CN112819274A - Financial voucher sample generation method and device and related method - Google Patents
Financial voucher sample generation method and device and related method Download PDFInfo
- Publication number
- CN112819274A CN112819274A CN202011623819.5A CN202011623819A CN112819274A CN 112819274 A CN112819274 A CN 112819274A CN 202011623819 A CN202011623819 A CN 202011623819A CN 112819274 A CN112819274 A CN 112819274A
- Authority
- CN
- China
- Prior art keywords
- sample
- name
- image
- financial
- filling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Finance (AREA)
- Educational Administration (AREA)
- Technology Law (AREA)
- Game Theory and Decision Science (AREA)
- Accounting & Taxation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Character Input (AREA)
Abstract
The application discloses a financial voucher sample generation method, a financial voucher sample generation device and a related method. The generation method comprises the steps of randomly acquiring a bottom plate image sample from a bottom plate image set; determining a first location and a second location in a floor image sample; randomly obtaining a name sample according to the surname set and the name set; and, randomly generating a sample of the amount; and filling the name sample into the first position, and filling the amount sample into the second position to obtain the financial certificate sample. Therefore, the financial voucher samples are generated from the image bottom plate, the name of the person and the amount of money, and abundant and balanced financial voucher samples can be generated. Therefore, sufficient samples are provided for model training, the sample coverage range is wide, and the distribution is uniform. Therefore, the samples have a good training effect on the deep learning model, and the accuracy of image recognition of the financial document by using the deep learning model is improved.
Description
Technical Field
The present application relates to the field of image recognition, and in particular, to a method and an apparatus for generating a financial document sample, and a related method.
Background
Deep learning is currently widely used in various industries. In the financial industry, the deep learning model can be used for image recognition of financial documents. Specifically, the financial instrument includes at least one of an invoice, a check, a deposit receipt, and a passbook. Before the financial documents are identified by the deep learning model, a large number of sample images are required to train the deep learning model. However, some financial documents have fewer image samples, and in addition, because the sample images of some financial documents are matched with the supervision requirements, the sample images of some financial documents cannot be used in model training, so that the number of samples of the financial documents is deficient. And if the number of samples is insufficient, the precision and accuracy of the trained model are difficult to ensure, and the use effect of the model is influenced.
Disclosure of Invention
In order to solve the above technical problems, the present application provides a method, an apparatus and a related method for generating financial document samples, so as to expand the number of financial document samples.
In order to achieve the above purpose, the technical solutions provided in the embodiments of the present application are as follows:
the embodiment of the application provides a method for generating a financial voucher sample, which is characterized by comprising the following steps:
randomly acquiring a bottom plate image sample from a bottom plate image set;
determining a first location and a second location in the floor image sample; the first position is a position to be filled with a name on the bottom plate image sample, and the second position is a position to be filled with money on the bottom plate image sample;
randomly obtaining a name sample according to the surname set and the name set; and, randomly generating a sample of the amount;
filling the name sample into the first position, and filling the amount sample into the second position to obtain an image sample;
and obtaining a financial certificate sample according to the image sample.
Optionally, the method further comprises:
shifting the coordinates of the first position and the coordinates of the second position by using a Gaussian function to obtain a first shifted position corresponding to the first position and a second shifted position corresponding to the second position;
the filling the name sample into the first position and the amount sample into the second position comprises:
and filling the name sample into the first offset position, and filling the amount sample into the second offset position.
Optionally, the randomly determining a sample of the names of the people according to the surname set and the first name set includes:
uniformly and randomly sampling in the surname set to obtain surname samples; obtaining name character strings in the name set through logarithmic Gaussian sampling according to word frequency; the name character string comprises at least one Chinese character;
and splicing the surname sample and the name character string to obtain the name sample.
Optionally, the randomly generating the monetary sample comprises:
generating a random number with the digit number of N, wherein N is any one integer from 1 to 9;
when the amount sample is generated every time, directly taking the random number as the amount sample according to a first probability, presetting the random number according to a second probability, and taking the processed random number as the amount sample; the sum of the first probability and the second probability is 1.
Optionally, the performing, with the second probability, a preset process on the random number includes: the random number is divided by 100 to obtain a random number comprising two decimal places.
