CN109284750A - Bank slip recognition method and device, electronic equipment and storage medium - Google Patents
Bank slip recognition method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN109284750A CN109284750A CN201810923451.0A CN201810923451A CN109284750A CN 109284750 A CN109284750 A CN 109284750A CN 201810923451 A CN201810923451 A CN 201810923451A CN 109284750 A CN109284750 A CN 109284750A
- Authority
- CN
- China
- Prior art keywords
- character
- line
- recognized
- images
- region
- 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
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
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- 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
- G06V10/267—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 by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/146—Aligning or centring of the image pick-up or image-field
- G06V30/1475—Inclination or skew detection or correction of characters or of image to be recognised
- G06V30/1478—Inclination or skew detection or correction of characters or of image to be recognised of characters or characters lines
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Character Input (AREA)
Abstract
This disclosure relates to bank slip recognition method and device, electronic equipment and storage medium.This method comprises: carrying out character machining to images to be recognized, at least one character zone in the images to be recognized is determined；Character recognition is carried out at least one described character zone, determines the character content at least one described character zone；Based on the character content in the corresponding ticket templates data of the images to be recognized and at least one described character zone, bank slip recognition result is obtained.The disclosure can all character contents in automatic identification bill, carry out ticket processing and examination without artificial, substantially increase bank slip recognition efficiency and accuracy.
Description
Technical field
This disclosure relates to technical field of computer vision more particularly to a kind of bank slip recognition method and device, electronic equipment
And storage medium.
Background technique
It is desirable to which manually bill is audited and is checked.For example, motor vehicle sale uniform invoice is purchasing unit
Payment voucher needs that manually motor vehicle sale uniform invoice is audited and checked at present.The audit and examination of bill need
A large amount of manpower, working efficiency is low, and error-prone.
Summary of the invention
In view of this, the present disclosure proposes a kind of bank slip recognition technical solutions.
According to the one side of the disclosure, a kind of bank slip recognition method is provided, comprising:
Character machining is carried out to images to be recognized, determines at least one character zone in the images to be recognized；
Character recognition is carried out at least one described character zone, is determined in the character at least one described character zone
Hold；
Based on the character in the corresponding ticket templates data of the images to be recognized and at least one described character zone
Content obtains bank slip recognition result.
In one possible implementation, before carrying out character machining to images to be recognized, the method also includes:
In the case where the images to be recognized has inclination or distortion, the images to be recognized is corrected, is obtained at correction
The images to be recognized after reason；
It is described that character machining is carried out to images to be recognized, comprising:
Character machining is carried out to the images to be recognized after the correction process.
In one possible implementation, the correction images to be recognized, after obtaining correction process it is described to
Identify image, comprising:
Determine the representative points coordinate of the images to be recognized；
According to the representative points coordinate of the images to be recognized and the initial vertax coordinate of the images to be recognized, determine
The corresponding projection matrix of the images to be recognized；
Projective transformation is carried out to the images to be recognized according to the projection matrix, it is described wait know after obtaining correction process
Other image.
In one possible implementation, described that character machining is carried out to images to be recognized, determine the figure to be identified
At least one character zone as in, comprising:
Character machining is carried out to the images to be recognized by first nerves network, is determined in the images to be recognized extremely
A few character zone.
In one possible implementation, described that character recognition is carried out at least one described character zone, determine institute
State the character content at least one character zone, comprising:
At least two character zones for being less than first threshold to the distance of horizontal direction merge, and obtain at least one text
Current row region；
Character recognition is carried out at least one described line of text region, obtains the word at least one described line of text region
Accord with content.
In one possible implementation, before carrying out character recognition at least one described line of text region, institute
State method further include:
Based on the size of each line of text region in the horizontal direction at least one described line of text region, to it is described extremely
A few line of text region is screened, at least one target text row region is obtained；
It is described that character recognition is carried out at least one described line of text region, it obtains at least one described line of text region
Character content, comprising:
Character recognition is carried out at least one described target text row region, obtains at least one described target text row area
The word content in each target text row region in domain.
In one possible implementation, described based on each line of text region at least one described line of text region
Size in the horizontal direction screens at least one described line of text region, obtains at least one target text row area
Domain, comprising:
The size that horizontal direction is removed from least one described line of text region is less than the line of text region of second threshold,
Obtain at least one target text row region；And/or
The ratio that the size of horizontal direction and the size of vertical direction are removed from least one described line of text region is small
In the line of text region of third threshold value, at least one target text row region is obtained.
In one possible implementation, described that character recognition is carried out at least one described line of text region, it obtains
Character content at least one described line of text region, comprising:
Feature extraction processing is carried out to the line of text region, obtains the characteristic pattern in the line of text region；
Processing is decoded to the characteristic pattern, obtains sequence label, wherein the sequence label includes at least one mark
Label, the corresponding character of each label；
Based on the sequence label, the character content in the line of text region is obtained.
In one possible implementation, the length of the sequence label is corresponding with the width of the characteristic pattern.
In one possible implementation, it is based on the sequence label, obtains the character content in the line of text region,
Include:
Based at least one label for corresponding to space character in the sequence label, the sequence label is divided at least
Two subsequences；
Based on the label for including in each subsequence at least two subsequence, determine that each subsequence is corresponding
Character content；
Putting in order based at least two subsequence connects the corresponding character content of at least two subsequence,
Obtain the character content in the line of text region.
In one possible implementation, based on the mark for including in each subsequence at least two subsequence
Label, determine the corresponding character content of each subsequence, comprising:
At least two adjacent label identical in the subsequence is merged, the sub- sequence after obtaining merging treatment
Column；
Based on the label for including in the subsequence after merging treatment, the corresponding character content of the subsequence is determined.
In one possible implementation, the method also includes:
Based on the character content at least one described character zone, the corresponding ticket templates of the images to be recognized are determined
Data.
In one possible implementation, described to be based on the corresponding ticket templates data of the images to be recognized and institute
The character content at least one character zone is stated, bank slip recognition result is obtained, comprising:
Based on the corresponding ticket templates data of the images to be recognized, the class for the information that the images to be recognized includes is determined
Other and position；
The classification for the information for including based on the images to be recognized and position, determine in the character in the images to be recognized
Classification belonging to holding；
It is merged to same category of character content is belonged to, obtains the bank slip recognition result.
According to the one side of the disclosure, a kind of bank slip recognition device is provided, comprising:
Character machining module determines in the images to be recognized at least for carrying out character machining to images to be recognized
One character zone；
Character recognition module, for carrying out character recognition at least one described character zone, determine it is described at least one
Character content in character zone；
First determining module, for based on the corresponding ticket templates data of the images to be recognized and it is described at least one
Character content in character zone obtains bank slip recognition result.
In one possible implementation, described device further include:
Correction module, for correcting the figure to be identified in the case where the images to be recognized has inclination or distortion
Picture, the images to be recognized after obtaining correction process；
The character machining module is used for:
Character machining is carried out to the images to be recognized after the correction process.
In one possible implementation, the correction module includes:
First determines submodule, for determining the representative points coordinate of the images to be recognized；
Second determine submodule, for according to the images to be recognized representative points coordinate and the images to be recognized
Initial vertax coordinate, determine the corresponding projection matrix of the images to be recognized；
Correction module is corrected for carrying out projective transformation to the images to be recognized according to the projection matrix
The images to be recognized that treated.
In one possible implementation, the character machining module is used for:
Character machining is carried out to the images to be recognized by first nerves network, is determined in the images to be recognized extremely
A few character zone.
In one possible implementation, the character recognition module includes:
First merges submodule, and at least two character zones for being less than first threshold for the distance to horizontal direction carry out
Merge, obtains at least one line of text region；
Character recognition submodule, for carrying out character recognition at least one described line of text region, obtain it is described at least
Character content in one line of text region.
