CN108717545A - A kind of bank slip recognition method and system based on mobile phone photograph - Google Patents
A kind of bank slip recognition method and system based on mobile phone photograph Download PDFInfo
- Publication number
- CN108717545A CN108717545A CN201810482124.6A CN201810482124A CN108717545A CN 108717545 A CN108717545 A CN 108717545A CN 201810482124 A CN201810482124 A CN 201810482124A CN 108717545 A CN108717545 A CN 108717545A
- Authority
- CN
- China
- Prior art keywords
- invoice
- key message
- bill
- keyword
- type
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/273—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion removing elements interfering with the pattern to be recognised
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Character Input (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of bank slip recognition method based on mobile phone photograph comprising following steps：After intelligent identifying system in S1, mobile phone learns a plurality of types of bills, the key message of all types of bills is stored, establishes bill key message database；S2, the scanning of various mixing bills is become by electronic edition image by mobile phone photograph, is uploaded to intelligent identifying system and obtains keyword, for the picture for tilting and rotating, intelligent identifying system automatic identification simultaneously corrects；S3, obtained electronic edition image is compared according to the information that scanning obtains with the key message of storage or keyword, obtains the bill type of the bill, S4, the invoice of None- identified class or tax bureau's examination mistake is recognized after image procossing.The present invention need not be manually entered manually, do not had to arrange bill type, greatly improved the efficiency and accuracy, saved cost and time, liberated manpower.
Description
Technical field
The present invention relates to bank slip recognition method and technology fields, more particularly to a kind of bank slip recognition side based on mobile phone photograph
Method and system.
Background technology
As Structures of Tax system battalion in China's changes the implementation of increasing, present value-added tax is the presently most important turnover tax tax in China
Kind, the taxation range of value-added tax further covers second and third industry till now from the most of secondary industry covered originally
Most industries.
The administration of collection of present value-added tax is stringenter, while VAT invoice amount largely increases, manual typing it is too slow and
Check it is true and false take very much, and inefficiency, error rate is high.A greater variety of bills are there is also this problem simultaneously, than
Such as various bank receipts, machine dismisses ticket, train ticket, and quota invoice etc. is all traditional-handwork typing.And Enterprises ' Financial Workers exist
After the certification deduction for completing bill, it is also necessary to the work such as the scanning of row document, data inputting, artificial check and correction.Traditional manual entry
Mode, user need to put into a large amount of human cost and time cost, have not only raised operation cost, but also input speed is difficult to
It is promoted, error rate is difficult to decrease, and to improving business processing timeliness, enterprise service quality brings many negative effects.
But only identify that a kind of bill does not meet the service condition in reality yet, usual enterprise have multiple-bill need into
Account, such as value-added tax bill, machine dismiss ticket, quota invoice train ticket, bank money etc..Therefore modern information technologies hand is utilized
It is imperative that section develops a mixed system for sweeping bank slip recognition.
Invention content
In order to overcome the deficiencies of existing technologies, the present invention provides a kind of bank slip recognition method based on mobile phone photograph and is
System, is identified multiple types bill mixed sweep and discrimination is very high, saves human cost and time cost improves effect
Rate.
Specifically, the present invention provides a kind of bank slip recognition method based on mobile phone photograph comprising following steps：
After S1, intelligent identifying system learn a plurality of types of bills, to the key message of all types of bills into
Row storage identifies the different key message of all types of bills and dismisses ticket, train ticket and quota invoice for bank money, machine and determines
Adopted keyword establishes bill key message database, bill key message by the way that constantly study stores during scanning bill
Database includes recognition sequence list, Keyword List, key message list and corresponding bill type list, key column
Table, key message list and corresponding bill type list are one-to-one, the following tables of bill key message database
It is described：
S2, clear electronic edition image upload invoice is generated by mobile phone photograph to intelligent identifying system, for the electricity of upload
Sub- domain picture, intelligent identifying system carry out intelligent edge detection automatically, remove in electronic image with bill irrelevant portions, retain ticket
According to part itself, for the picture for tilting and rotating, intelligent identifying system automated intelligent is identified and is corrected, for different brands type
Number mobile phone photograph caused by electronic image cause not of uniform size the case where, intelligent identifying system is by electronic edition Image Adjusting to setting
Optimal size, it is excessively bright or excessively dark for picture when taking pictures, intelligent identifying system by electronic edition Image Adjusting to setting most
Excellent dim degree；
S3, obtained electronic edition image is compared according to the key message or keyword that scan obtained information and storage
It is right, the bill type of the bill is obtained, comparison sequence is carried out according to the sequence of recognition sequence list, if bill type is increment
Tax invoice, then checked, and is such as checked successfully, then examination result is back to intelligent recognition terminal shows, such as examination is lost
It loses, is then classified as the invoice checking wrong class；If bill type is the invoice type except VAT invoice, by the invoice
Invoice type return directly to intelligent recognition terminal and shown, if the invoice type of the None- identified invoice, by institute
The invoice for stating None- identified invoice type is classified as None- identified class and returns to recognition result；
S4, the invoice of None- identified class or the wrong class of examination is recognized after image procossing, at described image
The method of reason is determined according to the concrete reason of None- identified, locking key message position is specifically included, according to pixel
Coordinate carry out stripping and slicing, eliminate red chapter, removal lines or machine learning training carried out to incomplete number；
S5, step S1-S3 to None- identified class or after checking the secondary identification of invoice of wrong class, is being repeated, acquisition is finally
Bill type and the corresponding key message of the bill type.