Optionally, the obtaining a financial document sample according to the image sample includes: performing at least one of the following image enhancement processing on the image sample to obtain a processed image sample:
clipping, zooming, Gaussian blur, brightness adjustment, contrast adjustment, hue adjustment, illumination adjustment, image distortion, image clipping or view angle adjustment;
and taking the processed image sample as a financial certificate sample.
The embodiment of the application provides a financial document sample generating device, the device includes:
the bottom plate acquisition module is used for randomly acquiring bottom plate image samples from the bottom plate image set;
a position determination module to determine a first position and a second position in the floor image sample; the first position is a position to be filled with a name on the bottom plate image sample, and the second position is a position to be filled with money on the bottom plate image sample;
the content generation module is used for randomly acquiring a name sample according to the surname set and the name set; and, randomly generating a sample of the amount;
the filling module is used for filling the name sample into the first position and filling the amount sample into the second position to obtain an image sample;
and the sample obtaining module is used for obtaining the financial voucher sample according to the image sample.
Optionally, the apparatus further comprises:
a position offset module, configured to offset the coordinates of the first position and the coordinates of the second position by using a gaussian function, so as to obtain a first offset position corresponding to the first position and a second offset position corresponding to the second position;
the filling module is specifically configured to: and filling the name sample into the first offset position, and filling the amount sample into the second offset position.
The embodiment of the application further provides a deep learning model training method, which is characterized by comprising the following steps:
combining m actually collected financial voucher samples and the corresponding financial information thereof, n financial voucher samples generated according to the financial voucher sample generation method and the corresponding financial information thereof into a group of training data; both m and n are integers;
and training the deep learning model to be trained by utilizing the training data.
The embodiment of the application also provides a generation method, a generation device and a related method, wherein the method comprises the following steps:
inputting an image of a financial instrument into the deep learning model of claim 9, obtaining financial information corresponding to the financial instrument.
According to the technical scheme, the method has the following beneficial effects:
the embodiment of the application provides a financial document sample generation method, a model training method and an identification method. In the financial voucher sample generation method, a bottom plate image sample is randomly acquired from a bottom plate image set; determining a first location and a second location in a floor image sample; the first position is the position of a name to be filled in the bottom plate image sample, and the second position is the position of a sum to be filled in the bottom plate image sample; randomly obtaining a name sample according to the surname set and the name set; and, randomly generating a sample of the amount; filling the name sample into the first position, and filling the amount sample into the second position to obtain an image sample; and obtaining the financial certificate sample according to the image sample. Therefore, the financial voucher samples are generated from the image bottom plate, the name of the person and the amount of money, and abundant and balanced financial voucher samples can be generated. Therefore, sufficient samples are provided for model training, the sample coverage range is wide, and the distribution is uniform. Therefore, the samples have a good training effect on the deep learning model, and the accuracy of image recognition of the financial document by using the deep learning model is improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a flowchart illustrating a method for generating a sample financial document according to an embodiment of the present disclosure;
fig. 2 is a schematic structural diagram of a financial document sample generation apparatus according to an embodiment of the present disclosure.
Detailed Description
In order to help better understand the scheme provided by the embodiment of the present application, before describing the method provided by the embodiment of the present application, a scenario of an application of the scheme of the embodiment of the present application is described.
Deep learning is currently widely used in various industries. In the financial industry, the deep learning model can be used for image recognition of financial documents. The financial instrument includes at least one of an invoice, a check, a deposit slip, and a passbook. Before the financial documents are identified by the deep learning model, a large number of sample images are required to train the deep learning model. However, some financial documents have fewer image samples, and in addition, because the sample images of some financial documents are matched with the supervision requirements, the sample images of some financial documents cannot be used in model training, so that the number of samples of the financial documents is deficient. And if the number of samples is insufficient, the precision and accuracy of the trained model are difficult to ensure, and the use effect of the model is influenced.
In order to solve the technical problem, an embodiment of the application provides a financial document sample generation method, a model training method and an identification method. In the financial voucher sample generation method, a bottom plate image sample is randomly acquired from a bottom plate image set; determining a first location and a second location in a floor image sample; the first position is the position of a name to be filled in the bottom plate image sample, and the second position is the position of a sum to be filled in the bottom plate image sample; randomly obtaining a name sample according to the surname set and the name set; and, randomly generating a sample of the amount; filling the name sample into the first position, and filling the amount sample into the second position to obtain an image sample; and obtaining the financial certificate sample according to the image sample. Therefore, the financial voucher samples are generated from the image bottom plate, the name of the person and the amount of money, and abundant and balanced financial voucher samples can be generated. Therefore, sufficient samples are provided for model training, the sample coverage range is wide, and the distribution is uniform. Therefore, the samples have a good training effect on the deep learning model, and the accuracy of image recognition of the financial document by using the deep learning model is improved.
