CN109726783A - A kind of invoice acquisition management system and method based on OCR image recognition technology - Google Patents
A kind of invoice acquisition management system and method based on OCR image recognition technology Download PDFInfo
- Publication number
- CN109726783A CN109726783A CN201811620377.1A CN201811620377A CN109726783A CN 109726783 A CN109726783 A CN 109726783A CN 201811620377 A CN201811620377 A CN 201811620377A CN 109726783 A CN109726783 A CN 109726783A
- Authority
- CN
- China
- Prior art keywords
- invoice
- image recognition
- ocr image
- information
- identification
- 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
Landscapes
- Character Discrimination (AREA)
- Character Input (AREA)
Abstract
The present invention relates to a kind of invoice acquisition management systems and method based on OCR image recognition technology, the invoice information acquisition unit of the system acquires invoice pictorial information data, OCR image identification unit is based on OCR image recognition technology and artificial intelligence deep learning algorithm is combined to carry out identification reading to invoice pictorial information data to obtain OCR image recognition result, invoice information verification processing unit identify and tax bureau's VAT invoice true and false is called to check platform automatically when identification result is VAT invoice to the invoice type of OCR image recognition result carrying out true and false examination, invoice information MMU memory management unit stores the OCR image recognition result that identification result is non-VAT invoice while will check the OCR image recognition result that result is genuine VAT invoice and repeat carrying out respective stored after collecting verification , invoice information classify display unit to invoice information data carry out classification displaying processing, synthetically realize invoice information data acquisition, identification, examination, processing and storage overall process.
Description
Technical field
The present invention relates to invoice information acquisition management technical fields, and in particular to a kind of based on OCR image recognition technology
Invoice acquisition management system system and method.
Background technique
With the continuous development of science and technology, automatic intelligent technology, artificial intelligence technology are brought much to people's lives
Convenience.And in information access process, language and text are even more that we obtain that information is most basic, most important approach.Scheming
As identification technology field, there are one very important subdivision field-OCR (Optical Character Recognition,
Optical character identification), refer to and check the character printed on paper by optical device, its shape is determined by the mode for detecting dark, bright
Then shape is translated into the process of computword by shape with character identifying method, be exactly reading of the computer to text.But one
Denier text is showed in the form of picture, just obtains to us and processing text has added many troubles.This aspect performance
For the text for being claimed picture format in digital world by storage due to specific reasons；It on the other hand is that we see in real life
All physical aspects text.So we need to extract these texts and information by OCR technique.So OCR
Image recognition technology starts to be paid attention to and develop.
However in invoice information data capture management technical field, traditional working forms, such as people are still sticked to
Typing invoice information data are for submitting an expense account or authenticate in systems, are by manually according to invoice content by invoice information number
According to being entered into system, a large amount of human and material resources and time are thus spent, and there are also there is the wind for input error occur
Danger, error rate is high, and this method is extremely inconvenient.The instrument for having gradually appeared some scanning recognitions later, can help slightly to mitigate
Manually typing invoice information data bring cost payout, but purchase machine it is mostly more expensive, still spend compared with
Greatly, and its problems such as that generally there is also scanning recognition speed is slow, using area is limited, so this method is unable to get effectively
Large-scale promotion.
Summary of the invention
The present invention in the prior art using traditional approach carry out invoice information data acquire human and material resources, financial resources and
Time consumption is big and error rate is high, carries out invoice information data acquisition instrument valuableness and identification speed using traditional scanning recognition equipment
The problems such as degree is slowly, using area is limited provides a kind of invoice acquisition management system based on OCR image recognition technology, this is
System designs simple, structure optimization, and dexterously introduces OCR image recognition technology and identify to invoice information data, identifies
Speed is fast and accuracy is high, which also integrates and carry out corresponding authenticity verification, storage and management and hair to invoice information data
The classification of ticket information is shown, efficiently easily manages invoice data, practicability is extremely strong.The present invention also provides one kind to be schemed based on OCR
As the invoice acquisition management method of identification technology.
Technical scheme is as follows:
A kind of invoice acquisition management system based on OCR image recognition technology, including the acquisition of sequentially connected invoice information
Unit, OCR image identification unit, invoice information verification processing unit, invoice information MMU memory management unit and invoice information classification
Display unit, the invoice information acquisition unit acquire invoice pictorial information data, and the OCR image identification unit is based on OCR
Image recognition technology simultaneously carries out identification reading to the invoice pictorial information data in conjunction with artificial intelligence deep learning algorithm to obtain
OCR image recognition result is obtained, the invoice information verification processing unit carries out the invoice type of the OCR image recognition result
Identify and tax bureau's VAT invoice true and false is called to check the platform progress true and false automatically when identification result is VAT invoice and looks into
It tests, the invoice information MMU memory management unit storage identification result is that the OCR image recognition result of non-VAT invoice simultaneously will
Examination result is that the OCR image recognition result of genuine VAT invoice repeat carrying out respective stored after collecting verification, described
The invoice information data that invoice information classification display unit stores the invoice information MMU memory management unit carry out classification displaying
Processing.
Preferably, the invoice information verification processing unit includes the invoice type identification module interconnected and invoice letter
Breath examination module, the invoice type identification module are connected to the OCR image identification unit and the invoice information storage tube
It manages between unit, invoice information examination module one end is remotely connected to tax bureau's VAT invoice true and false examination platform simultaneously
The other end is connected with the invoice information MMU memory management unit, and the invoice type identification module is to the OCR image recognition knot
The invoice type of fruit identify and be sent to invoice information MMU memory management unit when identification result is non-VAT invoice depositing
Storage is sent to invoice information when identifying is VAT invoice and checks module, and the invoice information examination module is in the invoice
The identification result of type identification module calls tax bureau's VAT invoice true and false to check platform automatically and carries out when being VAT invoice
The true and false checks and is sent to invoice information MMU memory management unit when examination is true and carries out repeating to collect verification and respective stored.