Preferably, step S3 specifically includes following steps：
S31, key message is directly extracted to obtained electronic edition image, if it directly can extract key message first
Obtained key message will be scanned commonly to send out with the value-added tax in the key message list stored in bill key message database
The key message row of ticket, roll type bill, value-added tax electronics common invoice, motor vehicle sale uniform invoice or VAT invoice
It is compared, if the invoice belongs to value-added tax common invoice, roll type bill, value-added tax electronics common invoice, motor vehicle sale
One kind in uniform invoice or VAT invoice, then checked, and invoice type and the invoice are returned if checking successfully
The corresponding key message of type, such as examination failure then are classified as the invoice to check wrong class and return invoice type and corresponding
Key message；If the invoice is not belonging to value-added tax common invoice, roll type bill, value-added tax electronics common invoice, motor vehicle pin
One kind in uniform invoice or VAT invoice is sold, then carries out keyword extraction and obtained according to the keyword extracted to be somebody's turn to do
The corresponding key message of keyword simultaneously enters step S32；
S32, by the bank bill in the Keyword List stored in the keyword extracted and bill key message database
According to key column compared, if the invoice belongs to bank money, according to the pass for including in keyword recognition keyword
Key information, surrender of bills type and corresponding key message enter step S33 if the invoice is not belonging to bank money；
S33, the keyword extracted and the machine in the Keyword List that is stored in bill key message database are dismissed
The key column of ticket is compared, if the invoice, which belongs to machine, dismisses ticket, according to the pass for including in keyword recognition keyword
Key information, surrender of bills type and corresponding key message enter step S34 if the invoice, which is not belonging to machine, dismisses ticket；
S34, by the train ticket in the Keyword List stored in the keyword extracted and bill key message database
Key column compared, if the invoice belongs to train ticket, according to include in keyword recognition keyword key believe
Breath, surrender of bills type and corresponding key message enter step S35 if the invoice is not belonging to train ticket；
S35, the quota in the Keyword List stored in the keyword extracted and bill key message database is sent out
The key column of ticket is compared, if the invoice belongs to quota invoice, according to the pass for including in keyword recognition keyword
Key information, surrender of bills type and corresponding key message enter step S36 if the invoice is not belonging to quota invoice；
If the invoice type of S36, the None- identified invoice, nothing is classified as by the invoice of the None- identified invoice type
Method identifies class and returns to recognition result.
Preferably, it is specially the number for easily identifying mistake to carry out machine learning training to incomplete number
It practises, the number for easily identifying mistake includes 6 and 8,1 and 0,5 and 9 and 2 and 0.
Study, which is carried out, preferably for the number for easily identifying mistake specifically includes following steps：
Pretreatment：It finds the ROI section subgraph of image and carries out the normalized of size；
Feature extraction converts image to feature vector；
Classification and Identification carries out classification processing using k- nearest neighbour classifications method, finally completes identification work according to classification results,
Number to easily identifying mistake accurately identifies.
Preferably, the feature extraction the specific steps are：After picture is opened, noise reduction process is carried out, then by it
Gray processing is finally arranged a threshold value and is saved in its binaryzation in the array of one 32*32, each point is a pixel
Value, by this 1024 (32*32) a numerical value, is converted into the vector of (1,1024).
Preferably, it is complete for keyword is sent to the State Tax Administration that the method that VAT invoice is checked is carried out in S3
State's VAT invoice examination platform checks the true and false.
Preferably, a kind of based on the mixed bank slip recognition system swept of mobile phone comprising scanning means, identification terminal and intelligence
Identifying system, the scanning means and identification terminal communicate with the intelligent identifying system connect respectively,
The intelligent identifying system includes picture processing unit, for handling picture；
Key message extraction unit, for carrying out key message extraction to picture according to related algorithm；
Recognition unit obtains bill type for carrying out bank slip recognition according to key message；
Inspection unit, for checking VAT invoice；
Communication unit, for being communicated with the intelligent terminal.
Preferably, further include machine learning unit, for incomplete number carry out machine learning training be specially for
The number for easily identifying mistake is learnt, and the number for easily identifying mistake includes 6 and 8,1 and 0,5 and 9 and 2 and 0.
Preferably, the scanning means is mobile phone, and the intelligent identifying system is cell phone application.
Compared with prior art, the invention has the advantages that：
The intelligent identifying system that the present invention uses can realize that mobile phone carries out identification of taking pictures to bill, need not manually by hand
Input does not have to arrange bill type, and Enterprises ' Financial Workers do not have to after completing the certification deduction of bill, it is also necessary to which row document is swept
Retouch, data inputting, the work such as artificial check and correction, greatly improve the efficiency and accuracy, saved cost and time, liberated people
Power.
Compared with prior art, the present invention maximum leap is the identification of taking pictures for realizing mobile phone to multiple-bill, it is not
For single a certain bank slip recognition, the type of identification is more abundant, more intelligently, has saved time cost, has improved effect
Rate,
Secondly recognition correct rate greatly promotes, and is identified for being identified as whole of nominal value for the first time, for tilting and
The picture of rotation, intelligent identifying system automatic identification and can correct, and wrong bill, intelligent identifying system pair are identified to identification
It carries out image procossing, and locking key message position carries out stripping and slicing according to the coordinate of pixel, eliminates red chapter, removes lines, right
Incomplete number carries out machine learning training, is recognized.To improve recognition correct rate.
