target object.

[0009] In a possible implementation, determining the area of ​​the target object based on the total number of pixels of the image to be reviewed and the target object includes: determining a scale based on preset specification parameters of the image to be reviewed and the target object, the preset specification parameters being used to represent the regular size of the target object; determining the number of pixels per unit area in the image to be reviewed based on the scale and the image to be reviewed; and determining the area of ​​the target object based on the total number of pixels of the target object and the number of pixels per unit area in the image to be reviewed.

[0010] In a possible implementation, determining the scale based on the preset specification parameters of the image to be reviewed and the target object includes: determining the measurement specification parameters of the target object based on the image to be reviewed, the measurement specification parameters being used to characterize the size of the target object in the image to be reviewed; and determining the scale based on the preset specification parameters and measurement specification parameters of the target object.

[0011] In a possible implementation, determining the number of pixels per unit area in the picture to be reviewed based on the scale and the picture to be reviewed includes: determining position information of a first coordinate point and a second coordinate point based on the picture to be reviewed, the first coordinate point and the second coordinate point being any two coordinate points in the picture to be reviewed; determining first area information based on the position information of the first coordinate point and the second coordinate point; determining second area information based on the position information of the first coordinate point and the second coordinate point and the scale; and determining the number of pixels per unit area in the picture to be reviewed based on the first area information and the second area information.

[0012] In a possible implementation, the compliance review of the target object based on the feasibility study report, the object knowledge base, the prompt word and the large language generation model includes: determining the area information to be reviewed and the quantity information to be reviewed of the target object based on the feasibility study report and the large language generation model; determining the quantity threshold and area threshold of the target object based on the feasibility study report, the object knowledge base and the large language generation model; and performing a compliance review of the target object based on the prompt word, the large language generation model and the area information to be reviewed, the quantity information to be reviewed, the area threshold and the quantity threshold of the target object.

[0013] In a second aspect, an embodiment of the present application provides a material compliance review device, the device comprising: an acquisition module, used to acquire a picture to be reviewed in a feasibility study report, the picture to be reviewed including at least one target object; a first construction module, used to construct an object knowledge base based on the picture to be reviewed and a segmentation model, using optical character recognition technology, the object knowledge base including a mapping relationship between the target object, the area of ​​the target object and the number of the picture to be reviewed; a second construction module, used to construct a prompt word based on the object knowledge base; a review module, used to perform a compliance review on the target object based on the feasibility study report, the object knowledge base, the prompt word and a large language generation model.

[0014] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method described in the first aspect or any one of the implementation methods thereof is implemented.

[0015] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect or any one of the implementation methods thereof is implemented.

[0016] In a fifth aspect, an embodiment of the present application provides a computer program product. When the computer program product runs on an electronic device, the electronic device executes the method described in the first aspect or any one of the implementation methods thereof.

[0017] Compared with the prior art, the embodiments of the present application have the following beneficial effects: obtaining the image to be reviewed in the feasibility study report, the image to be reviewed includes at least one target object; based on the image to be reviewed and the segmentation model, using optical character recognition technology, constructing an object knowledge base, the object knowledge base includes the mapping relationship between the target object, the area of ​​the target object and the number of the image to be reviewed; based on the object knowledge base, constructing prompt words; based on the feasibility study report, the object knowledge base, the prompt words and the large language generation model, the target object is subject to compliance review, thus realizing the automated compliance review of the materials in the image of the feasibility study report. Compared with the current compliance review of the objects in the image of the feasibility study report through manual review, the efficiency and accuracy of the compliance review of materials are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 A flowchart of a material compliance review method provided in one embodiment of the present application;

[0020] Figure 2 A flowchart of an implementation method of S120 in a material compliance review method provided in an embodiment of the present application;

[0021] Figure 3 A flowchart of an implementation method of S240 in a material compliance review method provided in an embodiment of the present application;

[0022] Figure 4 A flowchart of an implementation method of S320 in a material compliance review method provided in an embodiment of the present application;

[0023] Figure 5 A flowchart of an implementation method of S140 in a material compliance review method provided in an embodiment of the present application;

[0024] Figure 6 A structural block diagram of a material compliance review device provided in one embodiment of the present application;

[0025] Figure 7 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0026] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0027] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0028] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0029] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0030] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0031] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0032] With the widespread use of AI generation tools, the text and images in the feasibility study report do not match, and the obvious illogical places can be solved by using the inductive, concluding and understanding capabilities of the large language model. Although the large language model has strong advantages in natural language understanding and text generation, there are tools such as OCR text extraction to achieve text extraction when encountering pictures in the feasibility study report. However, it is impossible to judge whether the size of objects in the pictures in the feasibility study report is appropriate for compliance. Currently, it can only be achieved through manual review, resulting in low efficiency and accuracy of the review.