Preferably, the invoice information MMU memory management unit includes that invoice information interconnected repeats to collect verification module
With invoice information database, the invoice information repetition collects verification module and is connected with invoice information examination module, the invoice
Information database is connected between the invoice type identification module and invoice information classification display unit, invoice information weight
Involution twin check module receives the OCR image recognition result that examination result is genuine VAT invoice and carries out repeating to collect verification,
The invoice information database receives the OCR image recognition result that identification result is non-VAT invoice and is directly stored simultaneously
Reception repetition collects the OCR image recognition result that verification result is not duplicate VAT invoice and is stored.
Preferably, the OCR image identification unit combination convolutional neural networks deep learning algorithm, Recognition with Recurrent Neural Network are deep
Degree learning algorithm and timing sorting algorithm carry out identification to invoice pictorial information data and read to obtain OCR image recognition knot
Fruit.
Preferably, the OCR image recognition result is the structured message data of json format.
Preferably, the invoice type identification module is based on invoice codes and/or invoice number and combines official, the tax bureau
Invoice type judgment rule the invoice type of the OCR image recognition result is identified.
Preferably, the invoice information acquisition unit include but is not limited to be based on mobile terminal and/or the end PC by taking pictures on
It passes, the mode that photograph album uploads and picture library uploads carries out invoice information acquisition.
A kind of invoice acquisition management method based on OCR image recognition technology, it is sharp after acquiring invoice pictorial information data
With OCR image recognition technology combination artificial intelligence deep learning algorithm to the invoice pictorial information data carry out identification read with
OCR image recognition result is obtained, then the invoice type of the OCR image recognition result is identified, is to increase in identification result
Call tax bureau's VAT invoice true and false examination platform to carry out the true and false examination when value tax invoice, again when checking result and being true automatically
The OCR image recognition result of corresponding VAT invoice is carried out to repeat to collect verification, and will when verifying result and being not repeat
OCR image recognition result carries out respective stored；Or, when identification result is non-VAT invoice, to the OCR of non-VAT invoice
Image recognition result is directly stored；Then classification displaying processing is carried out to invoice information data.
Preferably, the method utilizes OCR image recognition technology combination convolution mind after acquiring invoice pictorial information data
Through network depth learning algorithm, Recognition with Recurrent Neural Network deep learning algorithm and timing sorting algorithm to invoice pictorial information data
Identification is carried out to read to obtain OCR image recognition result.
Preferably, the invoice type of the OCR image recognition result is identified specifically: based on invoice codes and/
Or invoice number and in conjunction with the invoice type judgment rule of official, the tax bureau to the invoice type of the OCR image recognition result into
Row identifies；
And/or the acquisition invoice pictorial information data include but is not limited to be based on mobile terminal and/or the end PC by taking pictures
It uploads, the mode that photograph album uploads and picture library uploads.
Technical effect of the invention is as follows:
The present invention relates to a kind of invoice acquisition management systems based on OCR image recognition technology, including invoice information to adopt
Collect unit, OCR image identification unit, invoice information verification processing unit, invoice information MMU memory management unit and invoice information point
Class display unit is based on after invoice information acquisition unit has acquired invoice pictorial information data by OCR image identification unit
OCR image recognition technology simultaneously combines artificial intelligence deep learning algorithm to carry out identification reading to invoice pictorial information data to obtain
OCR image recognition result, identification process more fast accurate, invoice information verification processing unit is to OCR image recognition result
Invoice type identify and calls the examination of tax bureau's VAT invoice true and false flat automatically when identification result is VAT invoice
Platform carries out true and false examination, and by true and false examination result feedback into system, invoice information MMU memory management unit receives identification result
It directly carries out storing while receiving examination result being genuine VAT invoice for the OCR image recognition result of non-VAT invoice
OCR image recognition result repeat carrying out respective stored after collecting verification, directly stores to non-VAT invoice, to value-added tax
Invoice finally obtains the invoice information data of the authenticated true and false then again based on whether repeating to collect and right after carrying out true and false examination
The invoice information data of the VAT invoice are stored, as a result, VAT invoice information and the success of non-VAT invoice information
The classification storage and authenticity of VAT invoice is verified, last invoice information classification display unit is to invoice information storage tube
The invoice information data of reason unit storage carry out classification displaying processing, further show invoice type, true from false of bills information, hair
Ticket managing detailed catalogue etc., the system synthetically realize the overall process of acquisition, identification, examination, processing and the storage of invoice information data,
And the OCR image recognition technology based on intelligent algorithm is innovatively introduced, and combines artificial intelligence deep learning
Algorithm effectively increases the speed that invoice pictorial information acquires ten times, more greatly reduces the manual work amount of staff, quasi-
It is really efficient, convenient and efficient, then pass through remotely connection tax bureau's VAT invoice true and false examination platform and carried out very based on key element
Puppet examination, fully ensures that the authenticity of VAT invoice, also carries out repeating to collect verification to VAT invoice before being stored,
It fully ensures that the uniqueness of VAT invoice, may finally quickly help user to collect the invoice in hand, can be used for invoice
Reimbursement can be used for the scenes such as the invoice acquisition certification of finance, can greatly reduce the time manually collected in this way, effectively subtract
The time of light Enterprises ' Financial Workers examination true from false of bills, the effective and reasonable a variety of financial risks evaded about invoice business are saved
Operation cost of enterprises.