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, embodiments accompanying the drawings are described in detail below.
Referring to fig. 1, a method for generating a financial document sample according to an embodiment of the present application is shown. As shown in fig. 1, the method for generating a financial document sample according to the embodiment of the present application includes the following steps S101 to S105:
s101: a floor image sample is randomly acquired from a floor image collection.
S102: determining a first location and a second location in a floor image sample; the first position is the position of the name to be filled in the bottom plate image sample, and the second position is the position of the amount to be filled in the bottom plate image sample.
S103: randomly obtaining a name sample according to the surname set and the name set; and, randomly generating a sample of the amount.
S104: and filling the name sample into the first position, and filling the amount sample into the second position to obtain the image sample.
S105: and obtaining the financial certificate sample according to the image sample.
It should be noted that step S103 in the embodiment of the present application may be executed after step S101 and step S102. In the case of determining the content to be filled in the floor image sample, step S103 in this application may be executed simultaneously with step S101 and step S102, or may be executed before step S101 and step S102, and the embodiment of this application is not limited herein.
In the embodiment of the application, a plurality of types of financial instrument sample backplanes are included in the backplane image collection. In order to avoid the influence of the size of the floor image on the training result, as a possible implementation, after the floor image sample is obtained, the floor image sample may also be scaled to the size of the target sample image. It can be understood that, because there may be differences in the sizes of the floor images in the floor image set, and these differences may have an effect on the deep learning model trained in the embodiment of the present application, the present application avoids such an effect by unifying the sizes of the plate images, so that the training effect of these samples on the deep learning model is better.
It should be noted that, the floor image sample in the embodiment of the present application may include only the first position and the second position, and may also include other positions, and the embodiment of the present application is not limited herein. As an example, the floor image sample in the embodiment of the present application may include a third position, where the credential number is to be filled in on the floor image sample. The floor image sample in the embodiment of the present application may further include a fourth position, where the fourth position is a position to be filled with an account on the floor image sample. As a possible implementation manner, the credential number and the account number in the embodiment of the present application may be obtained by randomly generating a numeric character string with a preset length and then adding a separator at a preset position. Because account numbers often have business relevance in actually collected financial samples, certain digits in a batch of samples are often fixed. Therefore, the certificate number and the account number in the embodiment have good randomness, and the identification range of the deep learning model in the embodiment of the application can be enlarged.
The following describes steps S102 to S104 of the financial document sample generation method according to an embodiment of the present application in detail:
(1) position determination
In an embodiment of the present application, as a possible implementation manner, the method for generating a financial document sample further includes: and offsetting the coordinates of the first position and the coordinates of the second position by using a Gaussian function to obtain a first offset position corresponding to the first position and a second offset position corresponding to the second position. At this time, in the embodiment of the present application, the name sample is filled in the first location, and the amount sample is filled in the second location, specifically: the person name sample is filled into the first offset location and the amount sample is filled into the second offset location. Specifically, the coordinates x, y for each location may be:
x＝x0+dx
y＝y0+dy
x～N(μ,d)
y～N(μ,d)
where x0, y0 are coordinates predefined by the content, N is a Gaussian distribution with mean μ and variance d.
In practical applications, when the financial document is printed, the position of the content of the document may be disturbed due to the printing problem. Therefore, to mimic this perturbation, embodiments of the present application offset the coordinates of the first and second locations by a gaussian function. Therefore, the diversity of the financial voucher samples in the embodiment of the application is improved, and the training effect of the samples on the deep learning model is better.