[0033] In order to solve the above technical problems, this application proposes a material compliance review method, which obtains the picture to be reviewed in the feasibility study report, and the picture to be reviewed includes at least one target object; based on the picture to be reviewed and the segmentation model, optical character recognition technology is used to build an object knowledge base, and the object knowledge base includes the mapping relationship between the target object, the area of ​​the target object and the number of the picture to be reviewed; based on the object knowledge base, prompt words are built; based on the feasibility study report, the object knowledge base, the prompt words and the large language generation model, the target object is subject to compliance review. The technical solution provided by this application can automatically conduct compliance review of materials in the pictures of the feasibility study report through the large language generation model, the segmentation model and the optical character recognition technology, thereby improving the efficiency and accuracy of material compliance review.

[0034] For ease of understanding, the technical solution of the present application will be described in detail below with reference to the accompanying drawings.

[0035] Figure 1 A flow chart of a material compliance review method provided in one embodiment of the present application. Figure 1 As shown, the method includes:

[0036] S110, obtaining a picture to be reviewed in the feasibility study report, where the picture to be reviewed includes at least one target object.

[0037] A feasibility study report is a report that is submitted for approval by decision makers and competent authorities before engaging in an economic activity (such as investment) by conducting specific investigations, research, and analysis on various factors, ranging from economics, technology, production, supply and marketing to various social environments and laws, to determine favorable and unfavorable factors, whether the project is feasible, and estimating the success rate, economic benefits and social effects.

[0038] In specific implementation, the feasibility study report is usually accompanied by multiple pictures to show the shape, size, color and other information of the object to be reviewed. Any picture containing at least one object to be reviewed (ie, target object) is selected from the feasibility report as the picture to be reviewed.

[0039] S120, based on the image to be reviewed and the segmentation model, an object knowledge base is constructed using optical character recognition technology, where the object knowledge base includes a mapping relationship between a target object, an area of ​​the target object, and a serial number of the image to be reviewed.

[0040] As an example, Figure 2 A flowchart of an implementation method of S120 in a material compliance review method provided in an embodiment of the present application includes:

[0041] S210, extracting text information from the image to be reviewed based on optical character recognition technology, where the text information includes the object name of the target object and the number of the image to be reviewed.

[0042] In specific implementations, the pictures to be reviewed include not only the target object, but also the text information corresponding to the target object, such as the object name and introduction of the target object. At the same time, the pictures in the feasibility study report have corresponding numbers, and the text information in the pictures to be reviewed can be extracted through optical character recognition technology.

[0043] It should be noted that the specific implementation method of extracting text information from the image to be reviewed based on optical character recognition technology can refer to the method of extracting text information using optical character recognition technology in the prior art, and this application will not go into details.

[0044] S220, based on the segmentation model, segment the target object from the image to be reviewed, and determine the total number of pixels of the target object.

[0045] In the specific implementation, the segmentation model (SAM) is used to segment the target object from the image to be examined, and the total number of pixels contained in the target object is obtained according to the correspondence between the text information in the image to be examined and the segmented target object, which is recorded as G. pixel Among them, SAM is an advanced computer vision model designed to achieve arbitrary object segmentation in images.

[0046] It should be noted that the method of using a segmentation model to segment the target object from the image to be reviewed and determine the total number of pixels of the target object can refer to the common methods in the prior art, and this application will not go into details.

[0047] S230, matching text information with a target object based on the image to be reviewed.

[0048] In a possible implementation, based on the image to be reviewed, the positions of text detection frames corresponding to multiple text information are determined; the distance between each text detection frame and the target object is calculated; and based on the distance between each text detection frame and the target object, the text information and the target object are matched.

[0049] As an example, the image to be reviewed includes multiple target objects and the object name corresponding to each target object. The positions of the text detection frames corresponding to the multiple object names are determined in the image to be reviewed, the distance between each text detection frame and each target object is calculated, and the object name corresponding to the text detection frame closest to the target object is matched with the target object to establish a mapping relationship between the two.

