CN100362525C - Method for gathering and recording business card information in mobile phone by using image recognition - Google Patents
Method for gathering and recording business card information in mobile phone by using image recognition Download PDFInfo
- Publication number
- CN100362525C CN100362525C CNB2005100264922A CN200510026492A CN100362525C CN 100362525 C CN100362525 C CN 100362525C CN B2005100264922 A CNB2005100264922 A CN B2005100264922A CN 200510026492 A CN200510026492 A CN 200510026492A CN 100362525 C CN100362525 C CN 100362525C
- Authority
- CN
- China
- Prior art keywords
- mobile phone
- image
- information
- name
- carry out
- 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.)
- Active
Links
Images
Landscapes
- Character Input (AREA)
- Character Discrimination (AREA)
- Telephone Function (AREA)
Abstract
The present invention relates to a method for collecting and entering business card information in a mobile phone by using image recognition, wherein the method comprises the following steps that a mobile phone obtains image information of business cards through a taking lens; the mobile phone preprocesses the images; the mobile phone carries out a layout analysis and region segmentation to the preprocessed images; the mobile phone carries out character recognition to each area after the segmentation; the mobile phone carries out data identification and an information analysis to the character recognition content, and stores the information in a telephone directory in the mobile phone. The method adopting the image recognition to collect and enter the business card information in the mobile phone has wide application foundation and prospect. Meanwhile, the information collection is independently completed by the mobile phone with a photographic function, which is in no need of any other subsidiary hardware. Furthermore, the present invention solves the shortcoming that the speed is slow when a user manually enters the business card information in the telephone directory. The present invention has selection schemes in automatic and manual modes, which increases the accuracy rate of the information collection and the optional enter speed of the information selection and achieves better application effect.
Description
Technical field
The present invention relates to field of mobile phones, particularly handset image information Recognition process field specifically is meant the method for using image recognition to carry out the collection and the typing of card information in a kind of mobile phone.
Background technology
As everyone knows, the business card that printing is made is the product of commercial age, is a kind of carrier that people's personal information exchanges, and has been subjected to being extensive use of of society.But finding required information efficiently from the business card that hundreds of printing is made for convenience, but is not the part nothing the matter, need quite take some time.The business card electronization can well be addressed this problem, can finish work such as retrieval, classification fast, user-friendly.Numerous card printers or business card electronization software has appearred for this reason in the market, card printer is a kind of Special Equipment that will print the business card electronization of making, and business card electronization software to be business card after utilizing image recognition technology to image conversion carry out information analysis and gather being entered into the cover of one in Computer Database software solution method.But the instream factor of these two kinds of schemes is not high, mainly has the following disadvantages:
(1) use not convenient because the collection of card information and preserve and need be equipped with again in addition a special computer hardware and maybe need to use a computer and finish this work and seem very inconvenient, so their practical value fails extensively to be adopted by the consumer;
(2) because of the printing different style of business card, the automatic information collecting effect of card printer is not good enough, finish this work by computer software and then also need to be equipped with special scanner, and to the unloading of the card information of follow-up acquisition, use and also need to store the use that could make things convenient for the user to go out in the equipment such as PDA or mobile phone into, have the problem of storage format (field) coupling.
How can conveniently finish the electronic process to business card, and can reach at any time the information of retrieving easily, using after the business card electronization, be to need the problem that solves at present.
Summary of the invention
The objective of the invention is to have overcome above-mentioned shortcoming of the prior art, the method for using image recognition to carry out the collection and the typing of card information in a kind of convenient to use, information acquisition typing accuracy is higher, stable performance, applicability are stronger mobile phone is provided.
In order to realize above-mentioned purpose, it is as follows to use image recognition to carry out the method for the collection of card information and typing in the mobile phone of the present invention:
The method of using image recognition to carry out the collection and the typing of card information in this mobile phone, its principal feature is that described method may further comprise the steps:
(1) mobile phone obtains the image information of business card by taking lens;
(2) mobile phone carries out pre-service to image;
(3) mobile phone carries out printed page analysis and Region Segmentation to pretreated image;
(4) mobile phone carries out literal identification to cutting apart each zone, back;
(5) mobile phone carries out data identification and information analysis to the literal content identified;
(6) data are deposited in the telephone directory of mobile phone.
The image information of obtaining business card of using image recognition to carry out the method for the collection of card information and typing in this mobile phone may further comprise the steps:
(1) taking lens of mobile phone is arranged to high pixel resolution screening-mode;
(2) will put in order the complete shooting typing of name cards data.
The pre-service of image of using image recognition to carry out the method for the collection of card information and typing in this mobile phone may further comprise the steps:
(1) image being carried out binary conversion treatment or grey level handles;
(2) with the row of the edge of image or literal to being that benchmark carries out degree of tilt and proofreaies and correct.
Printed page analysis and the Region Segmentation of using image recognition to carry out the method for the collection of card information and typing in this mobile phone may further comprise the steps:
(1) space of a whole page template of business card is set in advance in mobile phone;
(2) image is used decision tree carry out template matches;
(3) Region Segmentation of carrying out image layout according to the result of template matches is handled.
