Block chain-based employee self-service reimbursement system and control method thereof
By processing employee-uploaded expense voucher images using discrete calibration and visual descriptor technology, and combining this with blockchain technology, the problem of non-standard image quality for expense vouchers has been solved, achieving efficient, accurate, automated processing and secure storage.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SICHUAN VOCATIONAL & TECHN COLLEGE
- Filing Date
- 2023-12-20
- Publication Date
- 2026-04-17
AI Technical Summary
In the employee self-service expense reimbursement system, the non-standard image quality of reimbursement vouchers leads to low processing efficiency.
By acquiring images of expense reimbursement vouchers uploaded by employees, discrete calibration, feature difference coefficient determination, visual descriptor extraction, and automatic classification verification are performed. Blockchain technology is used to ensure secure transmission and storage, thereby achieving automated processing.
It improves the accuracy and efficiency of image processing for reimbursement vouchers, reduces manual intervention, and ensures data security and transparency.
Smart Images
Figure CN121882902A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of financial reimbursement technology, and more specifically, to a blockchain-based employee self-service reimbursement system and its control method. Background Technology
[0002] Blockchain is a distributed database technology originally designed for cryptocurrencies like Bitcoin. It is a network of multiple computer nodes that collectively maintain and manage a decentralized database. Blockchain has no central authority; all transactions and data are distributed across the network's nodes. Blockchain uses cryptographic techniques to ensure data security, transaction privacy, and integrity. Blockchain can execute automated smart contracts, which are computer programs that execute automatically when specific conditions are met.
[0003] A blockchain-based employee self-service expense reimbursement system can provide organizations with a highly transparent, secure, and automated expense reimbursement solution. The system architecture includes a blockchain network, user interface, and smart contracts. Smart contracts define which types of expenses are reimbursable, approval rules, and how funds are transferred to employees. Employees submit reimbursement requests, which are then converted into transactions and sent to the blockchain network. After verification through the smart contract, the reimbursement approval process is initiated. Blockchain-based employee self-service reimbursement offers transparency, tamper-proofing, and automated processing. Smart contracts and automated processes reduce human intervention and improve efficiency. However, in existing technologies, when employees upload reimbursement vouchers through reimbursement nodes on the blockchain for review, issues such as non-standard and low-quality images due to equipment and photography limitations lead to low processing efficiency for these uploaded vouchers. Summary of the Invention
[0004] This application provides a blockchain-based employee self-service expense reimbursement system and its control method to solve the technical problem of low processing efficiency of expense reimbursement voucher images uploaded by employees in the employee self-service expense reimbursement system.
[0005] To solve the above-mentioned technical problems, this application adopts the following technical solution:
[0006] Firstly, this application provides a control method for a blockchain-based employee self-service expense reimbursement system, comprising the following steps:
[0007] Obtain the reimbursement vouchers uploaded by employees to the self-service reimbursement center through the blockchain reimbursement node, and obtain the reimbursement voucher image;
[0008] Determine the discrepancy degree of the reimbursement voucher corresponding to the reimbursement voucher image, and adjust the reimbursement voucher image using the discrepancy degree to obtain an adjusted reimbursement voucher image.
[0009] Determine the horizontal feature difference coefficient and the vertical feature difference coefficient corresponding to the adjusted image of the reimbursement voucher, determine the adjusted and sharpened image of the reimbursement voucher through the horizontal feature difference coefficient and the vertical feature difference coefficient, and then extract the visual description sub-boundary of the voucher from the adjusted and sharpened image of the reimbursement voucher.
[0010] By using the bounds of the visual descriptor of the voucher, the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher are obtained, and the visual distribution similarity basis is determined by the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher.
[0011] The reimbursement vouchers uploaded by employees are automatically classified and verified based on the visual description sub-boundary of the voucher and the visual distribution similarity basis.
[0012] In some embodiments, determining the discrepancy degree of the reimbursement voucher corresponding to the reimbursement voucher image specifically includes:
[0013] The reimbursement voucher image is converted to grayscale to obtain a grayscale image of the reimbursement voucher.
[0014] Contour detection is performed on the grayscale image of the reimbursement voucher to determine the contour circumscribed matrix;
[0015] The discrete adjustment degree of the reimbursement voucher is obtained based on the outline outer matrix.
[0016] In some embodiments, determining the sharpened image of the reimbursement voucher using the horizontal feature difference coefficient and the vertical feature difference coefficient specifically includes:
[0017] The frequency and phase of the cosine function are determined by the discrete adjustment degree of the expense reimbursement voucher;
[0018] The horizontal and vertical difference ratios of the Gaussian function are determined by the horizontal feature difference coefficient and the vertical feature difference coefficient.
[0019] The adjusted and sharpened image of the expense voucher is obtained by convolving the frequency and phase of the cosine function, the horizontal feature difference coefficient, the vertical feature difference coefficient, the horizontal and vertical difference ratio of the Gaussian function, and the adjusted image of the expense voucher.