[0050] As another example, the serial number of the picture to be reviewed is extracted, and a mapping relationship between the serial number of the picture to be reviewed and each target object in the picture to be reviewed is established.

[0051] S240, determining the area of ​​the target object based on the total number of pixels of the image to be examined and the target object.

[0052] As an example, Figure 3 A flowchart of an implementation method of S240 in a material compliance review method provided in an embodiment of the present application includes:

[0053] S310, determining a scale based on the image to be reviewed and preset specification parameters of the target object, where the preset specification parameters are used to represent a regular size of the target object.

[0054] In a possible implementation, based on the image to be reviewed, measurement specification parameters of the target object are determined, and the measurement specification parameters are used to characterize the size of the target object in the image to be reviewed; based on the preset specification parameters and the measurement specification parameters of the target object, a scale is determined.

[0055] In the specific implementation, we first determine the actual conventional size of the target object based on experience and common sense, that is, the preset specification parameters of the target object, and use the image processing model to measure the target object in the image to be reviewed to obtain the size of the target object in the image to be reviewed, that is, the measured specification parameters of the target object, and use the ratio of the measured specification parameters of the target object to the preset specification parameters as the scale.

[0056] It should be noted that the method of measuring the target object in the image to be reviewed by using the image processing model in this embodiment can refer to the image processing method in the prior art, and this application will not elaborate on this.

[0057] S320, determining the number of pixels per unit area in the image to be reviewed based on the scale and the image to be reviewed.

[0058] As an example, Figure 4 A flowchart of an implementation method of S320 in a material compliance review method provided in an embodiment of the present application includes:

[0059] S410: Determine position information of a first coordinate point and a second coordinate point based on the picture to be reviewed, where the first coordinate point and the second coordinate point are any two coordinate points in the picture to be reviewed.

[0060] In the specific implementation, any two coordinate points P1 (P 1x ,P 1y ) and P2(P 2x ,P 2y ).

[0061] S420: Determine first area information based on position information of the first coordinate point and the second coordinate point.

[0062] In the specific implementation, the distance L1 between the first coordinate point P1 and the second coordinate point P2 is first calculated by the first calculation formula, and then the first area information is determined based on L1 and the second calculation formula, that is, the square area (total number of pixels) with L1 as the side length in the image to be reviewed.

[0063] As an example, the first calculation formula is Where L1 is the distance between the first coordinate point and the second coordinate point, P1 is the first coordinate point, P2 is the second coordinate point, and P 1x is the horizontal coordinate of the first coordinate point, P 2x is the horizontal coordinate of the second coordinate point, P 1y is the ordinate of the first coordinate point, P 2y is the ordinate of the second coordinate point.

[0064] As an example, the second calculation formula is Among them, S p1p2 Represents the first area.

[0065] S430: Determine second area information based on the position information of the first coordinate point and the second coordinate point, and the scale.

[0066] In a specific implementation, the second area information, ie, the actual area information (total number of pixels) of the first coordinate point and the second coordinate point is determined by the third calculation formula in combination with the position information of the first coordinate point and the second coordinate point and the scale.

[0067] As an example, the third calculation formula is Among them, S p1p2实际 represents the second area, P1 is the first coordinate point, P2 is the second coordinate point, and K is the scale.

[0068] S440: Determine the number of pixels per unit area in the image to be reviewed based on the first area information and the second area information.

[0069] In a specific implementation, the fourth calculation formula is used to combine the first area information and the second area information to obtain the number of pixels per unit area in the image to be reviewed.

[0070] As an example, the fourth calculation formula is Among them, U represents the number of pixels per unit area in the image to be reviewed, S p1p2 represents the first area, S p1p2实际 represents the second area, P1 is the first coordinate point, P2 is the second coordinate point, and K is the scale.

[0071] S330, determining the area of ​​the target object based on the total number of pixels of the target object and the number of pixels per unit area in the image to be examined.

[0072] In a specific implementation, the area of ​​the target object is determined by the fifth calculation formula based on the total number of pixels of the target object and the number of pixels per unit area in the image to be examined, wherein the total number of pixels of the target object is obtained in step S220.

[0073] As an example, the fifth calculation formula is S Gname =G pixel *U, where S Gname Represents the area of ​​the target object, G pixel It represents the total number of pixels of the target object, and U represents the number of pixels per unit area in the image to be reviewed.