The invoice acquisition management method based on OCR image recognition technology that the invention further relates to a kind of, can be adopted by several
Mode set utilizes OCR image recognition technology combination artificial intelligence deep learning algorithm to described after acquiring invoice pictorial information data
Invoice pictorial information data carry out identification and read to obtain OCR image recognition result, to the invoice of the OCR image recognition result
Type is identified, and is called tax bureau's VAT invoice true and false to check platform automatically when identification result is VAT invoice and is carried out
True and false examination carries out repeating to collect core to the OCR image recognition result of corresponding VAT invoice again when checking result and being true
It looks into, and OCR image recognition result is subjected to respective stored when verifying result and being not repeat；Or, being non-increment in identification result
When tax invoice, the OCR image recognition result of non-VAT invoice is directly stored；Finally function is shown using invoice information
Classification displaying is carried out to invoice information data, this method innovatively introduces the OCR image recognition based on intelligent algorithm
Technology, and artificial intelligence deep learning algorithm is combined, the speed that invoice pictorial information acquires ten times is effectively increased, and comprehensive
The overall process of acquisition, identification, examination, processing and the storage of invoice information data is realized on ground, quickly user may finally be helped to return
Collect the invoice in hand, can be used for the reimbursement of invoice, can be used for the scenes such as the invoice acquisition certification of finance, it in this way can be big
Reduce the manual work amount of staff, the financial resources that use manpower and material resources sparingly and time greatly and then save operation cost of enterprises, moreover it is possible to
Effectively mitigate the time of Enterprises ' Financial Workers examination true from false of bills and efficiently easily manages and store invoice information data.
Detailed description of the invention
Fig. 1: for the present invention is based on the structural block diagrams of the invoice acquisition management system of OCR image recognition technology.
Fig. 2: for the present invention is based on the preferred structure block diagrams of the invoice acquisition management system of OCR image recognition technology.
Fig. 3: for the present invention is based on the preferred flow charts of the invoice acquisition management method of OCR image recognition technology.
Specific embodiment
Further the present invention is described in detail with reference to the accompanying drawing.
The present invention relates to a kind of invoice acquisition management systems based on OCR image recognition technology, as shown in Figure 1, including
Sequentially connected invoice information acquisition unit, OCR image identification unit, invoice information verification processing unit, invoice information storage
Administrative unit and invoice information classification display unit, invoice information acquisition unit acquires invoice pictorial information data, specific preferred
Ground, the invoice information acquisition unit include but is not limited to be based on mobile terminal and/or the end PC by upload of taking pictures, photograph album upload with
And the various ways such as picture library upload carry out invoice information acquisition and obtain invoice pictorial information, system is receiving invoice pictorial information
It calls OCR image identification unit to be based on OCR image recognition technology afterwards and combines artificial intelligence deep learning algorithm to invoice picture
Information data carries out identification and reads to obtain OCR image recognition result, identification process more fast accurate, at invoice information verifying
Reason unit identify to the invoice type of OCR image recognition result and calls tax automatically when identification result is VAT invoice
The business office VAT invoice true and false checks platform and carries out true and false examination, and by true and false examination result feedback into system, invoice information
MMU memory management unit storage identification result is that examination result is simultaneously really to increase by the OCR image recognition result of non-VAT invoice
The OCR image recognition result of value tax invoice repeat carrying out respective stored after collecting verification, directly deposits to non-VAT invoice
Storage finally obtains the invoice information data of the authenticated true and false then again based on whether weight after carrying out true and false examination to VAT invoice
Involution collection and the invoice information data of the VAT invoice are stored, as a result, VAT invoice information and non-value-added tax hair
The authenticity of the storage of ticket information successful classification and VAT invoice is verified, and invoice information classifies display unit to invoice information
The invoice information data of MMU memory management unit storage carry out classification displaying processing, may further show invoice type, invoice
True and false information, invoice managing detailed catalogue etc., the system synthetically realize the acquisition of invoice information data, identification, examination, handle and deposit
The overall process of storage, and the OCR image recognition technology based on intelligent algorithm is innovatively introduced, and combine artificial intelligence
Energy deep learning algorithm effectively increases the speed that invoice pictorial information acquires ten times, more greatly reduces the hand of staff
Dynamic workload, it is precise and high efficiency, convenient and efficient, then platform is checked by remotely connection tax bureau's VAT invoice true and false and is based on key
Element carries out true and false examination, fully ensures that the authenticity of VAT invoice, also carries out weight to VAT invoice before being stored
Involution twin check fully ensures that the uniqueness of VAT invoice, may finally quickly help user to collect the invoice in hand, can be with
It for the reimbursement of invoice, can be used for the scenes such as the invoice acquisition certification of finance, can greatly reduce manually collect in this way
Time effectively mitigates the time of Enterprises ' Financial Workers examination true from false of bills, the effective and reasonable a variety of wealth evaded about invoice business
Business risk, saves operation cost of enterprises.