The intelligent identifying system that the present invention uses can realize mobile phone photograph intelligent recognition bill, need not return to company again
Reimbursement facilitates office worker's travel and routine office work to purchase, and provides authentic data for finance reimbursement, checks invoice whenever and wherever possible and close
Rule property is inquired true from false of bills, has been saved cost and time, improves efficiency, has liberated manpower.
Description of the drawings
Fig. 1 is the flow diagram of the present invention.
Specific implementation mode
Below with reference to the attached drawing exemplary embodiment that the present invention will be described in detail, feature and aspect.It is identical attached in attached drawing
Icon note indicates functionally the same or similar element.Although the various aspects of embodiment are shown in the accompanying drawings, unless special
It does not point out, it is not necessary to attached drawing drawn to scale.
A kind of bank slip recognition method based on mobile phone photograph of the present invention comprising following steps：
After S1, intelligent identifying system learn a plurality of types of bills, to the key message of all types of bills into
Row storage identifies the different key message of all types of bills and dismisses ticket, train ticket and quota invoice for bank money, machine and determines
Adopted keyword establishes bill key message database, bill key message by the way that constantly study stores during scanning bill
Database includes recognition sequence list, Keyword List, key message list and corresponding bill type list, key column
Table, key message list and corresponding bill type list are one-to-one.
Specifically, described in the following table of bill key message database：
Specific learning process is to scan a large amount of bills, and the key message of bill is distinguished, and the key of bill is believed
Breath is associated with actual bill type, and is directed to certain specific invoice definition of keywords, such as bank money, machine are dismissed
Ticket, train ticket and quota invoice, this few class invoice good keyword defined in learning process, and by keyword and key message
It is corresponding, in identification, as long as can scan pickup arrives keyword, the key message of needs can be extracted from keyword.It changes
Yan Zhi is to arrive keyword as long as can scan, it will be able to close comprising the key message needed in the keyword that certain bills define
The key message that keyword includes is obtained in key word.The study of database is based on largely scanning, and in practical applications, also may be used
Directly to define above-mentioned list, implant data library or increase further types of invoice type implant data library.
S2, the scanning of various mixing bills is become by electronic edition image by mobile phone, is uploaded to intelligent identifying system and obtains pass
Key word, for the picture for tilting and rotating, intelligent identifying system automatic identification simultaneously corrects.Electronic edition image can be black white image
It can also be coloured image.
S3, obtained electronic edition image is compared according to the key message or keyword that scan obtained information and storage
It is right, the bill type of the bill is obtained, comparison sequence is carried out according to the sequence of recognition sequence list, if bill type is identification
The first kind in sequence list and the second class invoice, (first kind and the second class invoice in recognition sequence list belong to rise in value
Tax invoice, below with VAT invoice replacement), then it is checked, is such as checked successfully, then examination result is back to intelligent recognition
Terminal is shown that the invoice is then classified as checking wrong class by such as examination failure；If bill type is except VAT invoice
Invoice type, then the invoice type of the invoice is returned directly into intelligent recognition terminal and shown, if None- identified should
The invoice of the None- identified invoice type is then classified as None- identified class and returns to recognition result by the invoice type of invoice.
The information obtained according to scanning to obtained electronic edition image is the keyword or key message defined before, is swept
Retouch to obtain mainly comprising the following steps for information and the Quick Response Code of the invoice of scanning positioned, and to the content of Quick Response Code storage inside into
Row Quick Response Code parses, and obtains information hiding inside Quick Response Code, is compared according to corresponding sequence after obtaining the information, judges
The invoice type of invoice.
Preferably, step S3 specifically includes following steps：
S31, key message is directly extracted to obtained electronic edition image, if it directly can extract key message first
Obtained key message will be scanned commonly to send out with the value-added tax in the key message list stored in bill key message database
The key message row of ticket, roll type bill, value-added tax electronics common invoice, motor vehicle sale uniform invoice or VAT invoice
It is compared, if the invoice belongs to value-added tax common invoice, roll type bill, value-added tax electronics common invoice, motor vehicle sale
One kind in uniform invoice or VAT invoice, then checked, and invoice type and the invoice are returned if checking successfully
The corresponding key message of type, such as examination failure then are classified as the invoice to check wrong class and return invoice type and corresponding
Key message；If the invoice is not belonging to value-added tax common invoice, roll type bill, value-added tax electronics common invoice, motor vehicle pin
One kind in uniform invoice or VAT invoice is sold, then carries out keyword extraction and obtained according to the keyword extracted to be somebody's turn to do
The corresponding key message of keyword simultaneously enters step S32；
S32, by the bank bill in the Keyword List stored in the keyword extracted and bill key message database
According to key column compared, if the invoice belongs to bank money, according to the pass for including in keyword recognition keyword
Key information, surrender of bills type and corresponding key message enter step S33 if the invoice is not belonging to bank money；
S33, the keyword extracted and the machine in the Keyword List that is stored in bill key message database are dismissed
The key column of ticket is compared, if the invoice, which belongs to machine, dismisses ticket, according to the pass for including in keyword recognition keyword
Key information, surrender of bills type and corresponding key message enter step S34 if the invoice, which is not belonging to machine, dismisses ticket；
S34, by the train ticket in the Keyword List stored in the keyword extracted and bill key message database
Key column compared, if the invoice belongs to train ticket, according to include in keyword recognition keyword key believe
Breath, surrender of bills type and corresponding key message enter step S35 if the invoice is not belonging to train ticket；
S35, the quota in the Keyword List stored in the keyword extracted and bill key message database is sent out
The key column of ticket is compared, if the invoice belongs to quota invoice, according to the pass for including in keyword recognition keyword
Key information, surrender of bills type and corresponding key message enter step S36 if the invoice is not belonging to quota invoice；
If the invoice type of S36, the None- identified invoice, nothing is classified as by the invoice of the None- identified invoice type
Method identifies class and returns to recognition result.