In one possible implementation, described device further include:
Screening module, for based on each line of text region at least one described line of text region in the horizontal direction
Size screens at least one described line of text region, obtains at least one target text row region；
The character recognition submodule is used for:
Character recognition is carried out at least one described target text row region, obtains at least one described target text row area
The word content in each target text row region in domain.
In one possible implementation, the screening module is used for:
The size that horizontal direction is removed from least one described line of text region is less than the line of text region of second threshold,
Obtain at least one target text row region；And/or
The ratio that the size of horizontal direction and the size of vertical direction are removed from least one described line of text region is small
In the line of text region of third threshold value, at least one target text row region is obtained.
In one possible implementation, the character recognition submodule includes:
Feature extraction unit obtains the line of text region for carrying out feature extraction processing to the line of text region
Characteristic pattern；
Decoding unit obtains sequence label for being decoded processing to the characteristic pattern, wherein the sequence label
Including at least one label, the corresponding character of each label；
Obtaining unit obtains the character content in the line of text region for being based on the sequence label.
In one possible implementation, the length of the sequence label is corresponding with the width of the characteristic pattern.
In one possible implementation, the obtaining unit includes:
Divide subelement, for based in the sequence label correspond to space character at least one label, by the mark
Label sequences segmentation is at least two subsequences；
Subelement is determined, for determining institute based on the label for including in each subsequence at least two subsequence
State the corresponding character content of each subsequence；
Subelement is connected, connects at least two subsequence for putting in order based at least two subsequence
Corresponding character content obtains the character content in the line of text region.
In one possible implementation, the determining subelement is used for:
At least two adjacent label identical in the subsequence is merged, the sub- sequence after obtaining merging treatment
Column；
Based on the label for including in the subsequence after merging treatment, the corresponding character content of the subsequence is determined.
In one possible implementation, described device further include:
Second determining module, for determining described to be identified based on the character content at least one described character zone
The corresponding ticket templates data of image.
In one possible implementation, first determining module includes:
Third determines submodule, for being based on the corresponding ticket templates data of the images to be recognized, determines described wait know
The classification for the information that other image includes and position；
4th determines submodule, the classification of the information for including based on the images to be recognized and position, determine described in
Classification belonging to character content in images to be recognized；
Second merges submodule, for merging to belonging to same category of character content, obtains the bank slip recognition
As a result.
According to another aspect of the present disclosure, a kind of electronic equipment is provided, comprising: processor；It can for storage processor
The memory executed instruction；Wherein, the processor is configured to executing the above method.
According to another aspect of the present disclosure, a kind of computer readable storage medium is provided, computer journey is stored thereon with
Sequence instruction, wherein the computer program instructions realize the above method when being executed by processor.
In the disclosure in some terms, the bill is motor vehicle invoice.
For the bank slip recognition method and device of all aspects of this disclosure by carrying out character machining to images to be recognized, determining should
At least one character zone in images to be recognized, at least one character zone carry out character recognition, determine this at least one
Character content in a character zone is based on the corresponding ticket templates data of the images to be recognized and at least one character area
Character content in domain obtains bank slip recognition as a result, thus, it is possible to all character contents in automatic identification bill, without artificial
Ticket processing and examination are carried out, bank slip recognition efficiency and accuracy are substantially increased.
According to below with reference to the accompanying drawings to detailed description of illustrative embodiments, the other feature and aspect of the disclosure will become
It is clear.
Detailed description of the invention
Comprising in the description and constituting the attached drawing of part of specification and specification together illustrates the disclosure
Exemplary embodiment, feature and aspect, and for explaining the principles of this disclosure.
Fig. 1 shows the flow chart of the bank slip recognition method according to one embodiment of the disclosure.
Fig. 2 shows the illustrative flow charts according to the bank slip recognition method of one embodiment of the disclosure.
Fig. 3 shows an illustrative stream of the positive images to be recognized of bank slip recognition method lieutenant colonel according to one embodiment of the disclosure
Cheng Tu.
Fig. 4 shows the schematic diagram of the image in the presence of inclination or distortion.
Fig. 5 shows the schematic diagram of the image after correction process.
Fig. 6 shows in the bank slip recognition method according to one embodiment of the disclosure and carries out character at least one character zone
One illustrative flow chart of identification.
Fig. 7 shows in the bank slip recognition method according to one embodiment of the disclosure and carries out word at least one line of text region
Accord with an illustrative flow chart of identification.
Fig. 8 shows in the bank slip recognition method according to one embodiment of the disclosure and obtains line of text region based on sequence label
The illustrative flow chart of the one of character content.
Fig. 9 is shown in the bank slip recognition method according to one embodiment of the disclosure based on every sub- sequence at least two subsequences
The label for including in column determines an illustrative flow chart of the corresponding character content of each subsequence.
Figure 10 is shown in the bank slip recognition method according to one embodiment of the disclosure based on the corresponding bill mould of images to be recognized
Character content in plate data and at least one character zone obtains an illustrative flow chart of bank slip recognition result.
Figure 11 shows the block diagram of the bank slip recognition device according to one embodiment of the disclosure.
Figure 12 shows an illustrative block diagram of the bank slip recognition device according to one embodiment of the disclosure.
Figure 13 is a kind of block diagram of device 800 for bank slip recognition shown according to an exemplary embodiment.
Figure 14 is a kind of block diagram of device 1900 for bank slip recognition shown according to an exemplary embodiment.
Specific embodiment
Various exemplary embodiments, feature and the aspect of the disclosure are described in detail below with reference to attached drawing.It is identical in attached drawing
Appended drawing reference indicate element functionally identical or similar.Although the various aspects of embodiment are shown in the attached drawings, remove
It non-specifically points out, it is not necessary to attached drawing drawn to scale.
Dedicated word " exemplary " means " being used as example, embodiment or illustrative " herein.Here as " exemplary "
Illustrated any embodiment should not necessarily be construed as preferred or advantageous over other embodiments.
In addition, giving numerous details in specific embodiment below to better illustrate the disclosure.
It will be appreciated by those skilled in the art that without certain details, the disclosure equally be can be implemented.In some instances, for
Method, means, element and circuit well known to those skilled in the art are not described in detail, in order to highlight the purport of the disclosure.
It should be understood that the embodiment of the present disclosure can be applied to the identification of various bills, for example, motor vehicle invoice, increment invoice,
Driver's license etc., the embodiment of the present disclosure does not limit this.
Fig. 1 shows the flow chart of the bank slip recognition method according to one embodiment of the disclosure.This method can be applied to mobile phone,
In the terminal devices such as PC (Personal Computer, personal computer) or tablet computer, server also can be applied to
In, it is not limited thereto.Bill in the present embodiment can be invoice, the bill of lading, draft or check etc., be not limited thereto.
For example, bill can sell uniform invoice for motor vehicle.As shown in Figure 1, the method comprising the steps of S11 to step S13.
In step s 11, character machining is carried out to images to be recognized, determines at least one character area in images to be recognized
Domain.
In the embodiments of the present disclosure, images to be recognized can be any image for needing to carry out bank slip recognition.Figure to be identified
It include document field as in.For example, the partial region in images to be recognized belongs to document field；For another example, in images to be recognized
All areas belong to document field.
In one possible implementation, before carrying out character machining to images to be recognized, this method can also be wrapped
It includes: obtaining images to be recognized.For example, images to be recognized can be obtained by taking pictures to bill.
In one possible implementation, character machining can be carried out to images to be recognized using deep learning algorithm,
Determine at least one character zone in images to be recognized.