Fig. 2 is that the present invention is based on the preferred structure block diagrams of the invoice acquisition management system of OCR image recognition technology, preferably
Ground, invoice information verification processing unit further comprise the invoice type identification module interconnected and invoice information examination mould
Block, invoice type identification module are connected between OCR image identification unit and invoice information MMU memory management unit, and invoice information is looked into
It tests module one end and is remotely connected to the tax bureau's VAT invoice true and false examination platform other end and invoice information storage management simultaneously
Unit is connected, and invoice type identifies that module carries out identification to the invoice type of OCR image recognition result and is non-in identification result
It is sent to the storage of invoice information MMU memory management unit when VAT invoice, is sent to invoice information when identifying is VAT invoice
Module is checked, invoice information checks module and calls tax automatically when invoice type identifies that the identification result of module is VAT invoice
Business office VAT invoice true and false examination platform carries out true and false examination and is sent to invoice information storage management list when examination is true
Member carries out repeating to collect verification and respective stored, further, as shown in Fig. 2, in invoice information verification processing unit into one
Step includes on the basis of the invoice type interconnected identifies module and invoice information examination module, and invoice information is deposited in the system
Storage administrative unit preferably includes invoice information interconnected and repeats to collect verification module and invoice information database, invoice information
Repetition, which collects, verifies module and invoice information examination module and is connected, invoice information database be connected to invoice type identify module with
Invoice information is classified between display unit, and it is genuine VAT invoice that invoice information, which repeats to collect and verifies module to receive examination result,
OCR image recognition result carry out repeating to collect verifications, it is non-VAT invoice that invoice information database, which receives identification result,
OCR image recognition result directly carries out storing while receiving repetition and collects the OCR for verifying that result is not duplicate VAT invoice
Image recognition result is stored, and the system structure advanced optimized is more perfect, and each component finely divides the work, cooperation, is had
Effect improves the speed that invoice pictorial information acquires ten times, more greatly reduces the manual work amount of staff, synthetically real
Show the overall process of acquisition, identification, examination, processing and the storage of invoice information data, it is precise and high efficiency, convenient and efficient.
It further understands, invoice information MMU memory management unit is mainly the invoice data for returning to OCR image recognition result
Examination result is that genuine, feedback (is sent to invoice information when identifying is VAT invoice to look into system after examination
Test module) invoice information data result carries out preservation processing.Invoice type on the market can substantially be divided into value-added tax at present
Invoice and non-VAT invoice two major classes identify two processes to hair by the OCR image recognition processes and invoice type of front
After ticket is classified, invoice information MMU memory management unit especially invoice information database can be direct for non-VAT invoice
It saves, since VAT invoice can judge the uniqueness of invoice according to invoice codes and invoice number, so for value-added tax
Invoice can inquire whether have existed this invoice, i.e. invoice information storage management list in library according to the code number of invoice
Member collects verification module using invoice information repetition and is made whether the verification for repeating to collect to the VAT invoice after the examination true and false,
Then according to query result, the judgement for repeating to collect can be carried out to invoice, and then allow invoice information database to query result
Invoice structured message not repeat to collect is saved.
And invoice information classification display unit mainly (stores the invoice information after collecting in invoice information database
Invoice information data) displaying of classification isloation state is carried out for checking that displaying uses, for example show invoice type, true from false of bills letter
Breath, invoice managing detailed catalogue etc..System had finally not only obtained invoice picture, but also the invoice structured message data after being parsed,
Invoice information display unit of classifying can show invoice content according to these, and user can will be in invoice picture when checking
Appearance is compared with the final parsing result of this system, it is ensured that accuracy, the integrality of information meet the demand of user.
Preferably, the OCR image identification unit combination convolutional neural networks deep learning algorithm (calculate by CNN identification service
Method), (CTC identification service is calculated for Recognition with Recurrent Neural Network deep learning algorithm (RNN identify service algorithm) and timing sorting algorithm
Method) identification reading is carried out to obtain OCR image recognition result to invoice pictorial information data, OCR image identification unit is based on people
The OCR image recognition technology of work intelligent algorithm, and combine tri- kinds of artificial intelligence deep learning algorithms of CNN, RNN and CTC and read
Text information on invoice picture substantially increases speed and efficiency that the identification of invoice pictorial information is read, and preferably, described
OCR image recognition result is the structuring invoice information data of json format.
Preferably, hair of the invoice type identification module based on invoice codes and/or invoice number and combination official, the tax bureau
Fare ticket type type judgment rule identifies the invoice type of OCR image recognition result, because OCR identification service is only to picture
Text information be read out and can not distinguish the true and false of invoice, so invoice type is reflected after obtaining OCR image recognition result
Cover half block obtains invoice codes and/or invoice number first, then judges invoice according to the judgment method that the tax authority announces
Type, if it is judged that then passing through invoice key element information (the OCR image that OCR image recognition obtains for VAT invoice
Recognition result), then call tax authority's value-added tax true and false examination interface (tax bureau's VAT invoice true and false checks platform) right
Invoice key element information carries out verification, finally obtains the invoice structured message data for testing the true and false.