The space of a whole page template of business card of using image recognition to carry out the method for the collection of card information and typing in this mobile phone comprises:
(1) from top to bottom arrange successively T1---unit, name and academic title, three of addresses, and two address blocks are arranged;
(2) from top to bottom arrange successively T2---unit, name and academic title, three of addresses, and an address block is arranged;
(3) T3---name and academic title are put into topmost, and an address block is arranged;
(4) T4---name and academic title are put into topmost, and two address blocks are arranged;
(5) T5---units chunk is at topmost, and name and academic title's piece and address are arranged side by side;
(6) T6---name and academic title are at topmost, and unit, address also are listed in the bottom;
(7) T7---remove other modes outside the above type-setting mode.
The use decision tree of using image recognition to carry out the method for the collection of card information and typing in this mobile phone carries out template matches and may further comprise the steps:
(1) mobile phone preestablishes the blank spaces threshold value according to user's operation;
(2) to image from top to bottom vertical direction carry out projection;
(3) judge that whether blank spaces in the image is greater than described blank spaces threshold value;
(4) if not, then matching template is T7, and carries out manual interactive mode and determine the layout area dividing processing;
(5) otherwise, then find first;
(6) judge whether vertical direction can continue projection;
(7) if not, then whether the determined level direction can continue projection, if horizontal direction cannot continue projection, then matching template is T7, and carries out manual interactive mode and determine the layout area dividing processing; If horizontal direction can continue projection, then matching template belongs to T5 or T6, and further employing level and vertical projection are finished template matches;
(8) if, then whether the determined level direction can continue projection, if horizontal direction cannot continue projection, then matching template belongs to T2 or T3, and further employing level and vertical projection are finished template matches; If horizontal direction can continue projection, then matching template belongs to T1 or T4, and further employing level and vertical projection are finished template matches.
The manual interactive mode of using image recognition to carry out the method for the collection of card information and typing in this mobile phone determines that the layout area dividing processing may further comprise the steps:
(1) selects the current field name that needs identification;
(2) select the size of the identification frame of this field;
(3) select the laggard mobile phone operation interface of going into, and the operation of directionkeys is come mobile identified region, carry out the setting of identified region according to the user;
(4) repeating step (1) is to (3), up to required identification content is all determined.
The collection of using image recognition to carry out card information in this mobile phone is discerned with the literal of the method for typing and be may further comprise the steps:
(1) line of text character segmentation: significantly image is cut into row earlier, isolates single character again from image line;
(2) character features extracts: extract statistical nature or architectural feature from single character picture, comprise refinement and normalization;
(3) literal identification is handled: the highest character type of symbol similarity finds from the feature database that study obtains and waits to become literate.
Use in this mobile phone image recognition carry out the collection of card information and typing method the literal content identified is carried out data identification and information analysis may further comprise the steps:
(1) recognition result with name piece, units chunk, academic title's piece is directly corresponding with name, organization and out of Memory field in the cell-phone telephone book;
(2) content of text in the address block is carried out field and extract processing.
The structure of using image recognition to carry out the content of text in the address block of method of the collection of card information and typing in this mobile phone is:
＜address block 〉: :=＜field 1〉|＜field 2〉|＜field 3〉...
＜field i 〉: :=＜first introductory word〉|＜delimiter〉|＜content 〉
＜content 〉: :=＜character 1〉|＜character 2〉|＜character 3〉...
The delimiter set=': ', '. ', ' '
Elder generation's introductory word, content are for example had the phone field by the decision of field kind:
Guide's set of words=and ' phone ', ' TEL ', ' PHONE ' ...
The content character set=numerical digit, ' (', ') ', '-, ' ', '. ',
Described field extracts to handle and may further comprise the steps:
(1) the fault-tolerant coupling of first introductory word;
(2) skip delimiter;
(3) judge according to the content character set whether content is legal;
(4) if content legality is the basis at the special-purpose dictionary of business card and rule base then, proofreads recognition result automatically and revise manifest error;
(5) otherwise, then get back to step (1).
Adopted the method for using image recognition to carry out the collection and the typing of card information in the mobile phone of this invention, because collection and input method to card information that the present invention sets forth are finished in conjunction with camera cell phone and image recognition technology, simultaneously along with development in science and technology is maked rapid progress, camera cell phone is along with the continuous lifting of pixel function, the camera cell phone that millions of pixels occurred at present, the photo effect of its shooting almost can match in excellence or beauty with the photo that mm professional camera special is taken, utilize camera cell phone picture shot resolution also to reach the requirement of image recognition processing simultaneously, become possibility in conjunction with these 2 collection and the typings of finishing card information, and had the following advantages:
(1) because mobile phone has reached greatly to be popularized, this technology at first has application foundation and application prospect widely;
(2) collection of information is independently finished by camera cell phone, need not any other additional computer hardware;
(3) information of gathering is kept in the phone directory of mobile phone, the user can conveniently in officely what is the need for and use when wanting, and has solved the manual typing card information of user to the slow shortcoming of phone directory medium velocity;
(4) automatically and the collection selection scheme of manual mode, further improve the accuracy rate and the optional information acquisition input speed of information acquisition, reached the better application effect.
Description of drawings
Fig. 1 carries out the process flow diagram of the method for the collection of card information and typing for using image recognition in the mobile phone of the present invention.