(2) Name sample generation
In this embodiment of the present application, as a possible implementation manner, randomly determining a sample of a personal name according to a surname set and a first name set may include: uniformly and randomly sampling in the surname set to obtain surname samples; obtaining name character strings in the name set through logarithmic Gaussian sampling according to word frequency; the name character string comprises at least one Chinese character; and splicing the surname sample and the first name character string to obtain a name sample. In addition, because names are highly abundant in practice, in order to improve the recognition range of the deep learning model to names in the present application, the name samples generated in the embodiments of the present application have good uniformity and a wide range, and can be sufficiently trained. Meanwhile, the probability of occurrence of the Chinese characters in the name samples generated by the embodiment of the application conforms to the lognormal distribution, which is consistent with the distribution of the Chinese characters in practical application, so that the name samples of the embodiment of the application cannot influence the identification of the model provided by the embodiment of the application to other content samples (such as money samples).
(3) Monetary sample generation
In this embodiment, as a possible implementation manner, randomly generating the monetary sample may include: generating a random number with the digit number of N, wherein N is any integer from 1 to 9; when the amount sample is generated every time, the random number is directly used as the amount sample according to the first probability, the random number is subjected to preset processing according to the second probability, and the processed random number is used as the amount sample; the sum of the first probability and the second probability is 1. It should be noted that, because the identification of the money sample is important in practical application, in order to improve the accuracy of the deep learning model in the present application to the money sample, the money samples generated in the embodiment of the present application are uniformly distributed, and the number of the samples is large, so that the accuracy of the deep learning model to the identification of the money sample can be improved.
Further, in order to improve the recognition range of the deep learning model for the money sample, in this embodiment of the application, the pre-setting the random number with the second probability may include: the random number is divided by 100 to obtain a random number comprising two decimal places. Therefore, the amount samples in the embodiment of the present application may include data that is less common in real samples and also includes angles and quantiles, so that the identification range of the deep learning model in the embodiment of the present application to the amount samples is expanded.
In an embodiment, to train the deep learning model further, the obtaining of financial document samples from image samples in an embodiment of the present application includes: performing at least one of the following image enhancement processing on the image sample to obtain a processed image sample: clipping, zooming, Gaussian blur, brightness adjustment, contrast adjustment, hue adjustment, illumination adjustment, image distortion, image clipping or view angle adjustment; and using the processed image sample as a financial certificate sample.
In summary, the method provided by the embodiment of the application generates the financial document samples according to different characteristics of the image bottom plate, the name and the amount of money, and can generate abundant and balanced financial document samples. Therefore, sufficient samples are provided for model training, the sample coverage range is wide, and the distribution is uniform. Therefore, the samples have a good training effect on the deep learning model, and the accuracy of image recognition of the financial document by using the deep learning model is improved.
According to the financial voucher sample generation method provided by the embodiment, the embodiment of the application also provides a financial voucher sample generation device. Referring to fig. 2, the figure is a schematic structural diagram of a financial document sample generating device according to an embodiment of the present application. As shown in fig. 2, an apparatus for generating a financial document sample according to an embodiment of the present application includes:
a bottom plate obtaining module 100, configured to randomly obtain a bottom plate image sample from a bottom plate image set;
a position determination module 200 for determining a first position and a second position in a floor image sample; the first position is the position of a name to be filled in the bottom plate image sample, and the second position is the position of a sum to be filled in the bottom plate image sample;
the content generation module 300 is configured to randomly obtain a name sample according to the surname set and the first name set; and, randomly generating a sample of the amount;
the filling module 400 is used for filling the name sample into the first position and filling the amount sample into the second position to obtain an image sample;
a sample obtaining module 500, configured to obtain a financial document sample according to the image sample.
As a possible implementation manner, an apparatus provided in an embodiment of the present application further includes: the position offset module is used for offsetting the coordinates of the first position and the coordinates of the second position by using a Gaussian function to obtain a first offset position corresponding to the first position and a second offset position corresponding to the second position; the fill-in module is specifically configured to: the person name sample is filled into the first offset location and the amount sample is filled into the second offset location.
To sum up, the device that this application embodiment provided generates the financial document sample to the image bottom plate, the different characteristics in the three aspect of name and amount of money, can produce abundant and balanced financial document sample. Therefore, sufficient samples are provided for model training, the sample coverage range is wide, and the distribution is uniform. Therefore, the samples have a good training effect on the deep learning model, and the accuracy of image recognition of the financial document by using the deep learning model is improved.