S4, the invoice of None- identified class or tax bureau's examination mistake is recognized after image procossing, the figure
The method of picture processing is determined according to the concrete reason of None- identified, locking key message position is specifically included, according to picture
The coordinate of vegetarian refreshments carries out stripping and slicing, eliminates red chapter, removal lines or carries out machine learning training to incomplete number.
Preferably, it is specially the number for easily identifying mistake to carry out machine learning training to incomplete number
It practises, the number for easily identifying mistake includes 6 and 8,1 and 0,5 and 9 and 2 and 0.
Study, which is carried out, preferably for the number for easily identifying mistake specifically includes following steps：
Pretreatment：It finds the ROI section subgraph of image and carries out the normalized of size；
Feature extraction converts image to feature vector；
Classification and Identification carries out classification processing using k- nearest neighbour classifications method, finally completes identification work according to classification results,
Number to easily identifying mistake accurately identifies.
Preferably, the feature extraction the specific steps are：After picture is opened, noise reduction process is carried out, then by it
Gray processing is finally arranged a threshold value and is saved in its binaryzation in the array of one 32*32, each point is a pixel
Value, by this 1024 (32*32) a numerical value, is converted into the vector of (1,1024).
Preferably, it is complete for keyword is sent to the State Tax Administration that the method that VAT invoice is checked is carried out in S3
State's VAT invoice examination platform checks the true and false.
Preferably, a kind of based on the mixed bank slip recognition system swept of mobile phone comprising scanning means, identification terminal and intelligence
Identifying system, the scanning means and identification terminal communicate with the intelligent identifying system connect respectively,
The intelligent identifying system includes picture processing unit, for handling picture；
Key message extraction unit, for carrying out key message extraction to picture according to related algorithm；
Recognition unit obtains bill type for carrying out bank slip recognition according to key message；
Inspection unit, for checking VAT invoice；
Communication unit, for being communicated with the intelligent terminal.
Preferably, further include machine learning unit, for incomplete number carry out machine learning training be specially for
The number for easily identifying mistake is learnt, and the number for easily identifying mistake includes 6 and 8,1 and 0,5 and 9 and 2 and 0.
Preferably, the scanning means is mobile phone.Clear electronic color image, which is generated, by mobile phone photograph uploads invoice extremely
Intelligent identifying system, for the electronic color image of upload, intelligent identifying system carries out intelligent edge detection automatically, removes electronics
With bill irrelevant portions in image, retain bill part itself, for the picture for tilting and rotating, intelligent identifying system can be certainly
The case where dynamic intelligent recognition simultaneously corrects, cause not of uniform size for electronic image caused by the mobile phone photograph of different brands model, intelligence
Energy identifying system carries out adjustment to best size, and excessively bright or excessively dark for picture when taking pictures, intelligent identifying system carries out
Adjustment is handled to best dim degree by intelligent recognition, is extracted key message, is checked, and examination result is returned to
Mobile phone terminal is shown.For identifying that wrong situation, intelligent identifying system carry out it image procossing caused by picture problem, lock
Determine key message position, stripping and slicing is carried out according to the coordinate of pixel, eliminates red chapter, removes lines, to incomplete digital carry out machine
Device learning training, is recognized.
Specific embodiment 1
By taking a value-added tax VAT invoice as an example, the key message of the VAT invoice of acquisition is scanned
For：Invoice codes：5XXX1XX1XX, invoice number：XXXX5XX4, date：20171027, the amount of money：88288.29.
Specific embodiment 2
By taking a value-added tax common invoice as an example, the key message for scanning the common invoice of acquisition is：Invoice codes：
5XXX17XXX0, invoice number：0XXX4XX8, date：20171017, verification examination code：551000.
Specific embodiment 3
By taking a value-added tax electronics common invoice as an example, the key message for scanning the common invoice of acquisition is：Invoice generation
Code：
01XXXXXX0111, invoice number：17XXXX54, date：20171017, verification examination code：3XXXX7.
Specific embodiment 4
By taking a bank money as an example, the key message for scanning the bank money of acquisition is：Bank Name：Chinese agriculture
Bank, bill name：Enterprise's Internetbank service charge, beneficiary：XXXX Co., Ltds of the areas XX of Chongqing City, paying party：Sichuan XXXXXX
Co., Ltd, date：20180206, the amount of money：10.00 remarks：Enterprise's Internetbank transaction procedure takes.
Specific embodiment 5
By taking a car machine dismisses ticket as an example, the keyword that machine dismisses ticket is：Machine dismisses ticket, and key message is：The amount of money：
195.00。
Specific embodiment 6
By taking a train ticket as an example, the keyword of train ticket is：Railway, 12306, hard seat, soft seat, commercial seat, first block,
Coach seat, soft sleeper, hard berth key message are：Departure place：West Beijing, destination：Zhengzhou, date：20170818, the amount of money：
93.00。
Specific embodiment 7
By taking a quota invoice as an example, the keyword of quota invoice is quota invoice, and key message is：The amount of money：100.00.