Fig. 2 uses decision tree to carry out the process flow diagram of template matches for of the present invention to business card image.
Fig. 3 is of the present invention to business card image Region Segmentation result schematic diagram.
Fig. 4 carries out the process flow diagram that field extracts processing for of the present invention to address block.
Fig. 5 (a) (b) (c) (d) (e) be that manual interactive information of the present invention is gathered the identifying synoptic diagram (f).
Embodiment
In order more to be expressly understood technology contents of the present invention, describe in detail especially exemplified by following examples.
The present invention at first is arranged to camera cell phone the exposal model of high pixel resolution, the business card of needs electronization is taken one by one be entered in the mobile phone.Business card image is utilized the positional information of the business card space of a whole page and the attribute of the business card knowledge analysis business card space of a whole page, card information is cut apart, dividing method can be selected to adopt and cuts apart automatically or provided by user interactions, afterwards character recognition is carried out in each zone of cutting apart, and recognition result done aftertreatment based on business card and word, from recognition result, extract useful information at last and store in each territory section of cellphone directory, finish the collection of card information and the automatic typing process of telephone directory information
See also shown in Figure 1ly, use image recognition to carry out the collection of card information in this mobile phone and the method for typing may further comprise the steps:
(1) mobile phone obtains the image information of business card by taking lens, may further comprise the steps:
(a) taking lens of mobile phone is arranged to high pixel resolution screening-mode;
(b) will put in order the complete shooting typing of name cards data.
(2) mobile phone carries out pre-service to image, may further comprise the steps:
(a) image being carried out binary conversion treatment or grey level handles;
(b) with the row of the edge of image or literal to being that benchmark carries out degree of tilt and proofreaies and correct.
In the middle of practical application, for obtaining system handles image preferably, improve follow-up analysis and recognition effect, when taking business card, should be noted that some with camera cell phone, the one, camera cell phone is arranged to higher pixel resolution screening-mode, the 2nd, guarantee the complete shooting typing of whole name cards data, to make things convenient for the follow-up automatic analysis of system to the business card space of a whole page.Obtain carrying out the necessary image pre-service behind the business card image, mainly being divided into two steps finishes, one is that image is carried out binary conversion treatment (or grey level processing) to reduce the interference of background noise, can also use other method that reduces image noise in this process, purpose is the outstanding useful data information that needs processing.Afterwards image is carried out degree of tilt and proofreaies and correct, with the row of the edge of business card or literal to being that benchmark is proofreaied and correct, so that follow-up printed page analysis and recognition efficiency are better.Just can begin the core process of real card information analysis and processing after finishing the image pre-service.
(3) mobile phone carries out printed page analysis and Region Segmentation to pretreated image, may further comprise the steps:
(a) space of a whole page template of business card is set in advance in mobile phone, comprises:
(i) from top to bottom arrange successively T1---unit, name and academic title, three of addresses, and two address blocks are arranged;
(ii) T2---from top to bottom arrange successively unit, name and academic title, three of addresses, and an address block is arranged;
(iii) T3---name and academic title are put into topmost, and an address block is arranged;
(iv) T4---name and academic title are put into topmost, and two address blocks are arranged;
(v) T5---units chunk is at topmost, and name and academic title's piece and address are arranged side by side;
(vi) T6---name and academic title are at topmost, and unit, address also are listed in the bottom;
(vii) T7---remove other modes outside the above type-setting mode.
(b) image is used decision tree carry out template matches, sees also shown in Figure 2ly, may further comprise the steps:
(i) mobile phone preestablishes the blank spaces threshold value according to user's operation;
(ii) to image from top to bottom vertical direction carry out projection;
Judge that (iii) whether blank spaces in the image is greater than described blank spaces threshold value;
(iv) if not, then matching template is T7, and carries out manual interactive mode and determine the layout area dividing processing;
Otherwise (v), then find first;
(judge vi) whether vertical direction can continue projection;
(vii) if not, then whether the determined level direction can continue projection, if horizontal direction cannot continue projection, then matching template is T7, and carries out manual interactive mode and determine the layout area dividing processing; If horizontal direction can continue projection, then matching template belongs to T5 or T6, and further employing level and vertical projection are finished template matches;
(viii) if, then whether the determined level direction can continue projection, if horizontal direction cannot continue projection, then matching template belongs to T2 or T3, and further employing level and vertical projection are finished template matches; If horizontal direction can continue projection, then matching template belongs to T1 or T4, and further employing level and vertical projection are finished template matches.
Wherein, manually interactive mode determines that the layout area dividing processing may further comprise the steps:
(1) selects the current field name that needs identification;
(2) select the size of the identification frame of this field;
(3) select the laggard mobile phone operation interface of going into, and the operation of directionkeys is come mobile identified region, carry out the setting of identified region according to the user;
(4) repeating step (1) is to (3), up to required identification content is all determined.
