CN103617422A - A social relation management method based on business card recognition - Google Patents
A social relation management method based on business card recognition Download PDFInfo
- Publication number
- CN103617422A CN103617422A CN201310521182.2A CN201310521182A CN103617422A CN 103617422 A CN103617422 A CN 103617422A CN 201310521182 A CN201310521182 A CN 201310521182A CN 103617422 A CN103617422 A CN 103617422A
- Authority
- CN
- China
- Prior art keywords
- contact person
- address
- node
- user
- business card
- 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
Images
Landscapes
- Character Input (AREA)
Abstract
Provided is a social relation management method based on business card recognition. The social relation management method comprises four steps: a step 1 of entering business card information, performing image acquisition on the business card with a camera or a scanner; dividing character blocks according to the characteristics of the business card image and performing character recognition with an OCR engine; performing word-dividing processing according to a key field; classifying and entering extracted information; filling the extracted information in corresponding forms; and interacting with a user in order to perform artificial confirmation and adjustment on possibly existed incapably-recognized information, a step 2 of establishing a social relation network, a step 3 of achieving intelligent retrieval, and a step 4 completing mobile terminal synchronization.
Description
Technical field
The present invention relates to by card information Rapid input computing machine and by the management method of the formed social networks of business card.
Background technology
Economic fast development impels human communication day by day frequent, a slight business card, and what carrying is actually resource, is business opportunity, even can be described as benefit.
Although mobile phone at present, PDA (Personal Digital Assistant), panel computer, notebook even Desktop PC all can be installed corresponding name card management system, but it is the most at all, also sixty-four dollar question that the management method that these systems adopt could not solve one well, be social networks management---the complicated interpersonal relation being produced by a large amount of contact persons, the fit person who usually allows people find with the fastest speed to want cooperation or ask for help.
The method for managing name cards of main flow all adopts the mode of grouping to carry out business card classifying at present.For example, contact person is divided into groups by keywords such as " friend ", " client ", " colleague ", " leaders ".This grouping management mode is confined to traditional custom, not only not directly perceived, and cannot provide a suitable contact approach based on interpersonal relation, can only be for " people " and can not be for " thing ".In interpersonal relation, usually exist a kind of situation, someone had side of the edge, and was unfamiliar with, and the information such as its name or phone number or work unit are just known wherein clip information, such as only know he (she) in city one hospital work, but other information are difficult to remember, while having thousands of contact persons in the address list of " I ", are difficult to find this contact person, therefore, be necessary to provide more intelligentized method to manage.The friend of side of the edge usually occurs in certain feature occasion, or more close friend is relevant with own relation with one, or is introduced by this friend, therefore, by certain Intelligentize query, can search for by following the clues, finds this friend.Such as: user wants to go to a hospital to see a doctor now, wishes to relate to Yi Ge hospital.But in the conventional mode, he can only first locate a suitable grouping, another checks whether there is friend doctor, cannot realize fuzzy location.And more common situation is: user has friend's doctor business card, but not yet done with it, need one or more friends' recommendation.At this moment, traditional grouping management mode just cannot provide a suitable recommendation approach, and needs the user effort plenty of time to go thinking, arrangement.And artificial arrangement is incomplete often, the thinkable recommendation approach of institute is also not necessarily most suitable.This is also a large number of users problem of headache the most, is also current various business cards and the insurmountable problem of social networks management method.In addition, for the friend's circle of oneself and the intimate degree of oneself, by conventional business card and social networks management method, also cannot provide, therefore be difficult to carry out efficient doings.
End is got up, and emphasis of the present invention has solved the innovative problems of three aspects:: (1) has solved the Rapid input of card information based on computer vision technique and semantic segmenting method; (2) intelligence of social networks location and optimum recommendation approach; (3) intelligent synchronization between mobile terminal and system database.
Summary of the invention
The present invention has overcome the business card of current employing and the problems referred to above that social networks management method exists, realized the intelligence location of card information automatic input based on computer vision and social networks, saved the artificial time of considering of user, and a most suitable contact channel and recommendation approach can be provided, provide optimum, contact scheme the most reliably.
The technical solution adopted for the present invention to solve the technical problems comprises four steps:
Step 1. typing card information
Utilize camera or scanner to carry out business card image collection, according to the feature of business card image, mark off character block and utilize OCR engine to carry out character recognition, according to critical field, carry out word segmentation processing, to the information categorization typing of extracting, inserted corresponding list, last and user carries out alternately, the None-identified information that may exist being carried out to manual confirmation and adjustment.Described flow process should comprise following operation:
1.1, gray processing.First to the business card image of camera or scanner collection, adopt method of weighted mean to carry out gray processing:
f(i,j)=0.30R(i,j)+0.59G(i,j)+0.11B(i,j)
Wherein (i, j) represents pixel coordinate, and in coloured image, three components of red, green, blue are respectively R (i, j), G (i, j), B (i, j), and f (i, j) is the gray-scale value of this point.
1.2, edge extracting.In conjunction with Sobel operator and LOG (Laplacian of Gauss) operator, form and revise LOG algorithm, choose following Sobel operator (S
xfor the center of horizontal direction divide poor, S
yfor the center of vertical direction, divide poorly, choose wherein higher value as gradient S)
As the precondition of rim detection, reduce unnecessary zero cross point.Recycling, if minor function is as wave filter, is carried out LOG detection.
LOG operator template is as follows
The digital form that LOG operator template is wherein LOG, does convolution using it as interior collecting image, the mean square deviation that σ is Gaussian distribution.
1.3, binaryzation.Calculated threshold, and gray-scale map is carried out to binaryzation.If gray-scale map height is h, wide is w, by following formula, tries to achieve threshold value, i.e. average gray Threshold:
According to threshold value Threshold, carry out binaryzation again:
1.4, tilt detection and rectification.By Hough transformation (Hough Transform), detect the edge line of business card frame, correct at the angle of inclination that obtains name panel region and judge business card.
Because the straight-line equation of y=kx+b form cannot represent the straight line (c is constant, i.e. the straight line parallel with x axle, slope k → ∞) of x=c form.Therefore adopt parametric equation ρ=x*cos θ+y*sin θ here, wherein by choosing p
1(x
1, y
1), p
2(x
2, y
2) two monitoring points, by following formula, can obtain tiltangleθ:
According to tiltangleθ, former figure and bianry image are carried out to affined transformation (using homogeneous matrix representation) simultaneously, P'=PR, general-θ substitution 2D rotation matrix obtains correcting matrix
By point
After rectification, obtain.Concrete operation is as follows:
x'=xcosθ+ysinθ
y'=-xsinθ+ycosθ
1.5, image is cut apart.According to the feature of business card image, mark off character block.Comprise that step is as follows:
1.5.1, definition surveys density D ensity, for current pixel up and down and the clinodiagonal quantity of black picture element in the neighbor in totally 8 directions, computing formula is as follows:
I ∈ [x-1, x+1] ∩ N wherein
*, j ∈ [y-1, y+1] ∩ N
*, (x, y) is current sensing point coordinate, N
*for positive integer collection.