Compared with prior art, the invention has the advantages that：
The intelligent identifying system that the present invention uses can realize that mobile phone carries out identification of taking pictures to bill, need not manually by hand
Input does not have to arrange bill type, and Enterprises ' Financial Workers do not have to after completing the certification deduction of bill, it is also necessary to which row document is swept
Retouch, data inputting, the work such as artificial check and correction, greatly improve the efficiency and accuracy, saved cost and time, liberated people
Power.
Compared with prior art, the present invention maximum leap is the identification of taking pictures for realizing mobile phone to multiple-bill, it is not
For single a certain bank slip recognition, the type of identification is more abundant, more intelligently, has saved time cost, has improved effect
Rate,
Secondly recognition correct rate greatly promotes, and is identified for being identified as whole of nominal value for the first time, for tilting and
The picture of rotation, intelligent identifying system automatic identification and can correct, and wrong bill, intelligent identifying system pair are identified to identification
It carries out image procossing, and locking key message position carries out stripping and slicing according to the coordinate of pixel, eliminates red chapter, removes lines, right
Incomplete number carries out machine learning training, is recognized.To improve recognition correct rate.
The intelligent identifying system that the present invention uses can realize mobile phone photograph intelligent recognition bill, need not return to company again
Reimbursement facilitates office worker's travel and routine office work to purchase, and provides authentic data for finance reimbursement, checks invoice whenever and wherever possible and close
Rule property is inquired true from false of bills, has been saved cost and time, improves efficiency, has liberated manpower.
Finally it should be noted that：Above-described embodiments are merely to illustrate the technical scheme, rather than to it
Limitation；Although the present invention is described in detail referring to the foregoing embodiments, it will be understood by those of ordinary skill in the art that：
It can still modify to the technical solution recorded in previous embodiment, or to which part or all technical features into
Row equivalent replacement；And these modifications or substitutions, it does not separate the essence of the corresponding technical solution various embodiments of the present invention technical side
The range of case.
Claims (10)
1. a kind of bank slip recognition method based on mobile phone photograph, it is characterised in that：It includes the following steps：
The intelligent identifying system that S1, interior of mobile phone are arranged carries out automatic identification to a plurality of types of bills and intellectual analysis learns
Afterwards, the key message of all types of bills is stored, identify the different key message of all types of bills and for bank money,
Machine dismisses ticket, train ticket and quota invoice definition of keywords, by the way that constantly training stores in the identification process to bill, builds
Vertical bill key message database, bill key message database include recognition sequence list, Keyword List, key message row
Table and corresponding bill type list, Keyword List, key message list and corresponding bill type list are one by one
It is corresponding, described in the following table of bill key message database：
S2, it clear electronic edition image is generated by mobile phone photograph is uploaded to intelligent identifying system, for the electronic edition image of upload,
Intelligent identifying system carries out intelligent edge detection automatically, removes in electronic image with bill irrelevant portions, retains bill portion itself
Point, for the picture for tilting and rotating, intelligent identifying system automated intelligent is identified and is corrected, for the mobile phone of different brands model
The case where cause not of uniform size of electronic image caused by taking pictures, intelligent identifying system is by the optimal big of electronic edition Image Adjusting to setting
It is small, it is excessively bright or excessively dark for picture when taking pictures, intelligent identifying system by electronic edition Image Adjusting to setting optimal dim degree；
S3, obtained electronic edition image is compared according to the information that scanning obtains with the key message of storage or keyword,
The bill type of the bill is obtained, comparison sequence is carried out according to the sequence of recognition sequence list, if bill type is that identification is suitable
The invoice of the first kind and the second class, then checked in sequence table, is such as checked successfully, then examination result is back to intelligent recognition
Terminal is shown that the invoice is then classified as checking wrong class by such as examination failure；If bill type is the first kind and the second class
Invoice except invoice type, then the invoice type of the invoice is returned directly into intelligent recognition terminal and shown, if
The invoice of the None- identified invoice type is then classified as None- identified class and returns to identification by the invoice type of the None- identified invoice
As a result；
S4, the invoice of None- identified class or the wrong class of examination is recognized after image procossing, described image processing
Method is determined according to the concrete reason of None- identified, and the specific method of graphics process includes locking key message position, root
Stripping and slicing is carried out, red chapter, removal lines are eliminated or machine learning training is carried out to incomplete number according to the coordinate of pixel；
S5, after the secondary identification of invoice to None- identified class or the wrong class of examination, repeat step S1-S3, obtain final ticket
According to type and the corresponding key message of the bill type.