In the middle of practical application, the printed page analysis of image can adopt automatic and manual dual mode to carry out, and it is to utilize the positional information of the business card space of a whole page and the attribute of the business card knowledge analysis business card space of a whole page that the space of a whole page is analyzed automatically, and the business card space of a whole page is carried out dividing processing.On the other hand,, reject the unwanted information of user, also can finish this work by user interactions in order to improve analyze and discern ageing.Because of business card has been concentrated numerous information on little area, various information tend to distribute on certain position, thereby make the space of a whole page can roughly be divided into units chunk, name piece, academic title's piece, address block etc., and the approximate location of piece is also regular, as: name is often put together with the academic title, address block is many in the bottom of business card, and units chunk is mostly on top or the middle part.Except that containing above-mentioned essential information, other decorative elements also often arranged on the business card, mainly contain two kinds: diagram and horizontal subdivision line for attractive in appearance.The sign of the icon unit of mostly being wherein, the horizontal subdivision line plays the content segmentation effect.An important feature of the business card space of a whole page is: the relative position of each piece is comparatively stable.Obtain classification results according to nearly 200 various business cards of collecting as following table.
The classification of the table 1 business card space of a whole page
| The model numbering | Space of a whole page model | Explanation | Shared percentage |
| T1 | Unit name and 1 address 2, academic title address | From top to bottom arrange successively unit, name and academic title, three of addresses, and two address blocks are arranged.This simulation is heavy maximum. | 44.5% |
| T2 | Unit name and academic title address | Substantially the same, just address block only has one. | 23.4% |
| T3 | Name and academic title's unit address | Name and academic title are put into topmost, and an address block is arranged. | 1.66% |
| T4 | Name and academic title's unit address 1 address 2 | With 3 similar, just many address blocks. | 0.78% |
| T5 | Unit name and academic title address | Units chunk is at topmost, and name and academic title's piece and address are arranged side by side. | 4.68% |
| T6 | Name and academic title's unit address | Name and academic title are at topmost, and unit, address also are listed in the bottom. | 0.78% |
| T7 | Other | This class comprises: 1, personalization is too strong, constitutes a class by itself; 2, the space of a whole page is very compact, can't divide; 3, piecemeal is too careless, and version is used with anyhow. | 24.2％ |
After above template classification had been arranged, we can adopt the method for template matches to solve the problem of printed page analysis.The seven top class spaces of a whole page as template (T1, T2, T3, T4, T5, T6, T7), to the input business card mate, see that it belongs to any.The most direct a kind of idea is the order coupling, will mate seven times ageing can't acceptance of coupling (because bottom operation is the projection of pixel, more time-consuming) under the worst case but be made in like this.Considering has similarity between each template, can realize template matches with decision tree.But determining to use some fine feature elimination of level separator bars and illustrated influence again behind the template type, and name and academic title are being separated.The generation that when system's empty information does not reach thresholding, may make the mistake, so when wrong generation, blank spaces d
0By the user can be mutual provide.Can know that from table 2 printed page analysis effect the justifying accuracy can reach 91%, consider the diversity of business card form, this automatic printed page analysis effect still is more satisfactory, simultaneously ageingly also can reach the requirement of handling in the mobile phone.See also shown in Figure 3ly, it is to use of this algorithm to handle the example synoptic diagram.
Table 2 printed page analysis effect
| Units chunk | Address block | Name academic title piece | Separate name and academic title | Justifying is right entirely | |
| Accuracy | 95％ | 93％ | 92％ | 94％ | 91％ |
| Refuse the branch rate | 2％ | 0 | 0 | 0 | 0 |
| Error rate | 3％ | 7％ | 8％ | 6％ | 9％ |
(c) Region Segmentation of carrying out image layout according to the result of template matches is handled.
(4) mobile phone carries out literal identification to cutting apart each zone, back, may further comprise the steps:
(a) line of text character segmentation: significantly image is cut into row earlier, isolates single character again from image line;
(b) character features extracts: extract statistical nature or architectural feature from single character picture, comprise refinement and normalization;
(c) literal identification is handled: the highest character type of symbol similarity finds from the feature database that study obtains and waits to become literate.
(5) mobile phone carries out data identification and information analysis to the literal content identified, may further comprise the steps:
(a) recognition result with name piece, units chunk, academic title's piece is directly corresponding with name, organization and out of Memory field in the cell-phone telephone book;
(b) content of text in the address block is carried out field and extract to handle, see also shown in Figure 4ly, may further comprise the steps:
(i) the fault-tolerant coupling of first introductory word;
(ii) skip delimiter;
(iii) judge according to the content character set whether content is legal;
If (iv) content legality, then special-purpose dictionary and the rule base at business card is the basis, proofreads recognition result automatically and revises manifest error;
Otherwise (v), then get back to step (i).