1.5.2, the image of name panel region is converted into density matrix, remove remaining noise.Method of operating is as follows: judge one by one the density of each pixel, when Density<2, homography element is designated as 0, is used as noise processed.When Density>=2, homography element is designated as 1, shows that this pixel is a character block part.
1.5.3, pass through conversion formula
Two-value achromatic sheet after correcting is converted into following two-dimensional array form:
1.5.4, according to density matrix, character block region, location (so far having two kinds of strategies to be applied to respectively quick recognition mode and accurate recognition mode), then cuts apart business card image according to region.
Quick recognition strategy: judge line by line density matrix, the ratio that in every a line, " 1 " element accounts for unit number, over certain threshold value, is considered as line of text, is considered as blank lower than this threshold value.
Accurate recognition strategy: judge line by line density matrix, continuous " 0 " element in row is connected to " detection line ", determine whether text filed according to the depth of " detection line " initiating terminal, difference in length and terminal position feature.Remove the detection line that length is less than threshold value; Then mark article one and survey the end of line and the top that second is surveyed line; To all row survey with mark after, the region being surrounded by mark is text filed.
1.5.5, definition R
i∈ d (i, y) | y ∈ [0, h] ∩ N
*,
and if only if
time record j value.One group of region [j of last gained
1, j
2], [j
2, j
3], [j
3, j
4] ... be some character blocks, wherein h represents line number, and w represents columns, N
*for positive integer collection, R
ifor the capable element of i in the two-dimensional array of step 3, Sum
jfor R
iinstitute's all elements sum of being expert at.
1.6, character recognition.Utilize OCR (Optical Character Recognition) technology to extract the information of each character block.Here adopt the OCR module of MODI (Microsoft Office Document Imaging) to identify one by one the image being partitioned into, each character block image is all become to one group of word.To character block [j
1, j
2], [j
2, j
3], [j
3, j
4] ... call one by one OCR identification engine, obtain the character set C corresponding with each region
1,2, C
2,3, C
3,4
1.7, word segmentation processing.According to critical field, carry out word segmentation processing, to the information categorization typing of extracting.This need to set up keywords database and semantic base, for example: company, position, address, telephone number, email, road, number etc.Definition of keywords set W={ company, position, address, telephone number, email ... and definition of keywords semantic base, such as Address, Addr, address, contact address etc. is the near synonym of address, semantically, be consistent, set up corresponding mapping relations, form semantic base.
According to keyword, based on contextual feature, extract the corresponding informance in every group of word, insert corresponding list, complete word segmentation processing.Concrete steps are as follows:
1.7.1, find ": " separator, utilize separator to define keyword and content.If not in set of keywords, transferring to user to determine whether to be regarded as keyword, the keyword defining is indexed in set.By this step, realize the learning functionality of participle strategy.
1.7.2, then match key word Key=C
x,y∩ W, by character string
insert in the form item of Key by name.When
time, by character set C
x,yinsert " unfiled " form item.
Strategy based on semantic base comprises:
Location name strategy (Chinese): get rid of other semantemes, number of characters, between 2~4, is followed by position, title, or with " name " class label;
Locating cellphone number strategy: with 11 pure digi-tal character strings of 13,15,18 beginnings, or with " mobile phone " class label;
Location telephone number strategy: 7~8 pure digi-tal character strings, or with area code beginning, comprise the separators such as bracket, hyphen, space, or with " phone " class label;
Location Business Name strategy: get rid of other semantemes, occur " ”,“ group of company ", each institutional settings keyword, or with " company " class label;
Positioning address strategy: there is the character string that mix with numeral " ”,“ district, ”,“ road, ”,“ town, ”,“ township, ”,“ county, province ”,“ city ”,“ Zhuan”,“ unit ", " chamber ", or with " address " class label;
Location postcode strategy: 6 pure digi-tal character strings, or with " postcode " class label;
Location mailbox strategy: there is numeral, letter, " ", ". " character, or with " mailbox " class label;
Location network address strategy: there is " http ", ". ", " www ", " com ", " cn ", " edu " character, or with " network address " class label.
1.8, manual synchronizing.After participle completes, need carry out alternately, the None-identified information that may exist being carried out to manual confirmation and adjustment with user, " unfiled " form item be carried out to manual sort, after assorting process finishes, automatic learning be entered to semantic base.
Step 2. is set up social networks network
Its step is as follows:
2.1, by move the relation that contact person's node is set up two contact persons with mouse.By the node A newly adding
jappend to existing node A
i(wanting the contact person of opening relationships) is upper, after definite operation both sides, utilize the form of line to represent the L that is related to setting up between two contact persons
i,j, it represents A
iwith A
jbe familiar with each other (two-way).
2.1.1(optional) if newly-established contact person's node A
ithe first contact people of user O (i.e. " I "), by new node A
ibe connected to the upper contact circuit OA that sets up of user's node O
i, its pass is L
0, i.
2.1.2(optional) if newly-established contact person's node A
jthat user O is by friend A
ithis contact person of cicada, and the other side is not familiar with in the situation of user O, by new node A
jbe connected to user friend's node A
iupper foundation contact circuit A
ia
j, its pass is L
i,j.
2.1.3(optional) if the former O of user before this passes through friend A
ithe contact person A learning
jwith user friend A
ithe first contact people (unidirectional), and after further cooperation and exchange, user O and this contact person A
jbecome the first contact people (two-way), by this contact person's node A
jbe connected to user's node O upper, form and the direct internuncial pathway OA of user's node
j, its pass is L
0, j.
2.2, for each, be related to L
i,jby using weights K
i,jthe power that represents relation.L
i,jweights K
i,jdepend on contact person A
iwith contact person A
jget to know time T (unit: day) and contact number of times Count.Circular is as follows:
2.3, according to number of degrees D and weights K, draw " cohesion " Intimacy of " I " and each contact person.The number of degrees D of agreement user node O
0=0 by path OA
ion be related to L
0, iconnected node A
inumber of degrees D
i=1+D
0; Pass through path A
ia
jon be related to L
i,jconnected node A
jnumber of degrees D
j=1+D
i." I " and node A
j" cohesion " computing formula be:
Intimacy
j=D
j+K
i,j,D
i∈N
*,K
i,j∈[0,1)
The integral part of " cohesion " Intimacy has represented the number of degrees in six degrees of separation, and fraction part is related to that by each under the equal number of degrees Further Division opens.Therefore, " cohesion " can reflect the close and distant relation that arrives each node centered by " I ", wherein N preferably
*for positive integer collection.
Step 3. realizes intelligent retrieval
The search key of inputting by user and the personal information of typing are carried out fuzzy matching, filter out possible object listing, and its concrete steps are as follows:
3.1, the keyword of establishing user input is character string WORD, and element is wherein Char, business card data collection DATA, if
and
return to its corresponding contact person A
(Char ∩ DATA).
3.2, user, from the list of returning, according to the actual requirements, selects objectives, navigates to a certain contact person A
i.
3.3, according to the weights K being related on L, by shortest path first (adopting dijkstra's algorithm here), obtain the communication approach of " the most effective " between " I " and target, by the most intimate (or minimum) contact person, relate to the contact person who needs most.