The invention further relates to a kind of the invoice acquisition management method based on OCR image recognition technology, process as shown in Figure 3
Figure, this method utilize OCR image recognition technology combination artificial intelligence deep learning algorithm after acquiring invoice pictorial information data
Identification is carried out to invoice pictorial information data to read to obtain OCR image recognition result, then to the OCR image recognition result
Invoice type is identified, and tax bureau's VAT invoice true and false is called to check platform automatically when identification result is VAT invoice
True and false examination is carried out, when examination result is that fictitious time prompts VAT invoice examination failure, when checking result and being true again to corresponding
The OCR image recognition result of VAT invoice carry out repeating to collect verifications, and prompt to rise in value when verifying result and being repeated
Tax invoice has repeated to collect, and OCR image recognition result is carried out respective stored when verifying result and being not repeat；Or, identifying
When being as a result non-VAT invoice, the OCR image recognition result of non-VAT invoice is directly stored；Then invoice is believed
Breath data carry out classification displaying processing, and this method innovatively introduces the OCR image recognition technology based on intelligent algorithm,
And artificial intelligence deep learning algorithm is combined, the speed that invoice pictorial information acquires ten times is effectively increased, and synthetically real
The overall process of acquisition, identification, examination, processing and the storage of existing invoice information data, quickly may finally help user to collect hand
In invoice, can be used for the reimbursement of invoice, can be used for finance invoice acquisition certification etc. scenes, can subtract significantly in this way
The manual work amount of staff, the financial resources that use manpower and material resources sparingly and time are lacked and then have saved operation cost of enterprises, moreover it is possible to effectively
Mitigate the time of Enterprises ' Financial Workers examination true from false of bills and efficiently easily manages and store invoice information data, applied field
Scape diversification and unrestricted, being capable of effectively large-scale promotion.
Preferably, this method utilizes OCR image recognition technology combination convolutional Neural after acquiring invoice pictorial information data
Network depth learning algorithm (CNN identify service algorithm), Recognition with Recurrent Neural Network deep learning algorithm (RNN identifies service algorithm) with
And timing sorting algorithm (CTC identifies service algorithm) carries out identification to invoice pictorial information data and reads to obtain the knowledge of OCR image
Not as a result, substantially increasing speed and efficiency that the identification of invoice pictorial information is read, and preferably, the OCR image recognition knot
Fruit is the structuring invoice information data of json format.
Preferably, the invoice type of the OCR image recognition result is identified specifically: based on invoice codes and/
Or invoice number and in conjunction with the invoice type judgment rule of official, the tax bureau to the invoice type of the OCR image recognition result into
Row identifies, because OCR identification service is only the text information of picture to be read out and can not be distinguished the true and false of invoice,
Invoice type identification module obtains invoice codes and/or invoice number first after obtaining OCR image recognition result, then basis
The judgment method that the tax authority announces judges the type of invoice；And/or acquisition invoice pictorial information data include but is not limited to base
Uploaded by upload of taking pictures, photograph album and the modes such as picture library uploads in mobile terminal and/or the end PC, certified invoice image data it is more
Source property.
This method can help user/user easily and quickly to invoice information number by the way that above-mentioned several functions are arranged
According to being collected, and it is easy to use, and user can be collected using mobile phone photograph outdoors, can be used in the place of office
Computer uploading pictures, as long as entire invoice information collection process 3 to the 5 seconds time, so that it may which completion collects an invoice, can
With the reimbursement work for invoice, can be used for the scenes such as the invoice acquisition certification of finance, be significantly reduced be manually entered, people
Work collects the time of invoice information, moreover it is possible to divide invoice type by special algorithm formula area, and effectively save Enterprises ' Financial Workers
The time of true from false of bills is examined, the comprehensive efficiency for promoting invoice and collecting reduces financial risk, saves operation cost of enterprises.
It should be pointed out that specific embodiment described above can make those skilled in the art that the present invention be more fully understood
It creates, but do not limit the invention in any way is created.Therefore, although this specification creates the present invention referring to drawings and examples
It makes and has been carried out detailed description, it will be understood by those skilled in the art, however, that still can modify to the invention
Or equivalent replacement, in short, the technical solution and its improvement of all spirit and scope for not departing from the invention, should all contain
It covers in the protection scope of the invention patent.
Claims (10)
1. a kind of invoice acquisition management system based on OCR image recognition technology, which is characterized in that including sequentially connected invoice
Information acquisition unit, OCR image identification unit, invoice information verification processing unit, invoice information MMU memory management unit and invoice
Information classification display unit, the invoice information acquisition unit acquire invoice pictorial information data, the OCR image identification unit
Identification reading is carried out to the invoice pictorial information data based on OCR image recognition technology and in conjunction with artificial intelligence deep learning algorithm
It takes to obtain OCR image recognition result, invoice class of the invoice information verification processing unit to the OCR image recognition result
Type identify and tax bureau's VAT invoice true and false is called to check platform automatically when identification result is VAT invoice carrying out
True and false examination, the invoice information MMU memory management unit storage identification result are the OCR image recognition result of non-VAT invoice
The OCR image recognition result that result is genuine VAT invoice will be checked simultaneously repeat accordingly being deposited after collecting verification
Storage, the invoice information data that the invoice information classification display unit stores the invoice information MMU memory management unit are divided
Class displaying processing.
2. the invoice acquisition management system according to claim 1 based on OCR image recognition technology, which is characterized in that institute
Stating invoice information verification processing unit includes the invoice type identification module interconnected and invoice information examination module, the hair
Ticket type identification module is connected between the OCR image identification unit and the invoice information MMU memory management unit, the hair
Ticket information examination module one end is remotely connected to the tax bureau's VAT invoice true and false examination platform other end and the invoice simultaneously
Information storage tube manages unit and is connected, and the invoice type identification module carries out the invoice type of the OCR image recognition result
Identify and be sent to when identification result is non-VAT invoice the storage of invoice information MMU memory management unit, is value-added tax identifying
Invoice information examination module, identification of the invoice information examination module in invoice type identification module are sent to when invoice
Automatically it calls tax bureau's VAT invoice true and false examination platform to carry out the true and false examination when being as a result VAT invoice and is in examination
Invoice information MMU memory management unit is sent to when true to carry out repeating to collect verification and respective stored.
