Tax optimization method based on artificial intelligence

Abstract

The invention provides a tax optimization method based on artificial intelligence, and relates to an image processing technology, and the method comprises the steps: obtaining the auditing information of a financial end for an electronic certificate set, generating an integration structure corresponding to the electronic certificate set according to the auditing information, generating a summary code corresponding to the integration structure, and carrying out the collection of the summary code. Generating a code combination image according to the summary code, sending the code combination image to a first user side corresponding to the electronic certificate set, responding to data uploading information of a second user side, obtaining an identification image of a certificate arrangement device shot by data acquisition equipment, obtaining label information and certificate information in the identification image, and sending the label information and the certificate information to the electronic certificate set; and generating voucher association data corresponding to the voucher information according to the label information, and sending the voucher association data to the financial terminal. The bill information and the user information can be associated.

Description

Technical Field

[0001] The present invention relates to image processing technology, and in particular to a tax optimization method based on artificial intelligence. Background Art

[0002] In the current financial management process, the collection, organization and aggregation of bills is a complex and tedious process. Due to the diverse sources of bills, including both paper bills and electronic bills, bill management becomes particularly difficult.

[0003] A Chinese patent with publication number CN113095307A discloses a method for automatically identifying financial voucher information. An image analysis module extracts image feature information from the image information of a paper financial voucher, and then the image analysis module analyzes the image feature information to obtain data information, which is then stored electronically. This invention only implements digital storage of the data information of the financial voucher, but cannot associate the voucher information with the user information. When the financial staff reviews or reimburses the voucher, they need to match the voucher information with the user before they can perform subsequent review or reimbursement operations, resulting in low processing efficiency.

[0004] Therefore, how to associate ticket information with user information becomes an urgent problem to be solved. Summary of the invention

[0005] The present invention provides a tax optimization method based on artificial intelligence, which can associate bill information with user information.

[0006] The first aspect of the present invention provides a tax optimization method based on artificial intelligence, comprising: Acquire the audit information of the electronic voucher set by the financial end, and generate the integration structure corresponding to the electronic voucher set according to the audit information; generating a summary code corresponding to the integrated structure, generating a coded combination image according to the summary code and sending it to a first user terminal corresponding to the electronic certificate set; In response to the data upload information of the second user terminal, obtaining the identification image taken by the data acquisition device of the voucher arrangement device; The tag information and the voucher information in the identified image are obtained, and the voucher-related data corresponding to the voucher information is generated according to the tag information and sent to the financial end.

[0007] Optionally, in a possible implementation of the first aspect, obtaining audit information of the electronic voucher set by the financial end, and generating an integrated structure corresponding to the electronic voucher set according to the audit information includes: Determining the compliance certificate selected by the financial end from the set of electronic certificates according to the audit information; Obtaining the bill type of each of the compliance vouchers, and determining that the compliance vouchers of the same type are the same voucher combination; Generate a credential node corresponding to each of the compliant credentials and a type node corresponding to each of the credential combinations, and connect the credential nodes and type nodes corresponding to each of the credential combinations to obtain a plurality of sub-combination structures; Generate multiple bill type groups selected by the financial end, each bill type group includes at least two bill types, and generate an integration node corresponding to each bill type group; The sub-combination structures corresponding to each of the bill type groups and the integration node are connected to obtain a plurality of integration structures corresponding to the electronic voucher set.

[0008] Optionally, in a possible implementation manner of the first aspect, generating a summary code corresponding to the integrated structure, generating a coded combination image according to the summary code and sending the coded combination image to a first user terminal corresponding to the electronic certificate set includes: Retrieving a blank arrangement template corresponding to the voucher arrangement device, wherein the blank arrangement template includes a label display area corresponding to each label component in the voucher arrangement device, and the label display area is a triangular area located at the top corner of the blank arrangement template; Determine other areas of the blank layout template except the label display area as layout areas, count the number of structures of each integrated structure, and determine the standard coding specification corresponding to the layout area according to the number of structures; Acquire the number of vouchers corresponding to each of the integrated structures, and offset the standard encoding specification according to the number of vouchers to obtain the encoding specification corresponding to each of the integrated structures; Generate a scan code corresponding to each integrated structure, adjust the scan code according to the coding specification to obtain a summary code, arrange each summary code in the arrangement area based on the coding arrangement strategy to obtain a coded combined image and send it to the first user end.

[0009] Optionally, in a possible implementation manner of the first aspect, counting the number of structures of each of the integrated structures, and determining the standard coding specification corresponding to the arrangement area according to the number of structures, includes: Obtaining the long side parallel direction and the short side parallel direction of the blank layout template, and determining the area edge of the label display area located inside the blank layout template as the target edge; Taking the line segment points of each target edge as the connection starting point and the connection end point, a plurality of mutually perpendicular connection lines are generated in the direction parallel to the long edge and the direction parallel to the short edge respectively; A plurality of rectangular areas are obtained according to the connecting lines, and the rectangular area with the largest area is determined as the target area; Counting the number of structures of each integrated structure, determining a division mode of the target area according to the number of structures, wherein the division mode includes an even-number division mode and an odd-number division mode, and dividing the target area according to the division mode to obtain a plurality of divided areas; The area specification of the divided area is determined to be the standard coding specification.

[0010] Optionally, in a possible implementation manner of the first aspect, a division mode of the target area is determined according to the number of structures, the division mode includes an even-number division mode and an odd-number division mode, and the target area is divided according to the division mode to obtain a plurality of divided areas, including: When the number of structures is an even number, determining that the division mode is an even number division mode, and obtaining multiple pairs of factor pairs corresponding to the number of structures according to the even number division mode; Determine the factor pair with the smallest difference as the target factor pair, the long side in the target area corresponds to the target value with the larger value in the target factor pair, and the short side in the target area corresponds to the target value with the smaller value in the target factor; Obtaining a first value according to the difference between the target value and the offset value, generating a plurality of first dividing lines perpendicular to the long side and the short side and corresponding to the first values ​​respectively, and dividing the target area according to the first dividing lines to obtain a plurality of divided areas; When the number of structures is an odd number, determining that the division mode is an odd number division mode, and obtaining the long side of the target area as the division side; A second value is obtained according to the difference between the structure quantity and the offset value, a plurality of second dividing lines perpendicular to the dividing edge and corresponding to the second value are generated, and the target area is divided according to the second dividing lines to obtain a plurality of divided areas.

[0011] Optionally, in a possible implementation manner of the first aspect, obtaining the number of credentials corresponding to each of the integrated structures, offsetting the standard encoding specification according to the number of credentials, and obtaining the encoding specification corresponding to each of the integrated structures includes: Determine the maximum voucher quantity as the standard quantity, and obtain the quantity difference between the standard quantity and the remaining voucher quantities; Retrieving a ratio matching table, traversing each of the quantity differences and a plurality of preset difference intervals in the ratio matching table, and determining that a preset reduction ratio corresponding to the preset difference interval in which each of the quantity differences is located is an offset reduction ratio; The standard coding specification is reduced and adjusted according to the offset reduction ratio to obtain the coding specification corresponding to each of the integrated structures.

[0012] Optionally, in a possible implementation manner of the first aspect, arranging the summary codes in the arrangement area based on a coding arrangement strategy to obtain a coded combined image and sending it to the first user terminal includes: Arranging the summary codes in order from large to small according to the coding specifications to obtain a coding arrangement sequence; Determine that the divided areas in the arrangement area that are in a row in a direction parallel to the long side are the same arrangement area group; Firstly, a plurality of divided areas in each of the arrangement area groups are arranged according to the direction parallel to the long sides, and then each of the arrangement area groups is arranged according to the direction parallel to the short sides to obtain an arrangement area sequence; Determine the correspondence between the summary codes and the divided areas having the same arrangement positions in the coding arrangement sequence and the arrangement area sequence; The center point of each of the divided areas is determined as a positioning point, the corresponding summary code is positioned based on the positioning point, and the summary code is arranged in the corresponding divided area to obtain a coded combined image which is sent to the first user terminal.

[0013] Optionally, in a possible implementation manner of the first aspect, acquiring tag information and credential information in the recognition image, and generating credential-related data corresponding to the credential information according to the tag information and sending the credential-related data to the financial end includes: Extracting the credential device outline in the recognition image, calling the information extraction template corresponding to the credential arrangement device, and acquiring the label information and credential information in the credential device outline according to the information extraction template; Retrieving a detailed filling table, wherein the detailed filling table includes a plurality of filling slots, each of the filling slots corresponds to a corresponding preset label; Determine a preset tag corresponding to the identification tag of the tag information, and fill the tag information into the filling slot corresponding to the preset tag to obtain a voucher detail list; The voucher details table is associated with the voucher information to obtain voucher associated data and sent to the financial end.

[0014] Optionally, in a possible implementation manner of the first aspect, extracting a credential device outline in the recognition image, calling an information extraction template corresponding to the credential arrangement device, and acquiring label information and credential information in the credential device outline according to the information extraction template includes: Acquire the positioning angle and positioning point of the outline of the credential device, position the information extraction template based on the positioning angle and positioning point, and superimpose the information extraction template on the top of the outline of the credential device; The tag information in the voucher device profile is obtained according to the information extraction area in the information extraction template, and the voucher information in the voucher device profile is obtained according to the voucher extraction area in the information extraction template.