[0020] In some embodiments, the extraction of the visual description sub-boundary of the reimbursement voucher from the sharpened image specifically includes:
[0021] The sharpened image of the reimbursement voucher is subjected to piecewise linear fitting to obtain a linear gamma correction curve;
[0022] The reimbursement voucher sharpened image is segmented and corrected using the linear gamma correction curve to obtain the corrected reimbursement voucher sharpened image.
[0023] The visual descriptor of the corrected reimbursement voucher is extracted by using a recurrent neural network to obtain the visual descriptor boundary of the voucher.
[0024] In some embodiments, automatically classifying and verifying expense vouchers uploaded by employees based on the voucher visual description subbound and the visual distribution similarity basis specifically includes:
[0025] The core boundary of the visual description sub-boundary of the voucher is determined based on the visual description sub-boundary and the visual distribution similarity basis.
[0026] Automatic reimbursement voucher classification is performed through the core interface of the voucher visual description sub-section to obtain the automatic classification result of the reimbursement voucher;
[0027] The reimbursement vouchers uploaded by employees are verified based on the automatic classification results of the reimbursement vouchers.
[0028] In some embodiments, determining the core bound of a credential visual descriptor based on the credential visual descriptor subbound and the visual distribution similarity cardinality means taking the credential visual descriptor subclass with the highest visual distribution similarity cardinality as the core bound of the credential visual descriptor.
[0029] In some embodiments, automatic reimbursement classification through the core boundary of the voucher visual descriptor involves inputting the core boundary of the voucher visual descriptor into a support vector machine for automatic reimbursement classification.
[0030] Secondly, this application provides a blockchain-based employee self-service expense reimbursement system, including a control unit, the control unit comprising:
[0031] The acquisition module is used to acquire reimbursement vouchers uploaded by employees to the self-service reimbursement center through the blockchain reimbursement node and obtain the reimbursement voucher image.
[0032] The calibration module is used to determine the discrete calibration degree of the reimbursement voucher corresponding to the reimbursement voucher image, and to calibrate the reimbursement voucher image by means of the discrete calibration degree of the reimbursement voucher to obtain the reimbursement voucher calibrated image.
[0033] The extraction module is used to determine the horizontal feature difference coefficient and the vertical feature difference coefficient corresponding to the reimbursement voucher adjustment image, determine the reimbursement voucher adjustment and sharpening image through the horizontal feature difference coefficient and the vertical feature difference coefficient, and then extract the voucher visual description sub-boundary from the reimbursement voucher adjustment and sharpening image;
[0034] The determination module is used to obtain the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher through the bounds of the visual descriptor of the voucher, and to determine the visual distribution similarity basis by the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher;
[0035] The classification and verification module is used to automatically classify and verify the expense vouchers uploaded by employees based on the visual description sub-boundary of the voucher and the visual distribution similarity basis.
[0036] Thirdly, this application provides a computer device, which includes a memory and a processor. The memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the control method of the blockchain-based employee self-service expense reimbursement system described above.
[0037] Fourthly, this application provides a computer-readable storage medium storing instructions or code that, when executed on a computer, cause the computer to implement the aforementioned control method for a blockchain-based employee self-service expense reimbursement system.
[0038] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:
[0039] In this application, firstly, blockchain technology ensures the secure transmission and storage of reimbursement vouchers. A blockchain-based system enables direct peer-to-peer interaction, reducing intermediate steps, accelerating the reimbursement process, and effectively improving review efficiency. Secondly, the voucher visual descriptor sub-boundary represents the set of all voucher visual descriptors of the reimbursement voucher image. Each voucher visual descriptor represents an image feature in the sharpened image of the reimbursement voucher. By utilizing the image feature information in the voucher visual descriptor sub-boundary, the processing speed of the reimbursement voucher image can be accelerated, and information on the sharpened image of the reimbursement voucher can be identified more accurately, thereby improving the accuracy and efficiency of reimbursement voucher image processing. Furthermore, visual distribution similarity basis can be used to measure the information entropy of visual description subclasses of vouchers. Therefore, by using visual distribution similarity basis, highly related visual description subclasses of vouchers can be measured, and it can help reimbursement classification and verification quickly find key image information, thereby improving the efficiency of reimbursement voucher image processing. Finally, based on the voucher visual description subclass boundaries and visual distribution similarity basis, the self-service reimbursement node can more efficiently identify the key feature information of employee reimbursement vouchers. In turn, through automated classification and blockchain, the processing speed of reimbursement voucher images can be accelerated, effectively improving the processing efficiency of the employee self-service reimbursement system for reimbursement voucher images uploaded by employees. Attached Figure Description
[0040] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0041] Figure 1 This is an exemplary flowchart of a control method for a blockchain-based employee self-service expense reimbursement system, as shown in some embodiments of this application.