As an example of the implementation, CTPN (Connectionist Text Proposal can be based on
Network, connect text candidates network) structure to images to be recognized carry out character machining, determine in images to be recognized at least
One character zone.It is mentioned for example, feature can be carried out to images to be recognized using miniature neural network (such as ShuffleNet)
It takes, and can determine at least one character zone in images to be recognized according to the feature of images to be recognized.Wherein, character zone
Shape can be perpendicular strip, that is, the height of character zone can be greater than width.
In one possible implementation, character machining is carried out to images to be recognized, determined in images to be recognized extremely
A few character zone, may include: the document field in determining images to be recognized；Character machining is carried out to document field, really
Determine at least one character zone in document field.
As an example of the implementation, object inspection can be carried out to images to be recognized using deep learning algorithm
It surveys, determines the document field in images to be recognized.For example, the feature of images to be recognized can be extracted, according to images to be recognized
Feature determines the candidate region in images to be recognized, determines that candidate region is the probability of document field, is ticket according to candidate region
According to the probability in region, area-of-interest is determined from candidate region, and returned based on area-of-interest, determine figure to be identified
Document field as in.In this example, before the feature for extracting images to be recognized, images to be recognized can be converted to finger
Scale cun, and then the images to be recognized input feature vector of specified size is extracted into network (Raw Feature Extraction
Network in), the feature of images to be recognized is exported.In this example, can be according to the feature of images to be recognized, it will be to be identified
Object region in image is used as candidate region, and each candidate region can be inputted candidate region network respectively
In (Region Proposal Network, RPN), exporting each candidate region by candidate region network is the general of document field
Rate.In this example, the candidate region in each candidate region for the maximum probability of document field can be determined as interested
Region, and can using Fast RCNN (Fast Regions with Convolutional Neural Network, quickly
Region convolutional neural networks) area-of-interest is returned, determine the document field in images to be recognized.
As an example of the implementation, character machining can be carried out to document field using degree learning algorithm, really
Determine at least one character zone in document field.In this example, character can be carried out to document field based on CTPN structure
Detection.For example, feature extraction can be carried out to document field using miniature neural network (such as ShuffleNet), and can root
The character zone in document field is determined according to the feature of document field.
In step s 12, character recognition is carried out at least one character zone, determined at least one character zone
Character content.
In one possible implementation, character machining is carried out to images to be recognized, determined in images to be recognized extremely
A few character zone, comprising: identify that image carries out character machining by first nerves network handles, determine in images to be recognized
At least one character zone.For example, first nerves network can be the neural network based on deep learning algorithm.
It, can be based on CNN (Convolutional Neural Network, volume in alternatively possible implementation
Product neural network) and CTC (Connectionist Temporal Classification, the classification of connection timing) method, to this
At least one character zone carries out character recognition, determines the character content at least one character zone.
In step s 13, based in the corresponding ticket templates data of images to be recognized and at least one character zone
Character content, obtain bank slip recognition result.
In one possible implementation, this method further include: based in the character at least one character zone
Hold, determines the corresponding ticket templates data of images to be recognized.For example, if character content includes " motor vehicle sale uniform invoice ",
It can then determine that the corresponding ticket templates data of images to be recognized are the ticket templates data that motor vehicle sells uniform invoice.Example
Such as, the ticket templates data of motor vehicle sale uniform invoice include that " machine plays code " " machine marking code " " identification number " " machine beats generation
Code " " machine marking code " " identification number " " purchaser's title " " ID card No. " " organization mechanism code " " purchaser's title " " body
Part card number " name fields such as " organization mechanism code " " invoice codes " " invoice number " and " type of vehicle ".
In one possible implementation, each format in the corresponding ticket templates data of available images to be recognized
Character content at least one character zone is matched with regular expression, obtains ticket by the regular expression of information
According to recognition result.
The present embodiment determines at least one character in the images to be recognized by carrying out character machining to images to be recognized
Region carries out character recognition at least one character zone, determines the character content at least one character zone, be based on
Character content in the corresponding ticket templates data of the images to be recognized and at least one character zone, obtains bank slip recognition
As a result, carrying out ticket processing and examination thus, it is possible to all character contents in automatic identification bill without artificial, greatly improving
Bank slip recognition efficiency and accuracy.
Fig. 2 shows the illustrative flow charts according to the bank slip recognition method of one embodiment of the disclosure.As shown in Fig. 2,
This method may include step S21 to step S24.
In the step s 21, in the case where images to be recognized has inclination or distortion, images to be recognized is corrected, school is obtained
Images to be recognized that just treated.
In this example, if images to be recognized has inclination or distortion, images to be recognized can be corrected, by inclination or is turned round
Bent images to be recognized is become a full member, and making the character in images to be recognized is horizontal direction.The example does not require in images to be recognized
Bill put just completely, to enhance the robustness of bank slip recognition.
In step S22, to after correction process images to be recognized carry out character machining, after determining correction process wait know
At least one character zone in other image.
In step S23, character recognition is carried out at least one character zone, is determined at least one character zone
Character content.
Wherein, the description to step S12 is seen above to step S23.
In step s 24, based in the corresponding ticket templates data of images to be recognized and at least one character zone
Character content, obtain bank slip recognition result.
Wherein, the description to step S13 is seen above to step S24.
Fig. 3 shows the positive images to be recognized of step S21 lieutenant colonel in the bank slip recognition method according to one embodiment of the disclosure, obtains
The illustrative flow chart of one of images to be recognized after correction process.As shown in figure 3, correction images to be recognized, obtains at correction
Images to be recognized after reason may include step S211 to step S213.
In step S211, the representative points coordinate of images to be recognized is determined.
Wherein, after the representative points coordinate representation correction process of images to be recognized images to be recognized apex coordinate.
It in one possible implementation, can be using pre- in the case where images to be recognized has inclination or distortion
Survey the representative points coordinate of model prediction images to be recognized.
In one possible implementation, in training prediction model, positive training image can be carried out random
Disturbance, and can be according to the apex coordinate before the random perturbation of training image and the apex coordinate after random perturbation, training prediction
Model.
In step S212, sat according to the representative points coordinate of images to be recognized and the initial vertax of images to be recognized
Mark, determines the corresponding projection matrix of images to be recognized.
In step S213, according to projection matrix to images to be recognized carry out projective transformation, after obtaining correction process to
Identify image.
In the embodiments of the present disclosure, pixel in images to be recognized after correction process can be determined according to projection matrix with
The corresponding relationship between the pixel in images to be recognized before correction process, so as to determine the figure to be identified after correction process
The pixel value of each pixel as in.
Fig. 4 shows the schematic diagram of the image in the presence of inclination or distortion.Fig. 5 shows the schematic diagram of the image after correction process.
Similarly with Fig. 4 and Fig. 5, projective transformation is carried out to the images to be recognized that there is inclination or distortion, after available correction process
Images to be recognized.
Fig. 6 shows an illustrative flow chart of the bank slip recognition method and step S12 according to one embodiment of the disclosure.Such as figure
Shown in 6, step S12 may include step S121 and step S122.
In step S121, at least two character zones for being less than first threshold to the distance of horizontal direction are merged,
Obtain at least one line of text region.
In the embodiments of the present disclosure, horizontal direction indicates the direction parallel with the presentation direction of character.For example, if character
Presentation direction is that laterally (such as from left to right or from right to left), then horizontal direction indicates laterally；If the presentation direction of character
For longitudinally (such as from top to bottom), then horizontal direction indicates longitudinal.
It in the embodiments of the present disclosure, can be true if the distance of the horizontal direction of two character zones is less than first threshold
The character content of fixed two character zones is related, therefore, can be less than at least the two of first threshold to the distance of horizontal direction
A character zone merges, and obtains line of text region.