3. the invoice acquisition management system according to claim 2 based on OCR image recognition technology, which is characterized in that institute
Stating invoice information MMU memory management unit includes that invoice information interconnected repeats to collect verification module and invoice information database,
The invoice information repetition collects verification module and is connected with invoice information examination module, and the invoice information database is connected to institute
It states between invoice type identification module and invoice information classification display unit, invoice information, which repeats to collect, verifies module reception
Examination result is that the OCR image recognition result of genuine VAT invoice carries out repeating to collect verification, the invoice information database
It receives the OCR image recognition result that identification result is non-VAT invoice and directly carries out storing while receiving repetition and collect verification and tie
Fruit is that the OCR image recognition result of not duplicate VAT invoice is stored.
4. according to claim 1 based on the invoice acquisition management system of OCR image recognition technology described in one of -3, feature exists
In, the OCR image identification unit combination convolutional neural networks deep learning algorithm, Recognition with Recurrent Neural Network deep learning algorithm with
And timing sorting algorithm carries out identification to invoice pictorial information data and reads to obtain OCR image recognition result.
5. the invoice acquisition management system according to claim 4 based on OCR image recognition technology, which is characterized in that institute
State the structured message data that OCR image recognition result is json format.
6. the invoice acquisition management system according to claim 5 based on OCR image recognition technology, which is characterized in that institute
Invoice type identification module is stated based on invoice codes and/or invoice number and combines the invoice type judgment rule of official, the tax bureau
The invoice type of the OCR image recognition result is identified.
7. the invoice acquisition management system according to claim 6 based on OCR image recognition technology, which is characterized in that institute
Stating invoice information acquisition unit includes but is not limited to be based on mobile terminal and/or the end PC to upload and scheme by upload of taking pictures, photograph album
The mode that library uploads carries out invoice information acquisition.
8. a kind of invoice acquisition management method based on OCR image recognition technology, which is characterized in that the method is in acquisition invoice
Using OCR image recognition technology combination artificial intelligence deep learning algorithm to the invoice pictorial information number after pictorial information data
It reads according to identification is carried out to obtain OCR image recognition result, then reflects to the invoice type of the OCR image recognition result
Not, it calls tax bureau's VAT invoice true and false to check platform automatically when identification result is VAT invoice and carries out true and false examination,
Again the OCR image recognition result of corresponding VAT invoice is carried out repeating to collect verification when checking result and being true, and in core
The fruit that comes to an end is that OCR image recognition result is carried out respective stored when not repeating；Or, when identification result is non-VAT invoice,
The OCR image recognition result of non-VAT invoice is directly stored；Then invoice information data are carried out at classification displaying
Reason.
9. the invoice acquisition management method according to claim 8 based on OCR image recognition technology, which is characterized in that institute
Method is stated to calculate after acquiring invoice pictorial information data using OCR image recognition technology combination convolutional neural networks deep learning
Method, Recognition with Recurrent Neural Network deep learning algorithm and timing sorting algorithm carry out identification to invoice pictorial information data and read to obtain
Obtain OCR image recognition result.
10. the invoice acquisition management method based on OCR image recognition technology according to claim 8 or claim 9, feature exist
In identifying to the invoice type of the OCR image recognition result specifically: simultaneously based on invoice codes and/or invoice number
The invoice type of the OCR image recognition result is identified in conjunction with the invoice type judgment rule of official, the tax bureau；
And/or the acquisition invoice pictorial information data include but is not limited to be based on mobile terminal and/or the end PC by taking pictures on
It passes, the mode that photograph album uploads and picture library uploads.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811620377.1A CN109726783A (en) | 2018-12-28 | 2018-12-28 | A kind of invoice acquisition management system and method based on OCR image recognition technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811620377.1A CN109726783A (en) | 2018-12-28 | 2018-12-28 | A kind of invoice acquisition management system and method based on OCR image recognition technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109726783A true CN109726783A (en) | 2019-05-07 |
Family
ID=66297438
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811620377.1A Pending CN109726783A (en) | 2018-12-28 | 2018-12-28 | A kind of invoice acquisition management system and method based on OCR image recognition technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109726783A (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110188743A (en) * | 2019-05-13 | 2019-08-30 | 武汉大学 | A kind of taxi invoice identifying system and method |