[0042] Figure 2 These are schematic diagrams of exemplary hardware and / or software of a control unit according to some embodiments of this application;
[0043] Figure 3 This is a schematic diagram of the structure of a computer device that implements a control method for a blockchain-based employee self-service expense reimbursement system, according to some embodiments of this application. Detailed Implementation
[0044] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.
[0045] This application provides a blockchain-based employee self-service expense reimbursement system and its control method. The core of the system involves acquiring expense vouchers uploaded by employees to the self-service expense reimbursement center via blockchain nodes, obtaining expense voucher images; determining the discrete adjustment degree of the expense voucher image; adjusting the expense voucher image using the discrete adjustment degree to obtain an adjusted expense voucher image; determining the horizontal and vertical feature difference coefficients corresponding to the adjusted expense voucher image; determining the adjusted and sharpened expense voucher image using the horizontal and vertical feature difference coefficients; extracting the expense voucher visual descriptor boundaries from the adjusted and sharpened expense voucher image; obtaining the typical quantity and discrete matrix of the expense voucher visual descriptor using the expense voucher visual descriptor boundaries; determining the visual distribution similarity basis using the typical quantity and discrete matrix of the expense voucher visual descriptor; and automatically classifying and verifying the expense vouchers uploaded by employees based on the expense voucher visual descriptor boundaries and the visual distribution similarity basis, effectively improving the processing efficiency of the employee self-service expense reimbursement system for employee-uploaded expense voucher images.
[0046] To better understand the above technical solutions, a detailed description of the solutions will be provided below in conjunction with the accompanying drawings and specific implementation methods. (Reference) Figure 1 The figure is an exemplary flowchart of a control method for a blockchain-based employee self-service expense reimbursement system according to some embodiments of this application. The control method 100 for the blockchain-based employee self-service expense reimbursement system mainly includes the following steps:
[0047] In step 101, the reimbursement vouchers uploaded by employees to the self-service reimbursement center through the blockchain reimbursement node are obtained, and the reimbursement voucher image is obtained.
[0048] In practice, employees first verify their identity through a blockchain identity verification system to obtain authorization to access the self-service reimbursement system. Employees then upload images of their reimbursement vouchers to reimbursement nodes on the blockchain network. During the upload, encryption technology ensures the security of the transmission. Smart contracts in the blockchain network are responsible for receiving and processing the uploaded reimbursement vouchers. Authorized personnel can query reimbursement vouchers in the self-service reimbursement center through the blockchain network and obtain the reimbursement voucher images by using the OpenCV image processing library.
[0049] It should be noted that the blockchain technology in this application ensures the secure transmission and storage of reimbursement vouchers, increases the credibility of reimbursement vouchers, reduces the possibility of potential fraud and errors, and enables direct peer-to-peer interaction, reducing intermediate steps, accelerating the reimbursement process, and effectively improving review efficiency.
[0050] In step 102, the discrepancy degree of the reimbursement voucher corresponding to the reimbursement voucher image is determined, and the reimbursement voucher image is adjusted using the discrepancy degree to obtain an adjusted reimbursement voucher image.
[0051] In some embodiments, determining the discrepancy degree of the reimbursement voucher corresponding to the reimbursement voucher image can be achieved in the following manner:
[0052] The reimbursement voucher image is converted to grayscale to obtain a grayscale image of the reimbursement voucher.
[0053] Contour detection is performed on the grayscale image of the reimbursement voucher to determine the contour circumscribed matrix;
[0054] The discrete adjustment degree of the reimbursement voucher is obtained based on the outline outer matrix.
[0055] In practice, firstly, the reimbursement voucher image is converted to grayscale to obtain a grayscale image of the reimbursement voucher. It should be noted that grayscale conversion is achieved by weighted averaging of the red, green, and blue channel values of each pixel in the reimbursement voucher image to obtain a single grayscale value. Grayscale conversion reduces the complexity and computational cost of reimbursement voucher image processing, and the removal of color information from the reimbursement voucher image reduces the risk of leakage of personal privacy and improves the privacy and security of the reimbursement voucher. Secondly, a contour detection algorithm is used to detect the contour of the reimbursement voucher grayscale image. For the detected grayscale image contour, the angle of the bounding rectangle of the reimbursement voucher grayscale image contour is calculated. The bounding rectangle is the smallest rectangle that can completely enclose the contour. The angle of this bounding rectangle is the discrete rotation angle of the reimbursement voucher. Based on the calculated angle of the bounding rectangle, this angle is the discrete adjustment degree of the reimbursement voucher.
[0056] In some embodiments, the reimbursement voucher image is adjusted using the discrepancy adjustment degree of the reimbursement voucher to obtain an adjusted reimbursement voucher image. Specifically, the reimbursement voucher image is rotated according to the discrepancy adjustment degree of the reimbursement voucher to obtain a rotated reimbursement voucher image, i.e., the adjusted reimbursement voucher image.