Here regard contact person's node A as point set in dijkstra's algorithm; Be related to that L regards the limit collection in algorithm as; The length on limit is corresponding K value; User's node O is as the source in algorithm.
Step 4. completes mobile terminal synchronization
Its step is as follows:
4.1, start service.PC server end requires to adopt the correlation technique design server programs such as Socket and multithreading, it can be resident service program, along with os starting, starts, also can start-up by hand, after startup, will connect with database server, then the request of AM automatic monitoring mobile terminal.
4.2, automatic configuration of IP.When user clicks the sync client program of mobile terminal, it will read server ip address, server end slogan PORT in the configuration file from mobile terminal; If cannot connect PC server end, it is by intelligent autoscan server.
Disposable IP scanning collocation strategy: terminal adds after the network segment of PC server place, its IP address of terminal program automatic acquisition, according to algorithm for text string location, obtain trizonal data before IP address, the IP address (as possible PC server IP address) that utilizes loop program autoscan to be formed by 0-255 and first three area data, if energy successful connection, be stored in configuration file, for automatic connection server next time.
4.3, system is set.The system setting function technical characterictic of mobile terminal is: when terminal program autoscan cannot connect upper PC server, can revise IP address, port numbers PORT by interactively, covering completely, difference when backup can also be set simultaneously and import cover two options.
4.4, backup mobile terminal.The backup functionality technical characterictic of mobile terminal is: when user clicks backup functionality button on mobile terminal, option in system arranges is for covering completely, program of mobile terminal read in the address list of this terminal i contact person (i=1,2 ... n), then i contact person is written in corresponding database by PC server, if there is this contact person in database, new data more, if do not exist, insert.When the option during if system arranges is difference, program of mobile terminal reads i contact person (i=1 in the address list of this terminal, 2,, n), then i contact person is written in corresponding database by PC server, if there is this contact person in database, ignore, if do not exist, insert.
4.5, import mobile terminal.The import feature technical characterictic of mobile terminal is: when user clicks import feature button on mobile terminal, option in system arranges is for covering completely, i contact person in PC server reading database address list (i=1,2 ... n), then i contact person is written in the address list of terminal by program of mobile terminal, if there is this contact person in address list, new data more, if do not exist, insert.If when the option in system arranges is difference, i contact person (i=1 in PC server reading database address list, 2,, n), then i contact person is written in the address list of terminal by program of mobile terminal, if there is this contact person in address list, ignore, if do not exist, insert.
Accompanying drawing explanation
Fig. 1 is business card typing process flow diagram.
When Fig. 2 is Density value, eight adjacent pixel location schematic diagram.
Fig. 3 is that Density value is given an example.
Fig. 4 is the density matrix that Fig. 3 changes.
Fig. 5 is business card image Processing Algorithm schematic diagram
Fig. 6 is card information word segmentation processing algorithm schematic diagram
Fig. 7 is a kind of social networks structural representation centered by " I ".
Fig. 8 is the concrete manifestation effect of result on social networks of retrieval.
Embodiment
Each process embodiment of the technical solution adopted in the present invention is as follows:
Step 1. typing card information
Utilize camera or scanner to carry out business card image collection, according to the feature of business card image, mark off character block and utilize OCR engine to carry out character recognition, according to critical field, carry out word segmentation processing, to the information categorization typing of extracting, inserted corresponding list, last and user carries out alternately, the None-identified information that may exist being carried out to manual confirmation and adjustment.
Whole process is exported from image to result, must through image input, image pre-treatment, character features extraction, matching identification, finally by manual synchronizing by the word correction of admitting one's mistake, deposit result in database, with reference to Fig. 1.
Described flow process should comprise following operation:
1.1, gray processing.First to the business card image of camera or scanner collection, adopt method of weighted mean to carry out gray processing.Because human eye is the highest to green sensitivity, to blue responsive minimum, therefore, use following formula can obtain more rational gray level image.
f(i,j)=0.30R(i,j)+0.59G(i,j)+0.11B(i,j)
Wherein (i, j) represents pixel coordinate, and in coloured image, three components of red, green, blue are respectively R (i, j), G (i, j), B (i, j), and f (i, j) is the gray-scale value of this point.
1.2, edge extracting.In conjunction with Sobel operator and LOG (Laplacian of Gauss) operator, form and revise LOG algorithm, choose following Sobel operator (S
xfor the center of horizontal direction divide poor, S
yfor the center of vertical direction, divide poorly, choose wherein higher value as gradient S)
As the precondition of rim detection, reduce unnecessary zero cross point.Recycling, if minor function is as wave filter, is carried out LOG detection.
LOG operator template is as follows
The digital form that LOG operator template is wherein LOG, does convolution using it as interior collecting image, the mean square deviation that σ is Gaussian distribution.
1.3, binaryzation.Calculated threshold, and gray-scale map is carried out to binaryzation.If gray-scale map height is h, wide is w, by following formula, tries to achieve threshold value, i.e. average gray Threshold:
According to threshold value Threshold, carry out binaryzation again:
1.4, tilt detection and rectification.By Hough transformation (Hough Transform), detect the edge line of business card frame, correct at the angle of inclination that obtains name panel region and judge business card.
Because the straight-line equation of y=kx+b form cannot represent the straight line (c is constant, i.e. the straight line parallel with x axle, slope k → ∞) of x=c form.Therefore adopt parametric equation ρ=x*cos θ+y*sin θ here, wherein by choosing p
1(x
1, y
1), p
2(x
2, y
2) two monitoring points, by following formula, can obtain tiltangleθ:
According to tiltangleθ, former figure and bianry image are carried out to affined transformation (using homogeneous matrix representation) simultaneously, P'=PR, general-θ substitution 2D rotation matrix obtains correcting matrix
By point
After rectification, obtain.Concrete operation is as follows:
x'=xcosθ+ysinθ
y'=-xsinθ+ycosθ
1.5, image is cut apart.According to the feature of business card image, mark off character block.Comprise that step is as follows:
1.5.1, definition surveys density D ensity, for current pixel up and down and the clinodiagonal quantity of black picture element in the neighbor in totally 8 directions, with reference to Fig. 2, computing formula is as follows:
I ∈ [x-1, x+1] ∩ N wherein
*, j ∈ [x-1, x+1] ∩ N
*, (x, y) is current sensing point coordinate, N
*for positive integer collection.
1.5.2, the image of name panel region is converted into density matrix, remove remaining noise.Method of operating is as follows: judge one by one the density of each pixel, when Density<2, homography element is designated as 0, is used as noise processed.When Density>=2, homography element is designated as 1, shows that this pixel is a character block part.In Fig. 3, the pixel Density=0 of upper right mark; The pixel Density=2 of upper left mark, it is converted into density matrix and sees Fig. 4.
1.5.3, pass through conversion formula
Two-value achromatic sheet after correcting is converted into following two-dimensional array form:
1.5.4, according to density matrix, character block region, location (this time have two kinds of strategies to differentiate to be applied to quick recognition mode with precisely recognition mode), then cuts apart business card image according to region.
Quick recognition strategy: judge line by line density matrix, the ratio that in every a line, " 1 " element accounts for unit number, over certain threshold value, is considered as line of text, is considered as blank lower than this threshold value.