[0015] Optionally, in a possible implementation manner of the first aspect, obtaining a positioning angle and a positioning position point of the outline of the credential device, positioning the information extraction template based on the positioning angle and the positioning position point, and superimposing the information extraction template on the outline of the credential device includes: Determine the center point of the outline of the credential device as the positioning position point, obtain the outline long side direction in the outline of the credential device, and determine the angle between the outline long side direction and the reference long side direction as the positioning angle; The center point of the information extraction template is located according to the positioning position point, and the information extraction template is rotated based on the positioning angle and the reference rotation direction, so that the information extraction template is superimposed on the outline of the credential device.

[0016] Optionally, in a possible implementation manner of the first aspect, acquiring tag information in the credential device profile according to the information extraction area in the information extraction template, and acquiring credential information in the credential device profile according to the credential extraction area in the information extraction template, includes: Extracting a first extraction character corresponding to the label sub-area in the information extraction area, and determining an identification label corresponding to the information extraction area according to the first extraction character; Acquire a second extracted character in the information extraction area except the first extracted character, and determine that the second extracted character is label information corresponding to the identification label; Extracting the voucher image in the voucher extraction area, and when the voucher image is a coded combination image, obtaining the voucher information according to the electronic voucher corresponding to the coded combination image; When the voucher image is a bill image, the current bill type of the bill image is acquired, and the voucher information is obtained according to the current bill type and the voucher image.

[0017] Optionally, in a possible implementation manner of the first aspect, when the credential image is a bill image, acquiring a current bill type of the bill image, and obtaining the credential information according to the current bill type and the credential image includes: Determine a preset vertex angle corresponding to each of the identification tags, wherein the preset vertex angle includes an upper vertex angle and a lower vertex angle; Acquire the center point of each of the information extraction areas as the target point, and determine the direction from the target point corresponding to the upper vertex angle to the target point corresponding to the lower vertex angle as the traversal direction; The text in the bill image is extracted in sequence based on the traversal direction, and the traversal is stopped when the text corresponding to the preset bill type is traversed, and the corresponding preset bill type is determined to be the current bill type, and the voucher information is obtained according to the current bill type and the voucher image.

[0018] Optionally, in a possible implementation manner of the first aspect, determining a preset tag corresponding to the identification tag of the tag information, and filling the tag information into a filling slot corresponding to the preset tag to obtain a voucher detail table includes: When the identification tag does not have corresponding tag information, determining the start upload time of the second user terminal, and obtaining the backtracking time period according to the start upload time and the current time; Acquire multiple historical recognition images in the retrospective time period, acquire historical recognition tags corresponding to each of the historical recognition images, and determine the historical recognition tag corresponding to the recognition tag as a target recognition tag; Determine that the historical tag information corresponding to the target identification tag is the tag information corresponding to the identification tag, and fill the tag information into the filling slot corresponding to the preset tag to obtain a voucher detail list.

[0019] A second aspect of the present invention provides a tax optimization system based on artificial intelligence, comprising: An acquisition module, used to acquire the audit information of the electronic voucher set by the financial end, and generate an integrated structure corresponding to the electronic voucher set according to the audit information; A generating module, used for generating a summary code corresponding to the integrated structure, generating a coded combination image according to the summary code and sending it to a first user terminal corresponding to the electronic voucher set; A response module, used to respond to the data upload information of the second user terminal and obtain the identification image taken by the data acquisition device on the voucher arrangement device; The sending module is used to obtain the label information and the voucher information in the identified image, generate the voucher-related data corresponding to the voucher information according to the label information, and send it to the financial end.

[0020] The beneficial effects of the present invention are as follows: 1. The present invention can organize the bill information of each user through the voucher processing device, realize the rapid classification and review of bills, and make the subsequent reimbursement process of finance more standardized and efficient. The present invention can integrate the electronic bills of the user to form a corresponding integrated structure. Through the integrated structure, the connection and type of each electronic bill can be clearly seen, and the bill management can be made more orderly, thereby improving the accuracy and processing efficiency of bill processing.

[0021] 2. The present invention can generate a summary code corresponding to the integrated structure. By scanning the summary code, one or more electronic invoices corresponding to the integrated structure can be quickly and accurately retrieved, thereby greatly simplifying the subsequent reimbursement or review process and improving work efficiency. The present invention can adjust the standard coding specifications accordingly in combination with the number of vouchers corresponding to the integrated structure, so that the adjusted coding specifications can intuitively reflect the number of vouchers contained in different integrated structures.

[0022] 3. The present invention can obtain an identification image containing complete bill information by photographing the voucher arrangement device through a data acquisition device, and can more accurately obtain detailed information of the bill by analyzing the identification image. The present invention can accurately extract label information and voucher information from the identification image by calling the information recognition template, and generate voucher-related data, which can improve the accuracy of bill processing and make the classification and management of bills more accurate and efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a flowchart of a tax optimization method based on artificial intelligence provided by an embodiment of the present invention; Figure 2 Is a schematic diagram of a connection line for determining a target area provided in an embodiment of the present invention; Figure 3 It is a schematic diagram of the target area for arranging the summary code according to an embodiment of the present invention; Figure 4 It is a schematic diagram of the divided areas obtained after dividing the target area when the number of integrated structures provided in the embodiment of the present invention is an even number; Figure 5 It is a schematic diagram of the divided areas obtained after dividing the target area when the number of integrated structures provided in the embodiment of the present invention is an odd number; Figure 6 It is a schematic diagram of the positioning angle when positioning the information extraction template provided in an embodiment of the present invention; Figure 7 It is a structural diagram of an artificial intelligence-based tax optimization system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0026] See also Figure 1 , is a schematic diagram of a tax optimization method based on artificial intelligence provided by an embodiment of the present invention, Figure 1 The execution subject of the method shown may be a software and / or hardware device. The execution subject of the present application may include but is not limited to at least one of the following: user equipment, network equipment, etc. Among them, user equipment may include but is not limited to computers, smart phones, personal digital assistants (PDA) and the electronic devices mentioned above. Network equipment may include but is not limited to a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers, wherein cloud computing is a type of distributed computing, a super virtual computer composed of a group of loosely coupled computers. This embodiment does not limit this. It includes steps S1 to S4, as follows: S1, obtaining the audit information of the electronic voucher set by the financial end, and generating an integrated structure corresponding to the electronic voucher set according to the audit information.

[0027] Among them, the financial end refers to the terminal held by the financial personnel, such as a computer, the electronic voucher set can be a collection of electronic invoices that need to be reimbursed, the audit information refers to the judgment information on whether each electronic voucher is compliant and can be reimbursed after the financial personnel review the electronic voucher set, and the integration structure refers to the corresponding structure after the electronic invoices that meet the reimbursement rules are integrated.

[0028] In actual applications, since the electronic invoices that users need to reimburse are of various formats and come from a wide range of sources, in order to quickly obtain and process these invoices, the users' electronic invoices can be integrated to form a corresponding integrated structure, so that the connection and type between the various invoices can be clearly seen through the integrated structure.

[0029] Specifically, since not all electronic invoices can be reimbursed, for example, invalid invoices such as invoices with missing information and invoices with irregular stamps cannot be reimbursed. In order to avoid invalid invoices from interfering with subsequent bill integration work, when users want to reimburse electronic invoices, they can first send the electronic voucher set containing all electronic invoices to the financial staff for review. During the review process, the financial staff can review the compliance and reimbursement eligibility of each electronic voucher, and obtain the review information corresponding to the electronic voucher set, so that the electronic invoices that have passed the review can be integrated according to the review information to obtain the corresponding integration structure.

[0030] In some embodiments, the specific implementation of step S1 may be: S11, determining the compliance certificate selected by the financial end from the electronic certificate set according to the audit information.

[0031] Among them, compliance vouchers refer to electronic receipts that can be used for subsequent invoice processing procedures.

[0032] Specifically, when the financial personnel on the financial side review the electronic invoices in the electronic voucher set, the electronic invoices that comply with the subsequent invoice processing process can be used as compliant vouchers. For example, if it is a reimbursement process, then the compliant voucher can be an electronic invoice that can be reimbursed. Therefore, based on the review information of the financial personnel, the electronic invoices that have passed the review can be marked as compliant vouchers, thereby determining the compliant vouchers in the electronic voucher set.

[0033] S12, obtaining the bill type of each of the compliance certificates, and determining that the compliance certificates of the same type are the same group of certificate combinations.

[0034] In actual applications, there are many types of electronic vouchers that comply with the invoice processing flow. For example, for the reimbursement process, the types of invoices that can be reimbursed include special value-added tax invoices, general value-added tax invoices, receipts, etc. The bill type corresponding to each compliant voucher can be obtained. In order to facilitate subsequent review, classification or reimbursement processing, the same type of compliant vouchers can be organized into a group as the same group of voucher combinations, so that the quantity and amount of different types of electronic vouchers can be more clearly understood, and subsequent bill management or reimbursement processing can be more effectively performed.

[0035] For example, if there is a batch of compliant documents, including 5 special VAT invoices, 3 ordinary VAT invoices and 2 receipts, then the 5 special VAT invoices can be regarded as the same voucher combination, the 3 ordinary VAT invoices can be regarded as the same voucher combination, and the 2 receipts can be regarded as the same voucher combination.

[0036] Among them, the bill type refers to the type of electronic certificate, and the certificate combination refers to the combination of compliant certificates of the same type.