Based on the financial instrument sample generation method and the financial instrument sample generation device provided by the above embodiments, the embodiment of the application further provides a deep learning model training method, and the method includes: combining m actually collected financial voucher samples and the corresponding financial information thereof, n financial voucher samples generated according to the financial voucher sample generation method provided by the embodiment and the corresponding financial information thereof into a set of training data; and training the deep learning model to be trained by utilizing the training data. It should be noted that m and n are integers.
Based on the method and the device for generating the financial voucher sample provided by the embodiment, the embodiment of the application further provides a method for identifying the financial voucher, and the method comprises the following steps: the image of the financial voucher is input into the deep learning model provided in the above embodiment, and the financial information corresponding to the financial voucher is obtained.
As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that all or part of the steps in the above embodiment methods can be implemented by software plus a necessary general hardware platform. Based on such understanding, the technical solution of the present application may be essentially or partially implemented in the form of a software product, which may be stored in a storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the method according to the embodiments or some parts of the embodiments of the present application.
It should be noted that, in the present specification, the embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments may be referred to each other. The method disclosed by the embodiment corresponds to the system disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the system part for description.
It should also be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A method of generating a sample of financial instruments, the method comprising:
randomly acquiring a bottom plate image sample from a bottom plate image set;
determining a first location and a second location in the floor image sample; the first position is a position to be filled with a name on the bottom plate image sample, and the second position is a position to be filled with money on the bottom plate image sample;
randomly obtaining a name sample according to the surname set and the name set; and, randomly generating a sample of the amount;
filling the name sample into the first position, and filling the amount sample into the second position to obtain an image sample;
and obtaining a financial certificate sample according to the image sample.
2. The method of claim 1, further comprising:
shifting the coordinates of the first position and the coordinates of the second position by using a Gaussian function to obtain a first shifted position corresponding to the first position and a second shifted position corresponding to the second position;
the filling the name sample into the first position and the amount sample into the second position comprises:
and filling the name sample into the first offset position, and filling the amount sample into the second offset position.
3. The method of claim 1, wherein randomly determining a sample of names from the set of last names and the set of first names comprises:
uniformly and randomly sampling in the surname set to obtain surname samples; obtaining name character strings in the name set through logarithmic Gaussian sampling according to word frequency; the name character string comprises at least one Chinese character;
and splicing the surname sample and the name character string to obtain the name sample.
4. The method of claim 1, wherein the randomly generating a sample of the amount comprises:
generating a random number with the digit number of N, wherein N is any one integer from 1 to 9;
when the amount sample is generated every time, directly taking the random number as the amount sample according to a first probability, presetting the random number according to a second probability, and taking the processed random number as the amount sample; the sum of the first probability and the second probability is 1.
5. The method of claim 4, wherein the pre-processing the random number with the second probability comprises: the random number is divided by 100 to obtain a random number comprising two decimal places.
6. The method of claim 1, wherein obtaining a sample of financial instruments from the image sample comprises: performing at least one of the following image enhancement processing on the image sample to obtain a processed image sample:
clipping, zooming, Gaussian blur, brightness adjustment, contrast adjustment, hue adjustment, illumination adjustment, image distortion, image clipping or view angle adjustment;
and taking the processed image sample as a financial certificate sample.
7. An apparatus for generating a sample of financial instruments, the apparatus comprising:
the bottom plate acquisition module is used for randomly acquiring bottom plate image samples from the bottom plate image set;
a position determination module to determine a first position and a second position in the floor image sample; the first position is a position to be filled with a name on the bottom plate image sample, and the second position is a position to be filled with money on the bottom plate image sample;
the content generation module is used for randomly acquiring a name sample according to the surname set and the name set; and, randomly generating a sample of the amount;
the filling module is used for filling the name sample into the first position and filling the amount sample into the second position to obtain an image sample;
and the sample obtaining module is used for obtaining the financial voucher sample according to the image sample.
8. The apparatus of claim 7, further comprising:
a position offset module, configured to offset the coordinates of the first position and the coordinates of the second position by using a gaussian function, so as to obtain a first offset position corresponding to the first position and a second offset position corresponding to the second position;
the filling module is specifically configured to: and filling the name sample into the first offset position, and filling the amount sample into the second offset position.