Accurate recognition strategy: judge line by line density matrix, continuous " 0 " element in row is connected to " detection line ", determine whether text filed according to the depth of " detection line " initiating terminal, difference in length and terminal position feature.Remove the detection line that length is less than threshold value; Then mark article one and survey the end of line and the top that second is surveyed line; To all row survey with mark after, the region being surrounded by mark is text filed.
1.5.5, definition R
i∈ d (i, y) | y ∈ [0, h] ∩ N
*,
and if only if
time record j value.One group of region [j of last gained
1, j
2], [j
2, j
3], [j
3, j
4] ... be some character blocks, wherein h represents line number, and w represents columns, N
*for positive integer collection, R
ifor the capable element of i in the two-dimensional array of step 3, Sum
jfor R
iinstitute's all elements sum of being expert at.
Because Chinese business card mostly is horizontal typesetting, between character block, be in vertical distribution, so the only employing level of locating is cut apart character block.If run into the business card of complicated space of a whole page pattern, need to carry out vertical segmentation, can segment again vertical character block by same method.
1.6, character recognition.Utilize OCR (Optical Character Recognition) technology to extract the information of each character block.Here adopt the OCR module of MODI (Microsoft Office Document Imaging) to identify one by one the image being partitioned into, each character block image is all become to one group of word.
To character block [j
1, j
2], [j
2, j
3], [j
3, j
4] ... call one by one OCR identification engine, obtain the character set C corresponding with each region
1,2, C
2,3, C
3,4...
Business card image is processed and is so far finished, and Fig. 5 has sketched this processing procedure and algorithm used.
1.7, word segmentation processing.According to critical field, carry out word segmentation processing, to the information categorization typing of extracting.This need to set up keywords database and semantic base, for example: company, position, address, telephone number, email, road, number etc.Definition of keywords set W={ company, position, address, telephone number, email ... and definition of keywords semantic base, such as Address, Addr, address, contact address etc. is the near synonym of address, semantically, be consistent, set up corresponding mapping relations, form the semantic base with self-learning function.
Then according to keyword, based on contextual feature, extract the corresponding informance in every group of word, insert corresponding list, complete word segmentation processing.Concrete steps are as follows:
1.7.1, find ": " separator, utilize separator to define keyword and content.If not in set of keywords, transferring to user to determine whether to be regarded as keyword, the keyword defining is indexed in set.By this step, realize the self-learning function of participle strategy.
1.7.2, then match key word Key=C
x,y∩ W, by character string
insert in the form item of Key by name.When
time, by character set C
x,yinsert " unfiled " form item.
Strategy based on semantic base comprises:
Location name strategy (Chinese): get rid of other semantemes, number of characters, between 2~4, is followed by position, title, or with " name " class label;
Locating cellphone number strategy: with 11 pure digi-tal character strings of 13,15,18 beginnings, or with " mobile phone " class label;
Location telephone number strategy: 7~8 pure digi-tal character strings, or with area code beginning, comprise the separators such as bracket, hyphen, space, or with " phone " class label;
Location Business Name strategy: get rid of other semantemes, occur " ”,“ group of company ", each institutional settings keyword, or with " company " class label;
Positioning address strategy: there is the character string that mix with numeral " ”,“ district, ”,“ road, ”,“ town, ”,“ township, ”,“ county, province ”,“ city ”,“ Zhuan”,“ unit ", " chamber ", or with " address " class label;
Location postcode strategy: 6 pure digi-tal character strings, or with " postcode " class label;
Location mailbox strategy: there is numeral, letter, " ", ". " character, or with " mailbox " class label;
Location network address strategy: there is " http ", ". ", " www ", " com ", " cn ", " edu " character, or with " network address " class label;
1.8 manual synchronizing.After participle completes, need carry out alternately, the None-identified information that may exist being carried out to manual confirmation and adjustment with user, " unfiled " form item be carried out to manual sort, after assorting process finishes, automatic learning be entered to semantic base.
Card information word segmentation processing so far finishes, and Fig. 6 has sketched its processing procedure and logic.
Step 2. is set up social networks network
According to the degree of six in sociology, separate theory, set up a kind of social networks centered by " I ".
Being embodied in based on " I " is starting point O, is arachnoid to the form of surrounding radiation, sees schematic diagram 7.
Familiarity between contact person and " me ", as foundation, is directly attached at " I " around with " I " directly related contact person Ai by line, becomes with " I " the closest ground floor contact person.By that analogy, by the contact person of every one deck to the direction OAi divergence contrary with " I ".
In addition, the power of relation also can be determined by the number of degrees D of six degree theories.1 degree relation i.e. relation between " I " and friend that I am familiar with, and 2 degree relations are the relation between the friend that is familiar with of the friend of " I " and " I ", by that analogy, are understandable that: 1 degree relation is certainly strong than 2 degree relations.
At this, introduce the concept of " cohesion " Intimacy and express " I " and each contact person's familiarity.It depends on that weights K(has represented the power of relation) and the theoretical number of degrees D of six degree.Here, using contact person as node, and provide visual interactive tool to divide into groups with opening relationships network to contact person.The step of setting up this network is as follows:
2.1, by mobile contact person's node, set up two contact persons' relation.By the node A newly adding
jappend to existing node A
i(wanting the contact person of opening relationships) is upper, after definite operation both sides, utilize the form of line to represent the L that is related to setting up between two contact persons
i,j.The all relations that occur are herein as L
i,j, show A
iwith A
jbe familiar with each other (two-way).
2.1.1(optional) if newly-established contact person's node A
ithe first contact people of user O (i.e. " I "), by new node A
ibe connected to the upper contact circuit OA that sets up of user's node O
i, its pass is L
0, i.
2.1.2(optional) if newly-established contact person's node A
jthat user O is by friend A
ithis contact person of cicada, and the other side is not familiar with in the situation of user O, by new node A
jbe connected to user friend's node A
iupper foundation contact circuit A
ia
j, its pass is L
i,j.
2.1.3(optional) if the former O of user before this passes through friend A
ithe contact person A learning
jwith user friend A
ithe first contact people (unidirectional), and after further cooperation and exchange, user O and this contact person A
jbecome the first contact people (two-way), by this contact person's node A
jbe connected to user's node O upper, form and the direct internuncial pathway OA of user's node
j, its pass is L
0, j.
2.2, for each, be related to that L is by representing the power of relation with weights K.L
i,jweights K
i,jdepend on contact person A
iwith contact person A
jget to know time T (unit: day) and contact number of times Count.Circular is as follows:
2.3, according to number of degrees D and weights K, draw " cohesion " Intimacy of " I " and each contact person.The number of degrees D of agreement user node O
0=0 by path OA
ion be related to L
0, iconnected node A
inumber of degrees D
i=1+D
0; Pass through path A
ia
jon be related to L
i,jconnected node A
jnumber of degrees D
j=1+D
i." I " and node A
j" cohesion " computing formula be:
Intimacy
j=D
j+K
i,j,D
i∈N
*,K
i,j∈[0,1)
The integral part of " cohesion " Intimacy has represented the number of degrees in six degrees of separation, and fraction part is related to that by each under the equal number of degrees Further Division opens, wherein N
*for positive integer collection.Therefore, " cohesion " can reflect the close and distant relation that arrives each node centered by " I " preferably.