In the middle of reality is used, in the identifying of carrying out literal, roughly comprise three contents: line of text character segmentation, character features are extracted and literal identification is handled.The row character segmentation is that image significantly is cut into row earlier, isolates the process of single character again from image line.Feature extraction is a most important ring in the whole link, it is a process of extracting statistical nature or architectural feature from single character picture, comprises for this reason and refinement (Thinning), normalization (Normalization) steps such as (with the font rule one that vary in size) done.The stability and the validity of the feature of extracting have directly determined the performance of discerning.The process that promptly finds and wait to become literate the highest character type of symbol similarity discerned in literal from the feature database that study obtains.By just having obtained the word content of each cut zone after the literal identification, next in order to realize that corresponding contents is filled in the phone directory in the corresponding field automatically, needs to consider the matching problem of related content.The recognition result of name piece, units chunk, academic title's piece can be directly corresponding with the fields such as name, organization and out of Memory in the phone directory, but with regard to complexity how address block, multinomial contents such as address, phone, fax, fax, Email, postcode are arranged, these contents can't be separated (Here it is front they are put together make the reason of printed page analysis) with printed page analysis, can only lean on the contextual understanding of recognition result.The understanding of general language is the comparison difficulty, but card information has tangible singularity.The most tangible is that each field is many with tangible first introductory word beginning, as: printed words such as common " phone ", " TEL ", " PHONE " before the telephone number.But experiment shows and only utilize first introductory word matching result and not really desirable that reason is: (1) although the individual character discrimination of character recognition program is very high, the first introductory word that a plurality of characters are formed then differ recognize surely right; (2) depend the wrong easily branch of first introductory word alone, for example: can regard telephone number to " TELEVISION STATION " as because of the existence of first introductory word " TEL ".Improved method is to carry out careful syntactic analysis, and the text of address block has following structure:
＜address block 〉: :=＜field 1〉|＜field 2〉|＜field 3〉...
＜field i 〉: :=＜first introductory word〉|＜delimiter〉|＜content 〉
＜content 〉: :=＜character 1〉|＜character 2〉|＜character 3〉...
The delimiter set=': ', '. ', ' '
Elder generation's introductory word, content are for example had the phone field by the decision of field kind:
Guide's set of words=and ' phone ', ' TEL ', ' PHONE ' ...
The content character set=numerical digit, ' (', ') ', '-', ' ', '. '
Note, adopted the fault-tolerant coupling of first introductory word here, found first introductory word as long as that is: there is the partial words of first introductory word just to think.Why can do like this is because the legitimacy of content also will be adjudicated in the back, needn't worry that first introductory word confused.This field extraction algorithm based on syntactic analysis has very strong robustness, can both correctly extract each field under the general situation of character recognition effect.Another effect that information after the identification is analyzed is based on special-purpose dictionary and rule base at business card, proofreads recognition result automatically, revises manifest error.Provide the example of an aftertreatment below, wherein underlined word recognizer has been admitted one's mistake, but all has been repaired by aftertreatment.
Before the processing: the address: the BeiJing, China revives the Luo Dao village outdoors
DizzyWords: 81.0631 (exchanges)
8I,2105 (direct dialing)
Cable address: 8681
Postal
BrocadeSign indicating number: 1000
S6
Handle the back: address field [BeiJing, China revives the Luo Dao village outdoors]
Phone field [81.0631 (exchange) 81.2105 (direct dialing)]
Cable address field [8681]
Postcode field [100086]
(6) data are deposited in the telephone directory of mobile phone.
See also again shown in Figure 5, for the business card of confirming as the T7 template type because there is complicated typesetting format in it, can't be according to normal identifying automatically, system provides user interactions identification to finish the identification input function of this category information simultaneously.
Concrete condition is as follows:
Open behind the phone directory and enter (a) behind the selection function button, select (b) business card Recognition and Acquisition to enter (c) operation-interface, can select automatic Recognition and Acquisition and manual identification collection this moment, whether adopts acquiescence d selecting automatic Recognition and Acquisition to point out
0Value is carried out the space of a whole page and is cut apart, and the user can adopt default value or select to revise d again when identification is wrong automatically
0Value is discerned automatically.When judging that according to automatic identification process this business card templates belongs to T7, also can point out simultaneously and enter this page automatically and carry out manual identification operation.Select manual identification to enter (d) and select the setting of manual interactive identification option, such as current selection identification field is filled in the name field of phone directory, and selects the size of identification frame, so that carry out identifying operation the most fast.Enter (e) operation-interface after selected, can move the setting that identified region is finished identified region by the direction of passage key, (d) and (e) can carry out repeatedly so that required identification content is all selected.After at least having an identified region chosen, can increase the function choosing-item of beginning Recognition and Acquisition information when turning back to (e) interface in the System menu automatically, after selecting the beginning identifying operation, system will be filled into corresponding field information in the corresponding contents of telephone directory as (f) automatically, if need to revise or change can be finished herein.
In this instructions, the present invention is described with reference to its certain embodiments.But, still can make various modifications and conversion obviously and not deviate from the spirit and scope of the present invention.Therefore, instructions and accompanying drawing are regarded in an illustrative, rather than a restrictive.
Claims (9)
1. the method for using image recognition to carry out the collection and the typing of card information in the mobile phone is characterized in that described method may further comprise the steps:
(1) mobile phone obtains the image information of business card by taking lens;
(2) mobile phone carries out pre-service to image;
(3) mobile phone carries out printed page analysis and Region Segmentation to pretreated image:
(a) space of a whole page template of business card is set in advance in mobile phone;
(b) image is used decision tree carry out template matches;
(c) Region Segmentation of carrying out image layout according to the result of template matches is handled;
(4) mobile phone carries out literal identification to cutting apart each zone, back;
(5) mobile phone carries out data identification and information analysis to the literal content identified;
(6) data are deposited in the telephone directory of mobile phone.