In step S122, character recognition is carried out at least one line of text region, obtains at least one line of text
Character content in region.
In one possible implementation, before carrying out character recognition at least one line of text region, the party
Method further include: based on the size of each line of text region in the horizontal direction at least one line of text region, at least to this
One line of text region is screened, at least one target text row region is obtained.At least one line of text region is carried out
Character recognition obtains the character content at least one line of text region, comprising: at least one target text row region into
Line character identification, obtains the word content in each target text row region at least one target text row region.
The implementation obtains at least one target text row area by screening at least one line of text region
Domain, and character recognition only is carried out to target text row region, thus, it is possible to improve the efficiency of character recognition, avoid to obviously not wrapping
Line of text region containing character carries out character recognition.
As an example of the implementation, based on each line of text region at least one line of text region in water
Square upward size, screens at least one line of text region, obtains at least one target text row region, wraps
Include: from least one line of text region remove horizontal direction size be less than second threshold line of text region, obtain to
A few target text row region.In this example, if the size of the horizontal direction in a certain line of text region is less than second threshold,
It can then determine and not include character in this article current row region.Size by removing horizontal direction is less than the line of text of second threshold
Region can remove the obvious line of text region for not including character, improve the efficiency of bank slip recognition.
As another example of the implementation, existed based on each line of text region at least one line of text region
Size in horizontal direction screens at least one line of text region, obtains at least one target text row region, packet
Include: the ratio that the size of horizontal direction and the size of vertical direction are removed from least one line of text region is less than third threshold
The line of text region of value obtains at least one target text row region.In the embodiments of the present disclosure, vertical direction indicates and level
The vertical direction in direction.In this example, if the size of the size and vertical direction of the horizontal direction in a certain line of text region
Ratio is less than third threshold value, then can determine and not include character in this article current row region.By remove horizontal direction size with
The ratio of the size of vertical direction is less than the line of text region of third threshold value, can remove the obvious line of text area for not including character
The efficiency of bank slip recognition is improved in domain.For example, third threshold value is equal to 0.5.
Fig. 7 shows an illustrative flow chart of the bank slip recognition method and step S122 according to one embodiment of the disclosure.Such as
Shown in Fig. 7, step S122 may include step S1221 to step S1223.
In step S1221, feature extraction processing is carried out to line of text region, obtains the characteristic pattern in line of text region.
In one possible implementation, line of text region can be carried out at feature extraction by nervus opticus network
Reason, obtains the characteristic pattern in line of text region.For example, pondization operation can be carried out by nervus opticus network, height is obtained as 1
Characteristic pattern.
As an example of the implementation, it is 32, width not less than 32 that line of text region can be adjusted to height
Input picture, input nervus opticus network.Nervus opticus network can successively be handled former height 32 by 4 pondization operations
It is 16,8,4,2,1 will highly be become by being finally filled with the convolutional layer that 0, convolution kernel is 2 using one, obtain height thus as 1
Characteristic pattern.
In one possible implementation, the width of the characteristic pattern in line of text region is equal to the width in line of text region.
In step S1222, processing is decoded to characteristic pattern, obtains sequence label, wherein sequence label includes at least
One label, the corresponding character of each label.
In one possible implementation, can be by characteristic pattern transposition, to channel, this dimension does full connection, by channel
Number is mapped as 5000 dimensions or so, then is decoded by CTC, determines the corresponding label of each pixel of characteristic pattern, so that it is determined that
Character content in line of text region.Wherein, CTC can be first using index normalization (Softmax) layer or other activation primitives
Characteristic pattern is normalized in layer, obtains probability distribution matrix, wherein and the line number of matrix is equal to the port number connected entirely,
Matrix column number is equal to the width of characteristic pattern, and the sum of probability of each column is equal to 1, and each of matrix element representation is corresponding
Location of pixels is the probability of each character.The serial number of the maximum value of each column can be obtained as the corresponding mark of the location of pixels
Label, to obtain length sequence label corresponding with the width of characteristic pattern.
In one possible implementation, the length of sequence label and the width of characteristic pattern are corresponding.In the realization side
In formula, the length value of sequence label is equal to the width value of characteristic pattern.For example, the width of characteristic pattern is 10 pixels, then sequence label
Including 10 labels, but the embodiment of the present disclosure is without being limited thereto.
In step S1223, it is based on sequence label, obtains the character content in line of text region.
In the embodiments of the present disclosure, the label for including according to the corresponding relationship and sequence label of label and character, can be with
Obtain the character content in line of text region.
Fig. 8 shows an illustrative flow chart of the bank slip recognition method and step S1223 according to one embodiment of the disclosure.Such as
Shown in Fig. 8, step S1223 may include step S12231 to step S12233.
In step S12231, based at least one label for corresponding to space character in sequence label, by sequence label point
It is segmented at least two subsequences.
In bill, space character is usually indicated to be spaced or be distinguished.The example corresponds to space by being based in sequence label
At least one label of symbol, is divided at least two subsequences for sequence label, can make not including space in each subsequence
Symbol.
In step S12232, based on the label for including in each subsequence at least two subsequences, every height is determined
The corresponding character content of sequence.
In the embodiments of the present disclosure, the label contained according to the corresponding relationship and sub-series of packets of label and character, can obtain
Obtain the corresponding character content of subsequence.
In step S12233, the connection corresponding word of at least two subsequences that puts in order based at least two subsequences
Content is accorded with, the character content in line of text region is obtained.
Fig. 9 shows an illustrative flow chart of the bank slip recognition method and step S12232 according to one embodiment of the disclosure.
As shown in figure 9, step S12232 may include step S122321 and step S122322.
In step S122321, at least two adjacent label identical in subsequence is merged, merging treatment is obtained
Subsequence afterwards.
In the embodiments of the present disclosure, if at least two adjacent labels are identical, which is closed
And thus, it is possible to avoid the same character in images to be recognized from corresponding to multiple labels in subsequence, so as to improve ticket
According to the accuracy of identification.
In step S122322, based on the label for including in the subsequence after merging treatment, the corresponding word of subsequence is determined
Accord with content.
In the embodiments of the present disclosure, include according in the subsequence after the corresponding relationship and merging treatment of label and character
Label, the corresponding character content of subsequence after can determining merging treatment.
Figure 10 shows an illustrative flow chart of the bank slip recognition method and step S13 according to one embodiment of the disclosure.Such as
Shown in Figure 10, step S13 may include step S131 to step S133.
In step S131, the corresponding ticket templates data of images to be recognized are based on, determine the letter that images to be recognized includes
The classification of breath and position.
In step S132, the classification for the information for including based on images to be recognized and position are determined in images to be recognized
Classification belonging to character content.
For example, character content " invoice codes " and the character content " 11100XXXXXXX " of " invoice codes " right belong to
Invoice codes classification.
In step S133, is merged to same category of character content is belonged to, obtain bank slip recognition result.
For example, character content " invoice codes " and the character content " 11100XXXXXXX " of " invoice codes " right belong to together
One classification can then carry out the character content " 11100XXXXXXX " of character content " invoice codes " and " invoice codes " right
Merge, obtains one in bank slip recognition result.
In one possible implementation, sub-category display can be carried out in bank slip recognition result.
It in one possible implementation, can be by bank slip recognition result and official after obtaining bank slip recognition result
Number formulary verifies the authenticity of bill according to comparing.
In one possible implementation, it can use bank slip recognition result and establish database, for each data management
System carries out large-scale search comparison.
Figure 11 shows the block diagram of the bank slip recognition device according to one embodiment of the disclosure.As shown in figure 11, which includes:
Character machining module 21 determines at least one character area in images to be recognized for carrying out character machining to images to be recognized
Domain；Character recognition module 22 determines at least one character zone for carrying out character recognition at least one character zone
In character content；First determining module 23, for based on the corresponding ticket templates data of images to be recognized and this at least one
Character content in a character zone obtains bank slip recognition result.
Figure 12 shows an illustrative block diagram of the bank slip recognition device according to one embodiment of the disclosure.It is as shown in figure 12:
In one possible implementation, device further include: correction module 24 inclines for existing in images to be recognized
Tiltedly or in the case where distortion, images to be recognized, the images to be recognized after obtaining correction process are corrected；Character machining module 21 is used
In: character machining is carried out to the images to be recognized after correction process.
In one possible implementation, correction module 24 includes: the first determining submodule 241, for determining wait know
The representative points coordinate of other image；Second determines submodule 242, for according to the representative points coordinate of images to be recognized and to
The initial vertax coordinate for identifying image, determines the corresponding projection matrix of images to be recognized；Correction module 243, for according to throwing
Shadow matrix carries out projective transformation to images to be recognized, the images to be recognized after obtaining correction process.
In one possible implementation, character machining module 21 is used for: being identified and is schemed by first nerves network handles
As carrying out character machining, at least one character zone in images to be recognized is determined.
In one possible implementation, character recognition module 22 includes: the first merging submodule 221, for water
Square to distance be less than first threshold at least two character zones merge, obtain at least one line of text region；Word
Symbol identification submodule 222 obtains at least one line of text area for carrying out character recognition at least one line of text region
Character content in domain.
In one possible implementation, the device further include: screening module 25, for being based at least one text
The size of each line of text region in the horizontal direction in row region is screened at least one line of text region, is obtained
At least one target text row region；Character recognition submodule 222 is used for: carrying out character at least one target text row region
Identification, obtains the word content in each target text row region at least one target text row region.
In one possible implementation, screening module 25 is used for: going to remove water from least one line of text region
Square to size be less than second threshold line of text region, obtain at least one target text row region；And/or from this to
The ratio of the size of the size and vertical direction of removal horizontal direction is less than the text of third threshold value in a few line of text region
Row region obtains at least one target text row region.
In one possible implementation, character recognition submodule 222 includes: feature extraction unit, for text
Row region carries out feature extraction processing, obtains the characteristic pattern in line of text region；Decoding unit, for being decoded place to characteristic pattern
Reason, obtains sequence label, wherein sequence label includes at least one label, the corresponding character of each label；Obtaining unit,
For being based on sequence label, the character content in line of text region is obtained.
In one possible implementation, the length of sequence label and the width of characteristic pattern are corresponding.
In one possible implementation, obtaining unit includes: segmentation subelement, for based on corresponding in sequence label
In at least one label of space character, sequence label is divided at least two subsequences；Subelement is determined, for based at least
The label for including in each subsequence in two subsequences determines the corresponding character content of each subsequence；Subelement is connected, is used
In the connection corresponding character content of at least two subsequences that puts in order based at least two subsequences, line of text region is obtained
Character content.
In one possible implementation, determine that subelement is used for: by least two adjacent mark identical in subsequence
Label merge, the subsequence after obtaining merging treatment；Based on the label for including in the subsequence after merging treatment, sub- sequence is determined
Arrange corresponding character content.
In one possible implementation, device further include: the second determining module 26 is used for based on this at least one
Character content in character zone determines the corresponding ticket templates data of images to be recognized.
In one possible implementation, the first determining module 23 includes: that third determines submodule 231, for being based on
The corresponding ticket templates data of images to be recognized, determine classification and the position of the information that images to be recognized includes；4th determines son
Module 232, the classification of the information for including based on images to be recognized and position determine the character content institute in images to be recognized
The classification of category；Second merges submodule 233, for merging to belonging to same category of character content, obtains bank slip recognition
As a result.
The present embodiment determines at least one character in the images to be recognized by carrying out character machining to images to be recognized
Region carries out character recognition at least one character zone, determines the character content at least one character zone, be based on
Character content in the corresponding ticket templates data of the images to be recognized and at least one character zone, obtains bank slip recognition
As a result, carrying out ticket processing and examination thus, it is possible to all character contents in automatic identification bill without artificial, greatly improving
Bank slip recognition efficiency and accuracy.
Figure 13 is a kind of block diagram of device 800 for bank slip recognition shown according to an exemplary embodiment.For example, dress
Setting 800 can be mobile phone, computer, digital broadcasting terminal, messaging device, game console, tablet device, medical treatment
Equipment, body-building equipment, personal digital assistant etc..
Referring to Fig.1 3, device 800 may include following one or more components: processing component 802, memory 804, power supply
Component 806, multimedia component 808, audio component 810, the interface 812 of input/output (I/O), sensor module 814, and
Communication component 816.
The integrated operation of the usual control device 800 of processing component 802, such as with display, telephone call, data communication, phase
Machine operation and record operate associated operation.Processing component 802 may include that one or more processors 820 refer to execute
It enables, to perform all or part of the steps of the methods described above.In addition, processing component 802 may include one or more modules, just
Interaction between processing component 802 and other assemblies.For example, processing component 802 may include multi-media module, it is more to facilitate
Interaction between media component 808 and processing component 802.
Memory 804 is configured as storing various types of data to support the operation in device 800.These data are shown
Example includes the instruction of any application or method for operating on device 800, contact data, and telephone book data disappears
Breath, picture, video etc..Memory 804 can be by any kind of volatibility or non-volatile memory device or their group
It closes and realizes, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM) is erasable to compile
Journey read-only memory (EPROM), programmable read only memory (PROM), read-only memory (ROM), magnetic memory, flash
Device, disk or CD.
Power supply module 806 provides electric power for the various assemblies of device 800.Power supply module 806 may include power management system
System, one or more power supplys and other with for device 800 generate, manage, and distribute the associated component of electric power.
Multimedia component 808 includes the screen of one output interface of offer between described device 800 and user.One
In a little embodiments, screen may include liquid crystal display (LCD) and touch panel (TP).If screen includes touch panel, screen
Curtain may be implemented as touch screen, to receive input signal from the user.Touch panel includes one or more touch sensings
Device is to sense the gesture on touch, slide, and touch panel.The touch sensor can not only sense touch or sliding action
Boundary, but also detect duration and pressure associated with the touch or slide operation.In some embodiments, more matchmakers
Body component 808 includes a front camera and/or rear camera.When device 800 is in operation mode, such as screening-mode or
When video mode, front camera and/or rear camera can receive external multi-medium data.Each front camera and
Rear camera can be a fixed optical lens system or have focusing and optical zoom capabilities.
Audio component 810 is configured as output and/or input audio signal.For example, audio component 810 includes a Mike
Wind (MIC), when device 800 is in operation mode, when such as call mode, recording mode, and voice recognition mode, microphone is matched
It is set to reception external audio signal.The received audio signal can be further stored in memory 804 or via communication set
Part 816 is sent.In some embodiments, audio component 810 further includes a loudspeaker, is used for output audio signal.
I/O interface 812 provides interface between processing component 802 and peripheral interface module, and above-mentioned peripheral interface module can
To be keyboard, click wheel, button etc..These buttons may include, but are not limited to: home button, volume button, start button and lock
Determine button.
Sensor module 814 includes one or more sensors, and the state for providing various aspects for device 800 is commented
Estimate.For example, sensor module 814 can detecte the state that opens/closes of device 800, and the relative positioning of component, for example, it is described
Component is the display and keypad of device 800, and sensor module 814 can be with 800 1 components of detection device 800 or device
Position change, the existence or non-existence that user contacts with device 800,800 orientation of device or acceleration/deceleration and device 800
Temperature change.Sensor module 814 may include proximity sensor, be configured to detect without any physical contact
Presence of nearby objects.Sensor module 814 can also include optical sensor, such as CMOS or ccd image sensor, at
As being used in application.In some embodiments, which can also include acceleration transducer, gyro sensors
Device, Magnetic Sensor, pressure sensor or temperature sensor.