2. the bank slip recognition method according to claim 1 based on mobile phone photograph, it is characterised in that：Step S3 is specifically included
Following steps：
S31, key message is directly extracted to obtained electronic edition image, will be swept first if it directly can extract key message
The key message retouched and value-added tax common invoice, volume in the key message list stored in bill key message database
The key message row progress of formula invoice, value-added tax electronics common invoice, motor vehicle sale uniform invoice or VAT invoice
Comparison, if the invoice belongs to value-added tax common invoice, roll type bill, value-added tax electronics common invoice, motor vehicle sale unification
One kind in invoice or VAT invoice, then checked, and invoice type and the invoice type are returned if checking successfully
Corresponding key message, such as examination failure, then be classified as the invoice to check wrong class and return to invoice type and corresponding key
Information；If the invoice is not belonging to value-added tax common invoice, roll type bill, value-added tax electronics common invoice, motor vehicle sale system
One kind in one invoice or VAT invoice then carries out keyword extraction and obtains the key according to the keyword extracted
The corresponding key message of word simultaneously enters step S32；
S32, by the bank money in the Keyword List stored in the keyword extracted and bill key message database
Key column is compared, if the invoice belongs to bank money, is believed according to the key for including in keyword recognition keyword
Breath, surrender of bills type and corresponding key message enter step S33 if the invoice is not belonging to bank money；
S33, the machine in the Keyword List stored in the keyword extracted and bill key message database is dismissed into ticket
Key column is compared, if the invoice, which belongs to machine, dismisses ticket, is believed according to the key for including in keyword recognition keyword
Breath, surrender of bills type and corresponding key message enter step S34 if the invoice, which is not belonging to machine, dismisses ticket；
S34, by the pass of the train ticket in the Keyword List stored in the keyword extracted and bill key message database
Key word row are compared, if the invoice belongs to train ticket, according to the key message for including in keyword recognition keyword, are returned
It returns bill type and corresponding key message enters step S35 if the invoice is not belonging to train ticket；
S35, by the quota invoice in the Keyword List stored in the keyword extracted and bill key message database
Key column is compared, if the invoice belongs to quota invoice, is believed according to the key for including in keyword recognition keyword
Breath, surrender of bills type and corresponding key message enter step S36 if the invoice is not belonging to quota invoice；
If the invoice type of S36, the None- identified invoice, the invoice of the None- identified invoice type is classified as not knowing
Other class simultaneously returns to recognition result.
3. the bank slip recognition method according to claim 1 based on mobile phone photograph, it is characterised in that：To incomplete number into
Row machine learning training is specially to learn for easily identifying the number of mistake, and the number for easily identifying mistake includes 6 Hes
8,1 and 0,5 and 9 and 2 and 0.
4. the bank slip recognition method according to claim 3 based on mobile phone photograph, it is characterised in that：For easily identifying mistake
Number accidentally carries out study and specifically includes following steps：
Pretreatment：It finds the ROI section subgraph of image and carries out the normalized of size；
Feature extraction converts image to feature vector；
Classification and Identification carries out classification processing using k- nearest neighbour classifications method, finally identification work is completed according to classification results, to holding
The number of mistake easy to identify is accurately identified.
5. the bank slip recognition method according to claim 4 based on mobile phone photograph, it is characterised in that：The feature extraction
The specific steps are：After image is opened, carry out noise reduction process, then by its gray processing, be finally arranged a threshold value will secondly
Value is saved in the array of a 32*32, each point is a pixel value, this 1024 (32*32) a numerical value is converted into
The vector of (1,1024).
6. the bank slip recognition method according to claim 1 based on mobile phone photograph, it is characterised in that：It is checked in S3
Method is that key message is sent to State Tax Administration's whole nation VAT invoice to check the platform examination true and false.
7. a kind of bank slip recognition system for bank slip recognition method described in claim 1, it is characterised in that：It includes scanning
Device, identification terminal and intelligent identifying system, the scanning means and identification terminal are logical with the intelligent identifying system respectively
News connection,
The intelligent identifying system includes picture processing unit, for handling picture；
Key message extraction unit, for carrying out key message extraction to picture according to keyword；
Recognition unit obtains bill type for carrying out bank slip recognition according to key message；
Inspection unit, for checking VAT invoice；
Communication unit, for being communicated with the intelligent terminal.
8. bank slip recognition system according to claim 7, it is characterised in that：Further include machine learning unit, for residual
It is specially to learn for easily identifying the number of mistake that scarce number, which carries out machine learning training, easily identifies the number of mistake
Word includes 6 and 8,1 and 0,5 and 9 and 2 and 0.
9. bank slip recognition system according to claim 8, it is characterised in that：For easily identifying the number of mistake
Habit specifically includes following steps：
Pretreatment：It finds the ROI section subgraph of image and carries out the normalized of size；
Feature extraction converts image to feature vector；
Classification and Identification carries out classification processing using k- nearest neighbour classifications method, finally identification work is completed according to classification results, to holding
The number of mistake easy to identify is accurately identified.