[0074] S250: construct an object knowledge base based on the target object, the text information, and the area of ​​the target object.

[0075] In a specific implementation, based on the target object segmented from the image to be reviewed, the text information extracted from the image to be reviewed, and the Figure 3 and Figure 4 The area of ​​the target object calculated by the method shown uses a large model, Langchain, and QLoRA to build an object knowledge base.

[0076] Among them, the big model is a large-scale language model, such as the chat generative pre-trained transformer (chatGPT) and the chat general language model (chatGLM); Langchain is a knowledge base retrieval method based on large models and vector retrieval, which is usually used to alleviate the generalization problem of large models; QLoRA freezes the pre-trained model weight parameters, freezes the original model parameters, adds additional network layers to the model, and only trains these newly added network layer parameters. Due to the small number of these newly added parameters, this not only significantly reduces the cost of fine-tuning the large model, but also achieves similar effects to full model fine-tuning.

[0077] S130, constructing prompt words based on the object knowledge base.

[0078] In a specific implementation, prompt words are constructed based on the review requirements for target objects in the to-be-reviewed pictures of the feasibility study report and the constructed object knowledge base.

[0079] As an example, the review requirement for the target object is to review whether the purchase quantity of the target object proposed in the feasibility study report is compliant. Then, based on the review requirement and the area of ​​the target object calculated by the above method, two prompt words are constructed. The first prompt word is whether there is a problem with the calculated area of ​​the target object, and the second prompt word is whether there is a problem with the purchase quantity of the target object.

[0080] S140, performs compliance review on the target object based on the feasibility study report, object knowledge base, prompt words and large language generation model.

[0081] In one possible implementation, after constructing the object knowledge base and prompt words of the target object, a compliance review is performed on the target object based on the text description of the target object in the feasibility report, the object knowledge base, prompt words and the large language generation model of the target object.

[0082] As an example, Figure 5 A flowchart of an implementation method of S140 in a material compliance review method provided in an embodiment of the present application includes:

[0083] S510, based on the feasibility study report and the large language generation model, determine the area information and quantity information of the target object to be reviewed.

[0084] In the specific implementation, based on the large language generation model, the text part corresponding to the target object is located in the feasibility study report, and the area information and quantity information to be reviewed of the target object are extracted from it, that is, the area and purchase quantity of the target object proposed in the feasibility study report.

[0085] As an example, the target object is an office desk. The area information to be reviewed of the target object includes the area of ​​the office desk calculated by the above method through the picture to be reviewed in the feasibility study report, for example, 4.5 square meters; the quantity information to be reviewed of the target object includes the planned purchase quantity of the office desks found in the feasibility study report, for example, 5.

[0086] S520, determining a quantity threshold and an area threshold of the target objects based on the feasibility study report, the object knowledge base, and the large language generation model.

[0087] In a specific implementation, the area threshold of the target object in reality is determined through the large language generation model, and the quantity threshold of the target object in the specific environment required in the feasibility study report is determined through the area information of the target object in the object knowledge base and the large language generation model.

[0088] As an example, when the target object is a desk, the area of ​​the desk calculated by the above method is 4.5 square meters. According to experience, the area of ​​a large desk is about 5 square meters. Therefore, the area threshold of the target object can be determined to be 5 square meters. When it is known from the feasibility study report that the specific environment in which the target object is placed is an office with a total area of ​​20 square meters, the large language generation model and the calculated area of ​​the desk can be used to determine the number threshold of the desk in the office to be 4 (20÷4.5≈4.44).

[0089] S530 , performing compliance review on the target object based on the prompt words, the large language generation model, the to-be-reviewed area information, to-be-reviewed quantity information, the area threshold, and the quantity threshold of the target object.

[0090] In a specific implementation, based on the pre-set prompt words and the large language generation model, it is determined whether the area to be reviewed of the target object exceeds the area threshold. If so, the target object is determined to be non-compliant; if not, it is further determined whether the number of target objects to be reviewed exceeds the number threshold. If so, the target object is determined to be non-compliant; if not, the target object is determined to be compliant.

