CN104346390B - A kind of method and device for forming character library - Google Patents

A kind of method and device for forming character library Download PDF

Info

Publication number
CN104346390B
CN104346390B CN201310332045.4A CN201310332045A CN104346390B CN 104346390 B CN104346390 B CN 104346390B CN 201310332045 A CN201310332045 A CN 201310332045A CN 104346390 B CN104346390 B CN 104346390B
Authority
CN
China
Prior art keywords
contour curve
character outline
character
contour
curve
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
Application number
CN201310332045.4A
Other languages
Chinese (zh)
Other versions
CN104346390A (en
Inventor
王玉欣
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
New Founder Holdings Development Co ltd
Original Assignee
Founder Information Industry Holdings Co Ltd
Peking University Founder Group Co Ltd
Beijing Founder Electronics Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Founder Information Industry Holdings Co Ltd, Peking University Founder Group Co Ltd, Beijing Founder Electronics Co Ltd filed Critical Founder Information Industry Holdings Co Ltd
Priority to CN201310332045.4A priority Critical patent/CN104346390B/en
Publication of CN104346390A publication Critical patent/CN104346390A/en
Application granted granted Critical
Publication of CN104346390B publication Critical patent/CN104346390B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/40Document-oriented image-based pattern recognition
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing
    • G06V10/22Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/40Extraction of image or video features
    • G06V10/44Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Controls And Circuits For Display Device (AREA)

Abstract

The present invention provides a kind of method for forming character library, comprises the following steps:1)Obtain word original text;2)The word original text is scanned into word manuscript base picture;3)Fitting is digitized to the word manuscript base picture, to form character outline;4)Automatic business processing is carried out to the character outline, to remove point unnecessary on contour curve and contour curve unnecessary in character outline;5)Dress up character library and test.Correspondingly, a kind of device for forming character library is also provided.The personal character library that the method and device of the present invention for forming character library is formed data volume compared with existing personal character library is smaller.