10. bank slip recognition system according to claim 7, it is characterised in that：The scanning means is mobile phone, the intelligence
Identifying system is cell phone application.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810482124.6A CN108717545B (en) | 2018-05-18 | 2018-05-18 | Bill identification method and system based on mobile phone photographing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810482124.6A CN108717545B (en) | 2018-05-18 | 2018-05-18 | Bill identification method and system based on mobile phone photographing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108717545A true CN108717545A (en) | 2018-10-30 |
| CN108717545B CN108717545B (en) | 2020-12-18 |
Family
ID=63900021
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810482124.6A Active CN108717545B (en) | 2018-05-18 | 2018-05-18 | Bill identification method and system based on mobile phone photographing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108717545B (en) |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109472919A (en) * | 2018-12-28 | 2019-03-15 | 远光软件股份有限公司 | A kind of bill takes over method and associated terminal and storage device |
| CN109726783A (en) * | 2018-12-28 | 2019-05-07 | 大象慧云信息技术有限公司 | A kind of invoice acquisition management system and method based on OCR image recognition technology |
| CN110147838A (en) * | 2019-05-20 | 2019-08-20 | 苏州微创关节医疗科技有限公司 | A kind of product specification typing, detection method and system |
| CN110334640A (en) * | 2019-06-28 | 2019-10-15 | 苏宁云计算有限公司 | A kind of ticket processing method and system |
| CN110427853A (en) * | 2019-07-24 | 2019-11-08 | 北京一诺前景财税科技有限公司 | A kind of method of smart tickets information extraction processing |
| CN110675546A (en) * | 2019-09-06 | 2020-01-10 | 深圳壹账通智能科技有限公司 | Invoice picture identification and verification method, system, equipment and readable storage medium |
| CN110675234A (en) * | 2019-08-23 | 2020-01-10 | 国信电子票据平台信息服务有限公司 | Electronic newspaper bill generation method and electronic equipment |
| CN111104853A (en) * | 2019-11-11 | 2020-05-05 | 中国建设银行股份有限公司 | Image information input method and device, electronic equipment and storage medium |
| CN111178345A (en) * | 2019-05-20 | 2020-05-19 | 京东方科技集团股份有限公司 | Bill analysis method, bill analysis device, computer equipment and medium |
| 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 |
| CN111275035A (en) * | 2018-12-04 | 2020-06-12 | 北京嘀嘀无限科技发展有限公司 | Method and system for identifying background information |
| CN111931473A (en) * | 2019-05-13 | 2020-11-13 | 阿里巴巴集团控股有限公司 | Bill processing method and device |
| 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 |
| CN112135002A (en) * | 2020-07-31 | 2020-12-25 | 钱微 | Bill filling system for financial management and working method thereof |
| CN112541461A (en) * | 2020-12-21 | 2021-03-23 | 四川新网银行股份有限公司 | Automatic auditing method and device for consumption credentials without fixed format template |
| CN112699860A (en) * | 2021-03-24 | 2021-04-23 | 成都新希望金融信息有限公司 | Method for automatically extracting and sorting effective information in personal tax APP operation video |
| US11030450B2 (en) * | 2018-05-31 | 2021-06-08 | Vatbox, Ltd. | System and method for determining originality of computer-generated images |
| 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 |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102208092A (en) * | 2011-05-25 | 2011-10-05 | 重庆市电力公司永川供电局 | Financial bill reimbursement automatic processing method |
| CN102750541A (en) * | 2011-04-22 | 2012-10-24 | 北京文通科技有限公司 | Document image classifying distinguishing method and device |
| CN104050450A (en) * | 2014-06-16 | 2014-09-17 | 西安通瑞新材料开发有限公司 | Vehicle license plate recognition method based on video |
| CN105046553A (en) * | 2015-07-09 | 2015-11-11 | 胡昭 | Cloud intelligent invoice recognition inspection system and method based on mobile phone |
| US20150339739A1 (en) * | 2012-04-26 | 2015-11-26 | Chengdu Santai Holding Group Co., Ltd. | Corporate bill selling system with anti-counterfeiting verification |
| CN105654072A (en) * | 2016-03-24 | 2016-06-08 | 哈尔滨工业大学 | Automatic character extraction and recognition system and method for low-resolution medical bill image |
| CN105809814A (en) * | 2014-12-30 | 2016-07-27 | 航天信息股份有限公司 | Invoice certification system supporting multiple invoice types and method |
| CN107977665A (en) * | 2017-12-15 | 2018-05-01 | 北京科摩仕捷科技有限公司 | The recognition methods of key message and computing device in a kind of invoice |
-
2018
- 2018-05-18 CN CN201810482124.6A patent/CN108717545B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102750541A (en) * | 2011-04-22 | 2012-10-24 | 北京文通科技有限公司 | Document image classifying distinguishing method and device |
| CN102208092A (en) * | 2011-05-25 | 2011-10-05 | 重庆市电力公司永川供电局 | Financial bill reimbursement automatic processing method |
| US20150339739A1 (en) * | 2012-04-26 | 2015-11-26 | Chengdu Santai Holding Group Co., Ltd. | Corporate bill selling system with anti-counterfeiting verification |
| CN104050450A (en) * | 2014-06-16 | 2014-09-17 | 西安通瑞新材料开发有限公司 | Vehicle license plate recognition method based on video |
| CN105809814A (en) * | 2014-12-30 | 2016-07-27 | 航天信息股份有限公司 | Invoice certification system supporting multiple invoice types and method |
| CN105046553A (en) * | 2015-07-09 | 2015-11-11 | 胡昭 | Cloud intelligent invoice recognition inspection system and method based on mobile phone |
| CN105654072A (en) * | 2016-03-24 | 2016-06-08 | 哈尔滨工业大学 | Automatic character extraction and recognition system and method for low-resolution medical bill image |
| CN107977665A (en) * | 2017-12-15 | 2018-05-01 | 北京科摩仕捷科技有限公司 | The recognition methods of key message and computing device in a kind of invoice |
Non-Patent Citations (1)
| Title |
|---|