Step 3. realizes intelligent retrieval
By the social networks centered by " I ", intelligent retrieval contacts approach with location the best.
The result of retrieval shows as by visual social network diagram: by the contact person's node on path and the highlighted demonstration of line, see schematic diagram 8.When user selects highlighted node contact person, contact person's essential information manifests on node side.Thus, user can check that user can check by the route of highlighted demonstration and can relate to needed contact person by which friend.Described retrieval flow should be realized:
● machine screening: fuzzy matching retrieving information, filters out possible object listing.
● manually locate: from object listing, manual positioning target.
● result feedback: obtain the communication approach of " the most effective " between " I " and target, relate to by the most intimate (or minimum) contact person the contact person who needs most.
During concrete enforcement, the search key of inputting by user and the personal information of typing are carried out fuzzy matching, filter out possible object listing, and step is as follows:
3.1, the keyword of establishing user input is character string WORD, and element is wherein Char, business card data collection DATA, if
meet
return to its corresponding contact person A
(Char ∩ DATA).
3.2, user, from the list of returning, according to the actual requirements, selects objectives, navigates to a certain contact person A
i.
3.3, according to the weights K being related on L, by shortest path first (adopting dijkstra's algorithm here), obtain the communication approach of " the most effective " between " I " and target, by the most intimate (or minimum) contact person, relate to the contact person who needs most.
Regard contact person's node A as point set in dijkstra's algorithm; Be related to that L regards the limit collection in algorithm as; The length on limit is corresponding K value; User's node O is as the source in algorithm.
Step 4. completes mobile terminal synchronization
This step by the card information in fulfillment database and mobile-terminal platform as the intelligent synchronization between the address list of the mobile phones such as Android, iOS.During concrete enforcement, should complete following steps:
4.1, start service.PC server end requires to adopt the correlation technique design server programs such as Socket and multithreading, it can be resident service program, along with os starting, starts, also can start-up by hand, after startup, will connect with database server, then the request of AM automatic monitoring mobile terminal.
4.2, automatic configuration of IP.When user clicks the sync client program of mobile terminal, it will read server ip address, server end slogan PORT in the configuration file from mobile terminal; If cannot connect PC server end, it is by autoscan server.
Disposable IP scanning collocation strategy: terminal adds after the network segment of PC server place, its IP address of terminal program automatic acquisition, according to algorithm for text string location, obtain trizonal data before IP address, the IP address (as possible PC server IP address) that utilizes loop program autoscan to be formed by 0-255 and first three area data, if energy successful connection, be stored in configuration file, for automatic connection server next time.
4.3, system is set.The system setting function technical characterictic of mobile terminal is: when terminal program autoscan cannot connect upper PC server, can revise IP address, port numbers PORT by interactively, two options of covering completely, difference when backup can also be set simultaneously and import.
4.4, backup mobile terminal.The backup functionality technical characterictic of mobile terminal is: when user clicks backup functionality button on mobile terminal, option in system arranges is for covering completely, program of mobile terminal read in the address list of this terminal i contact person (i=1,2 ... n), then i contact person is written in corresponding database by PC server, if there is this contact person in database, new data more, if do not exist, insert.If when the option in system arranges is difference, program of mobile terminal reads i contact person (i=1 in the address list of this terminal, 2,, n), then i contact person is written in corresponding database by PC server, if there is this contact person in database, ignore, if do not exist, insert.
4.5, import mobile terminal.The import feature technical characterictic of mobile terminal is: when user clicks import feature button on mobile terminal, option in system arranges is for covering completely, i contact person in PC server reading database address list (i=1,2 ... n), then i contact person is written in the address list of terminal by program of mobile terminal, if there is this contact person in address list, new data more, if do not exist, insert.If when the option in system arranges is difference, i contact person (i=1 in PC server reading database address list, 2,, n), then i contact person is written in the address list of terminal by program of mobile terminal, if there is this contact person in address list, ignore, if do not exist, insert.
Claims (6)
1. the social networks management method based on business card recognition, comprises following four steps:
Step 1, typing card information, utilize camera or scanner to carry out business card image collection, according to the feature of business card image, mark off character block and utilize OCR engine to carry out character recognition, according to critical field, carry out word segmentation processing, to the information categorization typing of extracting, inserted corresponding list, last and user carries out alternately, the None-identified information that may exist being carried out to manual confirmation and adjustment;
Step 2, set up social networks network;
Step 3, realize intelligent retrieval;
Step 4, complete mobile terminal synchronization.
2. a kind of social networks management method based on business card recognition as claimed in claim 1, is characterized in that:,
In step 1), according to the feature of business card image, divide character block and utilize OCR engine carry out character recognition concrete steps as follows:
1.1, gray processing.First to the business card image of camera or scanner collection, adopt method of weighted mean to carry out gray processing.Because human eye is the highest to green sensitivity, to blue responsive minimum, therefore use f (i, j)=0.30R (i, j)+0.59G (i, j)+0.11B (i, j) can obtain more rational gray level image.Wherein (i, j) represents pixel coordinate, and in coloured image, three components of red, green, blue are respectively R (i, j), G (i, j), B (i, j), and f (i, j) is the gray-scale value of this point;
1.2, edge extracting.In conjunction with Sobel operator and LOG (Laplacian of Gauss) operator, form and revise LOG algorithm, choose following Sobel operator (S
xfor the center of horizontal direction divide poor, S
yfor the center of vertical direction, divide poorly, choose wherein higher value as gradient S)
s=max{|S
x|, | S
y| as the precondition of rim detection, reduce unnecessary zero cross point.Recycling
this function, as wave filter, carries out LOG detection.LOG operator template is
it is the digital form of LOG, and it is done to convolution as interior collecting image.σ is the mean square deviation of Gaussian distribution;
1.3, binaryzation.Calculate threshold value gray-scale map is carried out to binaryzation.If gray-scale map height is h, wide is w, tries to achieve threshold value, i.e. average gray
according to threshold value Threshold, carry out binaryzation again:
1.4, tilt detection and rectification.By Hough transformation (Hough Transform), detect the edge line of business card frame, correct at the angle of inclination that obtains name panel region and judge business card.Because the straight-line equation of y=kx+b form cannot represent the straight line (c is constant, i.e. the straight line parallel with x axle, slope k → ∞) of x=c form.Therefore adopt parametric equation ρ=x*cos θ+y*sin θ here, wherein by choosing p
1(x
1, y
1), p
2(x
2, y
2) two monitoring points, can obtain pitch angle
according to tiltangleθ, former figure and bianry image are carried out to affined transformation (using homogeneous matrix representation) simultaneously, P'=PR, general-θ substitution 2D rotation matrix obtains correcting matrix
by point
after rectification, obtain.Concrete operation is:
1.5, image is cut apart.According to the feature of business card image, mark off character block.Comprise that step is as follows:
1.5.1, definition surveys density D ensity, for current pixel up and down and the clinodiagonal quantity of black picture element in the neighbor in totally 8 directions, computing formula is:
i ∈ [x-1, x+1] ∩ N wherein
*, j ∈ [x-1, x+1] ∩ N
*, (x, y) is current sensing point coordinate, N
*for positive integer collection;
1.5.2, the image of name panel region is converted into density matrix, remove remaining noise.Method of operating is as follows: judge one by one the density of each pixel, when Density<2, homography element is designated as 0, is used as noise processed.When Density>=2, homography element is designated as 1, shows that this pixel is a character block part;
1.5.3, pass through conversion formula
two-value achromatic sheet after correcting is converted into following two-dimensional array form:
1.5.4, according to density matrix, character block region, location (this time have two kinds of strategies to differentiate to be applied to quick recognition mode with precisely recognition mode), then cuts apart business card image according to region.Quick recognition strategy: judge line by line density matrix, the ratio that in every a line, " 1 " element accounts for unit number, over certain threshold value, is considered as line of text, is considered as blank lower than this threshold value.Accurate recognition strategy: judge line by line density matrix, continuous " 0 " element in row is connected to " detection line ", determine whether text filed according to the depth of " detection line " initiating terminal, difference in length and terminal position feature.Remove the detection line that length is less than threshold value; Then mark article one and survey the end of line and the top that second is surveyed line; To all row survey with mark after, the region being surrounded by mark is text filed;
1.5.5, definition R
i∈ d (i, y) | y ∈ [0, h] ∩ N
*,
and if only if
time record j value.One group of region [j of last gained
1, j
2], [j
2, j
3], [j
3, j
4] ... be some character blocks, wherein h represents line number, and w represents columns, N
*for positive integer collection, R
ifor the capable element of i in the two-dimensional array of step 3, Sum
jfor R
iinstitute's all elements sum of being expert at.