[0037] S13, generating a credential node corresponding to each of the compliant credentials and a type node corresponding to each of the credential combinations, and connecting the credential nodes and type nodes corresponding to each of the credential combinations to obtain a plurality of sub-combination structures.

[0038] Specifically, a node corresponding to each compliant credential, namely a credential node, and a node corresponding to each credential combination, namely a type node, can be generated. By connecting each credential node with the corresponding type node, multiple sub-combination structures can be obtained.

[0039] For example, if there is a batch of compliant vouchers, including 5 special VAT invoices, 3 ordinary VAT invoices and 2 receipts, then the type nodes corresponding to the special VAT invoice type, the type nodes corresponding to the ordinary VAT invoice type and the type nodes corresponding to the receipt type can be generated respectively, and voucher nodes corresponding to 5 special VAT invoices, 3 ordinary VAT invoices and 2 receipts can be generated. By connecting the voucher nodes corresponding to the 5 special VAT invoices with the type nodes corresponding to the special VAT invoice type, the sub-combination structure corresponding to the special VAT invoice type can be obtained. By connecting the voucher nodes corresponding to the 3 ordinary VAT invoices with the type nodes corresponding to the ordinary VAT invoice type, the sub-combination structure corresponding to the ordinary VAT invoice type can be obtained. By connecting the voucher nodes corresponding to the 2 receipts with the type nodes corresponding to the receipt type, the sub-combination structure corresponding to the receipt type can be obtained.

[0040] Among them, the credential node refers to the node corresponding to each compliant credential, the type node refers to the node corresponding to each credential combination, and the sub-combination structure refers to the sub-structure obtained by connecting the credential node and the type node corresponding to each credential combination.

[0041] S14, generating a plurality of bill type groups selected by the financial end, each of the bill type groups including at least two of the bill types, and generating an integration node corresponding to each of the bill type groups.

[0042] In actual applications, the financial end can select multiple bill types from the existing bill types to form different bill type groups according to actual needs. All bills in the same bill type group can be managed and operated in a unified manner later. Each bill type group contains at least two different types of bills. For example, the financial end can combine special value-added tax invoices and ordinary value-added tax invoices into a bill type group to classify sales-related bills. Another bill type group can include receipts, purchase invoices and payment vouchers to classify purchase and payment-related bills.

[0043] After determining each bill type group, a corresponding integration node can be generated for each bill type group. The integration node can serve as a representative of the bill type group. For example, for a bill type group that includes two bill types, namely, special value-added tax invoices and ordinary value-added tax invoices, an integration node corresponding to the bill type can be generated. For a bill type group that includes two bill types, namely, receipts and purchase invoices, an integration node corresponding to the bill type group can be generated.

[0044] The bill type group refers to a group consisting of multiple bill types, and the integration node refers to a node corresponding to the bill type group.

[0045] S15, connecting the sub-combination structures corresponding to each of the bill type groups and the integration node to obtain a plurality of integration structures corresponding to the electronic voucher set.

[0046] For example, for a bill type group that includes two bill types, namely, special value-added tax invoices and ordinary value-added tax invoices, the corresponding sub-combination structures include a sub-combination structure corresponding to the special value-added tax invoice type and a sub-combination structure corresponding to the ordinary value-added tax invoice type. By connecting these two sub-combination structures with the integration node corresponding to the bill type group, the integration structure corresponding to the bill type group can be obtained. Since the financial end may select multiple bill type groups, multiple corresponding integration structures can be obtained.

[0047] Through the above implementation, bill management can be made more orderly, and the possibility of confusion and errors can be reduced, thereby improving the accuracy of bill processing and improving processing efficiency.

[0048] S2, generating a summary code corresponding to the integrated structure, generating a coded combination image according to the summary code, and sending the image to a first user terminal corresponding to the electronic certificate set.

[0049] It can be understood that after obtaining the integrated structure, in order to facilitate the retrieval of the electronic invoice corresponding to the integrated structure, this solution can generate a summary code corresponding to the integrated structure. By scanning the summary code, one or more electronic invoices corresponding to the integrated structure can be quickly and accurately retrieved, thereby improving the efficiency of bill processing.

[0050] Since there may be multiple integration structures corresponding to the user, there may also be multiple corresponding summary codes. Multiple summary codes can be integrated to generate corresponding coding combination images. The user does not need to save multiple separate summary code images, but only needs to save one coding combination image. When making unified reimbursements later, the user can print out the coding combination image as part of the reimbursement materials. By scanning the printed coding combination image, all relevant electronic invoice information can be obtained, thereby improving reimbursement efficiency.

[0051] Among them, the summary code refers to the scanning code corresponding to the integrated structure, for example, it can be a QR code or a barcode, the coded combination image refers to an image composed of multiple summary codes, and the first user terminal refers to the terminal held by the user who makes electronic invoice reimbursement.

[0052] Based on the above embodiment, the specific implementation of step S2 may be: S21, calling a blank arrangement template corresponding to the voucher arrangement device, wherein the blank arrangement template includes a label display area corresponding to each label component in the voucher arrangement device, and the label display area is a triangular area located at the top corner of the blank arrangement template.

[0053] Among them, the voucher organizing device refers to a special invoice clamp device, which can be used to clamp invoices, and there are four triangular components in the four top corners of the device, and the basic classification information of the invoice can be filled in on the components. For example, the component in the upper left corner can be used to fill in the name and department information, the component in the upper right corner can be used to fill in the reason for the invoice, the component in the lower left corner can be used to fill in the department information, and the component in the lower right corner can be used to fill in the date and invoicing unit information. The blank layout template refers to the rectangular template corresponding to the voucher organizing device. The label component refers to the component on the voucher organizing device used for information filling. In order to avoid the label component blocking the invoice information, the label component is generally located at the four top corners of the voucher organizing device, and the label component can fix the bill. The label display area refers to the area corresponding to the label component.

[0054] In order to uniformly process the reimbursement or review of bills and improve the efficiency of bill processing, this solution can uniformly upload bills through photographed images when uploading bills. Specifically, for paper invoices, the paper invoice can be directly placed on the voucher sorting device for image capture, thereby realizing the uploading of relevant data. For electronic invoices, a corresponding coded combination image will be generated for the electronic invoice. By printing out the coded combination image and then placing it on the voucher sorting device for image capture, unified data upload can be realized, and the specifications during printing can be consistent with the specifications of the bill clamped by the voucher sorting device.

[0055] In some embodiments, when generating a coded combination image corresponding to an electronic receipt, in order to ensure that the coded combination image will not be obscured by the label component on the voucher organization device when uploading data, the coded combination image can be generated in combination with a blank layout template corresponding to the voucher organization device.

[0056] Specifically, a blank layout template corresponding to the voucher organization device can be retrieved, and the specifications of the blank layout template are the same as the specifications of the captured image corresponding to the voucher organization device, so that each summary code can be arranged according to the blank layout template to ensure that it will not be blocked by the label component, and the blank layout template includes a label display area corresponding to the label components in the four top corner areas of the device. The label display area can exclude interference areas that may block the coded combination image.

[0057] S22, determining other areas in the blank arrangement template except the label display area as arrangement areas, counting the number of structures of each integrated structure, and determining the standard coding specification corresponding to the arrangement area according to the number of structures.

[0058] Specifically, in order to ensure that each summary code is not blocked by the label component, the area other than the four label display areas in the blank layout template can be determined as the layout area, and the summary codes can be arranged accordingly in the layout area. Since the number of integrated structures corresponding to each user may be different, in order to be able to customize the coding configuration based on the actual needs of the user, the standard coding specifications corresponding to different users can be determined based on the number of integrated structures. By counting the number of integrated structures, the corresponding number of structures can be obtained. According to the number of structures, the optimal specifications for displaying each summary code in the layout area, that is, the standard coding specifications, can be determined.

[0059] Among them, the layout area refers to the area in the blank layout template except for the four label display areas, the number of structures refers to the number of integrated structures, and the standard coding specification refers to the optimal specification for displaying the summary code in the layout area under the number of structures.

[0060] In some embodiments, the step S22 of “counting the number of structures of each integrated structure, and determining the standard coding specification corresponding to the arrangement area according to the number of structures” includes the following steps: S221, obtaining the long side parallel direction and the short side parallel direction of the blank layout template, and determining the area edge in the label display area located inside the blank layout template as the target edge.

[0061] Specifically, since the blank layout template is a rectangular template, the direction parallel to the long side of the blank layout template, that is, the long side parallel direction, and the direction parallel to the short side of the blank layout template, that is, the short side parallel direction, can be obtained. Since the four label display areas are located in the four vertex areas of the blank layout template and are triangular areas, they contain three area edges, two of which are the long side and the short side of the blank layout template, and the third area edge connecting these two area edges is located inside the blank layout template. The area edge of the label display area located inside the blank layout template can be determined as the target edge.

[0062] Among them, the long side parallel direction refers to the direction parallel to the long side of the blank layout template, the short side parallel direction refers to the direction parallel to the short side of the blank layout template, the area edge refers to the three edges corresponding to the label display area, and the target edge refers to the area edge located inside the blank layout template.

[0063] S222: Using the line segment points of each target edge as the connection starting point and the connection end point, generate a plurality of mutually perpendicular connection lines in the direction parallel to the long edge and the direction parallel to the short edge.