| CN110288755A (en) * | 2019-05-21 | 2019-09-27 | 平安银行股份有限公司 | The invoice method of inspection, server and storage medium based on text identification |
| CN110363545A (en) * | 2019-07-23 | 2019-10-22 | 金在（北京）金融信息服务有限公司 | Management method, system and the readable storage medium storing program for executing of invoice information |
| CN110647824A (en) * | 2019-09-03 | 2020-01-03 | 四川大学 | Value-added tax invoice layout extraction method based on computer vision technology |
| CN110781141A (en) * | 2019-09-07 | 2020-02-11 | 北京中科云链信息技术有限公司 | Method for acquiring head-up of electronic ticket |
| CN111192019A (en) * | 2019-12-30 | 2020-05-22 | 武汉佰钧成技术有限责任公司 | Reimbursement processing method of target bill and related equipment |
| CN111199222A (en) * | 2019-12-30 | 2020-05-26 | 航天信息软件技术有限公司 | Bill management method and electronic equipment |
| CN111223230A (en) * | 2020-01-19 | 2020-06-02 | 河南电力物资有限公司 | Invoice file authenticity identification method based on CRNN algorithm |
| CN111401199A (en) * | 2020-03-10 | 2020-07-10 | 深圳航天信息有限公司 | Invoice identification method and system |
| CN111429645A (en) * | 2020-03-31 | 2020-07-17 | 重庆远见金税通信息系统技术有限公司 | True checking and weight checking system for bills |
| CN111461097A (en) * | 2020-03-18 | 2020-07-28 | 北京大米未来科技有限公司 | Method, apparatus, electronic device and medium for recognizing image information |
| CN111967458A (en) * | 2020-08-14 | 2020-11-20 | 中国工商银行股份有限公司 | Bill exchange method, related equipment and system |
| CN112053343A (en) * | 2020-09-02 | 2020-12-08 | 平安科技（深圳）有限公司 | User picture data processing method and device, computer equipment and storage medium |
| WO2020253113A1 (en) * | 2019-06-19 | 2020-12-24 | 深圳壹账通智能科技有限公司 | Invoice recording method, device, apparatus, and computer storage medium |
| CN112735055A (en) * | 2020-12-17 | 2021-04-30 | 航天信息股份有限公司 | Invoice prereviewing and collecting system |
| CN113011959A (en) * | 2021-05-24 | 2021-06-22 | 国能大渡河大数据服务有限公司 | Seven-expense intelligent auditing system and use method thereof |
| CN113065940A (en) * | 2021-04-27 | 2021-07-02 | 平安普惠企业管理有限公司 | Invoice reimbursement method, device, equipment and storage medium based on artificial intelligence |
| CN113114868A (en) * | 2021-04-16 | 2021-07-13 | 合肥新青罗数字技术有限公司 | OCR recognition device and system for intangible asset management |
| CN113240503A (en) * | 2021-04-08 | 2021-08-10 | 福建升腾资讯有限公司 | Reimbursement invoice management method, device and medium based on intelligent equipment |
| CN113344096A (en) * | 2021-06-22 | 2021-09-03 | 郑州信源信息技术股份有限公司 | Automatic bid document analysis method and system based on OCR technology |
| WO2022054136A1 (en) * | 2020-09-08 | 2022-03-17 | ファーストアカウンティング株式会社 | Data processing device, data processing method, and program |
| CN114821029A (en) * | 2022-05-16 | 2022-07-29 | 广东电网有限责任公司广州供电局 | OCR technology-based distribution network operation security ring identification method and system |
| CN115423586B (en) * | 2022-08-26 | 2023-09-29 | 重庆财经职业学院 | Financial invoice reimbursement uploading auditing system based on network |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150170255A1 (en) * | 2013-12-12 | 2015-06-18 | Ricoh Company, Ltd. | Information processing apparatus, information processing method and recording medium storing information processing program |
| CN106157100A (en) * | 2016-08-17 | 2016-11-23 | 广州市力融计算机技术有限公司 | Improvement contract managing bill level and the system and method for usefulness |
| CN108717545A (en) * | 2018-05-18 | 2018-10-30 | 北京大账房网络科技股份有限公司 | A kind of bank slip recognition method and system based on mobile phone photograph |
| CN108734849A (en) * | 2018-04-25 | 2018-11-02 | 新浪网技术(中国)有限公司 | A kind of automation invoice verification method and system |
| CN108734527A (en) * | 2018-05-18 | 2018-11-02 | 北京大账房网络科技股份有限公司 | The method and system of voucher are generated for the two-dimensional code scanning of VAT invoice |
| CN108765113A (en) * | 2018-05-17 | 2018-11-06 | 北京东港瑞宏科技有限公司 | A kind of examination of invoice, reimbursement, management system |
-
2018
- 2018-12-28 CN CN201811620377.1A patent/CN109726783A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150170255A1 (en) * | 2013-12-12 | 2015-06-18 | Ricoh Company, Ltd. | Information processing apparatus, information processing method and recording medium storing information processing program |
| CN106157100A (en) * | 2016-08-17 | 2016-11-23 | 广州市力融计算机技术有限公司 | Improvement contract managing bill level and the system and method for usefulness |
| CN108734849A (en) * | 2018-04-25 | 2018-11-02 | 新浪网技术(中国)有限公司 | A kind of automation invoice verification method and system |
| CN108765113A (en) * | 2018-05-17 | 2018-11-06 | 北京东港瑞宏科技有限公司 | A kind of examination of invoice, reimbursement, management system |
| CN108717545A (en) * | 2018-05-18 | 2018-10-30 | 北京大账房网络科技股份有限公司 | A kind of bank slip recognition method and system based on mobile phone photograph |
| CN108734527A (en) * | 2018-05-18 | 2018-11-02 | 北京大账房网络科技股份有限公司 | The method and system of voucher are generated for the two-dimensional code scanning of VAT invoice |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110188743A (en) * | 2019-05-13 | 2019-08-30 | 武汉大学 | A kind of taxi invoice identifying system and method |
| CN110288755A (en) * | 2019-05-21 | 2019-09-27 | 平安银行股份有限公司 | The invoice method of inspection, server and storage medium based on text identification |
| WO2020253113A1 (en) * | 2019-06-19 | 2020-12-24 | 深圳壹账通智能科技有限公司 | Invoice recording method, device, apparatus, and computer storage medium |