Description

A kind of method and device for forming character library
Technical field
The present invention relates to computer library software development technical field, and in particular to a kind of method and dress for forming character library Put.
Background technology
At present, the manufacturing process of computer library is roughly divided into following several stages:Design word original text;By designed word original text Scanning input computer;Fitting is digitized to the word original text after scanning;Manually repair word;Quality inspection;It is integrated into storehouse etc..Although using The efficiency that computer is digitized fitting to the word original text after scanning is very high, but just as computer can not substitute the work one of human brain Sample, it can be only done just level work, either character(For Chinese word library, letter that character refers to using in computer, Numeral, symbol or Chinese character)Quality, or structure, only artificial repair word by follow-up and could form satisfactory character library. Artificial to repair word be the very great system engineering of a workload, because simplified form of Chinese Character character library includes six or seven thousand Chinese characters, in traditional font Literal pool includes 14,000 Chinese characters, and GBK character libraries then include more than 20,000 Chinese characters.Each character after manually repairing word will also Check while a character, character are needed by strict quality inspection, i.e. quality inspection, during quality inspection, or even need point, one Check individual point because a character is well-done, except character outline it is smooth, it is rational in infrastructure in addition to, also have a whole set of skill Art specification needs to follow, for example in the distalmost end of curve an extreme point, each stroke need to be added to be retouched with minimum point State, to be reduced as far as storage information, improve reduction rate etc..
Computer library is developed so far, and the exploitation of " character library " product is also had deep love for calligraphy and Chinese character is being believed by more and more Personal concern interested is used in the diversification in breath epoch.Font manufacturer also experiences word during new font is developed Great demand of the storehouse outside traditional publication and distribution.Meanwhile the revolution of information propagation pattern that Internet era brings is to meter Calculation machine character library with bringing new problem --- the media transmission mode such as personal blog, personalized publication thing emerges in an endless stream, It is required that " computer font " this important transmitting carrier can serve " personalization " expression and personal individual character to a greater degree Show, therefore a kind of personal character library different from traditional computer character library occur.The personal character library is according to individual The demand of calligraphist and its fan colony and the caused personal computer being transformed into personal handwritten font truly Character library product, its birth indicate that " computer library " will enter the personalized epoch as a kind of " consumer products ".
But because personal character library is different from traditional computer library, it is primarily as " consumption is produced Product " position, the uniform technical specification that not must comply with, therefore current processing of the manufacturer for personal character library and oneself release Fine work character library be on Making programme it is discrepant, in order to reduce production cost, shorten the production cycle, existing personal character library Making programme reduce compared with traditional computer character library and manually repair word and quality inspection flow.Personal character library is firstly the need of author Required all characters are all write in word original text, then by word original text scanning input computer, after being digitized into fitting, directly integrated Into character library, if there is wrongly written character during this, it is only necessary to wrongly written character is modified, other equal retention figuresization fittings of character Initial data, and without manually repairing the link of word and quality inspection.As a result, in the personal character library formed character word Symbol profile is rough, and the point that each character outline of storage includes is too many, the data volume of the personal character library resulted in itself It is bigger(That is personal character library file is larger), it is more than 2 times of normal computer character library, wherein, that Brush calligraphy is formed Property character library data volume it is maximum, 3-4 times may be reached, because Brush calligraphy is typically write on rice paper to form word original text, through sweeping There are many burrs at the edge of the word manuscript base picture formed after retouching, not smooth enough, and the word formed after digitlization is fitted The point that symbol profile includes is especially more.Therefore, the big basic reason of personal character library data volume is the point that each character outline includes Too much, too close, character outline is not smooth enough.In the quality of character in ensureing character library and while do not improve production cost, how Effectively reduce personal character library file size, be that font manufacturer releases one that personal character library this " consumer products " faces The problem of urgent need to resolve.
The content of the invention
The technical problems to be solved by the invention are to be directed to the drawbacks described above in the presence of prior art, there is provided one kind is formed The method and device of character library, its personal character library formed data volume compared with existing personal character library are smaller.
Technical scheme used by solution present invention problem:
The method for forming character library comprises the following steps:
1)Obtain word original text;
2)The word original text is scanned into word manuscript base picture;
3)Fitting is digitized to the word manuscript base picture, to form character outline;
4)Automatic business processing is carried out to the character outline, it is bent to remove contour curve unnecessary in character outline and profile Unnecessary point on line;
5)Dress up character library and test.
Preferably, the step 2)In, the word manuscript base picture is bianry image.
Preferably, the step 3)Also include afterwards:
3A. density of setting threshold values, judgment step 3)In the character outline of middle formation in any region on all contour curves The density of point whether be more than the density threshold, in this way, then step 3B is performed, if not, performing step 4);
Point construction approximating functions of the 3B. according to step 3A in region on all contour curves, with corresponding to formation Matched curve, and all contour curves in the region are substituted with the matched curve, it is then back to step 3A.
Preferably, the step 4)In, removing contour curve unnecessary in character outline is specially:
Remove in character outline and the contour curve that is formed after fitting is digitized into by the noise of the word manuscript base picture;
And/or default first area threshold, if the area for the closed figure being made up of in character outline a plurality of contour curve Less than the area threshold, then a plurality of contour curve of the composition closed figure is removed.
Preferably, the step 4)In, removing point unnecessary on contour curve in character outline is specially:
Preset curvature radius threshold, if the radius of curvature minimum value of any bar contour curve is more than the song in character outline Rate radius threshold, then remove the control point on the contour curve;
And/or if a plurality of contour curve being sequentially connected is respectively positioned on same straight line in character outline, it is a plurality of to remove this The public point of every two connected contour curves in the contour curve being sequentially connected;
And/or pre-determined distance threshold value, if the distance between two end points in character outline on any bar contour curve are small In the distance threshold, then any one in two end points on the contour curve is removed;
And/or default second area threshold value, if a plurality of contour curve group being sequentially connected in any region in character outline Into the area of kick be less than the second area threshold value, then remove every wheel in a plurality of contour curve for forming the kick The public point of control point on wide curve and every two contour curves that are connected.
Present invention simultaneously provides a kind of device for forming character library, including:Acquiring unit, scanning element, digitlization fitting are single Member, automatic business processing unit and integration testing unit;
The acquiring unit is used to obtain word original text, and the word original text of acquisition is sent to scanning element;