[0091] As an example, the target object is an office desk. The area to be reviewed of the office desk is 4.5 square meters, the number to be reviewed is 5, the area threshold is 5 square meters, the number threshold is 4, and the first preset prompt word is whether there is a problem with the area to be reviewed of the target object, and the second prompt word is whether there is a problem with the number of purchases of the target object. Then, based on the large language generation model, firstly, based on the first prompt word, it is judged whether the area to be reviewed of the office desk exceeds the area threshold. Since the area to be reviewed of the office desk of 4.5 square meters is less than the area threshold of 5 square meters, there is no problem with the area to be reviewed of the office desk. Then, the second prompt word is used to judge whether the number of the target object to be reviewed exceeds the number threshold. Since the number of office desks to be reviewed of 5 exceeds the number threshold of 4, it is considered that the office desk in the picture of the feasibility study report is not compliant.

[0092] In addition, when there are multiple target objects in the image to be reviewed, compliance review is performed on the multiple target objects in turn according to the above method.

[0093] The technical solution provided by the present application obtains the picture to be reviewed in the feasibility study report, which includes at least one target object; based on the picture to be reviewed and the segmentation model, an object knowledge base is constructed by using optical character recognition technology, and the object knowledge base includes a mapping relationship between the target object, the area of ​​the target object and the number of the picture to be reviewed; based on the object knowledge base, prompt words are constructed; based on the feasibility study report, the object knowledge base, the prompt words and the large language generation model, a compliance review of the target object is performed, thereby realizing automated compliance review of materials in the picture of the feasibility study report through the large language generation model, the segmentation model and the optical character recognition technology, thereby improving the efficiency and accuracy of the compliance review of materials.

[0094] Figure 6 This is a structural block diagram of a material compliance review device provided in one embodiment of the present application. For the sake of convenience, only the parts related to the embodiment of the present application are shown. Figure 6 The material compliance review device 600 may include an acquisition module 601, a first construction module 602, a second construction module 603 and a review module 604.

[0095] In one implementation, the apparatus 600 may be used to implement the above Figure 1 For example, the acquisition module 601 is used to implement S110, the first construction module 602 is used to implement S120, the second construction module 603 is used to implement S130, and the review module 604 is used to implement S140.

[0096] In another implementation, the device 600 can also be used to implement the above Figure 2 For example, the first building block 602 is used to implement S210 to S250.

[0097] In another implementation, the device 600 can also be used to implement the above Figure 3 For example, the first building block 602 is used to implement S310 to S330.

[0098] In another implementation, the device 600 can also be used to implement the above Figure 4 For example, the first building block 602 is used to implement S410 to S440.

[0099] In another implementation, the device 600 can also be used to implement the above Figure 5 For example, the review module 604 is used to implement S510 to S530.

[0100] In an embodiment of the present application, a picture to be reviewed in a feasibility study report is obtained, wherein the picture to be reviewed includes at least one target object; based on the picture to be reviewed and a segmentation model, an object knowledge base is constructed using optical character recognition technology, wherein the object knowledge base includes a mapping relationship between the target object, the area of ​​the target object, and the serial number of the picture to be reviewed; based on the object knowledge base, prompt words are constructed; based on the feasibility study report, the object knowledge base, the prompt words, and a large language generation model, a compliance review is performed on the target object, thereby realizing an automated compliance review of materials in the picture of the feasibility study report through the large language generation model, the segmentation model, and the optical character recognition technology, thereby improving the efficiency and accuracy of the compliance review of materials.

[0101] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0102] Figure 7 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 7 As shown, the electronic device 7 of this embodiment includes: at least one processor 70 ( Figure 7 Only one is shown in the figure) a processor, a memory 71, and a computer program 72 stored in the memory 71 and executable on the at least one processor 70, the processor 70 implements the above-mentioned when executing the computer program 72 Figures 1 to 5 The steps in any method embodiment.

[0103] The electronic device 7 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The electronic device may include, but is not limited to, a processor 70 and a memory 71. Those skilled in the art will appreciate that Figure 7It is only an example of the electronic device 7 and does not constitute a limitation on the electronic device 7. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include input and output devices, network access devices, etc.

[0104] The processor 70 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0105] In some embodiments, the memory 71 may be an internal storage unit of the electronic device 7, such as a hard disk or memory of the electronic device 7. In other embodiments, the memory 71 may also be an external storage device of the electronic device 7, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 7. Further, the memory 71 may also include both an internal storage unit and an external storage device of the electronic device 7. The memory 71 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program. The memory 71 may also be used to temporarily store data that has been output or is to be output.

[0106] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0107] An embodiment of the present application also provides a network device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor implements the steps in any of the above-mentioned method embodiments when executing the computer program.