9. A deep learning model training method, the method comprising:
combining m actually collected financial instrument samples and corresponding financial information thereof, n financial instrument samples generated according to the financial instrument sample generation method of any one of claims 1-6 and corresponding financial information thereof into a set of training data; both m and n are integers;
and training the deep learning model to be trained by utilizing the training data.
10. A method for identifying financial instruments, the method comprising:
inputting an image of a financial instrument into the deep learning model of claim 9, obtaining financial information corresponding to the financial instrument.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011623819.5A CN112819274A (en) | 2020-12-30 | 2020-12-30 | Financial voucher sample generation method and device and related method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011623819.5A CN112819274A (en) | 2020-12-30 | 2020-12-30 | Financial voucher sample generation method and device and related method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112819274A true CN112819274A (en) | 2021-05-18 |
Family
ID=75854734
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011623819.5A Pending CN112819274A (en) | 2020-12-30 | 2020-12-30 | Financial voucher sample generation method and device and related method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112819274A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109753965A (en) * | 2018-12-14 | 2019-05-14 | 深圳壹账通智能科技有限公司 | A kind of document identification method and device, computer equipment based on deep learning |
| CN110163285A (en) * | 2019-05-23 | 2019-08-23 | 阳光保险集团股份有限公司 | Ticket recognition training sample synthetic method and computer storage medium |
| CN111931783A (en) * | 2020-08-18 | 2020-11-13 | 创新奇智(重庆)科技有限公司 | Training sample generation method, machine-readable code identification method and device |
-
2020
- 2020-12-30 CN CN202011623819.5A patent/CN112819274A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109753965A (en) * | 2018-12-14 | 2019-05-14 | 深圳壹账通智能科技有限公司 | A kind of document identification method and device, computer equipment based on deep learning |
| CN110163285A (en) * | 2019-05-23 | 2019-08-23 | 阳光保险集团股份有限公司 | Ticket recognition training sample synthetic method and computer storage medium |
| CN111931783A (en) * | 2020-08-18 | 2020-11-13 | 创新奇智(重庆)科技有限公司 | Training sample generation method, machine-readable code identification method and device |
Non-Patent Citations (1)
| Title |
|---|
| 刘欢: ""基于深度学习的发票图像文本检测与识别"", 中国优秀硕士论文全文数据库信息科技辑（月刊）, no. 03, pages 138 - 1181 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10402163B2 (en) | Intelligent data extraction | |
| CN112613501A (en) | Information auditing classification model construction method and information auditing method | |
| CN107918916A (en) | Self-service Claims Resolution application processing method, device, computer equipment and storage medium | |
| CN109391706A (en) | Domain name detection method, device, equipment and storage medium based on deep learning | |
| CN112614125B (en) | Method and device for detecting glass defects of mobile phone, computer equipment and storage medium | |
| JP2020170495A (en) | Single pixel attack sample generating method, device, facility, and storage medium | |
| CN111931780A (en) | Intelligent management method and equipment for accounting documents | |
| CN110956166A (en) | Bill marking method and device | |
| CN115758451A (en) | Data labeling method, device, equipment and storage medium based on artificial intelligence | |
| CN114511393B (en) | Financial data processing method and system | |
| WO2022126917A1 (en) | Deep learning-based face image evaluation method and apparatus, device, and medium | |
| CN111462388A (en) | Bill inspection method and device, terminal equipment and storage medium | |
| CN107168635A (en) | Information demonstrating method and device | |
| CN114219507A (en) | Qualification auditing method and device for traditional Chinese medicine supplier, electronic equipment and storage medium | |
| CN112464927B (en) | Information extraction method, device and system | |
| CN111274853A (en) | Image processing method and device | |
| CN112819274A (en) | Financial voucher sample generation method and device and related method | |
| CN111027533A (en) | Conversion method and system of point-to-read coordinates, terminal device and storage medium | |
| CN115661839A (en) | Bill filing method, device, terminal and storage medium | |
| CN114912051A (en) | Data processing method, device and computer readable storage medium | |
| CN115713248A (en) | Method for scoring and evaluating data for exchange | |
| CN113792570A (en) | Claims data processing method and device, electronic equipment and readable medium | |
| Tornés et al. | Receipt Dataset for Document Forgery Detection | |
| CN114218391B (en) | Sensitive information identification method based on deep learning technology | |
| CN117391076B (en) | Acquisition method and device of identification model of sensitive data, electronic equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |