CN109815949A - Invoice publicity method and system neural network based - Google Patents
Invoice publicity method and system neural network based Download PDFInfo
- Publication number
- CN109815949A CN109815949A CN201811564802.XA CN201811564802A CN109815949A CN 109815949 A CN109815949 A CN 109815949A CN 201811564802 A CN201811564802 A CN 201811564802A CN 109815949 A CN109815949 A CN 109815949A
- Authority
- CN
- China
- Prior art keywords
- invoice
- publicity
- type
- information
- neural network
- 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
Abstract
A kind of invoice publicity method and system neural network based, this method comprises: step 1: selection invoice sample is based on invoice sample training neural network model;Step 2: the invoice type to publicity invoice is identified according to neural network model;And step 3: the sensitive information to publicity invoice is hidden according to invoice type, and publicity waits for the non-sensitive information of publicity invoice.The present invention can be avoided privacy leakage when invoice publicity, is conducive to invoice and rationalizes publicity and application.
Description
Technical field
The present invention relates to field of computer technology, in particular to a kind of invoice publicity method neural network based and it is
System.
Background technique
The tax or other departments need to carry out invoice publicity under specific circumstances.Invoice information contains customer/enterprise
Bulk information, such as purchaser, pin side, the concerning taxes amount of money, date, commodity managing detailed catalogue, wherein the details of purchaser or pin side are again
It may include Taxpayer Identification Number, address, phone, bank of deposit and account etc..Which kind of, regardless of technology to carry out invoice publicity based on, all may be used
Can there are problems that user sensitive information is compromised.In addition, user is secret for protection privacy or business in more displayings application
Close reason is not intended to reveal excessive information, such as address, commodity detail, the amount of money.Therefore, existing invoice publicity method
It is not only easy to cause the privacy leakage of customer/enterprise, leads to potential security risk, and is unfavorable for invoice and rationalizes publicity and answer
With.
Summary of the invention
The object of the present invention is to provide a kind of invoice publicity method and systems neural network based, to avoid invoice publicity
When privacy leakage, so that invoice rationalizes publicity and application.
One aspect of the present invention proposes a kind of invoice publicity method neural network based, comprising:
Step 1: selection invoice sample is based on the invoice sample training neural network model;
Step 2: the invoice type to publicity invoice is identified according to the neural network model;And
Step 3: the sensitive information to publicity invoice being hidden according to the invoice type, and is sent out described in publicity to publicity
The non-sensitive information of ticket;
The invoice sample includes invoice image and invoice type;
It is described to include: according to the hiding sensitive information to publicity invoice of the invoice type
Step 301: obtaining the identity information of publicity object;
Step 302: the publicity object being determined according to the identity information and described between publicity invoice owner
Relationship;
Step 303: opening relationships template, the relationship templates include publicity object and between publicity invoice owner
Relationship and its corresponding sensitive information, according to the publicity object and described to publicity invoice owner determined in step 302
Between relationship, determine the sensitive information by the relationship templates;
Step 304: sensitive information region is determined according to the invoice type to formula invoice;
Step 305: processing is hidden to the sensitive information region.
Preferably, described to include: based on the invoice sample training neural network model
Using the invoice image as input, using the invoice type as label, the neural network model is carried out
Training.
Preferably, the step 2 includes:
Step 201: obtaining the invoice image to publicity invoice;
Step 202: identifying the text information in the invoice image, determined according to the text information described to publicity hair
The invoice type of ticket is as initial invoice type;
Step 203: using the invoice image to publicity invoice as input, institute being identified according to the neural network model
The invoice type to publicity invoice is stated as correction invoice type;
Step 204: judge whether the initial invoice type and the correction invoice type are consistent, if unanimously, with
The initial invoice type or the correction invoice type are as the invoice type to publicity invoice;If it is inconsistent,
Export default prompt information.
Preferably, the step 202 includes:
The text information in the invoice image is identified using optical character recognition method;
The text information is compared with predetermined keyword, if the text information includes predetermined keyword,
Using the corresponding invoice type of the predetermined keyword as the initial invoice type.
Preferably, the step 304 includes:
Establish invoice information template, invoice information template include invoice type, information corresponding to the invoice type and its
Position of the region in invoice image;
According to the invoice type and sensitive information to publicity invoice, determined based on the invoice information template described quick
Feel information region.
Another aspect of the present invention proposes a kind of invoice publicity system neural network based, comprising:
Memory is stored with computer executable instructions;
Processor, the processor run the computer executable instructions in the memory, execute following steps:
Step 1: selection invoice sample is based on the invoice sample training neural network model;
Step 2: the invoice type to publicity invoice is identified according to the neural network model;And
Step 3: the sensitive information to publicity invoice being hidden according to the invoice type, and is sent out described in publicity to publicity
The non-sensitive information of ticket;
The invoice sample includes invoice image and invoice type;
It is described to include: according to the hiding sensitive information to publicity invoice of the invoice type
Step 301: obtaining the identity information of publicity object;
Step 302: the publicity object being determined according to the identity information and described between publicity invoice owner
Relationship;
Step 303: opening relationships template, the relationship templates include publicity object and between publicity invoice owner
Relationship and its corresponding sensitive information, according to the publicity object and described to publicity invoice owner determined in step 302
Between relationship, determine the sensitive information by the relationship templates;
Step 304: sensitive information region is determined according to the invoice type to formula invoice;
Step 305: processing is hidden to the sensitive information region.
Preferably, described to include: based on the invoice sample training neural network model
Using the invoice image as input, using the invoice type as label, the neural network model is carried out
Training.
Preferably, the step 2 includes:
Step 201: obtaining the invoice image to publicity invoice;
Step 202: identifying the text information in the invoice image, determined according to the text information described to publicity hair
The invoice type of ticket is as initial invoice type;
Step 203: using the invoice image to publicity invoice as input, institute being identified according to the neural network model
The invoice type to publicity invoice is stated as correction invoice type;
Step 204: judge whether the initial invoice type and the correction invoice type are consistent, if unanimously, with
The initial invoice type or the correction invoice type are as the invoice type to publicity invoice;If it is inconsistent,
Export default prompt information.
Preferably, the step 202 includes:
The text information in the invoice image is identified using optical character recognition method;
The text information is compared with predetermined keyword, if the text information includes predetermined keyword,
Using the corresponding invoice type of the predetermined keyword as the initial invoice type.
Preferably, the step 304 includes:
Establish invoice information template, invoice information template include invoice type, information corresponding to the invoice type and its
Position of the region in invoice image;
According to the invoice type and sensitive information to publicity invoice, determined based on the invoice information template described quick
Feel information region.
The beneficial effects of the present invention are:
1, neural network model is established, and model is trained according to invoice sample big data, it is being optimized, be used for
Identify the neural network model of invoice type;Invoice type is identified using neural network model, effectively increases the accurate of identification
Property;It determines which information in invoice is sensitive information according to user identity, and place is hidden to the sensitive information in invoice
Reason both ensure that the safety of invoice publicity, and flexibly meet personalized, displaying application demand.
Method and system of the invention has other characteristics and advantages, these characteristics and advantages are attached from what is be incorporated herein
It will be apparent in figure and subsequent specific embodiment, or will be in the attached drawing being incorporated herein and subsequent specific reality
It applies in mode and is stated in detail, the drawings and the detailed description together serve to explain specific principles of the invention.
Detailed description of the invention
Exemplary embodiment of the present is described in more detail in conjunction with the accompanying drawings, of the invention is above-mentioned and other
Purpose, feature and advantage will be apparent, wherein in exemplary embodiments of the present invention, identical appended drawing reference is usual
Represent same parts.
Fig. 1 shows the flow chart of invoice publicity method neural network based according to an exemplary embodiment of the present invention;
Fig. 2 shows the example of invoice image according to an exemplary embodiment of the present invention;
Fig. 3 shows the invoice photo of exemplary embodiment of the present.
Specific embodiment
The present invention will be described in more detail below with reference to accompanying drawings.Although showing the preferred embodiment of the present invention in attached drawing,
However, it is to be appreciated that may be realized in various forms the present invention and should not be limited by the embodiments set forth herein.On the contrary, providing
These embodiments are of the invention more thorough and complete in order to make, and can will fully convey the scope of the invention to ability
The technical staff in domain.
Fig. 1 shows the flow chart of invoice publicity method neural network based according to an exemplary embodiment of the present invention, packet
It includes:
Step 1: selection invoice sample is based on invoice sample training neural network model;
Step 2: the invoice type to publicity invoice is identified according to neural network model;And
Step 3: the sensitive information to publicity invoice being hidden according to invoice type, and publicity waits for the non-sensitive letter of publicity invoice
Breath.