[0108] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0109] An embodiment of the present application provides a computer program product. When the computer program product runs on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0110] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0111] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0112] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0113] In the embodiments provided in the present application, it should be understood that the disclosed devices / network equipment and methods can be implemented in other ways. For example, the device / network equipment embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0114] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0115] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A material compliance review method, characterized in that: The method comprises: Acquire a picture to be reviewed in the feasibility study report, wherein the picture to be reviewed includes at least one target object; Based on the image to be reviewed and the segmentation model, an object knowledge base is constructed using optical character recognition technology, wherein the object knowledge base includes a mapping relationship between the target object, the area of ​​the target object, and the serial number of the image to be reviewed; Constructing prompt words based on the object knowledge base; Based on the feasibility study report, the object knowledge base, the prompt words and the large language generation model, a compliance review is performed on the target object.

2. The method according to claim 1, characterized in that The object knowledge base is constructed based on the image to be reviewed and the segmentation model using optical character recognition technology, including: Extracting text information from the image to be reviewed based on optical character recognition technology, wherein the text information includes the object name of the target object and the number of the image to be reviewed; Based on the segmentation model, segment the target object from the image to be reviewed, and determine the total number of pixels of the target object; Based on the image to be reviewed, matching the text information with the target object; Determining the area of ​​the target object based on the total number of pixels of the image to be reviewed and the target object; An object knowledge base is constructed based on the target object, the text information and the area of ​​the target object.

3. The method according to claim 2, characterized in that The matching of the text information with the target object based on the image to be reviewed includes: Based on the image to be reviewed, determining the position of the text detection box corresponding to the multiple text information; Calculating the distance between each text detection frame and the target object; The text information is matched with the target object based on the distance between each text detection frame and the target object.

4. The method according to claim 2, characterized in that: The determining the area of ​​the target object based on the total number of pixels of the image to be reviewed and the target object includes: Determining a scale based on the image to be reviewed and preset specification parameters of the target object, wherein the preset specification parameters are used to represent a conventional size of the target object; Determine the number of pixels per unit area in the picture to be reviewed based on the scale and the picture to be reviewed; The area of ​​the target object is determined based on the total number of pixels of the target object and the number of pixels per unit area in the image to be reviewed.

5. The method according to claim 4, characterized in that The determining of the scale based on the preset specification parameters of the image to be reviewed and the target object includes: Determining, based on the image to be reviewed, measurement specification parameters of the target object, where the measurement specification parameters are used to characterize the size of the target object in the image to be reviewed; The scale is determined based on preset specification parameters and measurement specification parameters of the target object.

6. The method according to claim 4, characterized in that The determining, based on the scale and the picture to be reviewed, the number of pixels per unit area in the picture to be reviewed includes: Based on the picture to be reviewed, determining position information of a first coordinate point and a second coordinate point, where the first coordinate point and the second coordinate point are any two coordinate points in the picture to be reviewed; Determine first area information based on the position information of the first coordinate point and the second coordinate point; Determine second area information based on the position information of the first coordinate point and the second coordinate point, and the scale; Based on the first area information and the second area information, the number of pixels of the unit area in the picture to be reviewed is determined.

7. The method according to claim 1, characterized in that The step of conducting compliance review on the target object based on the feasibility study report, the object knowledge base, the prompt words, and the large language generation model includes: Based on the feasibility study report and the large language generation model, determine the area information and quantity information of the target object to be reviewed; Determining a quantity threshold and an area threshold of the target objects based on the feasibility study report, the object knowledge base, and the large language generation model; Based on the prompt words, the large language generation model, and the area information to be reviewed, the quantity information to be reviewed, the area threshold, and the quantity threshold of the target object, a compliance review is performed on the target object.

8. A material compliance review device, characterized in that: The device comprises: An acquisition module, used to acquire a picture to be reviewed in the feasibility study report, wherein the picture to be reviewed includes at least one target object; A first construction module is used to construct an object knowledge base based on the image to be reviewed and the segmentation model by using optical character recognition technology, wherein the object knowledge base includes a mapping relationship between the target object, the area of ​​the target object and the serial number of the image to be reviewed; A second construction module is used to construct prompt words based on the object knowledge base; A review module is used to perform a compliance review on the target object based on the feasibility study report, the object knowledge base, the prompt words and the large language generation model.

9. An electronic 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 method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images