[0064] Specifically, each target edge has multiple corresponding line segment points. The line segment points on one of the target edges can be determined as the connection starting point, and the line segment points on the adjacent target edge can be determined as the connection end point. By connecting each connection starting point with the connection end point, multiple mutually perpendicular connection lines can be generated in the direction parallel to the long side and the direction parallel to the short side.

[0065] See also Figure 2 , which is a schematic diagram of a connection line for determining a target area provided by an embodiment of the present invention, such as Figure 2 As shown in , multiple line segment points on target edge 1 can be determined as connection starting points, and line segment points on target edge 2 and target edge 3 adjacent to target edge 1 can be determined as connection end points. By connecting the connection starting point on target edge 1 with the connection end point on target edge 2, multiple connection lines parallel to the parallel direction of the short sides can be generated. By connecting the connection starting point on target edge 1 with the connection end point on target edge 3, multiple connection lines parallel to the parallel direction of the long sides can be generated, and the connection lines parallel to the parallel direction of the short sides and the connection lines parallel to the parallel direction of the long sides are perpendicular to each other. Similarly, multiple line segment points on target edge 3 can be determined as connection starting points, and line segment points on target edge 1 and target edge 4 adjacent to target edge 3 can be determined as connection end points. Similarly, multiple mutually perpendicular connection lines can be generated in the parallel directions of the long sides and the parallel directions of the short sides, respectively.

[0066] Among them, the line segment point refers to the multiple points that constitute the target edge, the connection starting point refers to the line segment point where the connection line starts, the connection end point refers to the line segment point where the connection line ends, and the connection line refers to a line segment that is parallel to the long side or the short side and intersects with the two target edges.

[0067] S223, obtaining a plurality of rectangular areas according to the connecting lines, and determining the rectangular area with the largest area as the target area.

[0068] See also Figure 3 , which is a schematic diagram of a target area for arranging summary codes provided by an embodiment of the present invention, such as Figure 3 As shown in , multiple corresponding rectangular areas can be obtained according to each connecting line. It can be seen from the figure that the area of ​​the rectangular area formed by connecting line 1, connecting line 2, connecting line 3, and connecting line 4 is the largest, and the corresponding rectangular area can be determined as the target area.

[0069] The rectangular area refers to the area formed by four mutually perpendicular connecting lines, and the target area refers to the rectangular area with the largest area.

[0070] S224, counting the number of structures of each integrated structure, determining a division mode of the target area according to the number of structures, wherein the division mode includes an even-number division mode and an odd-number division mode, and dividing the target area according to the division mode to obtain a plurality of divided areas.

[0071] In practical applications, when the summary codes corresponding to multiple integrated structures need to be arranged in the target area, in order to maximize the use of the space in the target area, the target area can be divided according to the number of integrated structures. When dividing, in order to ensure that the summary codes corresponding to each integrated structure can present an orderly and regular layout in the target area, the target area can be divided according to the parity of the number of structures, thereby ensuring that the summary codes corresponding to each integrated structure can have sufficient arrangement space and that their layouts are balanced.

[0072] Specifically, the number of structures corresponding to the integrated structure can be obtained. In order to ensure that the specifications of each corresponding divided area can remain consistent and reach the maximum possible under the number of structures, the division mode corresponding to the target area can be determined according to the parity of the number of structures, and the target area can be divided accordingly according to different division modes to obtain multiple division areas. For example, if the number of structures is an even number, the division mode corresponding to the target area can be determined to be an even number division mode. If the number of structures is an odd number, the division mode corresponding to the target area can be determined to be an odd number division mode.

[0073] Among them, the division mode refers to the mode of dividing the target area into regions, the even-number division mode refers to the mode of dividing the target area into regions by combining two sides of the target area when the number of structures is an even number, the odd-number division mode refers to the mode of dividing the target area into regions by combining one side of the target area when the number of structures is an odd number, and the division area refers to the multiple small areas obtained after dividing the target area into regions.

[0074] Through the above implementation, the space of the target area can be utilized to the maximum extent, and it can be ensured that each summary code has enough space for clear display.

[0075] In some embodiments, the step S224 of "determining a division mode of the target area according to the number of structures, the division mode including an even-number division mode and an odd-number division mode, and dividing the target area according to the division mode to obtain a plurality of divided areas" includes the following steps: S2241: When the number of structures is an even number, determine that the division mode is an even number division mode, and obtain multiple pairs of factor pairs corresponding to the number of structures according to the even number division mode.

[0076] Specifically, when the number of structures is an even number, it can be determined that the corresponding division mode is an even number division mode, and multiple pairs of factor pairs corresponding to the number of structures can be obtained according to the even number division mode. For example, when the number of structures is 6, the corresponding factor pairs can be 1 and 6, 2 and 3.

[0077] A factor pair is a combination of numbers whose product equals the number of structures.

[0078] S2242, determining the factor pair with the smallest difference as the target factor pair, the long side in the target area corresponds to the larger target value in the target factor pair, and the short side in the target area corresponds to the smaller target value in the target factor pair.

[0079] Specifically, in order to ensure that the space is utilized to the maximum extent while maintaining the overall proportion of the region, the factor pair with the smallest difference can be determined as the target factor pair, and the long side in the target area can be corresponded to the target value with a large value, and the short side of the target area can be corresponded to the target value with a small value. For example, when the factor pairs are 1 and 6, 2 and 3, respectively, since the difference between 2 and 3 is the smallest, 2 and 3 can be used as the target factor pair, and since 3 is the target value with a large value in the target factor pair and 2 is the target value with a small value in the target factor pair, 3 can be corresponded to the long side in the target area, and 2 can be corresponded to the short side in the target area.

[0080] The target factor pair refers to the factor pair with the smallest difference, and the target value refers to the value included in the target factor pair.

[0081] S2243, obtaining a first value according to the difference between the target value and the offset value, generating a plurality of first dividing lines perpendicular to the long side and the short side and corresponding to the first values ​​respectively, and dividing the target area according to the first dividing lines to obtain a plurality of divided areas.

[0082] The offset value is 1, the first value refers to the difference between the target value and the offset value, and the first dividing line refers to the line for dividing the target area when the number of structures is an even number.

[0083] Specifically, the target value corresponding to the long side of the target area can be obtained, and the first value corresponding to the long side of the target area can be obtained according to the difference between the target value and the offset value. For example, the target value corresponding to the long side of the target area can be 3, then the first value corresponding to the long side of the target area is 2, and two first dividing lines perpendicular to the long side of the target area can be generated, and then the target value corresponding to the short side of the target area can be obtained. Similarly, the first value corresponding to the short side of the target area can be obtained according to the difference between the target value and the offset value. For example, the target value corresponding to the short side of the target area can be 2, then the first value corresponding to the short side of the target area is 1, and a first dividing line perpendicular to the short side of the target area can be generated. The target area can be divided into a number of divided areas through multiple dividing lines.

[0084] See also Figure 4 , is a schematic diagram of divided areas obtained after dividing a target area when the number of integrated structures provided in an embodiment of the present invention is an even number. When the number of structures is 6, the first value corresponding to the long side of the target area is 2, then 2 first dividing lines perpendicular to the long side of the target area can be generated, and the first value corresponding to the short side of the target area is 1, then 1 first dividing line perpendicular to the short side of the target area can be generated, and the target area can be divided into 6 divided areas of equal size through 3 first dividing lines.

[0085] S2244: When the number of structures is an odd number, determine that the division mode is an odd number division mode, and obtain the long side of the target area as a division side.

[0086] Specifically, when the number of structures is an odd number, for example, 3, the division mode corresponding to the target area can be determined as an odd division mode. In the odd division mode, in order to ensure that the specifications of the divided area can be maximized as much as possible, the long side of the target area can be determined as the division side. By dividing along the long side, it can be ensured that each divided area has sufficient space.

[0087] S2245, obtaining a second value according to the difference between the structure quantity and the offset value, generating a plurality of second dividing lines perpendicular to the dividing edge and corresponding to the second value, and dividing the target area according to the second dividing lines to obtain a plurality of divided areas.

[0088] Specifically, when dividing the target area along the long side, the number of dividing lines when dividing the long side, i.e., the second value, can be obtained based on the difference between the number of structures and the offset value. Combined with the second value, multiple dividing lines perpendicular to the dividing side can be generated, i.e., the second dividing lines. The target area can be divided into multiple divided areas of equal specifications through the second dividing lines.

[0089] See also Figure 5 , which is a schematic diagram of divided regions obtained after dividing a target region when the number of integrated structures provided by an embodiment of the present invention is an odd number, such as Figure 5 As shown in , when the number of structures is 3, the corresponding second value is 2. When dividing the target area, the long side of the target area can be determined as the dividing side, and two second dividing lines perpendicular to the dividing side are generated. The target area can be divided into three divided areas through the two second dividing lines.

[0090] The second value refers to the number of dividing lines corresponding to the target area when the number of structures is an odd number, and the second dividing line refers to the line for dividing the target area when the number of structures is an odd number. S225, determining that the area specification of the divided area is the standard coding specification.

[0091] Specifically, after obtaining multiple divided areas, the area specifications corresponding to the divided areas can be determined as standard coding specifications. The standard coding specifications can be considered as the maximum specifications corresponding to each summary code under the number of structures, and the standard coding specifications can be adjusted accordingly later.