The embodiment of the present invention is trained neural network model according to invoice sample big data, being optimized, be used for
Identify the neural network model of invoice type;Invoice type is identified using neural network model, effectively increases the accurate of identification
Property;It determines which information in invoice is sensitive information according to user identity, and place is hidden to the sensitive information in invoice
Reason both ensure that the safety of invoice publicity, and flexibly meet personalized, displaying application demand.
Below with reference to Fig. 1 detailed description of the present invention exemplary embodiment.As shown in Figure 1, exemplary reality according to the present invention
The invoice publicity method neural network based of example is applied, specifically includes the following steps:
Step 1: selection invoice sample is based on the invoice sample training neural network model.
Invoice sample includes invoice image and its corresponding invoice type.Wherein, invoice type may include: value-added tax hair
Ticket, motor vehicle invoice for sales, taxi invoice and various other bills, receipt etc..Table 1 shows the tool of invoice sample 1-4
Body information.
1 invoice sample instantiation of table
Invoice sample | Invoice image | Invoice type |
Sample 1 | original-05d0000046fd118d.jpg | VAT invoice |
Sample 2 | original-05d0000123fd166d.jpg | Motor vehicle invoice for sales |
Sample 3 | original-05d0000003fd123d.jpg | Motor vehicle invoice for sales |
Sample 4 | original-05d0000103fd137d.jpg | Taxi invoice |
…… | …… | …… |
In table 1, invoice image is image represented by picture name of the suffix for jpg, and invoice type can be by user
It is manually entered, it can also be when user uploads invoice picture by system automatically generated.It is sent out for example, user uploads in value-added tax system
Ticket image, then corresponding invoice type can be VAT invoice.Fig. 2 shows the example of an invoice image.
After getting enough invoice samples, these invoice sample training neural network models can be based on.Specifically
Ground, the corresponding pixel matrix of available invoice image, using pixel matrix as the input of neural network model, invoice class
Type is trained neural network model as label.
The pixel matrix of invoice image may include the corresponding pixel value of each pixel in invoice image, such as send out
Ticket image is the image of 1000*800, then the corresponding pixel matrix of invoice image can be the matrix of 1000*800.Pixel
Value can be there are many manifestation mode, such as can be gray value, or the GRB value of color space, YUV value etc..
Step 2: the invoice type to publicity invoice is identified according to neural network model.
Step 2 specifically includes the following steps:
Step 201: obtaining the invoice image to publicity invoice.
Invoice image to publicity invoice can be stored in server, in advance by uploading to publicity invoice owner or by opening
Ticket position uploads.Invoice image can be a frame or multiple image in the video of the photo of invoice, invoice etc..
In addition, some interference informations are usually present in the photo of invoice, video, for example, not only including in photo, video
The information of invoice also includes ambient condition information.Fig. 3 shows an invoice photo, in addition to comprising invoice in invoice photo, also
Parts of images comprising staff.Since external environment information may have an impact to sensitive information region is determined,
Therefore it needs first to extract invoice image from original invoice photo, video image frame.It in embodiments of the present invention, can be with
By the rectangular area in image-recognizing method identification invoice photo, video frame, the part in rectangular area is extracted as invoice
Image.
Step 202: identifying the text information in the invoice image, determined according to the text information described to publicity hair
The invoice type of ticket is as initial invoice type.
It can use OCR (Optical Character Recognition, optical character identification) technology identification invoice figure
Text information as in.Text conversion in paper document can be become black and white lattice using optical mode by OCR technique
Image file, and can by identification software by the text conversion in image at text formatting, further compiled for word processor
Collect processing.The text information in invoice image can be identified by OCR technique, and then true according to the text information in invoice image
Determine the type of invoice.
Optionally, the text information in invoice image can be compared with predetermined keyword, if the text is believed
Breath includes predetermined keyword, then confirms that the invoice type to publicity invoice is the corresponding invoice type of the predetermined keyword, and
As initial invoice type.For example, predetermined keyword can be " motor vehicle " for motor vehicle invoice, work as invoice
When text information in image includes " motor vehicle ", it is believed that the invoice type to publicity invoice is motor vehicle invoice.It can be with
The predetermined keyword corresponding to various invoice types is previously set, so that comparison uses.