2. the method for using image recognition to carry out the collection and the typing of card information in the mobile phone according to claim 1 is characterized in that the described image information of obtaining business card may further comprise the steps:
(1) taking lens of mobile phone is arranged to high pixel resolution screening-mode;
(2) will put in order the complete shooting typing of name cards data.
3. the method for using image recognition to carry out the collection and the typing of card information in the mobile phone according to claim 1 and 2 is characterized in that the pre-service of described image may further comprise the steps:
(1) image being carried out binary conversion treatment or grey level handles;
(2) with the row of the edge of image or literal to being that benchmark carries out degree of tilt and proofreaies and correct.
4. the method for using image recognition to carry out the collection and the typing of card information in the mobile phone according to claim 1 is characterized in that the space of a whole page template of described business card comprises:
(1) from top to bottom arrange successively T1---unit, name and academic title, three of addresses, and two address blocks are arranged;
(2) from top to bottom arrange successively T2---unit, name and academic title, three of addresses, and an address block is arranged;
(3) T3---name and academic title are put into topmost, and an address block is arranged;
(4) T4---name and academic title are put into topmost, and two address blocks are arranged;
(5) T5---units chunk is at topmost, and name and academic title's piece and address are arranged side by side;
(6) T6---name and academic title are at topmost, and unit, address also are listed in the bottom;
(7) T7---remove other modes outside the above type-setting mode.
5. the method for using image recognition to carry out the collection and the typing of card information in the mobile phone according to claim 4 is characterized in that described use decision tree carries out template matches and may further comprise the steps:
(1) mobile phone preestablishes the blank spaces threshold value according to user's operation;
(2) to image from top to bottom vertical direction carry out projection;
(3) judge that whether blank spaces in the image is greater than described blank spaces threshold value;
(4) if not, then matching template is T7, and carries out manual interactive mode and determine the layout area dividing processing;
(5) otherwise, then find first;
(6) judge whether vertical direction can continue projection;
(7) if not, then whether the determined level direction can continue projection, if horizontal direction cannot continue projection, then matching template is T7, and carries out manual interactive mode and determine the layout area dividing processing; If horizontal direction can continue projection, then matching template belongs to T5 or T6, and further employing level and vertical projection are finished template matches;
(8) if, then whether the determined level direction can continue projection, if horizontal direction cannot continue projection, then matching template belongs to T2 or T3, and further employing level and vertical projection are finished template matches; If horizontal direction can continue projection, then matching template belongs to T1 or T4, and further employing level and vertical projection are finished template matches.
6. the method for using image recognition to carry out the collection and the typing of card information in the mobile phone according to claim 5 is characterized in that described manual interactive mode determines that the layout area dividing processing may further comprise the steps:
(1) selects the current field name that needs identification;
(2) select the size of the identification frame of this field;
(3) select the laggard mobile phone operation interface of going into, and the operation of directionkeys is come mobile identified region, carry out the setting of identified region according to the user;
(4) repeating step (1) is to (3), up to required identification content is all determined.
7. the method for using image recognition to carry out the collection and the typing of card information in the mobile phone according to claim 1 and 2 is characterized in that, described literal identification may further comprise the steps:
(1) line of text character segmentation: significantly image is cut into row earlier, isolates single character again from image line;
(2) character features extracts: extract statistical nature or architectural feature from single character picture, comprise refinement and normalization;
(3) literal identification is handled: the highest character type of symbol similarity finds from the feature database that study obtains and waits to become literate.
8. the method for using image recognition to carry out the collection and the typing of card information in the mobile phone according to claim 4 is characterized in that, described the literal content identified is carried out data identification and information analysis may further comprise the steps:
(1) recognition result with name piece, units chunk, academic title's piece is directly corresponding with name, organization and out of Memory field in the cell-phone telephone book;
(2) content of text in the address block is carried out field and extract processing.
9. the method for using image recognition to carry out the collection and the typing of card information in the mobile phone according to claim 8 is characterized in that the structure of the content of text in the described address block is:
＜address block 〉: :=＜field 1〉|＜field 2〉|＜field 3〉...
＜field i 〉: :=＜first introductory word〉|＜delimiter〉|＜content 〉
＜content 〉: :=＜character 1〉|＜character 2〉|＜character 3〉...
The delimiter set=': ', '. ', ' '
Elder generation's introductory word, content are for example had the phone field by the decision of field kind:
Guide's set of words=and ' phone ', ' TEL ', ' PHONE ' ...