| SHITER: "《OpenCV手写数字字符识别(基于k近邻算法)》", 3 December 2013 * |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11030450B2 (en) * | 2018-05-31 | 2021-06-08 | Vatbox, Ltd. | System and method for determining originality of computer-generated images |
| CN111275035B (en) * | 2018-12-04 | 2023-10-31 | 北京嘀嘀无限科技发展有限公司 | Method and system for identifying background information |
| CN111275035A (en) * | 2018-12-04 | 2020-06-12 | 北京嘀嘀无限科技发展有限公司 | Method and system for identifying background information |
| CN109726783A (en) * | 2018-12-28 | 2019-05-07 | 大象慧云信息技术有限公司 | A kind of invoice acquisition management system and method based on OCR image recognition technology |
| CN109472919A (en) * | 2018-12-28 | 2019-03-15 | 远光软件股份有限公司 | A kind of bill takes over method and associated terminal and storage device |
| CN111931473A (en) * | 2019-05-13 | 2020-11-13 | 阿里巴巴集团控股有限公司 | Bill processing method and device |
| CN110147838A (en) * | 2019-05-20 | 2019-08-20 | 苏州微创关节医疗科技有限公司 | A kind of product specification typing, detection method and system |
| WO2020233270A1 (en) * | 2019-05-20 | 2020-11-26 | 京东方科技集团股份有限公司 | Bill analyzing method and analyzing apparatus, computer device and medium |
| CN111178345A (en) * | 2019-05-20 | 2020-05-19 | 京东方科技集团股份有限公司 | Bill analysis method, bill analysis device, computer equipment and medium |
| CN110147838B (en) * | 2019-05-20 | 2021-07-02 | 苏州微创关节医疗科技有限公司 | Product specification inputting and detecting method and system |
| 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 |
| CN110334640A (en) * | 2019-06-28 | 2019-10-15 | 苏宁云计算有限公司 | A kind of ticket processing method and system |
| CN110427853A (en) * | 2019-07-24 | 2019-11-08 | 北京一诺前景财税科技有限公司 | A kind of method of smart tickets information extraction processing |
| CN110675234A (en) * | 2019-08-23 | 2020-01-10 | 国信电子票据平台信息服务有限公司 | Electronic newspaper bill generation method and electronic equipment |
| CN110675546A (en) * | 2019-09-06 | 2020-01-10 | 深圳壹账通智能科技有限公司 | Invoice picture identification and verification method, system, equipment and readable storage medium |
| CN111104853A (en) * | 2019-11-11 | 2020-05-05 | 中国建设银行股份有限公司 | Image information input method and device, electronic equipment and storage medium |
| 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 |
| CN112135002A (en) * | 2020-07-31 | 2020-12-25 | 钱微 | Bill filling system for financial management and working method thereof |
| CN112541461A (en) * | 2020-12-21 | 2021-03-23 | 四川新网银行股份有限公司 | Automatic auditing method and device for consumption credentials without fixed format template |
| CN112699860A (en) * | 2021-03-24 | 2021-04-23 | 成都新希望金融信息有限公司 | Method for automatically extracting and sorting effective information in personal tax APP operation video |
| CN112699860B (en) * | 2021-03-24 | 2021-06-22 | 成都新希望金融信息有限公司 | Method for automatically extracting and sorting effective information in personal tax APP operation video |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108717545B (en) | 2020-12-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108717545A (en) | A kind of bank slip recognition method and system based on mobile phone photograph | |
| CN108777021A (en) | It is a kind of to mix the bank slip recognition method and system swept based on scanner | |
| US9767379B2 (en) | Systems, methods and computer program products for determining document validity | |
| US11151369B2 (en) | Systems and methods for classifying payment documents during mobile image processing | |
| CN102800148B (en) | RMB sequence number identification method | |
| WO2021027336A1 (en) | Authentication method and apparatus based on seal and signature, and computer device | |
| US7983468B2 (en) | Method and system for extracting information from documents by document segregation | |
| CA2589947C (en) | Machine character recognition verification | |
| CN107016363A (en) | Bill images managing device, bill images management system and method | |
| US11694499B2 (en) | Systems and methods for updating an image registry for use in fraud detection related to financial documents | |
| CN102194275A (en) | Automatic ticket checking method for train tickets | |
| CN113095307A (en) | Automatic identification method for financial voucher information | |
| CN110276587A (en) | The method, apparatus of project examination calculates equipment and computer readable storage medium | |
| CN111462388A (en) | Bill inspection method and device, terminal equipment and storage medium | |
| CN117036073B (en) | Invoice auditing and automatic reimbursement system based on Internet | |
| CN114219507A (en) | Qualification auditing method and device for traditional Chinese medicine supplier, electronic equipment and storage medium | |
| US20210090086A1 (en) | Systems and methods for fraud detection for images of financial documents | |
| CN113066223A (en) | Automatic invoice verification method and device | |
| CN111881880A (en) | Bill text recognition method based on novel network | |
| CN111582115A (en) | Financial bill processing method, device and equipment and readable storage medium | |
| CN109460720A (en) | Ballot paper recognition methods based on convolutional neural networks | |
| CN113989790A (en) | Precision self-adaptive civil aviation passenger ticket identification method based on generation countermeasure | |
| CN113688834A (en) | Ticket recognition method, ticket recognition system and computer readable storage medium | |
| CN116030479A (en) | Automatic bill verification method and system based on OCR | |
| CN113052134A (en) | Online account opening method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address | ||
| CP03 | Change of name, title or address |
Address after: 501-018, floor 5, No. 15, wanquanzhuang Road, Haidian District, Beijing 100089 Patentee after: Dajingfang Network Technology Co.,Ltd. Address before: 100000 405, No. 15, wanquanzhuang Road, Haidian District, Beijing Patentee before: BEIJING DAZHANGFANG NETWORK TECHNOLOGY Co.,Ltd. |