1.6, character recognition.Utilize OCR (Optical Character Recognition) technology to extract the information of each character block.Here adopt the OCR module of MODI (Microsoft Office Document Imaging) to identify one by one the image being partitioned into, each character block image is all become to one group of word.To character block [j
1, j
2], [j
2, j
3], [j
3, j
4] ... call one by one OCR identification engine, obtain the character set C corresponding with each region
1,2, C
2,3, C
3,4
3. a kind of social networks management method based on business card recognition as claimed in claim 1, is characterized in that:
In step 1), according to critical field, carry out the concrete steps of word segmentation processing as follows:
1.7, word segmentation processing.According to critical field, carry out word segmentation processing, to the information categorization typing of extracting.First this need to set up keywords database and semantic base, for example: company, position, address, telephone number, email, road, number etc.Definition of keywords set W={ company, position, address, telephone number, email ... and definition of keywords semantic base, such as Address, Addr, address, contact address etc. is the near synonym of address, semantically, be consistent, set up corresponding mapping relations, form semantic base.Then according to keyword, based on contextual feature, extract the corresponding informance in every group of word, insert corresponding list, complete word segmentation processing.Concrete steps are as follows:
1.7.1, find ": " separator, utilize separator to define keyword and content.If not in set of keywords, transferring to user to determine whether to be regarded as keyword, the keyword defining is indexed in set.By this step, realize the learning functionality of participle strategy;
1.7.2, then match key word Key=C
x,y∩ W, by character string
insert in the form item of Key by name.When
time, by character set C
x,yinsert " unfiled " form item;
Strategy based on semantic base comprises:
Location name strategy (Chinese): get rid of other semantemes, number of characters, between 2~4, is followed by position, title, or with " name " class label;
Locating cellphone number strategy: with 11 pure digi-tal character strings of 13,15,18 beginnings, or with " mobile phone " class label;
Location telephone number strategy: 7~8 pure digi-tal character strings, or with area code beginning, comprise the separators such as bracket, hyphen, space, or with " phone " class label;
Location Business Name strategy: get rid of other semantemes, occur " ”,“ group of company ", each institutional settings keyword, or with " company " class label;
Positioning address strategy: there is the character string that mix with numeral " ”,“ district, ”,“ road, ”,“ town, ”,“ township, ”,“ county, province ”,“ city ”,“ Zhuan”,“ unit ", " chamber ", or with " address " class label;
Location postcode strategy: 6 pure digi-tal character strings, or with " postcode " class label;
Location mailbox strategy: there is numeral, letter, " ", ". " character, or with " mailbox " class label;
Location network address strategy: there is " http ", ". ", " www ", " com ", " cn ", " edu " character, or with " network address " class label.
4. a kind of social networks management method based on business card recognition as claimed in claim 1, is characterized in that:
Step 2) step of setting up social networks network in is as follows:
2.1, by move the relation that contact person's node is set up two contact persons with mouse.By the node A newly adding
jappend to existing node A
i(wanting the contact person of opening relationships) is upper, after definite operation both sides, utilize the form of line to represent the L that is related to setting up between two contact persons
i,j, it represents A
iwith A
jbe familiar with each other (two-way);
2.1.1, (optional): if newly-established contact person's node A
ithe first contact people of user O (i.e. " I "), by new node A
ibe connected to the upper contact circuit OA that sets up of user's node O
i, its pass is L
0, i;
2.1.2, (optional): if newly-established contact person's node A
jthat user O is by friend A
ithis contact person of cicada, and the other side is not familiar with in the situation of user O, by new node A
jbe connected to user friend's node A
iupper foundation contact circuit A
ia
j, its pass is L
i,j;
2.1.3, (optional): if the former O of user is before this by friend A
ithe contact person A learning
jwith user friend A
ithe first contact people (unidirectional), and after further cooperation and exchange, user O and this contact person A
jbecome the first contact people (two-way), by this contact person's node A
jbe connected to user's node O upper, form and the direct internuncial pathway OA of user's node
j, its pass is L
0, j;
2.2, for each, be related to L
i,jby using weights K
i,jthe power that represents relation.L
i,jweights K
i,jdepend on contact person A
iwith contact person A
jget to know time T (unit: day) and contact number of times Count.Circular is:
2.3, according to number of degrees D and weights K, draw " cohesion " Intimacy of " I " and each contact person.The number of degrees D of agreement user node O
0=0 by path OA
ion be related to L
0, iconnected node A
inumber of degrees D
i=1+D
0; Pass through path A
ia
jon be related to L
i,jconnected node A
jnumber of degrees D
j=1+D
i." I " and node A
j" cohesion " computing formula be: Intimacy
j=D
j+ K
i,j, D
i∈ N
*, K
i,j∈ [0,1), the integral part of " cohesion " Intimacy has represented the number of degrees in six degrees of separation, fraction part is related to that by each under the equal number of degrees Further Division opens.Therefore, " cohesion " can reflect the close and distant relation that arrives each node centered by " I ", wherein N preferably
*for positive integer collection.
5. a kind of social networks management method institute based on business card recognition as claimed in claim 1, is characterized in that:
The search key of inputting by user in step 3) and the personal information of typing are carried out fuzzy matching, filter out possible object listing intelligent retrieval, and its concrete steps are as follows:
3.1, the keyword of establishing user input is character string WORD, and element is wherein Char, business card data collection DATA, if
and
return to its corresponding contact person A
(Char ∩ DATA).