Communication component 816 is configured to facilitate the communication of wired or wireless way between device 800 and other equipment.Device
800 can access the wireless network based on communication standard, such as WiFi, 2G or 3G or their combination.In an exemplary implementation
In example, communication component 816 receives broadcast singal or broadcast related information from external broadcasting management system via broadcast channel.
In one exemplary embodiment, the communication component 816 further includes near-field communication (NFC) module, to promote short range communication.Example
Such as, NFC module can be based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra wide band (UWB) technology,
Bluetooth (BT) technology and other technologies are realized.
In the exemplary embodiment, device 800 can be believed by one or more application specific integrated circuit (ASIC), number
Number processor (DSP), digital signal processing appts (DSPD), programmable logic device (PLD), field programmable gate array
(FPGA), controller, microcontroller, microprocessor or other electronic components are realized, for executing the above method.
In the exemplary embodiment, a kind of non-volatile computer readable storage medium storing program for executing is additionally provided, for example including calculating
The memory 804 of machine program instruction, above-mentioned computer program instructions can be executed above-mentioned to complete by the processor 820 of device 800
Method.
Figure 14 is a kind of block diagram of device 1900 for bank slip recognition shown according to an exemplary embodiment.For example,
Device 1900 may be provided as a server.Referring to Fig.1 4, it further comprises one that device 1900, which includes processing component 1922,
A or multiple processors and memory resource represented by a memory 1932, can be by processing component 1922 for storing
The instruction of execution, such as application program.The application program stored in memory 1932 may include one or more every
One corresponds to the module of one group of instruction.In addition, processing component 1922 is configured as executing instruction, to execute the above method.
Device 1900 can also include that a power supply module 1926 be configured as the power management of executive device 1900, and one
Wired or wireless network interface 1950 is configured as device 1900 being connected to network and input and output (I/O) interface
1958.Device 1900 can be operated based on the operating system for being stored in memory 1932, such as Windows ServerTM, Mac
OS XTM, UnixTM, LinuxTM, FreeBSDTM or similar.
In the exemplary embodiment, a kind of non-volatile computer readable storage medium storing program for executing is additionally provided, for example including calculating
The memory 1932 of machine program instruction, above-mentioned computer program instructions can be executed by the processing component 1922 of device 1900 to complete
The above method.
The disclosure can be system, method and/or computer program product.Computer program product may include computer
Readable storage medium storing program for executing, containing for making processor realize the computer-readable program instructions of various aspects of the disclosure.
Computer readable storage medium, which can be, can keep and store the tangible of the instruction used by instruction execution equipment
Equipment.Computer readable storage medium for example can be-- but it is not limited to-- storage device electric, magnetic storage apparatus, optical storage
Equipment, electric magnetic storage apparatus, semiconductor memory apparatus or above-mentioned any appropriate combination.Computer readable storage medium
More specific example (non exhaustive list) includes: portable computer diskette, hard disk, random access memory (RAM), read-only deposits
It is reservoir (ROM), erasable programmable read only memory (EPROM or flash memory), static random access memory (SRAM), portable
Compact disk read-only memory (CD-ROM), digital versatile disc (DVD), memory stick, floppy disk, mechanical coding equipment, for example thereon
It is stored with punch card or groove internal projection structure and the above-mentioned any appropriate combination of instruction.Calculating used herein above
Machine readable storage medium storing program for executing is not interpreted that instantaneous signal itself, the electromagnetic wave of such as radio wave or other Free propagations lead to
It crosses the electromagnetic wave (for example, the light pulse for passing through fiber optic cables) of waveguide or the propagation of other transmission mediums or is transmitted by electric wire
Electric signal.
Computer-readable program instructions as described herein can be downloaded to from computer readable storage medium it is each calculate/
Processing equipment, or outer computer or outer is downloaded to by network, such as internet, local area network, wide area network and/or wireless network
Portion stores equipment.Network may include copper transmission cable, optical fiber transmission, wireless transmission, router, firewall, interchanger, gateway
Computer and/or Edge Server.Adapter or network interface in each calculating/processing equipment are received from network to be counted
Calculation machine readable program instructions, and the computer-readable program instructions are forwarded, for the meter being stored in each calculating/processing equipment
In calculation machine readable storage medium storing program for executing.
Computer program instructions for executing disclosure operation can be assembly instruction, instruction set architecture (ISA) instructs,
Machine instruction, machine-dependent instructions, microcode, firmware instructions, condition setup data or with one or more programming languages
The source code or object code that any combination is write, the programming language include the programming language-of object-oriented such as
Smalltalk, C++ etc., and conventional procedural programming languages-such as " C " language or similar programming language.Computer
Readable program instructions can be executed fully on the user computer, partly execute on the user computer, be only as one
Vertical software package executes, part executes on the remote computer or completely in remote computer on the user computer for part
Or it is executed on server.In situations involving remote computers, remote computer can include by the network-of any kind
Local area network (LAN) or wide area network (WAN)-are connected to subscriber computer, or, it may be connected to outer computer (such as using
ISP is connected by internet).In some embodiments, by utilizing computer-readable program instructions
Status information carrys out personalized customization electronic circuit, such as programmable logic circuit, field programmable gate array (FPGA) or can compile
Journey logic array (PLA), the electronic circuit can execute computer-readable program instructions, to realize each side of the disclosure
Face.
Referring herein to according to the flow chart of the method, apparatus (system) of the embodiment of the present disclosure and computer program product and/
Or block diagram describes various aspects of the disclosure.It should be appreciated that flowchart and or block diagram each box and flow chart and/
Or in block diagram each box combination, can be realized by computer-readable program instructions.
These computer-readable program instructions can be supplied to general purpose computer, special purpose computer or other programmable datas
The processor of processing unit, so that a kind of machine is produced, so that these instructions are passing through computer or other programmable datas
When the processor of processing unit executes, function specified in one or more boxes in implementation flow chart and/or block diagram is produced
The device of energy/movement.These computer-readable program instructions can also be stored in a computer-readable storage medium, these refer to
It enables so that computer, programmable data processing unit and/or other equipment work in a specific way, thus, it is stored with instruction
Computer-readable medium then includes a manufacture comprising in one or more boxes in implementation flow chart and/or block diagram
The instruction of the various aspects of defined function action.
Computer-readable program instructions can also be loaded into computer, other programmable data processing units or other
In equipment, so that series of operation steps are executed in computer, other programmable data processing units or other equipment, to produce
Raw computer implemented process, so that executed in computer, other programmable data processing units or other equipment
Instruct function action specified in one or more boxes in implementation flow chart and/or block diagram.
The flow chart and block diagram in the drawings show system, method and the computer journeys according to multiple embodiments of the disclosure
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
One module of table, program segment or a part of instruction, the module, program segment or a part of instruction include one or more use
The executable instruction of the logic function as defined in realizing.In some implementations as replacements, function marked in the box
It can occur in a different order than that indicated in the drawings.For example, two continuous boxes can actually be held substantially in parallel
Row, they can also be executed in the opposite order sometimes, and this depends on the function involved.It is also noted that block diagram and/or
The combination of each box in flow chart and the box in block diagram and or flow chart, can the function as defined in executing or dynamic
The dedicated hardware based system made is realized, or can be realized using a combination of dedicated hardware and computer instructions.