[0057] It should be noted that adjusting the discrepancy of the reimbursement voucher can help improve the quality of the reimbursement voucher image, making it clearer.
[0058] In step 103, the horizontal feature difference coefficient and the vertical feature difference coefficient corresponding to the reimbursement voucher adjustment image are determined, and the reimbursement voucher adjustment and sharpening image is determined by the horizontal feature difference coefficient and the vertical feature difference coefficient, and then the voucher visual description sub-boundary is extracted from the reimbursement voucher adjustment and sharpening image.
[0059] In some embodiments, the horizontal feature difference coefficient and the vertical feature difference coefficient corresponding to the reimbursement voucher adjustment image are determined. It should be noted that, in this application, the horizontal feature difference coefficient represents the degree of difference between image features in the horizontal direction of the reimbursement voucher adjustment image, and the vertical feature difference coefficient represents the degree of difference between image features in the vertical direction of the reimbursement voucher adjustment image. Specifically, the standard deviation of the reimbursement voucher adjustment image in the horizontal direction is used as the horizontal feature difference coefficient, and the standard deviation of the reimbursement voucher adjustment image in the vertical direction is used as the vertical feature difference coefficient.
[0060] Preferably, in some embodiments, determining the sharpened image of the reimbursement voucher using the horizontal feature difference coefficient and the vertical feature difference coefficient can be achieved in the following manner:
[0061] The frequency and phase of the cosine function are determined by the discrete adjustment degree of the expense reimbursement voucher;
[0062] The horizontal and vertical difference ratios of the Gaussian function are determined by the horizontal feature difference coefficient and the vertical feature difference coefficient.
[0063] A sharpened image of the expense reimbursement voucher is obtained by convolving the frequency and phase of the cosine function, the horizontal feature difference coefficient, the vertical feature difference coefficient, the aspect ratio of the Gaussian function, and the adjusted image of the expense reimbursement voucher. Specifically, this sharpened image of the expense reimbursement voucher can be determined according to the following formula:
[0064]
[0065] in, This indicates that the image of the reimbursement voucher has been sharpened and adjusted. This indicates the image of the expense reimbursement voucher being adjusted. This represents the coordinate variables in the image of the expense reimbursement voucher. This represents the ratio of the horizontal to vertical differences of the Gaussian function. This represents the convolution operation. Image indicating reimbursement voucher adjustment The coefficient of difference in horizontal characteristics Image indicating reimbursement voucher adjustment The vertical characteristic difference coefficient, Represents the cosine function. Represents the frequency of the cosine function. The phase of the cosine function is represented. It should be noted that in this application, the visual sharpness of the reimbursement voucher adjusted and sharpened image obtained by the above method is higher, which is helpful for subsequent processing steps. The aspect ratio of the Gaussian function is an indicator used to measure the shape difference of the Gaussian function.
[0066] In practice, a two-dimensional discrete Fourier transform is performed on the grayscale image of the reimbursement voucher to obtain the frequency domain information of the image. The discrete rotation angle corresponds to the phase information of the Fourier transform. The Fourier transform can convert the grayscale image of the reimbursement voucher from the time domain to the frequency domain, providing frequency and phase information. In the result of the Fourier transform, the frequency and phase of the cosine function are determined based on the position and phase information of the main frequency components. The ratio of the horizontal feature difference coefficient to the vertical feature difference coefficient is used as the horizontal and vertical difference ratio of the Gaussian function.
[0067] It should be noted that the sharpened images of expense reimbursement vouchers highlight specific features, enhancing the information that needs attention and facilitating better analysis and processing. The characteristics of blockchain ensure the immutability and transparency of the data, allowing users to trust the processed expense reimbursement voucher image data without worrying about data tampering or falsification.
[0068] In some embodiments, determining the visual description sub-boundary of a reimbursement voucher by adjusting and sharpening the image can be achieved in the following ways:
[0069] The sharpened image of the reimbursement voucher is subjected to piecewise linear fitting to obtain a linear gamma correction curve;
[0070] The reimbursement voucher sharpened image is segmented and corrected using the linear gamma correction curve to obtain the corrected reimbursement voucher sharpened image.
[0071] The visual descriptor of the corrected reimbursement voucher is extracted by using a recurrent neural network to obtain the visual descriptor boundary of the voucher.