| CN110363545A (en) * | 2019-07-23 | 2019-10-22 | 金在（北京）金融信息服务有限公司 | Management method, system and the readable storage medium storing program for executing of invoice information |
| CN110647824A (en) * | 2019-09-03 | 2020-01-03 | 四川大学 | Value-added tax invoice layout extraction method based on computer vision technology |
| CN110647824B (en) * | 2019-09-03 | 2022-06-28 | 四川大学 | Value-added tax invoice layout extraction method based on computer vision technology |
| CN110781141A (en) * | 2019-09-07 | 2020-02-11 | 北京中科云链信息技术有限公司 | Method for acquiring head-up of electronic ticket |
| CN111199222A (en) * | 2019-12-30 | 2020-05-26 | 航天信息软件技术有限公司 | Bill management method and electronic equipment |
| CN111192019A (en) * | 2019-12-30 | 2020-05-22 | 武汉佰钧成技术有限责任公司 | Reimbursement processing method of target bill and related equipment |
| CN111223230A (en) * | 2020-01-19 | 2020-06-02 | 河南电力物资有限公司 | Invoice file authenticity identification method based on CRNN algorithm |
| CN111401199A (en) * | 2020-03-10 | 2020-07-10 | 深圳航天信息有限公司 | Invoice identification method and system |
| CN111461097A (en) * | 2020-03-18 | 2020-07-28 | 北京大米未来科技有限公司 | Method, apparatus, electronic device and medium for recognizing image information |
| CN111429645A (en) * | 2020-03-31 | 2020-07-17 | 重庆远见金税通信息系统技术有限公司 | True checking and weight checking system for bills |
| CN111967458A (en) * | 2020-08-14 | 2020-11-20 | 中国工商银行股份有限公司 | Bill exchange method, related equipment and system |
| CN111967458B (en) * | 2020-08-14 | 2024-04-16 | 中国工商银行股份有限公司 | Bill exchange method, related equipment and system |
| CN112053343A (en) * | 2020-09-02 | 2020-12-08 | 平安科技（深圳）有限公司 | User picture data processing method and device, computer equipment and storage medium |
| WO2022054136A1 (en) * | 2020-09-08 | 2022-03-17 | ファーストアカウンティング株式会社 | Data processing device, data processing method, and program |
| CN112735055A (en) * | 2020-12-17 | 2021-04-30 | 航天信息股份有限公司 | Invoice prereviewing and collecting system |
| CN113240503A (en) * | 2021-04-08 | 2021-08-10 | 福建升腾资讯有限公司 | Reimbursement invoice management method, device and medium based on intelligent equipment |
| CN113114868A (en) * | 2021-04-16 | 2021-07-13 | 合肥新青罗数字技术有限公司 | OCR recognition device and system for intangible asset management |
| CN113065940A (en) * | 2021-04-27 | 2021-07-02 | 平安普惠企业管理有限公司 | Invoice reimbursement method, device, equipment and storage medium based on artificial intelligence |
| CN113065940B (en) * | 2021-04-27 | 2023-11-17 | 江苏环迅信息科技有限公司 | Method, device, equipment and storage medium for reimbursement of invoice based on artificial intelligence |
| CN113011959A (en) * | 2021-05-24 | 2021-06-22 | 国能大渡河大数据服务有限公司 | Seven-expense intelligent auditing system and use method thereof |
| CN113344096A (en) * | 2021-06-22 | 2021-09-03 | 郑州信源信息技术股份有限公司 | Automatic bid document analysis method and system based on OCR technology |
| CN114821029A (en) * | 2022-05-16 | 2022-07-29 | 广东电网有限责任公司广州供电局 | OCR technology-based distribution network operation security ring identification method and system |
| CN115423586B (en) * | 2022-08-26 | 2023-09-29 | 重庆财经职业学院 | Financial invoice reimbursement uploading auditing system based on network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109726783A (en) | A kind of invoice acquisition management system and method based on OCR image recognition technology | |
| CN109543690B (en) | Method and device for extracting information | |
| CN114862540B (en) | Bill auditing system and method thereof | |
| CN112613501A (en) | Information auditing classification model construction method and information auditing method | |
| CN114117171B (en) | Intelligent project file collecting method and system based on energized thinking | |
| CN111459975B (en) | Bill verification system and method for enterprise reimbursement | |
| US8233751B2 (en) | Method and system for simplified recordkeeping including transcription and voting based verification | |
| CN109377189B (en) | Real estate electronic ticket system | |
| CN101706872A (en) | Universal open type face identification system | |
| CN110490238A (en) | A kind of image processing method, device and storage medium | |
| JPH07262224A (en) | Preservation/processing method of document image | |
| CN113011959A (en) | Seven-expense intelligent auditing system and use method thereof | |
| CN114202755A (en) | Transaction background authenticity auditing method and system based on OCR (optical character recognition) and NLP (non-line segment) technologies | |
| CN102456130A (en) | Method and system for verifying user identity document by face | |
| CN108959349A (en) | A kind of financial audit circular for confirmation system | |
| CN113379526A (en) | Intelligent invoice reimbursement method and device, electronic equipment and computer storage medium | |
| CN111753744A (en) | Method, device and equipment for classifying bill images and readable storage medium | |
| CN109784833A (en) | A kind of generation method and equipment of income statement | |
| KR101169444B1 (en) | 2 dimension code searching and storing device | |
| CN111429645B (en) | True checking and weight checking system for bills | |
| CN115563597A (en) | Artificial intelligence operation system and operation method based on big data | |
| CN111339939B (en) | Attendance checking method and device based on image recognition | |
| CN105718972B (en) | A kind of information intelligent acquisition method | |
| CN115309705A (en) | Data integration classification system and method for automatically identifying basic data elements of urban information model platform | |
| CN111444868A (en) | Bill duplicate checking system and method based on smart phone |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190507 |