The presently disclosed embodiments is described above, above description is exemplary, and non-exclusive, and
It is not limited to disclosed each embodiment.Without departing from the scope and spirit of illustrated each embodiment, for this skill
Many modifications and changes are obvious for the those of ordinary skill in art field.The selection of term used herein, purport
In the principle, practical application or technological improvement to the technology in market for best explaining each embodiment, or lead this technology
Other those of ordinary skill in domain can understand each embodiment disclosed herein.
Claims (10)
1. a kind of bank slip recognition method characterized by comprising
Character machining is carried out to images to be recognized, determines at least one character zone in the images to be recognized；
Character recognition is carried out at least one described character zone, determines the character content at least one described character zone；
Based on the character content in the corresponding ticket templates data of the images to be recognized and at least one described character zone,
Obtain bank slip recognition result.
2. the method according to claim 1, wherein described carry out character knowledge at least one described character zone
Not, the character content at least one described character zone is determined, comprising:
At least two character zones for being less than first threshold to the distance of horizontal direction merge, and obtain at least one line of text
Region；
Character recognition is carried out at least one described line of text region, is obtained in the character at least one described line of text region
Hold.
3. according to the method described in claim 2, it is characterized in that, carrying out character knowledge at least one described line of text region
Before not, the method also includes:
Based on the size of each line of text region in the horizontal direction at least one described line of text region, to described at least one
A line of text region is screened, at least one target text row region is obtained；
It is described that character recognition is carried out at least one described line of text region, obtain the word at least one described line of text region
Accord with content, comprising:
Character recognition is carried out at least one described target text row region, is obtained at least one described target text row region
The word content in each target text row region.
4. according to the method in claim 2 or 3, which is characterized in that described to be carried out at least one described line of text region
Character recognition obtains the character content at least one described line of text region, comprising:
Feature extraction processing is carried out to the line of text region, obtains the characteristic pattern in the line of text region；
Processing is decoded to the characteristic pattern, obtains sequence label, wherein the sequence label includes at least one label,
The corresponding character of each label；
Based on the sequence label, the character content in the line of text region is obtained.
5. a kind of bank slip recognition device characterized by comprising
Character machining module determines at least one of described images to be recognized for carrying out character machining to images to be recognized
Character zone；
Character recognition module determines at least one described character for carrying out character recognition at least one described character zone
Character content in region；
First determining module, for being based on the corresponding ticket templates data of the images to be recognized and at least one described character
Character content in region obtains bank slip recognition result.
6. device according to claim 5, which is characterized in that the character recognition module includes:
First merges submodule, and at least two character zones for being less than first threshold for the distance to horizontal direction close
And obtain at least one line of text region；
Character recognition submodule, for carrying out character recognition at least one described line of text region, obtain it is described at least one
Character content in line of text region.
7. device according to claim 6, which is characterized in that described device further include:
Screening module, for based on the ruler of each line of text region in the horizontal direction at least one described line of text region
It is very little, at least one described line of text region is screened, at least one target text row region is obtained；
The character recognition submodule is used for:
Character recognition is carried out at least one described target text row region, is obtained at least one described target text row region
The word content in each target text row region.
8. device according to claim 6 or 7, which is characterized in that the character recognition submodule includes:
Feature extraction unit obtains the spy in the line of text region for carrying out feature extraction processing to the line of text region
Sign figure；
Decoding unit obtains sequence label for being decoded processing to the characteristic pattern, wherein the sequence label includes
At least one label, the corresponding character of each label；
Obtaining unit obtains the character content in the line of text region for being based on the sequence label.
9. a kind of electronic equipment characterized by comprising
Processor；
Memory for storage processor executable instruction；
Wherein, the processor is configured to method described in any one of perform claim requirement 1 to 4.
10. a kind of computer readable storage medium, is stored thereon with computer program instructions, which is characterized in that the computer
Method described in any one of Claims 1-4 is realized when program instruction is executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810923451.0A CN109284750A (en) | 2018-08-14 | 2018-08-14 | Bank slip recognition method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810923451.0A CN109284750A (en) | 2018-08-14 | 2018-08-14 | Bank slip recognition method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109284750A true CN109284750A (en) | 2019-01-29 |
Family
ID=65183055
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810923451.0A Pending CN109284750A (en) | 2018-08-14 | 2018-08-14 | Bank slip recognition method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109284750A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110033000A (en) * | 2019-03-21 | 2019-07-19 | 华中科技大学 | A kind of text detection and recognition methods of bill images |
| CN110363190A (en) * | 2019-07-26 | 2019-10-22 | 中国工商银行股份有限公司 | A kind of character recognition method, device and equipment |
| CN110490193A (en) * | 2019-07-24 | 2019-11-22 | 西安网算数据科技有限公司 | Single Text RegionDetection method and ticket contents recognition methods |
| CN110598686A (en) * | 2019-09-17 | 2019-12-20 | 携程计算机技术（上海）有限公司 | Invoice identification method, system, electronic equipment and medium |
| CN111275880A (en) * | 2019-05-09 | 2020-06-12 | 杭州睿琪软件有限公司 | Bill identification method, device, equipment and storage medium |
| CN111814785A (en) * | 2020-06-11 | 2020-10-23 | 浙江大华技术股份有限公司 | Invoice recognition method, training method of related model, related equipment and device |
| CN112149654A (en) * | 2020-09-23 | 2020-12-29 | 四川长虹电器股份有限公司 | Invoice text information identification method based on deep learning |
| CN112200177A (en) * | 2020-07-21 | 2021-01-08 | 山东文多网络科技有限公司 | Single number identification method and device based on bill picking scanning piece big data |
| CN112507782A (en) * | 2020-10-22 | 2021-03-16 | 广东省电信规划设计院有限公司 | Text image recognition method and device |
| CN112560754A (en) * | 2020-12-23 | 2021-03-26 | 北京百度网讯科技有限公司 | Bill information acquisition method, device, equipment and storage medium |
| CN113065544A (en) * | 2020-01-02 | 2021-07-02 | 阿里巴巴集团控股有限公司 | Character recognition method and device and electronic equipment |
| CN113191448A (en) * | 2021-05-17 | 2021-07-30 | 广东电网有限责任公司 | Auditing method, device and equipment based on picture identification and storage medium |
| CN113780038A (en) * | 2020-06-10 | 2021-12-10 | 深信服科技股份有限公司 | Picture auditing method and device, computing equipment and storage medium |
| CN114363485A (en) * | 2021-12-10 | 2022-04-15 | 中电科思仪科技(安徽)有限公司 | FPGA-based binocular camera image stitching processing device and method |