[0072] In practice, firstly, the grayscale range of the sharpened image of the reimbursement voucher is divided into several segments (e.g., 0-63, 64-127, 128-191, 192-255). Linear fitting is performed on the grayscale values within each segment to obtain a segmented linear correction curve. This segmented linear correction curve can be derived from pre-collected standard image data or determined based on actual needs. Secondly, the grayscale value of each pixel in the sharpened image of the reimbursement voucher is corrected segment by segment. The segment to which each pixel belongs is determined based on its grayscale value, and then the corresponding linear correction curve is applied. Different corrections can be applied to different parts of the sharpened image of the reimbursement voucher to obtain a corrected sharpened image of the reimbursement voucher for better visual effects. Finally, the corrected sharpened image of the reimbursement voucher is converted into a sequence number suitable for processing by a recurrent neural network. This can be achieved by dividing the sharpened image of the reimbursement voucher into a grid, and using each grid unit as a time step of the recurrent neural network. Step 1) Construct a recurrent neural network model. By training the recurrent neural network model, learn the patterns and features in the calibrated and sharpened image of the reimbursement voucher. Use labeled data to train the recurrent neural network model. The trained recurrent neural network model can be used to extract the voucher visual descriptor of the calibrated and sharpened image of the reimbursement voucher, that is, obtain the voucher visual descriptor bound. The voucher visual descriptor represents the image visual features of the calibrated and sharpened image of the reimbursement voucher, and the voucher visual descriptor bound represents the set of all voucher visual descriptors.
[0073] It should be noted that the contrast and brightness of the sharpened image of the reimbursement voucher can be adjusted through calibration, thereby improving the image quality and making it clearer and easier to read. By defining the visual description sub-boundaries of the voucher, the processing speed of the reimbursement voucher information is accelerated, the information on the sharpened image of the reimbursement voucher is identified more accurately, the possibility of misjudgment is reduced, and the accuracy and efficiency of reimbursement voucher image processing are improved.
[0074] In step 104, the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher are obtained through the bound of the visual descriptor of the voucher, and the visual distribution similarity basis is determined by the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher.
[0075] In some embodiments, the typical value and discrete matrix of the visual descriptor of the voucher are determined through the visual descriptor boundary of the voucher. It should be noted that in this application, there are multiple visual descriptor subclasses of vouchers in the visual descriptor boundary. The typical value of the visual descriptor of the voucher represents the representative visual descriptor of the voucher in the visual descriptor subclass, and the discrete matrix of the visual descriptor of the voucher represents the degree of dispersion among the visual descriptors of the voucher in ...
[0076] Preferably, in some embodiments, the determination of the visual distribution similarity basis from the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher can be carried out in the following manner:
[0077] Determine the prior probability of each type of visual descriptor in the visual descriptor bound of the voucher;
[0078] The visual distribution similarity basis is determined by the prior probability of the visual descriptor subclass, the typical value of the visual descriptor, and the discrete matrix of the visual descriptor. In specific implementation, the visual distribution similarity basis can be determined according to the following formula:
[0079]
[0080] in, Let represent the visual distribution similarity cardinality of the i-th type of visual descriptor within the visual descriptor subboundion, and let n represent the total number of visual descriptor subcategories within the visual descriptor subboundion. This represents the prior probability corresponding to the visual descriptor of the i-th type of credential. This represents the prior probability corresponding to the visual descriptor of the j-th type of credential. The typical quantity of the visual descriptor of the i-th type of voucher represents the visual descriptor of the voucher. The typical quantity of the visual descriptor of the j-th type of voucher represents the visual descriptor of the voucher. Let represent the discrete matrix of visual descriptors for the i-th type of voucher. This represents the discrete matrix of the visual descriptors of the j-th type of voucher, where T denotes the transpose operation. This indicates an intermediate reference variable and has no actual meaning. It should be noted that in this application, the visual distribution similarity basis represents the similarity basis of the voucher visual description subclass based on the probability distribution relative to the voucher visual description subbound. The larger the visual distribution similarity basis, the more similar the voucher visual description subclass is to the voucher visual description subbound, indicating that the information entropy of the voucher visual description subclass is greater.
[0081] In a specific implementation, the prior probability corresponding to each type of visual descriptor in the visual descriptor bounding box of the voucher can be obtained. Specifically, for each visual descriptor feature category, the frequency of occurrence of that type of visual descriptor in the entire visual descriptor bounding box of the voucher can be counted, and then divided by the total number of visual descriptor bounding boxes of the voucher to obtain the prior probability of that type of visual descriptor.
[0082] Additionally, it should be noted that visual distribution similarity basis can be used to measure highly correlated visual descriptors of vouchers, which can help reimbursement classification and verification quickly find key information and improve the efficiency of reimbursement voucher image processing.
[0083] In step 105, the expense reimbursement vouchers uploaded by employees are automatically classified and verified based on the voucher visual description sub-boundary and the visual distribution similarity basis.
[0084] In some embodiments, the automatic classification and verification of expense vouchers uploaded by employees based on the voucher visual description sub-boundary and the visual distribution similarity basis can be carried out in the following manner:
[0085] The core boundary of the visual description sub-boundary of the voucher is determined based on the visual description sub-boundary and the visual distribution similarity basis.
[0086] Automatic reimbursement voucher classification is performed through the core interface of the voucher visual description sub-section to obtain the automatic classification result of the reimbursement voucher;
[0087] The reimbursement vouchers uploaded by employees are verified based on the automatic classification results of the reimbursement vouchers.