The content character set=numerical digit, ' (', ') ', '-', ' ', '. ',
Described field extracts to handle and may further comprise the steps:
(1) the fault-tolerant coupling of first introductory word;
(2) skip delimiter;
(3) judge according to the content character set whether content is legal;
(4) if content legality is the basis at the special-purpose dictionary of business card and rule base then, proofreads recognition result automatically and revise manifest error;
(5) otherwise, then get back to step (1).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100264922A CN100362525C (en) | 2005-06-06 | 2005-06-06 | Method for gathering and recording business card information in mobile phone by using image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100264922A CN100362525C (en) | 2005-06-06 | 2005-06-06 | Method for gathering and recording business card information in mobile phone by using image recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1877598A CN1877598A (en) | 2006-12-13 |
| CN100362525C true CN100362525C (en) | 2008-01-16 |
Family
ID=37510028
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100264922A Active CN100362525C (en) | 2005-06-06 | 2005-06-06 | Method for gathering and recording business card information in mobile phone by using image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100362525C (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104298982A (en) * | 2013-07-16 | 2015-01-21 | 深圳市腾讯计算机***有限公司 | Text recognition method and device |
Families Citing this family (40)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101122953B (en) * | 2007-09-21 | 2010-11-17 | 北京大学 | Picture words segmentation method |
| CN101751433B (en) * | 2008-12-22 | 2012-10-17 | 汉王科技股份有限公司 | Method for classifying business card character clauses and device thereof |
| CN101702242B (en) * | 2009-11-23 | 2012-07-04 | 中兴通讯股份有限公司 | Method and terminal for automatically filling in document with fixed format |
| CN101739441B (en) * | 2009-12-01 | 2012-01-25 | 中国建设银行股份有限公司 | Method of image information input and system thereof |
| CN102236782A (en) * | 2010-04-30 | 2011-11-09 | 陈斌 | Character identification method for certificates |
| CN101887521B (en) * | 2010-06-22 | 2013-09-11 | 中兴通讯股份有限公司 | Method and terminal for rectifying deviation of file |
| CN101916258B (en) * | 2010-07-13 | 2012-04-18 | 上海合合信息科技发展有限公司 | Method and system for automatically correcting identification result through network query |
| CN102567005A (en) * | 2010-12-31 | 2012-07-11 | 上海量明科技发展有限公司 | Method, device and system for realizing classification of business cards by using instant messaging tool |
| CN102169541A (en) * | 2011-04-02 | 2011-08-31 | 郝震龙 | Character recognition input system using optical localization and method thereof |
| JP2013037462A (en) * | 2011-08-05 | 2013-02-21 | Sony Corp | Information processor and information processing method |
| CN102932403A (en) * | 2012-08-30 | 2013-02-13 | 上海合合信息科技发展有限公司 | Method and system for automatically synchronizing contact information in calling card to mobile device |
| CN102945367A (en) * | 2012-11-26 | 2013-02-27 | 昆山振天智能化设备有限公司 | Vehicle license plate recognition system |
| CN103065146A (en) * | 2012-12-24 | 2013-04-24 | 广东电网公司电力调度控制中心 | Character recognition method for power communication machine room dumb equipment signboards |
| CN103235946A (en) * | 2013-04-08 | 2013-08-07 | 上海合合信息科技发展有限公司 | Divulgence-preventive processing method for artificially identifying information of business cards |
| CN103488608A (en) * | 2013-09-27 | 2014-01-01 | 杨昕吉 | Method and system for quickly creating visual-style electronic business card |
| CN104679800A (en) * | 2013-12-03 | 2015-06-03 | 中兴通讯股份有限公司 | Data inputting method and device |
| CN103810477B (en) * | 2014-02-21 | 2017-04-12 | 中国人民银行印制科学技术研究所 | Fault tolerance locating system and fault tolerance locating method |
| CN105335957A (en) * | 2014-08-07 | 2016-02-17 | 江苏奥博洋信息技术有限公司 | Information fragmentization picture cutting method |
| CN104239853B (en) * | 2014-08-27 | 2018-04-27 | 北京捷通华声语音技术有限公司 | A kind for the treatment of method and apparatus of image |
| CN105718846A (en) * | 2014-12-03 | 2016-06-29 | 航天信息股份有限公司 | Bill information input method and apparatus |
| CN104537339B (en) * | 2014-12-19 | 2018-04-10 | 汉王科技股份有限公司 | Information identifying method and information identification system |
| CN104915114B (en) * | 2015-05-29 | 2018-10-19 | 小米科技有限责任公司 | Information recording method and device, intelligent terminal |
| CN105373791B (en) * | 2015-11-12 | 2018-12-14 | 中国建设银行股份有限公司 | Information processing method and information processing unit |
| CN105512654A (en) * | 2016-02-19 | 2016-04-20 | 杭州泰格医药科技股份有限公司 | Handheld data acquisition device for clinical test |
| 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 |
| CN106412196A (en) * | 2016-05-31 | 2017-02-15 | 深圳市永兴元科技有限公司 | Contact person information processing method and apparatus |
| CN107105167B (en) * | 2017-06-05 | 2020-05-15 | 广东小天才科技有限公司 | Method and device for shooting picture during scanning question and terminal equipment |
| CN107403140A (en) * | 2017-07-04 | 2017-11-28 | 南京律智诚专利技术开发有限公司 | A kind of card information identification and business card image matching management method |
| CN107491640A (en) * | 2017-08-12 | 2017-12-19 | 合肥全飞信息科技有限公司 | A kind of nail disease data query method based on cell phone application |
| CN107783953A (en) * | 2017-09-22 | 2018-03-09 | 平安普惠企业管理有限公司 | Information input method and terminal device |
| CN111213157A (en) * | 2017-10-11 | 2020-05-29 | 深圳传音通讯有限公司 | Express information input method and system based on intelligent terminal |
| CN108038504B (en) * | 2017-12-11 | 2019-12-27 | 深圳房讯通信息技术有限公司 | Method for analyzing content of house property certificate photo |