3.2, user, from the list of returning, according to the actual requirements, selects objectives, navigates to a certain contact person A
i.
3.3, according to the weights K being related on L, by shortest path first (adopting dijkstra's algorithm here), obtain the communication approach of " the most effective " between " I " and target, by the most intimate (or minimum) contact person, relate to the contact person who needs most.Here regard contact person's node A as point set in dijkstra's algorithm; Be related to that L regards the limit collection in algorithm as; The length on limit is corresponding K value; User's node O is as the source in algorithm.
6. a kind of social networks management method based on business card recognition as claimed in claim 1, is characterized in that: the step that completes mobile terminal synchronization in step 4) is as follows:
4.1, start service.PC server end requires to adopt the correlation technique design server programs such as Socket and multithreading, it can be resident service program, along with os starting, starts, also can start-up by hand, after startup, will connect with database server, then the request of AM automatic monitoring mobile terminal;
4.2, automatic configuration of IP.When user clicks the sync client program of mobile terminal, it will read server ip address, server end slogan PORT in the configuration file from mobile terminal; If cannot connect PC server end, it is by autoscan server.Disposable IP scanning collocation strategy: terminal adds after the network segment of PC server place, its IP address of terminal program automatic acquisition, according to algorithm for text string location, obtain trizonal data before IP address, the IP address (as possible PC server IP address) that utilizes loop program autoscan to be formed by 0-255 and first three area data, if energy successful connection, be stored in configuration file, for automatic connection server next time;
4.3, system is set.The system setting function technical characterictic of mobile terminal is: when terminal program autoscan cannot connect upper PC server, can revise IP address, port numbers PORT by interactively, two options of covering completely, difference when backup can also be set simultaneously and import;
4.4, backup mobile terminal.The backup functionality technical characterictic of mobile terminal is: when user clicks backup functionality button on mobile terminal, option in system arranges is for covering completely, program of mobile terminal read in the address list of this terminal i contact person (i=1,2 ..., n), then i contact person is written in corresponding database by PC server, if there is this contact person in database, new data more, if do not exist, insert.If when the option in system arranges is difference, program of mobile terminal reads i contact person (i=1 in the address list of this terminal, 2,, n), then i contact person is written in corresponding database by PC server, if there is this contact person in database, ignore, if do not exist, insert;
4.5, import mobile terminal.The import feature technical characterictic of mobile terminal is: when user clicks import feature button on mobile terminal, option in system arranges is for covering completely, i contact person in PC server reading database address list (i=1,2 ... n), then i contact person is written in the address list of terminal by program of mobile terminal, if there is this contact person in address list, new data more, if do not exist, insert.If when the option in system arranges is difference, i contact person (i=1 in PC server reading database address list, 2,, n), then i contact person is written in the address list of terminal by program of mobile terminal, if there is this contact person in address list, ignore, if do not exist, insert.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310521182.2A CN103617422B (en) | 2013-10-29 | 2013-10-29 | A social relation management method based on business card recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310521182.2A CN103617422B (en) | 2013-10-29 | 2013-10-29 | A social relation management method based on business card recognition |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103617422A true CN103617422A (en) | 2014-03-05 |
CN103617422B CN103617422B (en) | 2017-01-11 |
Family
ID=50168125
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310521182.2A Active CN103617422B (en) | 2013-10-29 | 2013-10-29 | A social relation management method based on business card recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103617422B (en) |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104020847A (en) * | 2014-05-15 | 2014-09-03 | 上海合合信息科技发展有限公司 | Character input method and device |
CN104504069A (en) * | 2014-12-22 | 2015-04-08 | 北京奇虎科技有限公司 | Building method and device for file index |
CN105490811A (en) * | 2015-11-25 | 2016-04-13 | 小米科技有限责任公司 | Method and device for obtaining signature information |
CN105631486A (en) * | 2014-10-27 | 2016-06-01 | 深圳Tcl数字技术有限公司 | image character recognition method and device |
CN105631752A (en) * | 2016-03-22 | 2016-06-01 | 南京信息工程大学 | Social network sampling generation algorithm for improving Dijkstra weight |
CN105718972A (en) * | 2016-01-15 | 2016-06-29 | 卢怀农 | Intelligent acquiring method of information |
CN105791503A (en) * | 2016-04-28 | 2016-07-20 | 乐视控股(北京)有限公司 | Method of storing business card information in address list and apparatus thereof |
CN106230968A (en) * | 2016-08-22 | 2016-12-14 | 杨玲 | Cross-platform file based on Apache Server transmission APP software system |
CN106326854A (en) * | 2016-08-19 | 2017-01-11 | 掌阅科技股份有限公司 | Open fixed-layout document paragraph identification method |
CN106709488A (en) * | 2016-12-20 | 2017-05-24 | 深圳市深信服电子科技有限公司 | Business card identification method and device |
CN107239453A (en) * | 2016-03-28 | 2017-10-10 | 平安科技(深圳)有限公司 | Information write-in method and device |
CN107402927A (en) * | 2016-05-19 | 2017-11-28 | 上海斯睿德信息技术有限公司 | A kind of enterprise's incidence relation topology method for building up and querying method based on graph model |
CN107679240A (en) * | 2017-10-27 | 2018-02-09 | 中国计量大学 | A kind of virtual identity method for digging |
CN108009546A (en) * | 2016-10-28 | 2018-05-08 | 北京京东尚科信息技术有限公司 | information identifying method and device |
CN108446697A (en) * | 2018-03-06 | 2018-08-24 | 平安科技(深圳)有限公司 | Image processing method, electronic device and storage medium |
CN108509783A (en) * | 2018-04-09 | 2018-09-07 | 陈静 | A kind of intelligent terminal application management system |
CN109358067A (en) * | 2018-11-30 | 2019-02-19 | 桂林电子科技大学 | Motor ring varistor defect detecting system based on computer vision and method |
CN111147361A (en) * | 2019-12-30 | 2020-05-12 | 论客科技(广州)有限公司 | Method, device and storage medium for adding mailbox account |
CN111414536A (en) * | 2020-03-17 | 2020-07-14 | 支付宝(杭州)信息技术有限公司 | Service processing method and device |
CN112199545A (en) * | 2020-11-23 | 2021-01-08 | 湖南蚁坊软件股份有限公司 | Keyword display method and device based on picture character positioning and storage medium |
CN112733834A (en) * | 2021-03-30 | 2021-04-30 | 恒银金融科技股份有限公司 | Character area positioning method based on non-matrix window mode |