[0088] In the above embodiments, determining the core boundary of a voucher visual description sub-boundary based on the voucher visual description sub-boundary and visual distribution similarity cardinality involves selecting the voucher visual description sub-class with the highest visual distribution similarity cardinality as the core boundary. Specifically, the visual distribution similarity cardinality of all voucher visual description sub-classes in the voucher visual description sub-boundary is compared, and the voucher visual description sub-class with the highest visual distribution similarity cardinality is selected. This voucher visual description sub-class is the most important voucher visual description sub-class with the most notification information in the voucher visual description sub-boundary, and is therefore representative. Consequently, this voucher visual description sub-class is selected as the core boundary of the voucher visual description sub-boundary. It should be noted that extracting the core boundary of the voucher visual description sub-boundary enables a faster understanding of the key information of reimbursement vouchers, accelerating the approval decision-making process. For a large amount of reimbursement voucher data, the core boundary of the voucher visual description sub-boundary allows the system to process more data without affecting approval efficiency due to increased data volume.
[0089] In the above embodiments, automatic reimbursement classification using the core boundary of the voucher visual descriptor involves inputting the core boundary of the voucher visual descriptor into a support vector machine for automatic reimbursement classification. It should be noted that in this application, there are two types of automatic reimbursement voucher classification results: one where the voucher image reaches the preset image clarity in the reimbursement system, and another where the voucher image does not reach the preset image clarity in the reimbursement system. Specifically, the core boundary of the voucher visual descriptor is input into a support vector machine for automatic reimbursement classification to obtain the automatic reimbursement voucher classification result. An optimal linear classification surface that can divide the voucher into two classes and maximize the margin can be constructed using the support vector machine to build an automatic reimbursement voucher classification function. The question then becomes how to use this automatic reimbursement voucher classification function to perform automatic reimbursement classification and obtain the automatic reimbursement voucher classification result.
[0090] In the above embodiments, the reimbursement vouchers uploaded by employees are verified according to the automatic classification results of the reimbursement vouchers. Specifically, when the image of the reimbursement voucher reaches the preset image clarity in the reimbursement system, the reimbursement voucher is subjected to subsequent reimbursement verification processing. The authenticity of the reimbursement voucher is verified based on the historical reimbursement data in the reimbursement system and the employee's personal credit information. When the image of the reimbursement voucher does not reach the preset image clarity in the reimbursement system, the reimbursement system sends an active withdrawal message to the employee, reminding the employee to upload the reimbursement voucher image again.
[0091] It should be noted that this application significantly accelerates the processing speed of reimbursement voucher images through automated classification and blockchain, making the reimbursement voucher processing process more accurate, reducing manual intervention, and improving the efficiency of reimbursement voucher image processing. The reimbursement voucher image processing results are recorded on the blockchain, improving the transparency of the reimbursement voucher image processing process and the security of the data.
[0092] In this application, firstly, blockchain technology ensures the secure transmission and storage of reimbursement vouchers, increasing their credibility. Blockchain-based systems enable direct peer-to-peer interaction, reducing intermediate steps, accelerating the reimbursement process, and effectively improving review efficiency. Secondly, adjusting the discreteness of reimbursement vouchers helps improve the quality of the adjusted images, making specific features in the sharpened images more prominent and enhancing the information that needs attention. This facilitates better analysis and processing of the sharpened images. Furthermore, correction improves the quality of the sharpened images, making them clearer and more readable. By defining the visual description sub-boundaries of the vouchers, the processing speed of the reimbursement voucher images is accelerated, information on the sharpened images is identified more accurately, the possibility of misjudgment is reduced, and the efficiency of the reimbursement process is improved. The accuracy and efficiency of reimbursement voucher image processing are improved. Then, by using visual distribution similarity basis, highly correlated voucher visual descriptors can be measured, which can help quickly find key information in reimbursement classification and verification, thus improving the efficiency of reimbursement voucher image processing. Finally, the extraction of the core boundaries of the voucher visual descriptors enables a faster understanding of the key information in the reimbursement voucher image, accelerating the processing speed. For large amounts of reimbursement voucher data, the core boundaries of the voucher visual descriptors allow the system to process more data without affecting the efficiency of reimbursement voucher image processing due to increased data volume. Through automated classification and blockchain, the speed of reimbursement voucher image processing is greatly accelerated, making the reimbursement voucher image processing process more accurate, reducing manual intervention, and improving the efficiency of reimbursement voucher image processing. This effectively improves the processing efficiency of the employee self-service reimbursement system for reimbursement voucher images uploaded by employees.
[0093] Furthermore, in another aspect of this application, in some embodiments, this application provides a blockchain-based employee self-service expense reimbursement system, which includes a control unit, as referenced. Figure 2 The figure is a schematic diagram of exemplary hardware and / or software of a control unit according to some embodiments of this application. The control unit 200 includes: an acquisition module 201, an adjustment module 202, an extraction module 203, a determination module 204, and a classification verification module 205, which are described below:
[0094] The acquisition module 201 in this application is mainly used to acquire the reimbursement vouchers uploaded by employees to the self-service reimbursement center through the blockchain reimbursement node, and obtain the reimbursement voucher image.
[0095] The calibration module 202 in this application is mainly used to determine the discrete calibration degree of the reimbursement voucher corresponding to the reimbursement voucher image, and to calibrate the reimbursement voucher image by the discrete calibration degree of the reimbursement voucher to obtain the reimbursement voucher calibrated image.
[0096] Extraction module 203, in this application, is mainly used to determine the horizontal feature difference coefficient and the vertical feature difference coefficient corresponding to the reimbursement voucher adjustment image, determine the reimbursement voucher adjustment and sharpening image through the horizontal feature difference coefficient and the vertical feature difference coefficient, and then extract the voucher visual description sub-boundary from the reimbursement voucher adjustment and sharpening image;
[0097] The determination module 204 in this application is mainly used to obtain the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher through the bounds of the visual descriptor of the voucher, and to determine the visual distribution similarity basis by the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher.
[0098] The classification verification module 205 in this application is mainly used to automatically classify and verify the expense vouchers uploaded by employees based on the visual description sub-boundary of the voucher and the visual distribution similarity basis.
[0099] The foregoing has detailed examples of a blockchain-based employee self-service expense reimbursement system and its control method provided in the embodiments of this application. It is understood that the corresponding apparatus, in order to achieve the above functions, includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0100] In some embodiments, this application also provides a computer device, the computer device including a memory and a processor, the memory for storing a computer program, and the processor for calling and running the computer program from the memory, so that the computer device executes the control method of the blockchain-based employee self-service expense reimbursement system described above.
[0101] In some embodiments, reference Figure 3 The dashed lines in the figure indicate that the unit or module is optional. This figure is a structural schematic diagram of a computer device for a control method of a blockchain-based employee self-service expense reimbursement system according to an embodiment of this application. The control method of the blockchain-based employee self-service expense reimbursement system described in the above embodiment can be... Figure 3 The computer device shown is used to implement this, and the computer device 300 includes at least one processor 301, a memory 302 and at least one communication unit 305. The computer device 300 may be a terminal device, a server or a chip.
[0102] Processor 301 can be a general-purpose processor or a special-purpose processor. For example, processor 301 can be a central processing unit (CPU). The CPU can be used to control computer device 300, execute software programs, and process data of software programs. Computer device 300 may also include a communication unit 305 for signal input (receiving) and output (transmitting).
[0103] For example, computer device 300 may be a chip, communication unit 305 may be the input and / or output circuit of the chip, or communication unit 305 may be the communication interface of the chip, and the chip may be a component of terminal device, network device or other device.
[0104] For example, computer device 300 may be a terminal device or a server, and communication unit 305 may be a transceiver of the terminal device or the server, or communication unit 305 may be a transceiver circuit of the terminal device or the server.
[0105] The computer device 300 may include one or more memories 302 storing a program 304. The program 304 can be executed by a processor 301 to generate instructions 303, causing the processor 301 to execute the method described in the above method embodiments according to the instructions 303. Optionally, the memory 302 may also store data (such as a target audit model). Optionally, the processor 301 may also read data stored in the memory 302, which may be stored at the same memory address as the program 304, or it may be stored at a different memory address than the program 304.
[0106] The processor 301 and memory 302 can be configured separately or integrated together, for example, integrated on the system on chip (SOC) of the terminal device.
[0107] It should be understood that each step of the above method embodiment can be completed by hardware logic circuits or software instructions in the processor 301. The processor 301 can be a central processing unit, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, such as discrete gates, transistor logic devices, or discrete hardware components.
[0108] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0109] For example, in some embodiments, this application also provides a computer-readable storage medium storing instructions or code that, when executed on a computer, cause the computer to implement the control method of the blockchain-based employee self-service expense reimbursement system described above.
[0110] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.
[0111] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the present invention.
[0112] The spirit and scope of this application. Thus, if these modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations.
Claims
1. A control method of a blockchain-based employee self-reimbursement system, characterized by, Includes the following steps: Obtain the reimbursement vouchers uploaded by employees to the self-service reimbursement center through the blockchain reimbursement node, and obtain the reimbursement voucher image; Determine the discrepancy degree of the reimbursement voucher corresponding to the reimbursement voucher image, and adjust the reimbursement voucher image using the discrepancy degree to obtain an adjusted reimbursement voucher image. Determine the horizontal feature difference coefficient and the vertical feature difference coefficient corresponding to the adjusted image of the reimbursement voucher, determine the adjusted and sharpened image of the reimbursement voucher through the horizontal feature difference coefficient and the vertical feature difference coefficient, and then extract the visual description sub-boundary of the voucher from the adjusted and sharpened image of the reimbursement voucher. By using the bounds of the visual descriptor of the voucher, the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher are obtained, and the visual distribution similarity basis is determined by the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher. The reimbursement vouchers uploaded by employees are automatically classified and verified based on the visual description sub-boundary of the voucher and the visual distribution similarity basis.