Step 203: using the invoice image to publicity invoice as input, being identified according to neural network model to publicity invoice
Invoice type as correction invoice type.
Since OCR technique is there are certain misclassification rate and failure rate, identify that the text information in invoice image may
Failure.In addition, invoice often will appear it is fuzzy, be stained, situations such as text is blocked by seal, so not necessarily by text information
It can correctly determine invoice type.Therefore, can be after being identified by OCR technique to invoice type, then pass through nerve
Network model further identifies invoice type.Specifically, using the invoice image to publicity invoice as input, according to mind
It is the invoice type that can recognize to publicity invoice through network model, as correction invoice type.
Step 204: judge whether initial invoice type and correction invoice type are consistent, if unanimously, with initial invoice
Type or correction invoice type are as the invoice type to publicity invoice.
If initial invoice type is consistent with correction invoice type, illustrate that invoice type is exactly to determine using OCR technique
Invoice type or the invoice type determined using neural network model;If it is inconsistent, information warning can be exported, such as
" invoice type and, come in row verification inconsistent using the determining invoice type of neural network model that are determined using OCR technique ",
By staff confirm invoice type, or be subject to neural network model determine invoice type.
Step 3: the sensitive information to publicity invoice being hidden according to invoice type, and publicity waits for the non-sensitive letter of publicity invoice
Breath.
Wherein, according to invoice type hide to publicity invoice sensitive information specifically includes the following steps:
Step 301: obtaining the identity information of publicity object.
In a kind of specifically optional embodiment, it can be actively entered by publicity object (user) when inquiring invoice
Identity information, such as name, account, identification card number etc..
In another optional embodiment, the identity letter of publicity object can also be obtained by way of automatic identification
Breath.For example, camera can be set on the terminal device that publicity object checks invoice, the image taken according to camera
Information or video information carry out recognition of face, determine face characteristic information, and carry out with the face characteristic information in database
It compares, then obtains the identity information of formula object.Alternatively, iris recognition, retina identification, fingerprint recognition etc. can also be passed through
Mode determines the identity information of publicity object.
Step 302: publicity object being determined according to identity information and to the relationship between publicity invoice owner.
For example, if publicity object is the owner to publicity invoice or the administrator to company where publicity invoice
Member, then can normally show invoice, not be hidden to any information therein.
If publicity object is to the proprietary relatives of publicity invoice or friend, or to the general of company where publicity invoice
Logical employee, or be the administrative staff of brother company, then it can be carried out to the partial information of publicity invoice as sensitive information
It hides.Publicity object is allowed to check for example, can be used as non-sensitive information to the address in publicity invoice, phone, the letter such as price
Breath does not allow publicity object to check as sensitive information.
If publicity object is stranger, can will be carried out to the more information of publicity invoice as sensitive information hidden
Hiding.For example, sensitive information may include to address, phone and the price etc. in publicity invoice.
Wherein, publicity object and can be by being set in advance to publicity invoice owner to the relationship between publicity invoice owner
It sets and is stored;Alternatively, can be according to storage in the terminal device (such as mobile phone) when publicity invoice owner uploads invoice
Contact information determine publicity object and to the relationship between publicity invoice owner;Alternatively, can also be by being sent out to publicity
Ticket user relationship management system comes upload, modification, identification of implementation relation etc..
Step 303: opening relationships template, relationship templates include publicity object and to the relationship between publicity invoice owner
And its corresponding sensitive information leads to according to the publicity object determined in step 302 and to the relationship between publicity invoice owner
It crosses relationship templates and determines sensitive information.
Sensitive information is the information that should not be disclosed in invoice, such as name, address, product name, the amount of money of user etc.,
General information other than sensitive information date of such as making out an invoice can carry out publicity.
Optionally, can preset relationship templates by backstage manager, relationship templates include publicity object with to public affairs
Show the relationship between invoice owner and its corresponding sensitive information.For example, relationship templates include following information:
Relationship: administrative staff, sensitive information: ---;
Relationship: common employee, sensitive information: price;
Relationship: stranger, sensitive information: address, phone, price.
Alternatively it is also possible to by presetting relationship templates to publicity invoice uploader.
In this step, according to the publicity object determined in step 302 and to the relationship between publicity invoice owner, lead to
Cross relationship templates i.e. and can determine the sensitive information for specific publicity object.