The scanning element is used to the word original text being scanned into word manuscript base picture, and the word manuscript base picture is sent to digitlization Fitting unit;
The digitlization fitting unit is used to be digitized fitting to the word manuscript base picture, to form character outline, and The character outline is sent to automatic business processing unit;
The automatic business processing unit is used to carry out automatic business processing to the character outline, more in character outline to remove Unnecessary point on remaining contour curve and contour curve, and the character outline after processing is sent to integration testing unit;
The integration testing unit is used to the character outline after the processing is dressed up into character library and tested.
Preferably, the word manuscript base picture that the scanning element is scanned into is bianry image.
Preferably, in addition to judging unit and fitting unit is approached;
Density threshold is preset with the judging unit, is appointed in the character outline formed for judging digitlization fitting unit Whether the density of the point in one region on all contour curves is more than the density threshold;
If not, judging that all contour curves in the region meet the requirements, and qualifying signal is sent to automation Processing unit, so that must automate processing unit carries out automatic business processing to satisfactory character outline;
In this way, then judge that all contour curves in the region are undesirable, and send unqualified signal to approaching Fitting unit;
The fitting unit that approaches is used for when receiving the unqualified signal, bent according to all profiles in the region Point construction approximating function on line, to form corresponding matched curve, and is substituted in the region with the matched curve and owned Whether contour curve, the density for then continuing to judge to put in the region by judging unit are more than the density threshold, until sentencing Untill disconnected unit judges that the matched curve in the region meets the requirements.
Preferably, contour curve unnecessary in the automatic business processing unit removal character outline is specially:
Remove after being digitized into fitting by the noise of the word manuscript base picture in the character outline that digitlization fitting unit is formed The contour curve formed;
And/or the first area threshold is preset with it, if by a plurality of wheel in the character outline that digitlization fitting unit is formed The area of the closed figure of wide curve composition is less than the area threshold, then a plurality of profile for removing the composition closed figure is bent Line.
Preferably, point unnecessary on contour curve in the automatic business processing unit removal character outline is specially:
Radius of curvature threshold value is preset with it, if any bar contour curve in the character outline that digitlization fitting unit is formed Radius of curvature minimum value be more than the radius of curvature threshold value, then remove the control point on the contour curve;
And/or if a plurality of contour curve being sequentially connected is respectively positioned on same straight line in character outline, it is a plurality of to remove this The public point of every two connected contour curves in the contour curve being sequentially connected;
And/or distance threshold is preset with it, if any bar profile is bent in the character outline that digitlization fitting unit is formed The distance between two end points on line are less than the distance threshold, then remove any in two end points on the contour curve It is individual;
And/or second area threshold value is preset with it, if any region in the character outline that digitlization fitting unit is formed The area of the kick of the interior a plurality of contour curve composition being sequentially connected is less than the second area threshold value, then removing composition, this is small The public point of control point in a plurality of contour curve of projection on every contour curve and every two contour curves that are connected.
Beneficial effect:
In the character library that the method and device of the present invention for forming character library is formed, character outline smoother, character wheel Contour curve negligible amounts in exterior feature(The point that i.e. character outline includes is less), therefore the data volume of character library is smaller.That is, The character library that is formed of the present invention is as a kind of personal character library, compared with existing personal character library, both can guarantee that wherein character Quality, and production cost is not improved, font file is also smaller, thus it is bigger than normal to efficiently solve existing personal character library file Problem.
Brief description of the drawings
Fig. 1 is the method flow diagram that personal character library is formed in the embodiment of the present invention 1;
Fig. 2 is the word manuscript base picture of Chinese character " ";
Fig. 3 is that the character outline formed after fitting is digitized to the picture of word manuscript base shown in Fig. 2;
Fig. 4 is that the character outline after automatic business processing is carried out to character outline shown in Fig. 3;
Fig. 5 is that the artificial character outline repaiied after word is carried out to character outline shown in Fig. 4;
Fig. 6 is the method flow diagram that personal character library is formed in the embodiment of the present invention 2;
Fig. 7 is the structural representation for the device that personal character library is formed in the embodiment of the present invention 3;
Fig. 8 is the structural representation for the device that personal character library is formed in the embodiment of the present invention 4.
Embodiment
To make those skilled in the art more fully understand technical scheme, with reference to the accompanying drawings and examples to this The method and device for inventing the formation character library is described in further detail.
It should be noted that character outline is made up of a plurality of contour curve, the contour curve is conic section, three times song Line or straight line.In the present invention, " point on contour curve " is not meant to that " point " one is positioned on " contour curve ", actually refers to It is " point corresponding to contour curve ".Because contour curve is recorded in a manner of point(Storage), the point includes end points And control point;The end points is positioned at the stem or afterbody of every contour curve, the starting point of as every contour curve or termination Point;The control point is used for the shape for controlling every contour curve, in most cases not on contour curve.Moreover, work as institute When to state contour curve be conic section, the corresponding control point of the contour curve and two end points, i.e. a control point and two End points records a conic section;When the contour curve is cubic curve, corresponding two control points and two of the contour curve Individual end points, i.e. two control points and two end points record a cubic curve;When the contour curve is straight line, the profile is bent Line only corresponds to two end points, i.e. two end points record straight lines.End points corresponding to two connected contour curves is three, i.e., The public point of one of end points two contour curve connected as this is corresponding with two contour curves respectively, this In, two contour curves, which are connected, refers to the starting point of wherein one contour curve or rising for terminating point and another contour curve Initial point or terminating point overlap and the situation of public point as two contour curves, without referring to two contour curves friendships The situation of fork.In the present invention, it is described it is a plurality of refer to two or more, such as a plurality of contour curve refers to two or two Contour curve more than bar;The line refers to line segment.
Embodiment 1:
As shown in figure 1, the present embodiment provides a kind of method for forming personal character library, comprise the following steps:
S101. word original text is obtained.
All characters needed for character library are required to write in advance in word original text, for example do simplified character library, it then follows GB2312- 1980, it is necessary to writing Chinese characters 6763.The sequential write of character is prior by designer in the format write and word original text of word original text Set.
S102. the word original text is scanned into word manuscript base picture.
Due to word original text be write on paper or other carriers on, it is necessary to by its scanning input computer.
Preferably, the word manuscript base picture is bianry image(binary image).Bianry image refers to each pixel only There is the digital picture of two probable values.For example, conventional black and white, monochrome image represent bianry image, its advantage is space-consuming It is few.Here, it is necessary to neatly adjust the resolution ratio of word manuscript base picture and other relevant parameters according to the size of word original text and definition, with Word manuscript base picture is enabled to reflect the overall picture of word original text without distortion.The word manuscript base picture that the word original text of Brush calligraphy " " is scanned into is as schemed Shown in 2.
S103. fitting is digitized to the word manuscript base picture, to form character outline.
The word manuscript base formed will automatically be scanned by Fontlab ScanFont softwares or other digitlization fitting softwares Picture(Bianry image)Digitlization is fitted to the character outline as close possible to word original text.The picture of word manuscript base shown in Fig. 2 is digitized into being fitted The character outline formed afterwards is as shown in Figure 3, it is seen that the contour curve in character outline shown in Fig. 3 is especially more, and character outline includes Point it is especially more, and character outline is not smooth enough, there is many kicks.
S104. according to the characteristics of the character outline of formation, automatic business processing is carried out to the character outline, to remove character Point unnecessary on unnecessary contour curve and contour curve in profile.Character of the character outline shown in Fig. 3 after automatic business processing Profile is as shown in Figure 4, it is seen that compared with Fig. 3, the contour curve in character outline shown in Fig. 4 is many less, what character outline included Point is also few a lot, and character outline is relatively smooth.
Preferably, removing contour curve unnecessary in character outline is specially:
Remove in character outline and the contour curve that is formed after fitting is digitized into by the noise of the word manuscript base picture;It is described Noise be by word original text be scanned into word manuscript base as when formed in word manuscript base picture, i.e., occur occurring in word manuscript base picture is outer Carry out pixel, it is generally produced by electronic interferences;
And/or default first area threshold, if the area for the closed figure being made up of in character outline a plurality of contour curve Less than the area threshold, then a plurality of contour curve of the composition closed figure is removed.When using Brush calligraphy written word original text, word It is possible that blank spot in stroke on original text, such as the blank spot that the upper right portion of " mouth " occurs by the left avertence of " " in Fig. 2, this Kind blank spot can form the closed figure being made up of a plurality of contour curve, and the Closed Graph after step s103 digitlization fitting The area of shape is all smaller, and the closed figure is as shown in figure 3, therefore can be by way of default first area threshold, by group The part contour curve for being less than the closed figure of the first area threshold into area removes, and removes the closing song for meeting above-mentioned condition Character outline after line is as shown in Figure 4;First area threshold can by those skilled in the art according to character outline in itself Stroke feature sets itself.
Preferably, removing point unnecessary on contour curve in character outline is specially:
Preset curvature radius threshold, if the radius of curvature minimum value of any bar contour curve is more than the song in character outline Rate radius threshold, then the control point on the contour curve is removed, so as to by class straight line in character outline(Or it is near linear) Conic section, cubic curve be revised as straight line;Conic section, the cubic curve of the class straight line refer respectively to radius of curvature Minimum value is more than conic section, the cubic curve of the radius of curvature threshold value, because the conic section of this species straight line, three times song Curvature of a curve is smaller(I.e. radius of curvature is larger), do not have camber, be revised as not influenceing whole character outline after straight line Perception, the quantity at control point can be also reduced, and then reduce the quantity for the point that character outline includes;The radius of curvature threshold value can be by Stroke feature sets itself of the those skilled in the art according to character outline in itself;
And/or if a plurality of contour curve being sequentially connected is respectively positioned on same straight line in character outline, it is a plurality of to remove this The public point of every two connected contour curves in the contour curve being sequentially connected;That is, the word manuscript base picture is through numeral Change in the character outline formed after fitting, it is understood that there may be a plurality of straight line is sequentially connected and the situation without knuckle, and it is more to remove this The public point of bar straight line may be such that a plurality of straight line becomes in alignment, if for example, three straight lines are sequentially connected and without knuckle When one share four end points, two of which is public point, and after two public points are removed, and can cause three straight lines Become in alignment, therefore reduce contour curve in character outline(For straight line when)Quantity and point quantity;
And/or pre-determined distance threshold value, if the distance between two end points in character outline on any bar contour curve are small In the distance threshold, then remove any one in two end points on the contour curve, with cause the contour curve and with Its connected another contour curve is changed into a contour curve(I.e. two contour curves merge into a contour curve), and this goes The end points removed is the public point of the contour curve and coupled another contour curve;It is if it should be noted that described Contour curve and coupled another contour curve are respectively conic section(Or cubic curve)And straight line, then it is public to remove its After end points, the conic section(Or cubic curve)Control point turn into merge after contour curve control point;If the profile Curve and coupled another contour curve are conic section, then after removing its public point, this two conic sections Control point only remain it is next, can be related to the sequential write of character as which remaining control point, also can be advance by designer Setting;If the contour curve and coupled another contour curve are cubic curve, after removing its public point, this The control point of two cubic curves is only left two, can be related to the sequential write of character as which remaining two control point, It can be preset by designer;If the contour curve and coupled another contour curve are respectively conic section and three Secondary curve, then after removing its public point, the control point of the contour curve formed may be one, it is also possible to be two, extremely In remaining several control points, and which remaining control point can be related to the sequential write of character, also can be pre- by designer First set;The distance threshold can be by stroke feature sets itself of the those skilled in the art according to character outline in itself;
And/or default second area threshold value, if a plurality of contour curve group being sequentially connected in any region in character outline Into the area of kick be less than the second area threshold value, then remove every wheel in a plurality of contour curve for forming the kick The public point of control point on wide curve and every two contour curves that are connected, to eliminate the kick.Such as the left side of " " in Fig. 2 Many burrs that the left-half of radical " mouth " occurs, the burr can be formed by a plurality of wheel after step s103 digitlization fitting The kick of wide curve composition, as shown in figure 3, composition area can be less than into second by way of default second area threshold value The control point of the part contour curve of the kick of area threshold removes with the public point of every two contour curves that are connected, and goes Except meeting, the character outline after the kick of above-mentioned condition is as shown in Figure 4;The second area threshold value can be by people in the art Stroke feature sets itself of the member according to character outline in itself.
S105. dress up character library and test.This step is prior art, is repeated no more.
It can be seen that the method that personal character library is formed described in the present embodiment is compared with prior art, step s104 merely add The quantity for the point that the quantity and character outline for the contour curve that can be reduced to a great extent in character outline include so that character Profile is relatively smooth, also eliminates the kick in the character outline formed by the burr in word original text, therefore the individual character formed It is smaller to change character library data volume compared with the personal character library that prior art is formed, solves the text that existing personal character library is faced The problem of part is too big.Through actual test, the file size for the personal character library that the present embodiment methods described is formed only has existing 2/3rds or so of property font file size.
The method that personal character library is formed described in the present embodiment can also be applied in the manufacturing process of active computer character library, Artificial the step of repairing word and quality inspection need to only be increased between step s104 and step s105.The formation of active computer character library Method is directly to carry out the word manuscript base picture after digitlization is fitted manually to repair word, for example, directly by character outline shown in Fig. 3 The form of character outline shown in Fig. 4 is modified as by artificial mode, workload is very big, and should by the present embodiment methods described After manufacturing process for active computer character library, due to adding the step of carrying out automatic business processing to character outline, therefore Character outline shown in Fig. 3 is carried out without directly manually to repair word, character outline shown in Fig. 4 after automatic business processing need to only be entered Pedestrian's work repaiies word, and Fig. 4 is compared with painting, the quantity for the point that the quantity of contour curve and character outline include in character outline It is few a lot, therefore the artificial workload for repairing word is considerably reduced using the present embodiment methods described, shorten computer library The generation cycle.
Embodiment 2:
In the present embodiment, the method for forming personal character library comprises the following steps:
S201-s203 is identical with the s101-s103 in embodiment 1, repeats no more.
S204. density of setting threshold value, all profiles are bent in any region in the character outline formed in judgment step s203 Whether the density of the point on line is more than the density threshold, in this way, then performs step s205, if not, performing step s206.
S205. the point construction approximating function according to step s204 in region on all contour curves, with formation pair The matched curve answered, and all contour curves in the region are substituted with the matched curve, it is then back to step s204.
S206-s207 is identical with the s104-s105 in embodiment 1, repeats no more.
It can be seen that if it find that the point being digitized into after fitting in the character outline that is formed in certain panel region is relatively intensive, then Approximating function is constructed based on the point in the region on all contour curves, to form corresponding matched curve, i.e., using song Line approaches the point being fitted in the region, the contour curve being fitted again in this way in the region, reaches and reduces the area The purpose put in domain.The error that curve approaches can be set according to the actual conditions of character outline in the region;The density The position and area in threshold value and the region can voluntarily be set by those skilled in the art according to the stroke feature of character outline in itself It is fixed.
Other method and effect in the present embodiment is all same as Example 1, repeats no more here.
Embodiment 3:
As shown in fig. 7, the present embodiment provides a kind of device for forming personal character library, including:Acquiring unit, scanning are single Member, digitlization fitting unit, automatic business processing unit and integration testing unit.
The acquiring unit is used to obtain word original text, and the word original text of acquisition is sent to scanning element.
The scanning element is used to the word original text being scanned into word manuscript base picture, and the word manuscript base picture is sent to digitlization Fitting unit.Preferably, the word manuscript base picture that the scanning element is scanned into is bianry image.
The digitlization fitting unit is used to be digitized fitting to the word manuscript base picture, to form character outline, and The character outline is sent to automatic business processing unit.
The automatic business processing unit is used to carry out automatic business processing to the character outline, more in character outline to remove Unnecessary point on remaining contour curve and contour curve, and the character outline after processing is sent to integration testing unit.
The integration testing unit is used to the character outline after the processing is dressed up into character library and tested.
Preferably, contour curve unnecessary in the automatic business processing unit removal character outline is specially:
Remove after being digitized into fitting by the noise of the word manuscript base picture in the character outline that digitlization fitting unit is formed The contour curve formed;
And/or the first area threshold is preset with it, if by a plurality of wheel in the character outline that digitlization fitting unit is formed The area of the closed figure of wide curve composition is less than the area threshold, then a plurality of profile for removing the composition closed figure is bent Line;First area threshold can be by stroke feature sets itself of the those skilled in the art according to character outline in itself.
Preferably, point unnecessary on contour curve in the automatic business processing unit removal character outline is specially:
Radius of curvature threshold value is preset with it, if any bar contour curve in the character outline that digitlization fitting unit is formed Radius of curvature minimum value be more than the radius of curvature threshold value, then remove the control point on the contour curve;The radius of curvature Threshold value can be by stroke feature sets itself of the those skilled in the art according to character outline in itself;
And/or if a plurality of contour curve being sequentially connected is respectively positioned on same straight line in character outline, it is a plurality of to remove this The public point of every two connected contour curves in the contour curve being sequentially connected;
And/or distance threshold is preset with it, if any bar profile is bent in the character outline that digitlization fitting unit is formed The distance between two end points on line are less than the distance threshold, then remove any in two end points on the contour curve It is individual;The distance threshold can be by stroke feature sets itself of the those skilled in the art according to character outline in itself;
And/or second area threshold value is preset with it, if any region in the character outline that digitlization fitting unit is formed The area of the kick of the interior a plurality of contour curve composition being sequentially connected is less than the second area threshold value, then removing composition, this is small The public point of control point in a plurality of contour curve of projection on every contour curve and every two contour curves that are connected;It is described Second area threshold value can be by stroke feature sets itself of the those skilled in the art according to character outline in itself.
Embodiment 4:
The present embodiment and the difference of embodiment 3 are:
In the present embodiment, the device for forming personal character library also includes judging unit and approaches fitting unit;
Density threshold is preset with the judging unit, is appointed in the character outline formed for judging digitlization fitting unit Whether the density of the point in one region on all contour curves is more than the density threshold;
If not, judging that all contour curves in the region meet the requirements, and qualifying signal is sent to automation Processing unit, so that must automate processing unit carries out automatic business processing to satisfactory character outline;
In this way, then judge that all contour curves in the region are undesirable, and send unqualified signal to approaching Fitting unit;
The fitting unit that approaches is used for when receiving the unqualified signal, bent according to all profiles in the region Point construction approximating function on line, to form corresponding matched curve, and is substituted in the region with the matched curve and owned Contour curve;
Whether the density for then continuing to judge to put in the region by judging unit is more than the density threshold, in this way, then Approach fitting unit to be again fitted the point in the region, that is, reconfigure approximating function, it is bent to form new fitting Line, and with the matched curve of the instead preceding formation of the new matched curve, and so on, until judging unit judges the area Matched curve in domain meets the requirements(The density put in i.e. described region is less than the density threshold)Untill, reduce institute to reach State the purpose put in region
The position and area in the density threshold and the region can by those skilled in the art according to character outline in itself Stroke feature sets itself.
Other method and effect in the present embodiment is all same as Example 3, repeats no more here.
It is understood that the principle that embodiment of above is intended to be merely illustrative of the present and the exemplary implementation that uses Mode, but the invention is not limited in this.For those skilled in the art, the essence of the present invention is not being departed from In the case of refreshing and essence, various changes and modifications can be made therein, and these variations and modifications are also considered as protection scope of the present invention.

Claims (8)

  1. A kind of 1. method for forming character library, it is characterised in that comprise the following steps:
    1) word original text is obtained;
    2) the word original text is scanned into word manuscript base picture;
    3) fitting is digitized to the word manuscript base picture, to form character outline;
    4) automatic business processing is carried out to the character outline, to remove on contour curve and contour curve unnecessary in character outline Unnecessary point;
    In the step 4), removing point unnecessary on contour curve in character outline is specially:
    Preset curvature radius threshold, if the radius of curvature minimum value of any bar contour curve is more than the curvature half in character outline Footpath threshold value, then remove the control point on the contour curve;
    And/or if a plurality of contour curve being sequentially connected is respectively positioned on same straight line in character outline, it is a plurality of successively to remove this The public point of every two connected contour curves in connected contour curve;
    And/or pre-determined distance threshold value, if the distance between two end points in character outline on any bar contour curve are less than institute Distance threshold is stated, then removes any one in two end points on the contour curve;
    And/or default second area threshold value, if a plurality of contour curve composition being sequentially connected in any region in character outline The area of kick is less than the second area threshold value, then removes every profile song in a plurality of contour curve for forming the kick The public point of control point on line and every two contour curves that are connected;
    5) dress up character library and test.
  2. 2. according to the method for claim 1, it is characterised in that in the step 2), the word manuscript base picture is bianry image.
  3. 3. according to the method for claim 1, it is characterised in that
    Also include after the step 3):
    3A. density of setting threshold values, judgment step 3) in point in the character outline that is formed in any region on all contour curves Density whether be more than the density threshold, in this way, then perform step 3B, if not, perform step 4);
    Point construction approximating functions of the 3B. according to step 3A in region on all contour curves, with fitting corresponding to formation Curve, and all contour curves in the region are substituted with the matched curve, it is then back to step 3A.
  4. 4. according to the method for claim 1, it is characterised in that in the step 4), remove wheel unnecessary in character outline Wide curve is specially:
    Remove in character outline and the contour curve that is formed after fitting is digitized into by the noise of the word manuscript base picture;
    And/or default first area threshold, if the area for the closed figure being made up of in character outline a plurality of contour curve is less than The area threshold, then remove a plurality of contour curve of the composition closed figure.
  5. A kind of 5. device for forming character library, it is characterised in that including:Acquiring unit, scanning element, digitlization fitting unit, from Dynamicization processing unit and integration testing unit;
    The acquiring unit is used to obtain word original text, and the word original text of acquisition is sent to scanning element;
    The scanning element is used to the word original text being scanned into word manuscript base picture, and the word manuscript base picture is sent to digitlization and is fitted Unit;
    The digitlization fitting unit is used to be digitized fitting to the word manuscript base picture, to form character outline, and by institute Character outline is stated to send to automatic business processing unit;
    The automatic business processing unit is used to carry out automatic business processing to the character outline, unnecessary in character outline to remove Unnecessary point on contour curve and contour curve, and the character outline after processing is sent to integration testing unit;
    The automatic business processing unit removes point unnecessary on contour curve in character outline:
    Radius of curvature threshold value is preset with it, if digitizing the song of any bar contour curve in the character outline that fitting unit is formed Rate radius minimum value is more than the radius of curvature threshold value, then removes the control point on the contour curve;
    And/or if a plurality of contour curve being sequentially connected is respectively positioned on same straight line in character outline, it is a plurality of successively to remove this The public point of every two connected contour curves in connected contour curve;
    And/or distance threshold is preset with it, if in the character outline that digitlization fitting unit is formed on any bar contour curve The distance between two end points be less than the distance threshold, then remove any one in two end points on the contour curve;
    And/or second area threshold value is preset with it, if more in any region in the character outline that digitlization fitting unit is formed The area of the kick for the contour curve composition that bar is sequentially connected is less than second area threshold value, then removes and form the more of the kick The public point of control point in bar contour curve on every contour curve and every two contour curves that are connected;
    The integration testing unit is used to the character outline after the processing is dressed up into character library and tested.
  6. 6. device according to claim 5, it is characterised in that the word manuscript base picture that the scanning element is scanned into is binary map Picture.
  7. 7. device according to claim 5, it is characterised in that
    Also include judging unit and approach fitting unit;
    Density threshold is preset with the judging unit, any area in the character outline formed for judging digitlization fitting unit Whether the density of the point in domain on all contour curves is more than the density threshold;
    If not, judging that all contour curves in the region meet the requirements, and qualifying signal is sent to automatic business processing Unit, so that must automate processing unit carries out automatic business processing to satisfactory character outline;
    In this way, then judge that all contour curves in the region are undesirable, and send unqualified signal to approach fitting Unit;
    The fitting unit that approaches is used for when receiving the unqualified signal, according on all contour curves in the region Point construction approximating function, to form corresponding matched curve, and with all profiles in the matched curve replacement region Whether curve, the density for then continuing to judge to put in the region by judging unit are more than the density threshold, until judging single Untill member judges that the matched curve in the region meets the requirements.
  8. 8. device according to claim 5, it is characterised in that the automatic business processing unit removes unnecessary in character outline Contour curve be specially:
    Remove in the character outline that digitlization fitting unit is formed and institute's shape after being fitted is digitized into by the noise of the word manuscript base picture Into contour curve;
    And/or the first area threshold is preset with it, if bent by a plurality of profile in the character outline that digitlization fitting unit is formed The area of the closed figure of line composition is less than the area threshold, then removes a plurality of contour curve of the composition closed figure.
CN201310332045.4A 2013-08-01 2013-08-01 A kind of method and device for forming character library Active CN104346390B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310332045.4A CN104346390B (en) 2013-08-01 2013-08-01 A kind of method and device for forming character library

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310332045.4A CN104346390B (en) 2013-08-01 2013-08-01 A kind of method and device for forming character library

Publications (2)

Publication Number Publication Date
CN104346390A CN104346390A (en) 2015-02-11
CN104346390B true CN104346390B (en) 2018-01-23

Family

ID=52502004

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310332045.4A Active CN104346390B (en) 2013-08-01 2013-08-01 A kind of method and device for forming character library

Country Status (1)

Country Link
CN (1) CN104346390B (en)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105760437A (en) * 2016-02-02 2016-07-13 刘敏 Word-stock creating and releasing method and system and font input method
TWI638283B (en) * 2016-04-11 2018-10-11 亞太戲網股份有限公司 Method and system for font creating by handwriting
CN106802800A (en) * 2016-12-30 2017-06-06 深圳芯智汇科技有限公司 The generation method and display device of graphical interfaces
CN107610200B (en) * 2017-10-10 2020-11-03 南京师范大学 Character library rapid generation method based on characteristic template
CN110134921B (en) * 2018-02-09 2020-12-04 北大方正集团有限公司 Method and device for checking whether font outline is deformed
CN110427886B (en) * 2019-08-02 2022-02-08 福建天晴数码有限公司 Automatic verification method and system for handwriting recognition
CN110532973B (en) * 2019-09-03 2022-02-01 海南阿凡题科技有限公司 Double-page text image identification and positioning segmentation method based on special anchor points
CN111898600A (en) * 2020-07-10 2020-11-06 浙江大华技术股份有限公司 Character outline extraction method and device, storage medium and electronic device
CN113052002B (en) * 2021-02-05 2024-05-14 广州八爪鱼教育科技有限公司 Method, device, equipment and storage medium for screening handwriting sampling points

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1588350A (en) * 2004-09-17 2005-03-02 华南理工大学 Treating method and its use for dynamic Chinese character word library containing writing time sequence information
CN1650345A (en) * 2002-02-25 2005-08-03 夏普株式会社 Character display apparatus and character display method, control program for controlling the character display method and recording medium recording the control program
CN101055565A (en) * 2007-06-15 2007-10-17 中国科学院软件研究所 Character library and font standard detection method
CN101894370A (en) * 2010-07-14 2010-11-24 苏州大学 Automatic generation method of shape parameter-adaptive oracle-bone inscription contour glyphs
CN103136769A (en) * 2011-12-02 2013-06-05 北京三星通信技术研究有限公司 Method and device of generation of writing style font of user

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1650345A (en) * 2002-02-25 2005-08-03 夏普株式会社 Character display apparatus and character display method, control program for controlling the character display method and recording medium recording the control program
CN1588350A (en) * 2004-09-17 2005-03-02 华南理工大学 Treating method and its use for dynamic Chinese character word library containing writing time sequence information
CN101055565A (en) * 2007-06-15 2007-10-17 中国科学院软件研究所 Character library and font standard detection method
CN101894370A (en) * 2010-07-14 2010-11-24 苏州大学 Automatic generation method of shape parameter-adaptive oracle-bone inscription contour glyphs
CN103136769A (en) * 2011-12-02 2013-06-05 北京三星通信技术研究有限公司 Method and device of generation of writing style font of user

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
书法碑帖图像去噪、轮廓拟合及纹理建模研究;张俊松;《中国博士学位论文全文数据库 哲学与人文科学辑》;20080715;第F088-4页 *
纳西象形文轮廓字体设计制作及植入研究;郭海;《中国优秀硕士学位论文全文数据库 信息科技辑》;20041215;第I138-865页 *
高精度曲线轮廓甲骨文字形生成***的研究与实现;王晓丽;《中国优秀硕士学位论文全文数据库 信息科技辑》;20110115;第I138-1168页 *

Also Published As

Publication number Publication date
CN104346390A (en) 2015-02-11

Similar Documents

Publication Publication Date Title
CN104346390B (en) A kind of method and device for forming character library
CN109117713B (en) Drawing layout analysis and character recognition method of full convolution neural network
US11636234B2 (en) Generating 3D models representing buildings
CN109684803B (en) Man-machine verification method based on gesture sliding
CN105654049A (en) Facial expression recognition method and device
CN102598002B (en) System and method for producing editable three-dimensional models
CN102609405A (en) Method for generating dynamic contract text and system
Karasneh et al. Extracting UML models from images
CN104866498A (en) Information processing method and device
CN103885942B (en) A kind of rapid translation device and method
CN104064182A (en) A voice recognition system and method based on classification rules
CN103699955A (en) Custom taxonomy based service model analysis method and device
CN103353866A (en) Three-dimensional model file format conversion method supporting XNA technology
CN116503173A (en) Rule engine and modeling method, modeling device and instruction processing method thereof
CN115828349A (en) Geometric model processing method and device, electronic equipment and storage medium
CN113867694B (en) Method and system for intelligently generating front-end code
CN103440197B (en) A kind of method automatically generating difference test report based on contrast test
CN102136154B (en) Cartoon manufacture method and device
CN105183678A (en) Communication method and apparatus of terminal interface
CN107807844A (en) Generation method, device and the computer-readable recording medium of traffic test scene
CN103455607B (en) Method for automatically converting waveform image file into preset waveform data file
CN107368423A (en) Interface automatic configuration system and the automatized script processing method carried out wherein
CN115758738A (en) Traffic accident simulation test method, data format conversion method, system, device and computer readable storage medium
Simona et al. Researches concerning risk assessing using Pareto diagram for design process of technological processes
CN105590337B (en) Joint setting method and setting equipment for 3D model

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CP03 Change of name, title or address

Address after: 100871, Beijing, Haidian District Cheng Fu Road 298, founder building, 9 floor

Patentee after: PEKING UNIVERSITY FOUNDER GROUP Co.,Ltd.

Patentee after: PKU FOUNDER INFORMATION INDUSTRY GROUP CO.,LTD.

Patentee after: BEIJING FOUNDER ELECTRONICS Co.,Ltd.

Address before: 100871, Beijing, Haidian District, Cheng Fu Road, No. 298, Zhongguancun Fangzheng building, 5 floor

Patentee before: PEKING UNIVERSITY FOUNDER GROUP Co.,Ltd.

Patentee before: FOUNDER INFORMATION INDUSTRY HOLDINGS Co.,Ltd.

Patentee before: BEIJING FOUNDER ELECTRONICS Co.,Ltd.

CP03 Change of name, title or address
TR01 Transfer of patent right

Effective date of registration: 20220919

Address after: 3007, Hengqin international financial center building, No. 58, Huajin street, Hengqin new area, Zhuhai, Guangdong 519031

Patentee after: New founder holdings development Co.,Ltd.

Patentee after: BEIJING FOUNDER ELECTRONICS Co.,Ltd.

Address before: 100871, Beijing, Haidian District Cheng Fu Road 298, founder building, 9 floor

Patentee before: PEKING UNIVERSITY FOUNDER GROUP Co.,Ltd.

Patentee before: PKU FOUNDER INFORMATION INDUSTRY GROUP CO.,LTD.

Patentee before: BEIJING FOUNDER ELECTRONICS Co.,Ltd.

TR01 Transfer of patent right