2. The method as described in claim 1, characterized in that, Determining the discrepancy degree of the reimbursement voucher corresponding to the image specifically includes: The reimbursement voucher image is converted to grayscale to obtain a grayscale image of the reimbursement voucher. Contour detection is performed on the grayscale image of the reimbursement voucher to determine the contour circumscribed matrix; The discrete adjustment degree of the reimbursement voucher is obtained based on the outline outer matrix.
3. The method as described in claim 1, characterized in that, Determining the sharpening and refining of the reimbursement voucher image using the horizontal feature difference coefficient and the vertical feature difference coefficient specifically includes: The frequency and phase of the cosine function are determined by the discrete adjustment degree of the expense reimbursement voucher; The horizontal and vertical difference ratios of the Gaussian function are determined by the horizontal feature difference coefficient and the vertical feature difference coefficient. The adjusted and sharpened image of the expense voucher is obtained by convolving the frequency and phase of the cosine function, the horizontal feature difference coefficient, the vertical feature difference coefficient, the horizontal and vertical difference ratio of the Gaussian function, and the adjusted image of the expense voucher.
4. The method as described in claim 1, characterized in that, The extraction of the visual description sub-boundary of the reimbursement voucher from the sharpened image specifically includes: The sharpened image of the reimbursement voucher is subjected to piecewise linear fitting to obtain a linear gamma correction curve; The reimbursement voucher sharpened image is segmented and corrected using the linear gamma correction curve to obtain the corrected reimbursement voucher sharpened image. The visual descriptor of the corrected reimbursement voucher is extracted by using a recurrent neural network to obtain the visual descriptor boundary of the voucher.
5. The method as described in claim 1, characterized in that, The automatic classification and verification of expense reimbursement vouchers uploaded by employees based on the visual description sub-boundary of the voucher and the visual distribution similarity basis specifically includes: The core boundary of the visual description sub-boundary of the voucher is determined based on the visual description sub-boundary and the visual distribution similarity basis. Automatic reimbursement voucher classification is performed through the core interface of the voucher visual description sub-section to obtain the automatic classification result of the reimbursement voucher; The reimbursement vouchers uploaded by employees are verified based on the automatic classification results of the reimbursement vouchers.
6. The method as described in claim 5, characterized in that, The core bound of a voucher visual description subclass is determined by taking the voucher visual description subclass with the highest visual distribution similarity as the core bound of the voucher visual description subclass.
7. The method as described in claim 5, characterized in that, Automatic reimbursement classification using the core bounds of the voucher visual descriptor involves inputting the core bounds of the voucher visual descriptor into a support vector machine for automatic reimbursement classification.
8. A blockchain-based employee self-service expense reimbursement system, characterized in that, The blockchain-based employee self-service expense reimbursement system includes a control unit, which comprises: The acquisition module is used to acquire reimbursement vouchers uploaded by employees to the self-service reimbursement center through the blockchain reimbursement node and obtain the reimbursement voucher image. The calibration module is used to determine the discrete calibration degree of the reimbursement voucher corresponding to the reimbursement voucher image, and to calibrate the reimbursement voucher image by means of the discrete calibration degree of the reimbursement voucher to obtain the reimbursement voucher calibrated image. The extraction module is used to determine the horizontal feature difference coefficient and the vertical feature difference coefficient corresponding to the adjusted image of the reimbursement voucher, determine the adjusted and sharpened image of the reimbursement voucher through the horizontal feature difference coefficient and the vertical feature difference coefficient, and then extract the visual description sub-boundary of the voucher from the adjusted and sharpened image of the reimbursement voucher. The determination module is used to obtain the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher through the bounds of the visual descriptor of the voucher, and to determine the visual distribution similarity basis by the typical quantity of the visual descriptor of the voucher and the discrete matrix of the visual descriptor of the voucher; The classification and verification module is used to automatically classify and verify the expense vouchers uploaded by employees based on the visual description sub-boundary of the voucher and the visual distribution similarity basis.
9. A computer device, characterized in that, The computer device includes a memory and a processor. The memory is used to store computer programs, and the processor is used to call and run the computer programs from the memory, causing the computer device to execute the control method of the blockchain-based employee self-service expense reimbursement system according to any one of claims 1 to 7.
10. A computer-readable storage medium storing instructions or code that, when executed on a computer, cause the computer to implement the control method for a blockchain-based employee self-service expense reimbursement system as described in any one of claims 1 to 7.