Step 304: sensitive information region is determined according to the invoice type to formula invoice.
In embodiments of the present invention, invoice information template is pre-established, invoice information template includes invoice type, the invoice
The position of information corresponding to type and its region in invoice image.For example, information model 1 includes:
Invoice type: VAT invoice;
Purchaser's title: the 18%-20% on invoice picture altitude direction, the 10%- on invoice image length direction
40%;
Purchaser's Taxpayer Identification Number: the 20%-22% on invoice picture altitude direction, on invoice image length direction
10%-40%;
Purchaser address and phone: the 22%-24% on invoice picture altitude direction, on invoice image length direction
10%-40%;Etc..
Can corresponding invoice information template be established for all kinds of invoices in advance.
The identified invoice type to formula invoice in step 2, has determined that in step 303 to formula invoice
Sensitive information be therefore in this step the sensitive information location that can determine to publicity invoice by invoice information template
Domain.
Step 305: processing is hidden to sensitive information region.It can be (such as black by blurring or color
Color or white) it blocks processing is hidden to sensitive information region.
According to above-described embodiment, it can be determined according to the identity of publicity object and need to hide which information in invoice, energy
The user of different rights is enough facilitated to check different information, certified invoice publicity is normally carried out, and can be owned by invoice
The corresponding sensitive information of person's self-setting class of subscriber, meets the individual demand of uploader.
The embodiment of the present invention also provides a kind of invoice publicity system neural network based characterized by comprising
Memory is stored with computer executable instructions;
Processor, the processor run the computer executable instructions in the memory, execute following steps:
Step 1: selection invoice sample is based on the invoice sample training neural network model;
Step 2: the invoice type to publicity invoice is identified according to the neural network model;And
Step 3: the sensitive information to publicity invoice being hidden according to the invoice type, and is sent out described in publicity to publicity
The non-sensitive information of ticket;
The invoice sample includes invoice image and invoice type;
It is described to include: according to the hiding sensitive information to publicity invoice of the invoice type
Step 301: obtaining the identity information of publicity object;
Step 302: the publicity object being determined according to the identity information and described between publicity invoice owner
Relationship;
Step 303: opening relationships template, the relationship templates include publicity object and between publicity invoice owner
Relationship and its corresponding sensitive information, according to the publicity object and described to publicity invoice owner determined in step 302
Between relationship, determine the sensitive information by the relationship templates;
Step 304: sensitive information region is determined according to the invoice type to formula invoice;
Step 305: processing is hidden to the sensitive information region.
In one example, described to include: based on the invoice sample training neural network model
Using the invoice image as input, using the invoice type as label, the neural network model is carried out
Training.
In one example, the step 2 includes:
Step 201: obtaining the invoice image to publicity invoice;
Step 202: identifying the text information in the invoice image, determined according to the text information described to publicity hair
The invoice type of ticket is as initial invoice type;
Step 203: using the invoice image to publicity invoice as input, institute being identified according to the neural network model
The invoice type to publicity invoice is stated as correction invoice type;
Step 204: judge whether the initial invoice type and the correction invoice type are consistent, if unanimously, with
The initial invoice type or the correction invoice type are as the invoice type to publicity invoice;If it is inconsistent,
Export default prompt information.
In one example, the step 202 includes:
The text information in the invoice image is identified using optical character recognition method;
The text information is compared with predetermined keyword, if the text information includes predetermined keyword,
Using the corresponding invoice type of the predetermined keyword as the initial invoice type.
In one example, the step 304 includes:
Establish invoice information template, invoice information template include invoice type, information corresponding to the invoice type and its
Position of the region in invoice image;
According to the invoice type and sensitive information to publicity invoice, determined based on the invoice information template described quick
Feel information region.
The apparatus embodiments described above are merely exemplary, wherein described, unit can as illustrated by the separation member
It is physically separated with being or may not be, component shown as a unit may or may not be physics list
Member, it can it is in one place, or may be distributed over multiple network units.It can be selected according to the actual needs
In some or all of the modules achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creativeness
Labour in the case where, it can understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It is realized by the mode of required general hardware platform is added, naturally it is also possible to which reality is come in conjunction with by way of hardware and software
It is existing.Based on this understanding, substantially the part that contributes to existing technology can be to calculate in other words for above-mentioned technical proposal
The form of machine product embodies, and it wherein includes the meter of computer usable program code that the present invention, which can be used in one or more,
The computer journey implemented in calculation machine usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
The form of sequence product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable invoice publicity equipment to produce
A raw machine, so that being generated by the instruction that the processor of computer or other programmable invoice publicity equipment executes for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable invoice publicity equipment with spy
In the computer-readable memory for determining mode floor, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions can also be loaded into computer or other programmable invoice publicity equipment, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
In a typical configuration, calculating equipment includes one or more processors (CPU), input/output interface, net
Network interface and memory.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flashRAM).Memory is showing for computer-readable medium
Example.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),
Digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devices
Or any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, it calculates
Machine readable medium does not include temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.
Various embodiments of the present invention are described above, above description is exemplary, and non-exclusive, and
It is not limited to disclosed each embodiment.Without departing from the scope and spirit of illustrated each embodiment, for this skill
Many modifications and changes are obvious for the those of ordinary skill in art field.
Claims (10)
1. a kind of invoice publicity method neural network based characterized by comprising
Step 1: selection invoice sample is based on the invoice sample training neural network model;
Step 2: the invoice type to publicity invoice is identified according to the neural network model;And
Step 3: the sensitive information to publicity invoice being hidden according to the invoice type, and to publicity invoice described in publicity
Non-sensitive information;
The invoice sample includes invoice image and invoice type;
It is described to include: according to the hiding sensitive information to publicity invoice of the invoice type
Step 301: obtaining the identity information of publicity object;
Step 302: the publicity object and the relationship between publicity invoice owner are determined according to the identity information;
Step 303: opening relationships template, the relationship templates include publicity object and to the relationship between publicity invoice owner
And its corresponding sensitive information, according to the publicity object and described between publicity invoice owner determined in step 302
Relationship, determine the sensitive information by the relationship templates;
Step 304: sensitive information region is determined according to the invoice type to formula invoice;
Step 305: processing is hidden to the sensitive information region.
2. invoice publicity method neural network based according to claim 1, which is characterized in that described to be based on the hair
Ticket sample training neural network model includes:
Using the invoice image as input, using the invoice type as label, the neural network model is trained.
3. invoice publicity method neural network based according to claim 1, which is characterized in that the step 2 includes:
Step 201: obtaining the invoice image to publicity invoice;
Step 202: identifying the text information in the invoice image, determined according to the text information described to publicity invoice
Invoice type is as initial invoice type;
Step 203: using the invoice image to publicity invoice as input, according to neural network model identification it is described to
The invoice type of publicity invoice is as correction invoice type;
Step 204: judge whether the initial invoice type and the correction invoice type are consistent, if unanimously, with described
Initial invoice type or the correction invoice type are as the invoice type to publicity invoice;If it is inconsistent, output
Default prompt information.
4. invoice publicity method neural network based according to claim 3, which is characterized in that step 202 packet
It includes:
The text information in the invoice image is identified using optical character recognition method;
The text information is compared with predetermined keyword, if the text information includes predetermined keyword, with institute
The corresponding invoice type of predetermined keyword is stated as the initial invoice type.
5. invoice publicity method neural network based according to claim 1, which is characterized in that step 304 packet
It includes:
Establish invoice information template, invoice information template includes invoice type, information and its place corresponding to the invoice type
Position of the region in invoice image;
According to the invoice type and sensitive information to publicity invoice, the sensitive letter is determined based on the invoice information template
Cease region.
6. a kind of invoice publicity system neural network based characterized by comprising
Memory is stored with computer executable instructions;
Processor, the processor run the computer executable instructions in the memory, execute following steps:
Step 1: selection invoice sample is based on the invoice sample training neural network model;
Step 2: the invoice type to publicity invoice is identified according to the neural network model;And
Step 3: the sensitive information to publicity invoice being hidden according to the invoice type, and to publicity invoice described in publicity
Non-sensitive information;
The invoice sample includes invoice image and invoice type;
It is described to include: according to the hiding sensitive information to publicity invoice of the invoice type
Step 301: obtaining the identity information of publicity object;
Step 302: the publicity object and the relationship between publicity invoice owner are determined according to the identity information;
Step 303: opening relationships template, the relationship templates include publicity object and to the relationship between publicity invoice owner
And its corresponding sensitive information, according to the publicity object and described between publicity invoice owner determined in step 302
Relationship, determine the sensitive information by the relationship templates;
Step 304: sensitive information region is determined according to the invoice type to formula invoice;
Step 305: processing is hidden to the sensitive information region.
7. invoice publicity system neural network based according to claim 6, which is characterized in that described to be based on the hair
Ticket sample training neural network model includes:
Using the invoice image as input, using the invoice type as label, the neural network model is trained.
8. invoice publicity system neural network based according to claim 6, which is characterized in that the step 2 includes:
Step 201: obtaining the invoice image to publicity invoice;
Step 202: identifying the text information in the invoice image, determined according to the text information described to publicity invoice
Invoice type is as initial invoice type;
Step 203: using the invoice image to publicity invoice as input, according to neural network model identification it is described to
The invoice type of publicity invoice is as correction invoice type;
Step 204: judge whether the initial invoice type and the correction invoice type are consistent, if unanimously, with described
Initial invoice type or the correction invoice type are as the invoice type to publicity invoice;If it is inconsistent, output
Default prompt information.
9. invoice publicity method neural network based according to claim 8, which is characterized in that step 202 packet
It includes:
The text information in the invoice image is identified using optical character recognition method;
The text information is compared with predetermined keyword, if the text information includes predetermined keyword, with institute
The corresponding invoice type of predetermined keyword is stated as the initial invoice type.
10. invoice publicity system neural network based according to claim 6, which is characterized in that step 304 packet
It includes:
Establish invoice information template, invoice information template includes invoice type, information and its place corresponding to the invoice type
Position of the region in invoice image;
According to the invoice type and sensitive information to publicity invoice, the sensitive letter is determined based on the invoice information template
Cease region.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811564802.XA CN109815949A (en) | 2018-12-20 | 2018-12-20 | Invoice publicity method and system neural network based |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811564802.XA CN109815949A (en) | 2018-12-20 | 2018-12-20 | Invoice publicity method and system neural network based |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109815949A true CN109815949A (en) | 2019-05-28 |
Family
ID=66601709
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811564802.XA Pending CN109815949A (en) | 2018-12-20 | 2018-12-20 | Invoice publicity method and system neural network based |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109815949A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110334596A (en) * | 2019-05-30 | 2019-10-15 | 平安科技(深圳)有限公司 | Invoice picture method of summary, electronic device and readable storage medium storing program for executing |
CN110333813A (en) * | 2019-05-30 | 2019-10-15 | 平安科技(深圳)有限公司 | Method, electronic device and the computer readable storage medium of invoice picture presentation |
CN110348441A (en) * | 2019-07-10 | 2019-10-18 | 深圳市华云中盛科技有限公司 | VAT invoice recognition methods, device, computer equipment and storage medium |
CN110598826A (en) * | 2019-09-03 | 2019-12-20 | 数字广东网络建设有限公司 | Electronic certificate classified display method, device and system and computer equipment |
CN111814539A (en) * | 2020-05-28 | 2020-10-23 | 平安科技(深圳)有限公司 | Character recognition method and device based on infrared light and ultraviolet light and computer equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102567764A (en) * | 2012-01-13 | 2012-07-11 | 中国工商银行股份有限公司 | Bill certificate and system for improving electronic image recognition efficiency |
CN107704770A (en) * | 2017-09-28 | 2018-02-16 | 平安普惠企业管理有限公司 | Sensitive information desensitization method, system, equipment and readable storage medium storing program for executing |
CN107766809A (en) * | 2017-10-09 | 2018-03-06 | 平安科技(深圳)有限公司 | Electronic installation, billing information recognition methods and computer-readable recording medium |
CN107798299A (en) * | 2017-10-09 | 2018-03-13 | 平安科技(深圳)有限公司 | Billing information recognition methods, electronic installation and readable storage medium storing program for executing |
CN108399597A (en) * | 2017-02-07 | 2018-08-14 | 深圳前海明磊融创科技有限公司 | Key message treating method and apparatus |
-
2018
- 2018-12-20 CN CN201811564802.XA patent/CN109815949A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102567764A (en) * | 2012-01-13 | 2012-07-11 | 中国工商银行股份有限公司 | Bill certificate and system for improving electronic image recognition efficiency |
CN108399597A (en) * | 2017-02-07 | 2018-08-14 | 深圳前海明磊融创科技有限公司 | Key message treating method and apparatus |
CN107704770A (en) * | 2017-09-28 | 2018-02-16 | 平安普惠企业管理有限公司 | Sensitive information desensitization method, system, equipment and readable storage medium storing program for executing |
CN107766809A (en) * | 2017-10-09 | 2018-03-06 | 平安科技(深圳)有限公司 | Electronic installation, billing information recognition methods and computer-readable recording medium |
CN107798299A (en) * | 2017-10-09 | 2018-03-13 | 平安科技(深圳)有限公司 | Billing information recognition methods, electronic installation and readable storage medium storing program for executing |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110334596A (en) * | 2019-05-30 | 2019-10-15 | 平安科技(深圳)有限公司 | Invoice picture method of summary, electronic device and readable storage medium storing program for executing |
CN110333813A (en) * | 2019-05-30 | 2019-10-15 | 平安科技(深圳)有限公司 | Method, electronic device and the computer readable storage medium of invoice picture presentation |
CN110334596B (en) * | 2019-05-30 | 2024-02-02 | 平安科技(深圳)有限公司 | Invoice picture summarizing method, electronic device and readable storage medium |
CN110348441A (en) * | 2019-07-10 | 2019-10-18 | 深圳市华云中盛科技有限公司 | VAT invoice recognition methods, device, computer equipment and storage medium |
CN110348441B (en) * | 2019-07-10 | 2021-08-17 | 深圳市华云中盛科技股份有限公司 | Value-added tax invoice identification method and device, computer equipment and storage medium |
CN110598826A (en) * | 2019-09-03 | 2019-12-20 | 数字广东网络建设有限公司 | Electronic certificate classified display method, device and system and computer equipment |
CN111814539A (en) * | 2020-05-28 | 2020-10-23 | 平安科技(深圳)有限公司 | Character recognition method and device based on infrared light and ultraviolet light and computer equipment |
WO2021139341A1 (en) * | 2020-05-28 | 2021-07-15 | 平安科技(深圳)有限公司 | Character recognition method and apparatus based on infrared light and ultraviolet light, and computer device |
CN111814539B (en) * | 2020-05-28 | 2023-07-21 | 平安科技(深圳)有限公司 | Character recognition method and device based on infrared light and ultraviolet light and computer equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109815949A (en) | Invoice publicity method and system neural network based | |
US11636550B2 (en) | Using machine learning to classify insurance card information | |
US8818107B2 (en) | Identification generation and authentication process application | |
US20190279170A1 (en) | Dynamic resource management associated with payment instrument exceptions processing | |
CN112613501A (en) | Information auditing classification model construction method and information auditing method | |
CN108876133A (en) | Risk assessment processing method, device, server and medium based on business information | |
US20150286860A1 (en) | Method and Device for Generating Data from a Printed Document | |
CN106104572A (en) | For identifying the system and method that document process and Business Works are integrated | |
US9824288B1 (en) | Programmable overlay for negotiable instrument electronic image processing | |
US11966971B2 (en) | Computer-implemented mortgage processing system and method for facilitating a mortgage fulfillment process | |
US11488419B2 (en) | Identity and liveness verification | |
US11715285B2 (en) | Methods and systems for providing images for facilitating communication | |
KR20200008467A (en) | System and method for food expenses management using employee card | |
CN114186275A (en) | Privacy protection method and device, computer equipment and storage medium | |
US20190259026A1 (en) | Anonymous Event Processing Using Secure Digital Information Vault | |
CN111145031B (en) | Insurance business customization method, device and system | |
KR102222064B1 (en) | A management support system that converts product order details from different forms into a unified one | |
KR20190085870A (en) | System for managing expense and method using therewith | |
CN109858275A (en) | Invoice publicity method and system neural network based | |
US20200013097A1 (en) | Connectivity Hub with Data-Hiding Features | |
US20230316790A1 (en) | Authentication of Age, Gender, and Other Biometric Data from Live Images of Users | |
US20210248615A1 (en) | Method and system for digitally onboarding customers for providing one or more solutions in real-time | |
US20230125814A1 (en) | Credit score management apparatus, credit score management method, and computer readable recording medium | |
KR20200111641A (en) | Harmful customer management system and harmful customer management program | |
CN114553684B (en) | Method, device, computer equipment and storage medium for network point operation and maintenance |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190528 |
|
RJ01 | Rejection of invention patent application after publication |