[0092] S23, obtaining the number of vouchers corresponding to each of the integrated structures, offsetting the standard encoding specification according to the number of vouchers, and obtaining the encoding specification corresponding to each of the integrated structures.

[0093] In actual applications, different integrated structures may contain different numbers of electronic invoices. If all integrated structures use the same standard coding specifications, the difference in quantity between them cannot be intuitively reflected. This can be achieved by making reasonable adjustments to the standard coding specifications, and when adjusting, the standard coding specifications can be adjusted accordingly based on the number of electronic invoices corresponding to the integrated structures, so that the adjusted coding specifications can intuitively reflect the number of electronic invoices contained in different integrated structures.

[0094] Specifically, first, the number of electronic receipts corresponding to each integrated structure can be obtained, that is, the number of vouchers. According to the number of vouchers, the standard encoding specifications are offset accordingly to obtain the encoding specifications corresponding to each integrated structure. The larger the number of vouchers, the larger the encoding specifications corresponding to the integrated structure can be, and the smaller the number of vouchers, the smaller the encoding specifications corresponding to the integrated structure can be. Among them, the number of vouchers refers to the number of electronic receipts corresponding to the integrated structure.

[0095] In some embodiments, step S23 includes S231 to S233, which are specifically as follows: S231, determining that the maximum voucher quantity is the standard quantity, and obtaining the quantity difference between the standard quantity and the remaining voucher quantities.

[0096] Among them, the standard quantity refers to the maximum voucher quantity, and the quantity difference refers to the difference between the standard quantity and other voucher quantities. According to the quantity difference, the gap between each integrated structure and the maximum voucher quantity can be determined, which can provide a basis for the subsequent adjustment of the standard coding specifications.

[0097] S232, calling a ratio matching table, traversing each of the quantity differences and a plurality of preset difference intervals in the ratio matching table, and determining a preset reduction ratio corresponding to the preset difference interval where each of the quantity differences is located as an offset reduction ratio.

[0098] Among them, the ratio matching table refers to a pre-configured table that can determine the reduction ratio corresponding to the quantity difference, the preset difference interval refers to a pre-configured interval containing multiple quantity differences, and the preset reduction ratio refers to a pre-configured ratio for reducing the encoding specification.

[0099] Specifically, when generating a proportion matching table, a large amount of historical data can be collected and analyzed. The historical data includes a large number of integrated structures and their corresponding voucher quantities, and each integrated structure in the historical data has a summary code of corresponding size. The difference between the voucher quantity corresponding to each integrated structure and the standard voucher quantity corresponding to the summary code of the standard specification is calculated, and a large number of historical quantity differences can be obtained. According to the distribution range of a large number of historical quantity differences, the historical quantity differences are divided into multiple preset difference intervals. These preset difference intervals can cover most possible difference situations. The specification size of the summary code corresponding to a large number of integrated structures is analyzed in combination with the voucher quantity of the integrated structure. A reasonable reduction ratio can be set for each preset difference interval. The larger the difference, the larger the reduction ratio can be. By arranging multiple preset difference intervals and the corresponding reduction ratios into a table, a corresponding proportion matching table can be obtained.

[0100] After obtaining the quantity difference corresponding to each integrated structure, a pre-configured ratio matching table can be called up. By comparing the quantity difference with multiple preset difference intervals in the ratio matching table, the preset difference interval in which the quantity difference is located can be determined, and the preset reduction ratio corresponding to the preset difference interval can be determined as the offset reduction ratio corresponding to the corresponding integrated structure. Subsequently, the summary code corresponding to the integrated structure can be adjusted accordingly according to the offset reduction ratio.

[0101] S233, reducing and adjusting the standard coding specification according to the offset reduction ratio to obtain the coding specification corresponding to each of the integrated structures.

[0102] Specifically, the standard coding specifications can be reduced and adjusted according to the determined offset reduction ratio to obtain the coding specifications corresponding to each integrated structure, thereby achieving targeted adjustments to the standard coding specifications of each integrated structure, and then the size of the coding specifications can intuitively reflect the number of vouchers.

[0103] Through the above implementation, the space of the target area can be used more effectively, and the number of vouchers can be intuitively reflected through the size of the coding specification.

[0104] S24, generating scanning codes corresponding to each of the integrated structures, adjusting the specifications of the scanning codes according to the coding specifications to obtain a summary code, arranging each of the summary codes in the arrangement area based on the coding arrangement strategy to obtain a coded combined image and sending it to the first user terminal.

[0105] Specifically, for each integrated structure, a unique corresponding scan code can be generated. The scan code can be a rectangular code for quickly identifying and reading the ticket information in the integrated structure. Then, the corresponding scan code can be adjusted according to the encoding specifications corresponding to each integrated structure to obtain the summary code corresponding to each integrated structure. The size of the summary code can be used to show the size of the voucher quantity corresponding to the integrated structure.

[0106] After obtaining the summary code corresponding to each integrated structure, in order to ensure that all the summary codes can be clearly identified and scanned in the layout area, the various summary codes can be arranged in the layout area according to the coding layout strategy, and the corresponding coding combination image can be obtained, and the generated coding combination image can be sent to the first user terminal to facilitate subsequent corresponding reimbursement or review operations.

[0107] When arranging each summary code in combination with the coding arrangement strategy, each summary code can be arranged accordingly to obtain a sequence corresponding to the summary code, and each divided area can be arranged accordingly to obtain a sequence corresponding to the divided area. The summary codes and divided areas at the same arrangement position in the two sequences are matched to achieve an orderly arrangement of each summary code.

[0108] The scan code refers to a unique corresponding identifiable graphic code generated for each integrated structure.

[0109] Through the above implementation, the subsequent reimbursement or review process can be made more standardized and the efficiency can be improved.

[0110] In some embodiments, the step S24 of "arranging each of the summary codes into the arrangement area based on the coding arrangement strategy to obtain a coded combined image and sending it to the first user terminal" includes the following steps: S241, arranging the summary codes in order from large to small according to the coding specification to obtain a coding arrangement sequence.

[0111] The coding arrangement sequence refers to a sequence obtained by arranging the summary codes in order from large to small according to the size of the coding specification.

[0112] S242, determining that the divided areas in the arrangement area that are located in a row in a direction parallel to the long side are the same arrangement area group.

[0113] For example, when the parallel direction of the long side corresponding to the arrangement area is consistent with the x-axis direction, the coordinate values ​​corresponding to the center points of each divided area can be obtained. When the y-axis coordinate values ​​corresponding to the center points of the divided areas are consistent, it can be considered that the corresponding divided areas are in the same row, and the corresponding divided areas can be determined as the same group of arrangement area groups.

[0114] The arrangement area group refers to a group consisting of divided areas located in the same row in the length direction.

[0115] S243, firstly arranging the multiple divided areas in each of the arrangement area groups according to the direction parallel to the long sides, and then arranging each of the arrangement area groups according to the direction parallel to the short sides to obtain an arrangement area sequence.

[0116] For example, in the arrangement area, there are two groups of arrangement area groups, namely arrangement area group A and arrangement area group B. In arrangement area group A, there are corresponding division area 1 and division area 2, and in arrangement area group B, there are corresponding division area 3 and division area 4. First, the division areas in each arrangement area group are arranged according to the direction parallel to the long side. It can be obtained that in arrangement area group A, division area 1 is arranged in front of division area 2, and in arrangement area group B, division area 3 is arranged in front of division area 4. Arrange the arrangement area group A and the arrangement area group B sequentially in the direction parallel to the short side to obtain an arrangement area sequence containing two arrangement area groups, and it can be determined that in the arrangement area sequence, the two division areas corresponding to the arrangement area group A are located in front of the two division areas corresponding to the arrangement area group B, that is, the first position is division area 1, the second position is division area 2, the third position is division area 3, and the fourth position is division area 4.

[0117] The arrangement area sequence refers to a sequence obtained by arranging each arrangement area group.

[0118] S244, determining the correspondence between the summary codes and the divided areas having the same arrangement positions in the coding arrangement sequence and the arrangement area sequence.

[0119] For example, in the coding arrangement sequence, the second position is summary code 2, and in the arrangement area sequence, the second position is partition area 2. Since the arrangement positions corresponding to summary code 2 and partition area 2 are the same, summary code 2 can be matched with partition area 2.

[0120] S245, determining the center point of each of the divided areas as a positioning point, positioning the corresponding summary code based on the positioning point, arranging the summary code in the corresponding divided area to obtain a coded combined image and sending it to the first user terminal.

[0121] Specifically, after the summary code is matched with the divided area, the center point of each divided area can be determined as the positioning point, and the summary code corresponding to the divided area can be positioned at the positioning point, so that each summary code can be arranged in the corresponding divided area to obtain the corresponding encoded combined image.

[0122] Through the above implementation, the arrangement area can be used more effectively, and it can be ensured that each summary code is accurately placed in the corresponding divided area, thereby improving the accuracy of scanning and recognition.

[0123] S3, responding to the data upload information of the second user terminal, obtaining the identification image taken by the data acquisition device on the voucher arrangement device.

[0124] Among them, the second user end refers to the terminal held by the user who makes bill reimbursement, the data upload information refers to the information uploaded by the bill voucher, the data acquisition device refers to a device that can centrally acquire images of invoice data, such as a camera that can take overhead photos, and the recognized image refers to the image obtained by taking an image of the voucher sorting device through the data acquisition device.