| CN114493475A (en) * | 2021-12-09 | 2022-05-13 | 国网浙江省电力有限公司 | Intelligent image examination method based on text recognition and target detection |
| CN117727059A (en) * | 2024-02-18 | 2024-03-19 | 蓝色火焰科技成都有限公司 | Method and device for checking automobile financial invoice information, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1339775A (en) * | 2000-08-22 | 2002-03-13 | 英业达集团(上海)电子技术有限公司 | Automatic identifying method and system for name card |
| CN1630302A (en) * | 2003-12-17 | 2005-06-22 | 大唐移动通信设备有限公司 | Handset capable of automatically recording characters and images, and method of recording and processing thereof |
| CN101771752A (en) * | 2009-12-29 | 2010-07-07 | 中兴通讯股份有限公司 | Mobile phone TV text information extraction method and mobile terminal with same |
| CN107563377A (en) * | 2017-08-30 | 2018-01-09 | 江苏实达迪美数据处理有限公司 | It is a kind of to detect localization method using the certificate key area of edge and character area |
| CN108229469A (en) * | 2017-11-22 | 2018-06-29 | 北京市商汤科技开发有限公司 | Recognition methods, device, storage medium, program product and the electronic equipment of word |
| CN108229299A (en) * | 2017-10-31 | 2018-06-29 | 北京市商汤科技开发有限公司 | The recognition methods of certificate and device, electronic equipment, computer storage media |
-
2018
- 2018-08-14 CN CN201810923451.0A patent/CN109284750A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1339775A (en) * | 2000-08-22 | 2002-03-13 | 英业达集团(上海)电子技术有限公司 | Automatic identifying method and system for name card |
| CN1630302A (en) * | 2003-12-17 | 2005-06-22 | 大唐移动通信设备有限公司 | Handset capable of automatically recording characters and images, and method of recording and processing thereof |
| CN101771752A (en) * | 2009-12-29 | 2010-07-07 | 中兴通讯股份有限公司 | Mobile phone TV text information extraction method and mobile terminal with same |
| CN107563377A (en) * | 2017-08-30 | 2018-01-09 | 江苏实达迪美数据处理有限公司 | It is a kind of to detect localization method using the certificate key area of edge and character area |
| CN108229299A (en) * | 2017-10-31 | 2018-06-29 | 北京市商汤科技开发有限公司 | The recognition methods of certificate and device, electronic equipment, computer storage media |
| CN108229469A (en) * | 2017-11-22 | 2018-06-29 | 北京市商汤科技开发有限公司 | Recognition methods, device, storage medium, program product and the electronic equipment of word |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110033000A (en) * | 2019-03-21 | 2019-07-19 | 华中科技大学 | A kind of text detection and recognition methods of bill images |
| CN111275880A (en) * | 2019-05-09 | 2020-06-12 | 杭州睿琪软件有限公司 | Bill identification method, device, equipment and storage medium |
| CN111489487A (en) * | 2019-05-09 | 2020-08-04 | 杭州睿琪软件有限公司 | Bill identification method, device, equipment and storage medium |
| US12008828B2 (en) | 2019-05-09 | 2024-06-11 | Hangzhou Glority Software Limited | Receipt identification method, apparatus, electronic device and computer-readable storage medium |
| CN110490193A (en) * | 2019-07-24 | 2019-11-22 | 西安网算数据科技有限公司 | Single Text RegionDetection method and ticket contents recognition methods |
| CN110490193B (en) * | 2019-07-24 | 2022-11-08 | 西安网算数据科技有限公司 | Single character area detection method and bill content identification method |
| CN110363190A (en) * | 2019-07-26 | 2019-10-22 | 中国工商银行股份有限公司 | A kind of character recognition method, device and equipment |
| CN110598686A (en) * | 2019-09-17 | 2019-12-20 | 携程计算机技术（上海）有限公司 | Invoice identification method, system, electronic equipment and medium |
| CN110598686B (en) * | 2019-09-17 | 2023-08-04 | 携程计算机技术（上海）有限公司 | Invoice identification method, system, electronic equipment and medium |
| CN113065544B (en) * | 2020-01-02 | 2024-05-10 | 阿里巴巴集团控股有限公司 | Character recognition method and device and electronic equipment |
| CN113065544A (en) * | 2020-01-02 | 2021-07-02 | 阿里巴巴集团控股有限公司 | Character recognition method and device and electronic equipment |
| CN113780038A (en) * | 2020-06-10 | 2021-12-10 | 深信服科技股份有限公司 | Picture auditing method and device, computing equipment and storage medium |
| CN111814785A (en) * | 2020-06-11 | 2020-10-23 | 浙江大华技术股份有限公司 | Invoice recognition method, training method of related model, related equipment and device |
| CN111814785B (en) * | 2020-06-11 | 2024-03-29 | 浙江大华技术股份有限公司 | Invoice recognition method, training method of relevant model, relevant equipment and device |
| CN112200177A (en) * | 2020-07-21 | 2021-01-08 | 山东文多网络科技有限公司 | Single number identification method and device based on bill picking scanning piece big data |
| CN112149654B (en) * | 2020-09-23 | 2022-08-02 | 四川长虹电器股份有限公司 | Invoice text information identification method based on deep learning |
| CN112149654A (en) * | 2020-09-23 | 2020-12-29 | 四川长虹电器股份有限公司 | Invoice text information identification method based on deep learning |
| CN112507782A (en) * | 2020-10-22 | 2021-03-16 | 广东省电信规划设计院有限公司 | Text image recognition method and device |
| CN112560754A (en) * | 2020-12-23 | 2021-03-26 | 北京百度网讯科技有限公司 | Bill information acquisition method, device, equipment and storage medium |
| CN113191448A (en) * | 2021-05-17 | 2021-07-30 | 广东电网有限责任公司 | Auditing method, device and equipment based on picture identification and storage medium |
| CN114493475A (en) * | 2021-12-09 | 2022-05-13 | 国网浙江省电力有限公司 | Intelligent image examination method based on text recognition and target detection |
| CN114363485A (en) * | 2021-12-10 | 2022-04-15 | 中电科思仪科技(安徽)有限公司 | FPGA-based binocular camera image stitching processing device and method |
| CN114363485B (en) * | 2021-12-10 | 2024-04-16 | 中电科思仪科技(安徽)有限公司 | Binocular camera image stitching processing device and method based on FPGA |
| CN117727059A (en) * | 2024-02-18 | 2024-03-19 | 蓝色火焰科技成都有限公司 | Method and device for checking automobile financial invoice information, electronic equipment and storage medium |
| CN117727059B (en) * | 2024-02-18 | 2024-05-03 | 蓝色火焰科技成都有限公司 | Method and device for checking automobile financial invoice information, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109284750A (en) | Bank slip recognition method and device, electronic equipment and storage medium | |
| CN111476284B (en) | Image recognition model training and image recognition method and device and electronic equipment | |
| US10666784B2 (en) | Intuitive computing methods and systems | |
| CN112307472B (en) | Abnormal user identification method and device based on intelligent decision and computer equipment | |
| CN107909065A (en) | The method and device blocked for detecting face | |
| CN109800325A (en) | Video recommendation method, device and computer readable storage medium | |
| US20150347860A1 (en) | Systems And Methods For Character Sequence Recognition With No Explicit Segmentation | |
| CN110287874A (en) | Target tracking method and device, electronic equipment and storage medium | |
| CN109344832A (en) | Image processing method and device, electronic equipment and storage medium | |
| CN109345553B (en) | Palm and key point detection method and device thereof, and terminal equipment | |
| TW202113757A (en) | Target object matching method and apparatus, electronic device and storage medium | |
| US20110244919A1 (en) | Methods and Systems for Determining Image Processing Operations Relevant to Particular Imagery | |
| CN109040664A (en) | video stream processing method and device, electronic equipment and storage medium | |
| CN111475613A (en) | Case classification method and device, computer equipment and storage medium | |
| CN109801270A (en) | Anchor point determines method and device, electronic equipment and storage medium | |
| JP2013509654A (en) | Sensor-based mobile search, related methods and systems | |
| CN108287921A (en) | Information recommendation method and device | |
| CN114885613A (en) | Tagging objects in augmented reality to track object data | |
| CN108985176A (en) | image generating method and device | |
| CN109615006A (en) | Character recognition method and device, electronic equipment and storage medium | |
| CN109635142A (en) | Image-selecting method and device, electronic equipment and storage medium | |
| CN108596093A (en) | The localization method and device of human face characteristic point | |
| CN111046184A (en) | Text risk identification method, device, server and storage medium | |
| CN108446658A (en) | The method and apparatus of facial image for identification | |
| CN109726709A (en) | Icon-based programming method and apparatus based on convolutional neural networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190129 |