| CN108090488A (en) * | 2017-12-18 | 2018-05-29 | 广东广业开元科技有限公司 | A kind of artificial intelligence labour union Message Entry System based on big data OCR Internet of Things |
| CN108415930B (en) * | 2018-01-19 | 2021-07-09 | 大象慧云信息技术有限公司 | Data analysis method and device |
| CN109492532A (en) * | 2018-10-09 | 2019-03-19 | 好活（昆山）网络科技有限公司 | A kind of methods, devices and systems of image AI identification |
| CN109635807A (en) * | 2018-10-16 | 2019-04-16 | 深圳壹账通智能科技有限公司 | Information input method, device, equipment and computer readable storage medium |
| CN111340031A (en) * | 2020-02-25 | 2020-06-26 | 杭州测质成科技有限公司 | Equipment almanac target information extraction and identification system based on image identification and method thereof |
| CN111753840A (en) * | 2020-06-18 | 2020-10-09 | 北京同城必应科技有限公司 | Ordering technology for business cards in same city logistics distribution |
| CN112466199A (en) * | 2020-11-26 | 2021-03-09 | 联盛（厦门）彩印有限公司 | Automatic typesetting method, system, equipment and storage medium for electronic tag hang tag |
| CN112929480A (en) * | 2021-02-10 | 2021-06-08 | 王晨旭 | Storage method and storage device of contact mode |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1561126A (en) * | 2004-02-18 | 2005-01-05 | 惠州Tcl移动通信有限公司 | Method for transmitting message on name card to telepone number book of mobile phone |
| CN1592340A (en) * | 2003-08-27 | 2005-03-09 | 致伸科技股份有限公司 | Mobile telephone system with calling card scanning identification function |
| US20050100216A1 (en) * | 2003-11-11 | 2005-05-12 | Sri International | Method and apparatus for capturing paper-based information on a mobile computing device |
| 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 |
-
2005
- 2005-06-06 CN CNB2005100264922A patent/CN100362525C/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1592340A (en) * | 2003-08-27 | 2005-03-09 | 致伸科技股份有限公司 | Mobile telephone system with calling card scanning identification function |
| US20050100216A1 (en) * | 2003-11-11 | 2005-05-12 | Sri International | Method and apparatus for capturing paper-based information on a mobile computing device |
| 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 |
| CN1561126A (en) * | 2004-02-18 | 2005-01-05 | 惠州Tcl移动通信有限公司 | Method for transmitting message on name card to telepone number book of mobile phone |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104298982A (en) * | 2013-07-16 | 2015-01-21 | 深圳市腾讯计算机***有限公司 | Text recognition method and device |
| CN104298982B (en) * | 2013-07-16 | 2019-03-08 | 深圳市腾讯计算机***有限公司 | A kind of character recognition method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1877598A (en) | 2006-12-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100362525C (en) | Method for gathering and recording business card information in mobile phone by using image recognition | |
| CN108960223B (en) | Method for automatically generating voucher based on intelligent bill identification | |
| US7734636B2 (en) | Systems and methods for electronic document genre classification using document grammars | |
| CN106681633B (en) | System and method for auxiliary information input control function of sliding operation of portable terminal equipment | |
| US7860312B2 (en) | System and method for identifying and labeling fields of text associated with scanned business documents | |
| CN101908136B (en) | Table identifying and processing method and system | |
| US11348330B2 (en) | Key value extraction from documents | |
| JP4577931B2 (en) | Document processing system and index information acquisition method | |
| CN105930159A (en) | Image-based interface code generation method and system | |
| CN1488120A (en) | Method, device and computer program for recognition of a handwritten character | |
| CN103577818A (en) | Method and device for recognizing image characters | |
| KR102373884B1 (en) | Image data processing method for searching images by text | |
| CN102945377B (en) | Obtain method and the device of content in papery notebook | |
| CN107403140A (en) | A kind of card information identification and business card image matching management method | |
| JP4991407B2 (en) | Information processing apparatus, control program thereof, computer-readable recording medium storing the control program, and control method | |
| CN103778141A (en) | Mixed PDF book catalogue automatic extracting algorithm | |
| CN105956098B (en) | A kind of correlating method and system of paper printed matter and e-sourcing | |
| US9323726B1 (en) | Optimizing a glyph-based file | |
| CN116092231A (en) | Ticket identification method, ticket identification device, terminal equipment and storage medium | |
| CN100485711C (en) | Computer identification and automatic inputting method for hand writing character font | |
| CN1459073A (en) | Method and arrangement for identifying and processing commands in digital images, where user marks the command, for example by encircling it | |
| JP4597644B2 (en) | Character recognition device, program and recording medium | |
| CN1336604A (en) | Method and system of digitizing ancient Chinese books and automatizing the content search | |
| CN105791503A (en) | Method of storing business card information in address list and apparatus thereof | |
| CN115774805A (en) | File intelligent query method and system based on digital processing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20201020 Address after: No.018, floor 8, building 6, yard 33, middle Xierqi Road, Haidian District, Beijing 100085 Patentee after: BEIJING XIAOMI MOBILE SOFTWARE Co.,Ltd. Address before: 200233 Shanghai City Guangxi Jinglu No. 7 Patentee before: Inventec Appliances (Shanghai) Co.,Ltd. |
|
| TR01 | Transfer of patent right |