[0125] In actual applications, when financial personnel manage bills, they need to deal with a large number of reimbursement bills. These bills often have different formats and contents. Financial personnel are confused and inefficient in managing them. For example, for paper bills, financial personnel need to manually enter and organize them. For electronic bills, although the entry process is simplified to a certain extent, due to their diverse formats and wide sources, they also bring considerable challenges to financial management. Therefore, in order to solve the confusion and inefficiency problems faced by financial personnel when dealing with a large number of reimbursement vouchers, this solution can organize the bill information of each user through a voucher processing device, realize rapid classification and review of bills, and make the subsequent financial reimbursement process more standardized and efficient.

[0126] Specifically, when the user needs to reimburse or review the bill, the corresponding data upload information can be sent. After responding to the data upload information sent by the user, the user can put the prepared bill to be reimbursed or the coded combination image into the voucher sorting device. Through the voucher sorting device, the bill information can be quickly uploaded, and each bill has a corresponding voucher sorting device, and each coded combination image also has a corresponding voucher sorting device. By placing the voucher sorting device under the data acquisition device and controlling the data acquisition device to take a bird's-eye view of the voucher sorting device, an identification image containing the bill information can be obtained. Through the identification image, the bill information can be quickly acquired.

[0127] S4, obtaining the tag information and the voucher information in the identified image, generating voucher-related data corresponding to the voucher information according to the tag information, and sending the generated voucher-related data to the financial terminal.

[0128] Among them, label information refers to the information filled in the label component. For example, if the identity information is filled in the label component in the upper left corner of the voucher sorting device, then the label information can be the user's name. Voucher information refers to the bill information extracted from the recognition image. If the recognition image is a photographed image corresponding to the encoded combination image, then the corresponding voucher information can be an electronic invoice. If the recognition image is an image of a paper version of the invoice, then the voucher information can be the image information of the corresponding paper invoice. Voucher-associated data refers to data that associates user-related information with corresponding bill information.

[0129] Specifically, the corresponding label information and credential information can be extracted from the recognized image. After obtaining these two types of information, the credential information corresponding to each bill can be associated with its corresponding identity information, invoice purpose and other label information to form a complete and accurate credential association data set. The credential association data set not only contains all the information of the bill, but also establishes a connection with the corresponding user through the label information, thereby realizing efficient organization of the bills of each user.

[0130] Based on the above embodiment, the specific implementation of step S4 may be: S41, extracting the voucher device outline in the recognition image, calling the information extraction template corresponding to the voucher arrangement device, and acquiring the label information and voucher information in the voucher device outline according to the information extraction template.

[0131] In actual applications, the user may place the credential organization device under the data acquisition equipment at different angles to take images. Therefore, the credential device in the obtained identification image may be rotated, tilted, etc. In order to accurately extract the label information and credential information in the identification image, the label information and credential information can be quickly extracted by calling the information extraction template corresponding to the credential organization device.

[0132] Specifically, first, the contour corresponding to the credential organization device, that is, the credential device contour, can be extracted from the recognition image through contour extraction technology such as edge detection algorithm, and the information extraction template corresponding to the credential organization device can be retrieved. In the information extraction template, there are areas for quickly extracting label information and areas for quickly extracting credential information. Therefore, the label information and credential information in the credential device contour can be obtained through the information extraction template.

[0133] Among them, the credential device contour refers to the contour corresponding to the credential organization device, and the information extraction template refers to a template that can extract credential information and label information. Since the shooting posture of the data acquisition device remains unchanged during image shooting, the specifications of the information extraction template can be consistent with the specifications of the credential device contour.

[0134] In some embodiments, step S41 includes S411 to S412, which are specifically as follows: S411, obtaining a positioning angle and a positioning point of the outline of the credential device, positioning the information extraction template based on the positioning angle and the positioning point, and superimposing the information extraction template on the outline of the credential device.

[0135] Since the credential organization device may be rotated or tilted, the outline of the credential device in the recognition image may also be rotated or tilted. Therefore, when quickly obtaining the label information and credential information in the outline of the credential device through the information extraction template, in order to ensure that valid information can be extracted, the angle of the information extraction template can be adjusted accordingly based on the position of the outline of the credential device and the corresponding tilt angle, and then the corresponding information extraction can be performed, thereby improving the accuracy of information acquisition.

[0136] Specifically, the inclination angle corresponding to the outline of the credential device, that is, the positioning angle, can be obtained. According to the relative position of the outline of the credential device in the recognition image, the corresponding positioning position point can be determined, the information extraction template is positioned according to the positioning position point, and the information extraction template is adjusted accordingly according to the positioning angle, so that the information extraction template can be superimposed on the top of the outline of the credential device.

[0137] The positioning angle refers to the tilt angle of the outline of the credential device, and the positioning position point refers to the position point corresponding to the outline of the credential device in the recognition image.

[0138] In some embodiments, step S411 includes S4111 to S4112, which are specifically as follows: S4111, determining the center point of the credential device contour as the positioning position point, obtaining the contour long side direction in the credential device contour, and determining the angle between the contour long side direction and the reference long side direction as the positioning angle.

[0139] Specifically, the center point of the credential device contour can be determined as the corresponding positioning position point. When determining the positioning angle, the corresponding reference long side direction can be pre-set. The corresponding positioning angle can be determined by the angle between the reference long side direction and the contour long side direction corresponding to the credential device contour, see Figure 6 , which is a schematic diagram of a positioning angle for positioning an information extraction template provided by an embodiment of the present invention, such as Figure 6 As shown in , the direction parallel to the horizontal direction can be determined as the reference long side direction, and the direction corresponding to the long side in the outline of the credential device can be determined as the outline long side direction. There is an angle between the outline long side direction and the reference long side direction, and the corresponding angle can be determined as the positioning angle.

[0140] The reference long side direction may be a horizontal direction.

[0141] S4112: locate the center point of the information extraction template according to the positioning position point, rotate the information extraction template based on the positioning angle and the reference rotation direction, and superimpose the information extraction template on the top of the outline of the credential device.

[0142] Specifically, after determining the positioning point, the center point of the information extraction template can be positioned at the positioning point, according to the pre-configured reference rotation direction, for example Figure 6 As shown in , the counterclockwise direction can be determined as the reference rotation direction. When the positioning angle is 30°, after the information extraction template is rotated 30° counterclockwise, the inclination angle of the information positioning template and the outline of the credential device are consistent. At this time, the information extraction template can be superimposed on the top of the outline of the credential device.

[0143] The reference rotation direction refers to the direction in which the information positioning template is rotated.

[0144] S412, acquiring tag information in the voucher device profile according to the information extraction area in the information extraction template, and acquiring voucher information in the voucher device profile according to the voucher extraction area in the information extraction template.

[0145] Specifically, since the information extraction template has an information extraction area that can extract label information and a credential extraction area that can extract credential information, after superimposing the information extraction template on the credential device outline, the label information in the credential device outline can be obtained from the information extraction area, and the credential information in the credential device outline can be obtained from the credential extraction area.

[0146] The information extraction area refers to the triangular area located at the four vertices of the information extraction template where the label information can be extracted, and the credential extraction area refers to the remaining area except the information extraction area.

[0147] In some embodiments, the specific implementation of step S412 may be: S4121, extracting the first extraction text corresponding to the label sub-area in the information extraction area, and determining the identification label corresponding to the information extraction area according to the first extraction text.

[0148] The label sub-area refers to an area where user information filled in the label component can be extracted, the first extracted text refers to the corresponding text in the label sub-area, and the identification label refers to an identifier that can identify the label information.

[0149] Specifically, there is an area in the information extraction area for extracting user information filled in the label component, namely the label sub-area, and the text corresponding to the label sub-area can be obtained and determined as the first extracted text, and the first extracted text can be determined as the information label of the information extraction area. For example, when the label component is a component for filling in the user's name, the corresponding label sub-area can be pre-configured as the area corresponding to the two characters of the name, then the corresponding first extracted text obtained can be the name, and the name can be determined as the identification label corresponding to the information extraction area.

[0150] S4122, obtaining a second extracted character in the information extraction area except the first extracted character, and determining that the second extracted character is label information corresponding to the identification label.

[0151] In actual applications, multiple characters can be extracted from the information extraction area, including the first extracted characters. The remaining characters except the first extracted characters can be determined as second extracted characters, and the second extracted characters can be determined as label information corresponding to the identification label. For example, when the multiple characters extracted from the information extraction area are the name Zhang San, since the name is the characters in the label sub-area, it can be determined as the first extracted characters. Except for the two characters of the name, the remaining characters Zhang San can be determined as the second extracted characters. Since Zhang San is the actual content filled in under the name label, it can be regarded as the label information corresponding to the identification label.

[0152] S4123, extracting the voucher image in the voucher extraction area, and when the voucher image is a coded combination image, obtaining the voucher information according to the electronic voucher corresponding to the coded combination image.

[0153] Specifically, the voucher image in the voucher extraction area can be extracted. When the voucher image is a coded image, one or more corresponding electronic vouchers can be obtained from the coded image, and corresponding voucher information can be obtained based on the obtained electronic vouchers.

[0154] The voucher image refers to a coded combination image or a photographed image corresponding to a paper ticket.

[0155] S4124: When the voucher image is a bill image, obtain the current bill type of the bill image, and obtain the voucher information according to the current bill type and the voucher image.

[0156] Specifically, when the voucher image is a bill image, the bill image can be identified to determine the current bill type corresponding to the bill image, and the voucher information corresponding to the paper bill can be obtained based on the bill image and its corresponding bill type.

[0157] The bill image refers to the image corresponding to the paper version of the bill, and the current bill type refers to the bill type corresponding to the currently recognized bill image.

[0158] In some embodiments, the specific implementation of step S4124 may be: S41241, determining a preset vertex angle corresponding to each of the identification tags, wherein the preset vertex angle includes an upper vertex angle and a lower vertex angle.

[0159] Specifically, since the information extraction area is located in the four vertex areas of the information extraction template, namely the upper left corner area, the lower left corner area, the upper right corner area, and the upper right corner area, after obtaining the identification labels corresponding to each information extraction area, the preset vertex angles corresponding to each identification label can be determined. For example, when the text in the bill image is arranged from left to right and from top to bottom, the preset vertex angle corresponding to the identification label in the upper left corner area can be the upper vertex angle, the preset vertex angle corresponding to the identification label in the lower left corner area can be the lower vertex angle, the preset vertex angle corresponding to the identification label in the upper right corner area can be the upper vertex angle, and the preset vertex angle corresponding to the identification label in the lower right corner area can be the lower vertex angle.

[0160] Among them, the preset vertex angle refers to the pre-configured vertex angle position associated with the information extraction area, the upper vertex angle refers to the vertex angle located at the top of the information extraction template, and the lower vertex angle refers to the vertex angle located at the bottom of the information extraction template.

[0161] S41242, obtaining the center point of each of the information extraction areas as the target point, and determining the direction from the target point corresponding to the upper vertex to the target point corresponding to the lower vertex as the traversal direction.

[0162] Among them, the target point refers to the center point of the information extraction area, and the traversal direction refers to the direction from the target point corresponding to the upper vertex to the target point corresponding to the lower vertex. Since the text corresponding to the bill type is generally located above the bill, the bill image is traversed according to the traversal direction to quickly determine the bill type corresponding to the bill image.

[0163] S41243, extract the text in the bill image in sequence based on the traversal direction, stop traversing when traversing to the text corresponding to the preset bill type, and determine that the corresponding preset bill type is the current bill type, and obtain the voucher information according to the current bill type and the voucher image.

[0164] The preset bill type refers to a pre-configured bill type, such as a special value-added tax invoice.

[0165] Specifically, the text in the bill image is extracted in sequence according to the traversal direction. When the text corresponding to the pre-configured bill type is recognized, the traversal can be stopped. For example, when the preset bill type is a special value-added tax invoice, if the words "special value-added tax invoice" are recognized, the traversal can be stopped, and the special value-added tax invoice can be determined as the current bill type. The voucher information can be obtained based on the current bill type and the corresponding voucher image.

[0166] Through the above implementation, tag information and credential information can be accurately located and extracted, thereby improving the efficiency of information extraction.

[0167] S42, calling a detailed filling table, wherein the detailed filling table includes a plurality of filling slots, each of which corresponds to a corresponding preset label.

[0168] In order to associate the tag information with the credential information, the detailed filling table can be called up. There are multiple filling slots in the detailed filling table for filling information. Each filling slot has a corresponding preset label. For example, the preset label corresponding to the filling slot for identity information can be a name label, and the preset label corresponding to the filling slot for bill type can be a type label.

[0169] Among them, the detailed filling table refers to a pre-configured table that can be filled with label information, the filling slot refers to a slot that can be filled with information, and the preset label refers to the pre-configured label corresponding to each filling slot.

[0170] S43, determining a preset tag corresponding to the identification tag of the tag information, and filling the tag information into the filling slot corresponding to the preset tag to obtain a voucher detail list.

[0171] Specifically, the tag information filled in the tag component has a corresponding identification tag. When the tag information is extracted, the identification tag corresponding to the tag information can be identified. For example, when the tag information is extracted, the two words "name" are extracted, so the name can be used as the corresponding identification tag. That is to say, when the tag information is the user's name, it can be determined that the corresponding identification tag is a name tag. Therefore, the preset tag corresponding to the identification tag can be determined, and the tag information corresponding to the identification tag is filled into the filling slot corresponding to the preset tag. The corresponding voucher detail table can be obtained. For example, the preset tag corresponding to the first filling slot in the detailed filling table is a name tag. Then, when it is determined that the identification tag corresponding to the tag information is also a name tag, the tag information can be filled into the first filling slot in the detailed filling table. The user identity information corresponding to the current ticket and the tag information such as the ticket type can be quickly determined through the voucher detail table.

[0172] The voucher details table refers to a table that records complete label information and its corresponding identification labels.

[0173] Based on the above embodiments, this solution also includes the following embodiments: A1, when the identification tag does not have corresponding tag information, determine the start upload time of the second user terminal, and obtain the backtracking time period according to the start upload time and the current time.

[0174] In actual applications, when a user needs to upload multiple tickets, he needs to fill in the corresponding information in the label components of multiple voucher sorting devices corresponding to the multiple tickets respectively, but the user may not necessarily fill in the complete information on each voucher sorting device. When there is no corresponding label information for the identification tag, it can be considered that in the corresponding identification image, the user may not have filled in the corresponding information, resulting in missing label information. When it is identified that the label information corresponding to the identification tag is missing, the start time when the user's first ticket is uploaded can be obtained, that is, the start upload time. According to the start upload time and the current time, the corresponding backtracking time period can be obtained, so that possible historical label data can be found within the backtracking time period to complete the missing label information.

[0175] The start upload time refers to the time when the ticket upload operation starts, the current time refers to the upload time corresponding to the current identification tag, and the backtracking time period refers to the time period from the start upload time to the current time.

[0176] A2, obtaining a plurality of historical recognition images in a retrospective time period, obtaining a historical recognition tag corresponding to each of the historical recognition images, and determining a historical recognition tag corresponding to the recognition tag as a target recognition tag.

[0177] Specifically, during the retrospective period, the data acquisition device collects multiple identification images, namely historical identification images, and each historical identification image also has a corresponding identification tag, namely a historical identification tag. The historical identification tag corresponding to the identification tag without label information can be determined as the target identification tag.

[0178] Among them, the historical identification image refers to the image obtained by the data acquisition device taking an image of the corresponding voucher sorting device within the retrospective time period, the historical identification tag refers to the identification tag corresponding to the historical identification image, and the target identification tag refers to the historical identification tag corresponding to the identification tag for which no label information exists.

[0179] A3, determining that the historical tag information corresponding to the target identification tag is the tag information corresponding to the identification tag, and filling the tag information into the filling slot corresponding to the preset tag to obtain a voucher detail list.

[0180] Specifically, the target identification tag has corresponding tag information, namely, historical tag information. The historical tag information can be determined as the tag information corresponding to the identification tag, and can be filled into the filling slot corresponding to the preset tag to obtain the corresponding voucher details table.

[0181] The historical tag information refers to the tag information corresponding to the historical identification tag.

[0182] S44, associating the voucher details table with the voucher information to obtain voucher association data and sending it to the financial end.

[0183] Specifically, by associating the voucher details table with the corresponding voucher information, the corresponding voucher association data can be obtained, and the voucher association data can be sent to the financial end. The financial personnel can quickly view the bill information uploaded by the user through the voucher association data, and can quickly obtain the specific information corresponding to the viewed bill, such as the user's identity information, bill type, etc.

[0184] Through the above implementation, the accuracy of bill processing can be improved, and the bills can be classified and managed more accurately and efficiently.

[0185] See also Figure 7 , is a schematic diagram of the structure of a tax optimization system based on artificial intelligence provided by an embodiment of the present invention. The data processing system of the tax optimization system based on artificial intelligence includes: An acquisition module, used to acquire the audit information of the electronic voucher set by the financial end, and generate an integrated structure corresponding to the electronic voucher set according to the audit information; A generating module, used for generating a summary code corresponding to the integrated structure, generating a coded combination image according to the summary code and sending it to a first user terminal corresponding to the electronic voucher set; A response module, used to respond to the data upload information of the second user terminal and obtain the identification image taken by the data acquisition device on the voucher arrangement device; The sending module is used to obtain the label information and the voucher information in the identified image, generate the voucher-related data corresponding to the voucher information according to the label information, and send it to the financial end.

[0186] Figure 7 The apparatus of the embodiment shown can be used to perform Figure 1 The implementation principles and technical effects of the steps in the method embodiment shown are similar and will not be repeated here.

[0187] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention. scope.

Claims

1. A tax optimization method based on artificial intelligence, characterized in that: include: Acquire the audit information of the electronic voucher set by the financial end, and generate the integration structure corresponding to the electronic voucher set according to the audit information; generating a summary code corresponding to the integrated structure, generating a coded combination image according to the summary code and sending it to a first user terminal corresponding to the electronic certificate set; In response to the data upload information of the second user terminal, obtaining the identification image taken by the data acquisition device of the voucher arrangement device; The tag information and the voucher information in the identified image are obtained, and the voucher-related data corresponding to the voucher information is generated according to the tag information and sent to the financial end.

2. The method according to claim 1, characterized in that Acquiring the audit information of the electronic voucher set by the financial end, and generating the integration structure corresponding to the electronic voucher set according to the audit information, including: Determining the compliance certificate selected by the financial end from the set of electronic certificates according to the audit information; Obtaining the bill type of each of the compliance vouchers, and determining that the compliance vouchers of the same type are the same voucher combination; Generate a credential node corresponding to each of the compliant credentials and a type node corresponding to each of the credential combinations, and connect the credential nodes and type nodes corresponding to each of the credential combinations to obtain a plurality of sub-combination structures; Generate multiple bill type groups selected by the financial end, each bill type group includes at least two bill types, and generate integration nodes corresponding to each bill type group; The sub-combination structures corresponding to each of the bill type groups and the integration node are connected to obtain a plurality of integration structures corresponding to the electronic voucher set.

3. The method according to claim 1, characterized in that Generating a summary code corresponding to the integrated structure, generating a coded combination image according to the summary code and sending it to a first user terminal corresponding to the electronic certificate set, comprising: Retrieving a blank arrangement template corresponding to the voucher arrangement device, wherein the blank arrangement template includes a label display area corresponding to each label component in the voucher arrangement device, and the label display area is a triangular area located at the top corner of the blank arrangement template; Determine other areas of the blank layout template except the label display area as layout areas, count the number of structures of each integrated structure, and determine the standard coding specification corresponding to the layout area according to the number of structures; Acquire the number of vouchers corresponding to each of the integrated structures, and offset the standard encoding specification according to the number of vouchers to obtain the encoding specification corresponding to each of the integrated structures; Generate a scan code corresponding to each integrated structure, adjust the scan code according to the coding specification to obtain a summary code, arrange each summary code in the arrangement area based on the coding arrangement strategy to obtain a coded combined image and send it to the first user end.

4. The method according to claim 3, characterized in that Counting the number of structures of each integrated structure, and determining the standard coding specification corresponding to the arrangement area according to the number of structures, including: Obtaining the long side parallel direction and the short side parallel direction of the blank layout template, and determining the area edge of the label display area located inside the blank layout template as the target edge; Taking the line segment points of each target edge as the connection starting point and the connection end point, a plurality of mutually perpendicular connection lines are generated in the direction parallel to the long edge and the direction parallel to the short edge respectively; A plurality of rectangular areas are obtained according to the connecting lines, and the rectangular area with the largest area is determined as the target area; Counting the number of structures of each integrated structure, determining a division mode of the target area according to the number of structures, wherein the division mode includes an even-number division mode and an odd-number division mode, and dividing the target area according to the division mode to obtain a plurality of divided areas; The area specification of the divided area is determined to be the standard coding specification.

5. The method according to claim 4, characterized in that Determine a division mode of the target area according to the number of structures, the division mode includes an even number division mode and an odd number division mode, divide the target area according to the division mode to obtain a plurality of division areas, including: When the number of structures is an even number, determining that the division mode is an even number division mode, and obtaining multiple pairs of factor pairs corresponding to the number of structures according to the even number division mode; Determine the factor pair with the smallest difference as the target factor pair, the long side in the target area corresponds to the target value with the larger value in the target factor pair, and the short side in the target area corresponds to the target value with the smaller value in the target factor pair; Obtaining a first value according to the difference between the target value and the offset value, generating a plurality of first dividing lines perpendicular to the long side and the short side and corresponding to the first values ​​respectively, and dividing the target area according to the first dividing lines to obtain a plurality of divided areas; When the number of structures is an odd number, determining that the division mode is an odd number division mode, and obtaining the long side of the target area as the division side; A second value is obtained according to the difference between the structure quantity and the offset value, a plurality of second dividing lines perpendicular to the dividing edge and corresponding to the second value are generated, and the target area is divided according to the second dividing lines to obtain a plurality of divided areas.

6. The method according to claim 3, characterized in that Acquiring the number of vouchers corresponding to each of the integrated structures, offsetting the standard encoding specification according to the number of vouchers, and obtaining the encoding specification corresponding to each of the integrated structures, including: Determine the maximum voucher quantity as the standard quantity, and obtain the quantity difference between the standard quantity and the remaining voucher quantities; Retrieving a ratio matching table, traversing each of the quantity differences and a plurality of preset difference intervals in the ratio matching table, and determining that a preset reduction ratio corresponding to the preset difference interval in which each of the quantity differences is located is an offset reduction ratio; The standard coding specification is reduced and adjusted according to the offset reduction ratio to obtain the coding specification corresponding to each of the integrated structures.

7. The method according to claim 4, characterized in that Arranging each of the summary codes in the arrangement area based on the coding arrangement strategy to obtain a coded combined image and sending it to the first user terminal includes: Arranging the summary codes in order from large to small according to the coding specifications to obtain a coding arrangement sequence; Determine that the divided areas in the arrangement area that are in a row in a direction parallel to the long side are the same arrangement area group; Firstly, a plurality of divided areas in each of the arrangement area groups are arranged according to the direction parallel to the long sides, and then each of the arrangement area groups is arranged according to the direction parallel to the short sides to obtain an arrangement area sequence; Determine the correspondence between the summary codes and the divided areas having the same arrangement positions in the coding arrangement sequence and the arrangement area sequence; The center point of each of the divided areas is determined as a positioning point, the corresponding summary code is positioned based on the positioning point, and the summary code is arranged in the corresponding divided area to obtain a coded combined image which is sent to the first user terminal.

8. The method according to claim 1, characterized in that Acquiring tag information and voucher information in the identified image, generating voucher-related data corresponding to the voucher information according to the tag information and sending the generated voucher-related data to the financial end, including: Extracting the credential device outline in the recognition image, calling the information extraction template corresponding to the credential arrangement device, and acquiring the label information and credential information in the credential device outline according to the information extraction template; Retrieving a detailed filling table, wherein the detailed filling table includes a plurality of filling slots, each of the filling slots corresponds to a corresponding preset label; Determine a preset tag corresponding to the identification tag of the tag information, and fill the tag information into the filling slot corresponding to the preset tag to obtain a voucher detail list; The voucher details table is associated with the voucher information to obtain voucher associated data and sent to the financial end.

9. The method according to claim 8, characterized in that Extracting the voucher device outline in the recognition image, calling the information extraction template corresponding to the voucher arrangement device, and obtaining the label information and voucher information in the voucher device outline according to the information extraction template, including: Acquire the positioning angle and positioning point of the outline of the credential device, position the information extraction template based on the positioning angle and positioning point, and superimpose the information extraction template on the top of the outline of the credential device; The tag information in the voucher device profile is obtained according to the information extraction area in the information extraction template, and the voucher information in the voucher device profile is obtained according to the voucher extraction area in the information extraction template.

10. The method according to claim 9, characterized in that Acquiring a positioning angle and a positioning point of the outline of the credential device, positioning the information extraction template based on the positioning angle and the positioning point, and superimposing the information extraction template on the outline of the credential device, including: Determine the center point of the outline of the credential device as the positioning position point, obtain the outline long side direction in the outline of the credential device, and determine the angle between the outline long side direction and the reference long side direction as the positioning angle; The center point of the information extraction template is located according to the positioning position point, and the information extraction template is rotated based on the positioning angle and the reference rotation direction, so that the information extraction template is superimposed on the outline of the credential device.

11. The method according to claim 9, characterized in that Acquiring tag information in the credential device profile according to the information extraction area in the information extraction template, and acquiring credential information in the credential device profile according to the credential extraction area in the information extraction template, including: Extracting a first extraction character corresponding to the label sub-area in the information extraction area, and determining an identification label corresponding to the information extraction area according to the first extraction character; Acquire a second extracted character in the information extraction area except the first extracted character, and determine that the second extracted character is label information corresponding to the identification label; Extracting the voucher image in the voucher extraction area, and when the voucher image is a coded combination image, obtaining the voucher information according to the electronic voucher corresponding to the coded combination image; When the voucher image is a bill image, the current bill type of the bill image is acquired, and the voucher information is obtained according to the current bill type and the voucher image.

12. The method according to claim 11, characterized in that When the voucher image is a bill image, obtaining the current bill type of the bill image, and obtaining the voucher information according to the current bill type and the voucher image, including: Determine a preset vertex angle corresponding to each of the identification tags, wherein the preset vertex angle includes an upper vertex angle and a lower vertex angle; Acquire the center point of each of the information extraction areas as the target point, and determine the direction from the target point corresponding to the upper vertex angle to the target point corresponding to the lower vertex angle as the traversal direction; The text in the bill image is extracted in sequence based on the traversal direction, and the traversal is stopped when the text corresponding to the preset bill type is traversed, and the corresponding preset bill type is determined to be the current bill type, and the voucher information is obtained according to the current bill type and the voucher image.

13. The method according to claim 8, characterized in that Determining a preset tag corresponding to the identification tag of the tag information, and filling the tag information into the filling slot corresponding to the preset tag to obtain a voucher detail table, including: When the identification tag does not have corresponding tag information, determining the start upload time of the second user terminal, and obtaining the backtracking time period according to the start upload time and the current time; Acquire multiple historical recognition images in the retrospective time period, acquire historical recognition tags corresponding to each of the historical recognition images, and determine the historical recognition tag corresponding to the recognition tag as a target recognition tag; Determine that the historical tag information corresponding to the target identification tag is the tag information corresponding to the identification tag, and fill the tag information into the filling slot corresponding to the preset tag to obtain a voucher detail list.

Images