CN113569738A (en) * | 2021-07-28 | 2021-10-29 | 深圳思谋信息科技有限公司 | Tabulation information extraction processing method and device, computer equipment and storage medium |
CN115170040A (en) * | 2022-09-08 | 2022-10-11 | 南方电网数字电网研究院有限公司 | Method and system for dynamically updating asset directory |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101882264A (en) * | 2009-11-27 | 2010-11-10 | 镇立新 | Method for adding contacts |
CN101916271A (en) * | 2010-08-10 | 2010-12-15 | 上海合合信息科技发展有限公司 | Method and system for automatically acquiring contact head image |
CN102622592A (en) * | 2012-01-19 | 2012-08-01 | 北京友录在线科技发展有限公司 | Name card recognition method based on cloud technology |
-
2013
- 2013-10-29 CN CN201310521182.2A patent/CN103617422B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101882264A (en) * | 2009-11-27 | 2010-11-10 | 镇立新 | Method for adding contacts |
CN101916271A (en) * | 2010-08-10 | 2010-12-15 | 上海合合信息科技发展有限公司 | Method and system for automatically acquiring contact head image |
CN102622592A (en) * | 2012-01-19 | 2012-08-01 | 北京友录在线科技发展有限公司 | Name card recognition method based on cloud technology |
Non-Patent Citations (2)
Title |
---|
张纯等: "中文商务名片识别系统的实现", 《中文信息学报》 * |
牛华网: "到底谁更实用 经纬名片通VS名片全能王", 《HTTP://WWW.NEWHUA.COM/2013/0105/191386.SHTML》 * |
Cited By (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104020847A (en) * | 2014-05-15 | 2014-09-03 | 上海合合信息科技发展有限公司 | Character input method and device |
CN105631486A (en) * | 2014-10-27 | 2016-06-01 | 深圳Tcl数字技术有限公司 | image character recognition method and device |
CN104504069A (en) * | 2014-12-22 | 2015-04-08 | 北京奇虎科技有限公司 | Building method and device for file index |
CN105490811A (en) * | 2015-11-25 | 2016-04-13 | 小米科技有限责任公司 | Method and device for obtaining signature information |
CN105718972A (en) * | 2016-01-15 | 2016-06-29 | 卢怀农 | Intelligent acquiring method of information |
CN105718972B (en) * | 2016-01-15 | 2019-01-29 | 卢怀农 | A kind of information intelligent acquisition method |
CN105631752A (en) * | 2016-03-22 | 2016-06-01 | 南京信息工程大学 | Social network sampling generation algorithm for improving Dijkstra weight |
CN107239453A (en) * | 2016-03-28 | 2017-10-10 | 平安科技(深圳)有限公司 | Information write-in method and device |
CN105791503A (en) * | 2016-04-28 | 2016-07-20 | 乐视控股(北京)有限公司 | Method of storing business card information in address list and apparatus thereof |
CN107402927A (en) * | 2016-05-19 | 2017-11-28 | 上海斯睿德信息技术有限公司 | A kind of enterprise's incidence relation topology method for building up and querying method based on graph model |
CN106326854A (en) * | 2016-08-19 | 2017-01-11 | 掌阅科技股份有限公司 | Open fixed-layout document paragraph identification method |
CN106326854B (en) * | 2016-08-19 | 2019-09-06 | 掌阅科技股份有限公司 | A kind of format document paragraph recognition methods |
CN106230968A (en) * | 2016-08-22 | 2016-12-14 | 杨玲 | Cross-platform file based on Apache Server transmission APP software system |
CN108009546A (en) * | 2016-10-28 | 2018-05-08 | 北京京东尚科信息技术有限公司 | information identifying method and device |
CN106709488A (en) * | 2016-12-20 | 2017-05-24 | 深圳市深信服电子科技有限公司 | Business card identification method and device |
CN107679240A (en) * | 2017-10-27 | 2018-02-09 | 中国计量大学 | A kind of virtual identity method for digging |
CN108446697A (en) * | 2018-03-06 | 2018-08-24 | 平安科技(深圳)有限公司 | Image processing method, electronic device and storage medium |
CN108446697B (en) * | 2018-03-06 | 2019-11-12 | 平安科技(深圳)有限公司 | Image processing method, electronic device and storage medium |
CN108509783A (en) * | 2018-04-09 | 2018-09-07 | 陈静 | A kind of intelligent terminal application management system |
CN109358067A (en) * | 2018-11-30 | 2019-02-19 | 桂林电子科技大学 | Motor ring varistor defect detecting system based on computer vision and method |
CN111147361A (en) * | 2019-12-30 | 2020-05-12 | 论客科技(广州)有限公司 | Method, device and storage medium for adding mailbox account |
CN111147361B (en) * | 2019-12-30 | 2022-06-07 | 论客科技(广州)有限公司 | Method, device and storage medium for adding mailbox account |
CN111414536A (en) * | 2020-03-17 | 2020-07-14 | 支付宝(杭州)信息技术有限公司 | Service processing method and device |
CN112199545A (en) * | 2020-11-23 | 2021-01-08 | 湖南蚁坊软件股份有限公司 | Keyword display method and device based on picture character positioning and storage medium |
CN112733834A (en) * | 2021-03-30 | 2021-04-30 | 恒银金融科技股份有限公司 | Character area positioning method based on non-matrix window mode |
CN112733834B (en) * | 2021-03-30 | 2021-07-06 | 恒银金融科技股份有限公司 | Character area positioning method based on non-matrix window mode |
CN113569738A (en) * | 2021-07-28 | 2021-10-29 | 深圳思谋信息科技有限公司 | Tabulation information extraction processing method and device, computer equipment and storage medium |
CN115170040A (en) * | 2022-09-08 | 2022-10-11 | 南方电网数字电网研究院有限公司 | Method and system for dynamically updating asset directory |
Also Published As
Publication number | Publication date |
---|---|
CN103617422B (en) | 2017-01-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103617422B (en) | A social relation management method based on business card recognition | |
CN110569832B (en) | Text real-time positioning and identifying method based on deep learning attention mechanism | |
US10013624B2 (en) | Text entity recognition | |
CN101297318B (en) | Data organization and access for mixed media document system | |
US8892595B2 (en) | Generating a discussion group in a social network based on similar source materials | |
US10990814B2 (en) | Converting an image into a structured table | |
US9020265B2 (en) | System and method of determining building numbers | |
CN102779140A (en) | Keyword acquiring method and device | |
CN103577818A (en) | Method and device for recognizing image characters | |
CN111191652A (en) | Certificate image identification method and device, electronic equipment and storage medium | |
CN112000834A (en) | Document processing method, device, system, electronic equipment and storage medium | |
CN115774805A (en) | File intelligent query method and system based on digital processing | |
CN114998905A (en) | Method, device and equipment for verifying complex structured document content | |
Lu et al. | Efficient descriptor of histogram of salient edge orientation map for finger vein recognition | |
US20190065579A1 (en) | Apparatus, method, and computer-readable medium for recognition of a digital document | |
CN109948598B (en) | Document layout intelligent analysis method and device | |
Guo et al. | Text line detection based on cost optimized local text line direction estimation | |
CN104899551A (en) | Form image classification method | |
Ma et al. | Simple and efficient registration of 3D point cloud and image data for an indoor mobile mapping system | |
CN108830217B (en) | Automatic signature distinguishing method based on fuzzy mean hash learning | |
Rahul et al. | Deep reader: Information extraction from document images via relation extraction and natural language | |
WO2016069005A1 (en) | Text line detection | |
US11704924B2 (en) | Method, apparatus, and computer-readable storage medium for recognizing characters in a digital document | |
Khlif | Multi-lingual scene text detection based on convolutional neural networks | |
JP2022045011A (en) | Image processing device, control method of image processing device, and program |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |