CN104182766A - Method for judging correctness of position relation of strokes of handwritten Chinese character based on template matching - Google Patents

Method for judging correctness of position relation of strokes of handwritten Chinese character based on template matching Download PDF

Info

Publication number
CN104182766A
CN104182766A CN201410439007.3A CN201410439007A CN104182766A CN 104182766 A CN104182766 A CN 104182766A CN 201410439007 A CN201410439007 A CN 201410439007A CN 104182766 A CN104182766 A CN 104182766A
Authority
CN
China
Prior art keywords
chinese character
parts
stroke
relation
user
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201410439007.3A
Other languages
Chinese (zh)
Inventor
白晓东
仇宏斌
赵胜伟
钱晓芬
姜杰
黄峰
李艺
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.)
Nanjing Normal University
Original Assignee
Nanjing Normal University
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 Nanjing Normal University filed Critical Nanjing Normal University
Priority to CN201410439007.3A priority Critical patent/CN104182766A/en
Publication of CN104182766A publication Critical patent/CN104182766A/en
Pending legal-status Critical Current

Links

Abstract

The invention discloses a method for judging the correctness of the position relation of strokes of a handwritten Chinese character based on template matching. The method comprises the specific steps of building a whole character feature set and a part feature set of a standard Chinese character; recording handwriting of the Chinese character handwritten by a user to obtain stroke features of the Chinese character; carrying out matching comparison on whole character features of the user Chinese character and whole character features of the standard Chinese character, building the one-to-one correspondence relation between the strokes of the user Chinese character and strokes of the standard Chinese character, confirming the attribute of each stroke of the user Chinese character, and distributing the strokes of the user Chinese character into parts to which the strokes belong; carrying out matching comparison on part features of the user Chinese character and part features of the standard Chinese character, judging whether the relative azimuth relation and the distance relation between the strokes in the parts of the user Chinese character are correct, and judging whether the relative azimuth relation and the distance relation between the parts of the user Chinese character are correct. The method can be used for judging whether the position relation between the strokes of the Chinese character handwritten by the user is correct and whether the position relation between the parts of the Chinese character handwritten by the user is correct and has important value in applications of judging the writing quality of the handwritten Chinese character and the like.

Description

Hand-written Chinese character strokes position relationship correction judgement method based on template matches
Technical field
The present invention relates to the technical field of utilizing computing machine to evaluate Chinese-character writing, relate in particular to a kind of hand-written Chinese character strokes position relationship correction judgement method based on template matches.
Background technology
Chinese character is the notation that records Chinese (Han)language, is the rap off instrument of emotion of the people of all use Chineses, and it has unique charm, is full of aesthetic feeling, is one of the longest word in the world.Chinese character is condensing the intelligence and wisdom of the Chinese nation, is carrying culture and the civilized history of the Chinese nation, is containing fine pursuit and the moral of the Chinese nation, is the most valuable cultural heritage of the Chinese nation.And be accompanied by the universal of computing machine, and there is revolutionary variation in the mode that people record Chinese character, and traditional paper pen is write and is replaced by computer typewriter gradually, and the chance of people's handwritten Chinese character is fewer and feweri, and writing ability generally declines.What how to promote compatriots' Chinese-character writing ability becomes a problem demanding prompt solution.
Current handwritten Chinese character is write automatic evaluation method and is mainly contained following several:
1, evaluate at the distribution relation of writing in space by horizontal, the longitudinal proportionate relationship, architectural feature and the handwritten Chinese character each point that judge handwritten Chinese character, as the method and apparatus of Chinese invention patent evaluating writing Chinese characters structure normative (publication number: CN101251891A).
2, adopt the method for machine learning, in single stroke to handwritten Chinese character, font, in the space structure of each several part, font, the writing style consistance of each several part is carried out aesthetic property evaluation, as Chinese invention patent " a kind of computer estimation method of Chinese character writing morphology beauty " (publication number: CN101295371A);
3, by artificial given Hanzi specimen is carried out to machine learning, then use image to process with the method for artificial intelligence handwritten Chinese character is carried out to similarity fuzzy Judgment, as Chinese invention patent " a kind of computer estimation method of Chinese-character writing aesthetics " (publication number: CN101295371A);
Although said method can be evaluated for layout relationship between handwritten Chinese character space structure, stroke in some aspects, but whether these methods can accurately not judge the relative position relation between user's handwritten Chinese character and standard Chinese character stroke or parts and write correctly, thereby cannot realize Chinese-character writing stroke position and be related to the judgement of correction judgement, reasonable structure and overall writing quality evaluation.
Summary of the invention
The object of the invention is to overcome the deficiencies in the prior art, a kind of hand-written Chinese character strokes position relationship correction judgement method based on template matches is provided.
The technical solution used in the present invention is as follows:
Hand-written Chinese character strokes position relationship correction judgement method based on template matches, specifically comprises the steps:
(1) whole word feature set and the component feature collection of Criterion Chinese character, component feature collection comprises in parts in the topological relation between stroke, parts distance relation five category features between the distance relation between stroke in the relative orientation relation between stroke, parts, relative orientation relation and parts between parts;
(2) written handwriting of recording user handwritten Chinese character, obtains its stroke feature;
(3) according to whole word feature, user's Chinese character is mated with standard Chinese character, set up one-to-one relationship between user's Chinese-character stroke and standard Chinese character stroke, confirm the attribute of the every unicursal of user's Chinese character, and stroke is assigned in parts;
(4) according to the distance relation between stroke in the relative orientation relation between stroke and parts in the topological relation between stroke in described parts, parts, user's Chinese character and standard Chinese character are compared, whether correct according to the position relationship between stroke in comparison result decision means;
(5) according to the relative orientation relation between described parts and the distance relation between parts, user's Chinese character and standard Chinese character are compared, whether correct according to the position relationship between comparison result decision means.
In described step (1), the whole word feature set of Criterion Chinese character and the method for component feature collection are as follows:
Build standard Chinese character feature set, wherein comprise whole word feature set and component feature collection, whole word feature set comprises and crossingly between type, absolute positional relation, relative position relation and the stroke of each stroke in this Chinese character is related to four kinds of features; Component feature collection comprises in parts in the topological relation between stroke, parts the five kinds of features of distance relation between the distance relation between stroke in the relative orientation relation between stroke, parts, relative orientation relation and parts between parts; Wherein in parts the topological relation between stroke comprise crossing, join, adjacent, non-adjacent four kinds, adjacently refer to that two strokes are non-intersect, do not join and centre does not have stroke to pass, if having other strokes in two strokes through being non-neighbour; Relative orientation relation in parts between stroke comprises left and right, upper and lower four kinds; Relative orientation relation between parts comprise semi-surrounding under upper and lower, left and right, upper left, upper right, upper left semi-surrounding, lower-left semi-surrounding, upper right semi-surrounding, bottom right semi-surrounding, upper bag, under wrap semi-surrounding, the right semi-surrounding of left bag, complete surround etc.
In described step (2), the written handwriting of recording user handwritten Chinese character, obtains its stroke feature method as follows:
The written handwriting of recording user handwritten Chinese character, according to the person's handwriting situation of movement of user's handwritten Chinese character on platform, obtains this handwritten Chinese character point set, and the point set forming from the first stroke of a Chinese character of starting writing in user writing process is defined as to a stroke; Traversal stroke point set, carries out noise reduction process to original point set, and the stroke after noise reduction is carried out to redundant classification.
The concrete steps of described step (3) are as follows:
Taking standard Chinese character as benchmark, successively handwritten Chinese character is carried out to following judgement:
A, the judgement of stroke type uniqueness; B, the judgement of absolute position uniqueness; C, crossing feature uniqueness judgement; D, the judgement of relative position similarity; E, by steps A-D remain unfulfilled corresponding relation set up user's hand-written Chinese character strokes, calculate respectively and in itself and standard Chinese character, do not complete corresponding relation and set up the stroke type similarity of stroke, absolute positional relation similarity, relative position relation similarity, and three kinds of similarities are carried out to overall treatment, set up the corresponding relation of user's hand-written Chinese character strokes and standard Chinese character stroke according to comprehensive similarity;
According to the division of parts class in the corresponding relation between handwritten Chinese character and standard Chinese character stroke and standard Chinese character, corresponding user's Chinese character stroke is assigned in different parts, obtain the component feature collection of user's Chinese character.
In described step (4), the method for topological relation, relative orientation relation and distance relation in comparison user's Chinese character and standard Chinese character parts between stroke is as follows:
The comparison of the topological relation in A, parts between stroke: the topological relation feature set in the relevant parts that define in traversal standard Chinese character between stroke, judge the topological relation between corresponding stroke in user's Hanzi component, the method for judgement is as follows: find immediate some p of two stroke original coordinates iand p j, at a p iboth sides respectively get a bit, put p i-1and p i+1, be this straight line y=ax+b of 2, get a p jthe point p of both sides j-1and p j+1, judging point p j-1with a p j+1whether in straight line both sides, if in both sides two-stroke intersect; Otherwise judged whether that other strokes are through their immediate points, if there is no other strokes through their immediate points, two-stroke is adjacent relation; Otherwise two-stroke is non-neighbour's relation.Judge that whether user's Chinese character is identical with the topological relation between stroke in standard Chinese character parts, if the same think that the topological relation between stroke is write correctly, otherwise clerical error.In comparison process, joining in user's Chinese character between stroke do not distinguish with neighbouring relations.
The comparison of the relative orientation relation in B, parts between stroke: according to stroke s in standard Chinese character parts jwith respect to stroke s iposition relation, judge stroke s in user's Hanzi component j' with respect to stroke s i' position relation, traversal stroke s j' unique point set, obtain stroke point and concentrate maximal value and the minimum value x of x and y coordinate jmin, x jmax, y jmin', y jmax' and stroke s jthe coordinate figure x of the x of ' central point and y jcenter', y jcenter', wherein x jcenter ′ = x j min ′ + x j max ′ 2 , y jcenter ′ = y j min ′ + y j max ′ 2 ; In like manner can obtain stroke s imaximal value and the minimum value x of ' a concentrated x and y coordinate imin', x imax', y imin', y imax' and stroke s i' x and the y coordinate figure x of central point icenter', y icenter', wherein between stroke, the method for position relation judgement is as follows:
1) left and right position relation: suppose stroke s j' at stroke s i' the right, judge x jmin'-x imax' >0 and ((ε 1for a certain threshold value) whether set up simultaneously, set up simultaneously and think stroke s j' at stroke s i' position relation write correctly, otherwise clerical error;
2) upper and lower position relation: suppose stroke s j' at stroke s i' top, judge y jmin'-y imax' >0 and ((ε 1for a certain threshold value) whether set up simultaneously, set up simultaneously and think stroke s j' at stroke s i' position relation write correctly, otherwise clerical error.Can obtain the relative orientation relationship characteristic collection between stroke in user's Hanzi component by said method, whether comparison standard Chinese character is identical with the relative orientation relation between corresponding stroke in user's Hanzi component, if the same think that the relative orientation relation between two strokes is write correctly in user's Chinese character, otherwise think clerical error;
The comparison of the distance relation in C, parts between stroke: calculate stroke s in standard Chinese character jwith stroke s idistance, computing method are: find immediate some p of two stroke original coordinates i(x i, y i) and p j(x j, y j), the distance of two points of calculating in like manner can obtain corresponding stroke s in user's Chinese character j' and stroke s i' distance d ', judgement | d-d ' | < ε 22for a certain threshold value) whether set up, if set up, think user's Chinese-character stroke s j' and stroke s i' distance relation write correctly, otherwise think clerical error; Whether the topological relation feature set in traversal standard Chinese character parts between stroke, utilize said method to judge to write without the distance relation between the stroke of overlapping relation in user's Hanzi component correct;
In described step (5), the relative orientation relation between comparison standard Chinese character and user's Hanzi component and the method for distance relation are as follows:
The comparison of the relative orientation relation between A, parts: according to parts c in standard Chinese character jwith respect to parts c iposition relation, judge parts c in user's Chinese character j' with respect to parts c i' position relation, traversal parts c j' unique point set, obtain parts point and concentrate maximal value and the minimum value x of x and y coordinate jmin', x jmax', y jmin', y jmax' and parts c jthe coordinate x of the x of ' central point and y jcenter', y jcenter', wherein x jcenter &prime; = x j min &prime; + x j max &prime; 2 , y jcenter &prime; = y j min &prime; + y j max &prime; 2 ; In like manner can obtain parts c imaximal value and the minimum value x of ' a concentrated x and y coordinate imin', x imax', y imin', y imax' and parts c i' x and the y coordinate x of central point icenter', y icenter', wherein x icenter &prime; = x i min &prime; + x i max &prime; 2 , y icenter &prime; = y i min &prime; + y i max &prime; 2 ; Each side is that the determination methods of relation is as follows:
1) upper and lower position relation: suppose parts c j' at parts c i' top, judge y jmin'-y imax' >0 and ((ε 3for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
2) left and right position relation: suppose parts c j' at parts c i' the right, judge x jmin'-x imax' >0 and ((ε 3for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
3) upper left position relation: suppose parts c j' at parts c i' lower left, judge y jmin'-y imax' >-ε 4, x imax'-x jmax' > ε 4and x imin'-x jmin' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
4) upper right position relation: suppose parts c j' at parts c i' lower right, judge y jmin'-y imax' >-ε 4, x jmax'-x imax' > ε 4and x jmin'-x imin' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
5) upper left semi-surrounding position relation: suppose parts c j' the left side and top by parts c i' semi-surrounding, judges y imax'-y jcenter' > ε 4, x imax'-x jcenter' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
6) lower-left semi-surrounding position relation: suppose parts c j' the left side with below by parts c i' semi-surrounding, judges y jcenter-' y imin' > ε 4, x imax'-x jcenter' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
7) upper right semi-surrounding position relation: suppose parts c j' the right and top by parts c i' semi-surrounding, judges y imax'-y jcenter' > ε 4, x jcenter'-x imin' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
8) bottom right semi-surrounding position relation: suppose parts c j' the right with below by parts c i' semi-surrounding, judges y jcenter'-y imin' > ε 4, x jcenter'-x imin' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
9) the semi-surrounding position relation under upper bag: suppose parts c j' the right and left and top by parts c i' semi-surrounding, judges y imax'-y jcenter' > ε 4((ε 4for a certain threshold value), x imin' <x jmin' and x imax' >x jmax' whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
10) the semi-surrounding position relation of wrapping under: suppose parts c j' the right and left with below by parts c i' semi-surrounding, judges y jcenter'-y imin' > ε 4((ε 4for a certain threshold value), x imin' <x jmin' and x imax' >x jmax' whether set up simultaneously, set up simultaneously and think parts c j' with parts c j' position relation write correctly, otherwise clerical error;
11) the right semi-surrounding position relation of left bag: suppose parts c j' both sides up and down and the left side by parts c i' semi-surrounding, judges x imax '-x jcenter' > ε 4((ε 4for a certain threshold value), y imin' <y jmin' and y imax' >y jmax' whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
12) entirely surround position relation: suppose parts c j' by parts c i' entirely surround, judge x imin' <x jmin' and x imax' >x jmax', y imin' <y jmin' and y imax' >y jmax' whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error;
Can obtain the relative orientation relationship characteristic collection between user's Hanzi component by said method, relative orientation relation in comparison standard Chinese character parts and user's Chinese character between corresponding component, if position relation is identical, thinks that the relative orientation relation between two parts is write correctly in user's Chinese character, otherwise think clerical error;
The comparison of the distance relation between B, parts: calculate parts c in standard Chinese character jwith parts c idistance, computing method are: traversal parts c junique point set, obtain parts point and concentrate maximal value and the minimum value x of x and y coordinate jmin, x jmax, y jmin, y jmax, in like manner can obtain parts c ipoint is concentrated maximal value and the minimum value x of x and y coordinate imin, x imax, y imin, y imax; Judge that according to the relative orientation relation between parts in standard Chinese character whether correct distance relation between user's Hanzi component write, determination methods is:
1) left and right position relation: suppose parts c jat parts c iright side, parts c jwith parts c idistance be d=x jmin-x imax;
2) upper and lower position relation: suppose parts c jat parts c itop, parts c jwith parts c idistance be d=y imin-y jmax;
3) upper left position relation: suppose parts c jat parts c ilower left, if x imin>x jmax, parts c jwith parts c idistance be otherwise parts c jwith parts c idistance be d=y jmin-y imax;
4) upper right position relation: suppose parts c jat parts c ilower right, if x jmin>x imax, parts c jwith parts c idistance be otherwise parts c jwith parts c idistance be d=y jmin-y imax;
Utilize said method to calculate corresponding component c in user's Chinese character j' with parts c i' distance d ', judgement | d-d ' | < ε 55for a certain threshold value) whether set up, if set up, think user's Hanzi component c jwith respect to parts c ibetween distance relation write correctly, otherwise think clerical error; In like manner can judge whether the distance relation between miscellaneous part is write correctly.
The present invention's beneficial effect compared with prior art:
(1) whether this method is correct by the position relationship of the each parts of computing machine automatic decision user handwritten Chinese character, each stroke, can realize judgement and remind after user's individual character has been write, and has directly perceived, efficient, ageing strong etc. a little.
(2) this method has realized the accurate judgement of the position relationship between user's hand-written Chinese character strokes and the position relationship between parts, evaluating handwritten Chinese character writing quality and writing in the application such as correctness, has important value.
Brief description of the drawings
Fig. 1 is process flow diagram of the present invention;
Fig. 2 is stroke s in handwritten Chinese character parts described in the invention iwith stroke s jbetween the Chinese character example of topological relation, (a)-(d) represent respectively to intersect, join, adjacent, non-neighbour's topological relation;
Fig. 3 is the Chinese character example of the relative orientation relation between division and the parts of handwritten Chinese character parts class described in the invention, that wherein (a)-(l) represents is respectively upper and lower, semi-surrounding under left and right, upper left, upper right, upper left semi-surrounding, lower-left semi-surrounding, upper right semi-surrounding, bottom right semi-surrounding, upper bag, under wrap semi-surrounding, the right semi-surrounding of left bag, the full position relation surrounding.
Embodiment
Below in conjunction with accompanying drawing, the present invention is elaborated.
As Fig. 1, a kind of hand-written Chinese character strokes position relationship confirmation method based on template matches, comprises the steps:
1, whole word feature set and the component feature collection of Criterion Chinese character.
Build standard Chinese character feature set, wherein comprise whole word feature set and component feature collection, whole word feature set comprises and crossingly between type, absolute positional relation, relative position relation and the stroke of each stroke in this Chinese character is related to four kinds of features; Component feature collection comprises in parts in the topological relation between stroke, parts the five kinds of features of distance relation between the distance relation between stroke in the relative orientation relation between stroke, parts, relative orientation relation and parts between parts; Wherein in parts the topological relation between stroke comprise crossing, join, adjacent, non-adjacent four kinds, as shown in Figure 2, adjacently refer to that two strokes are non-intersect, do not join and centre does not have stroke to pass, if having other strokes in two strokes through being non-neighbour; Relative orientation relation in parts between stroke comprises left and right, upper and lower four kinds; Relative orientation relation between parts comprise semi-surrounding under upper and lower, left and right, upper left, upper right, upper left semi-surrounding, lower-left semi-surrounding, upper right semi-surrounding, bottom right semi-surrounding, upper bag, under wrap semi-surrounding, the right semi-surrounding of left bag, complete surround etc.
2, the written handwriting of recording user handwritten Chinese character, obtains its stroke feature.
The written handwriting of recording user handwritten Chinese character, according to the person's handwriting situation of movement of user's handwritten Chinese character on platform, obtains this handwritten Chinese character point set, and the point set forming from the first stroke of a Chinese character of starting writing in user writing process is defined as to a stroke; Traversal stroke point set, carries out noise reduction process to original point set, and the stroke after noise reduction is carried out to redundant classification.
3, the one-to-one relationship of setting up handwritten Chinese character and the each stroke of standard Chinese character, is assigned to the stroke of user's Chinese character in associated components.
(1) with the whole word feature set of standard Chinese character, successively handwritten Chinese character is carried out to following judgement:
A, the judgement of stroke type uniqueness: separate the uniqueness stroke in handwritten Chinese character, mate with standard Chinese character stroke feature;
B, the judgement of absolute position uniqueness: set up handwritten Chinese character outsourcing frame, and by some outsourcing frame deciles, determine the residing position of each stroke, mate with standard Chinese character absolute position feature;
C, crossing feature uniqueness judgement: extract in handwritten Chinese character and there is the stroke of crossing feature with determining corresponding relation stroke, mate with standard Chinese character according to the uniqueness of correlationship;
D, relative position similarity judgement: the stroke that does not complete coupling is carried out to cluster by stroke type, calculate the relative position of each stroke and outsourcing frame summit in similar stroke, mate with standard Chinese character according to relative position relation;
E, by steps A-D remain unfulfilled corresponding relation set up user's hand-written Chinese character strokes, calculate respectively and in itself and standard Chinese character, do not complete corresponding relation and set up the stroke type similarity of stroke, absolute positional relation similarity, relative position relation similarity, and three kinds of similarities are carried out to overall treatment, set up the corresponding relation of user's hand-written Chinese character strokes and standard Chinese character stroke according to comprehensive similarity;
(2) according to the division of parts class in the corresponding relation between handwritten Chinese character and standard Chinese character stroke and standard Chinese character, corresponding user's Chinese character stroke is assigned in different parts, obtains the component feature collection of user's Chinese character, method is as follows:
The feature set C={c of parts class in traversal standard Chinese character 1(s 1, s 2..., s i), c 2(s i+1, s i+2..., s k) ..., c j(s k+1, s k+2..., s n), wherein s ifor the stroke in standard Chinese character, n, j is respectively the order of strokes observed in calligraphy of standard Chinese character stroke and the number of parts, according to the feature set C defining in standard Chinese character to user's pen for writing Chinese characters album
S '={ s 1' (p 1', p 2' ..., p i'), s 2' (p 1', p 2' ..., p i') ..., s j' (p 1', p 2' ..., p i' carry out parts and divide the feature set of final user Hanzi component class
C′={c 1′(s 1′,s 2′,...,s i′),c 2′(s i+1′,s i+2′,...,p k′),...,c j′(s k+1′,s k+2′,...,s n′},
Wherein j is user's Hanzi component number, and n is the order of strokes observed in calligraphy of user's Chinese-character stroke.
Distance relation feature set between the topological relation in the parts that 4, define in traversal standard Chinese character between stroke, relative orientation relation, the stroke between stroke, set up the corresponding feature set of user's Chinese character the processing of comparing, whether write correctly according to the position relationship between stroke in comparison result decision means, method is as follows:
(1) comparison of the topological relation between stroke in parts: the topological relation feature set T={t in the relevant parts that define in traversal standard Chinese character between stroke 1, t 2..., t i, wherein t i=1,2,3,4, t ivalue represent respectively crossing, join, adjacent, non-neighbour's relation, judge the topological relation between corresponding stroke in user's Hanzi component, the method for judgement is as follows: establish stroke S in user's Chinese character S ' h' point set be stroke S k' point set be first look for S h', S k' the immediate point of original coordinates, method is: recycle ratio is value, gets its reckling, now corresponding coordinate with be stroke s ' hwith stroke s ' kin immediate point.? both sides respectively get a bit: with be this straight line y=ax+b of 2, then got the point of both sides with this 2 point of substitution in y=ax+b, judges that whether they are in the both sides of straight line, and the method for employing is judgement with or with whether set up simultaneously, intersect if set up two-stroke, otherwise non-intersect; If two strokes are not crossing relations, judge closest approach distance whether set up, if set up s' hand s' kjoin, wherein, L disit is empirical value; If s' hand s' kalso do not join, judge s' hand s' kwhether adjacent, judgement according to being: if s ' hand s ' kadjacent, there is no other strokes through their immediate points, method is: establish s l' ≠ s' h, s l' ≠ s' k, get s l' end points and each turning point cross with be straight line y=ax+b, successively judgement i=1,2 ..., s-1, whether in the both sides of straight line, if there is the situation that is positioned at straight line both sides, illustrates s ' hand s ' kfrom; If there is no be positioned at the situation of straight line both sides, s ' is described hand s ' kadjacent.Can obtain the topological relation feature set T '={ t between stroke in user's Hanzi component according to said method 1', t 2' ..., t i', the topological relation in comparison standard Chinese character and user's Hanzi component between stroke, judge in user's Hanzi component that according to comparison result whether correct the topological relation between stroke write, specifically comparison process is: if t i=t 1'=1 sets up, and thinks that overlapping relation writes correctly, otherwise clerical error; If t i=2 or t i=3 and t i'=1 and t i'=4 set up simultaneously, and think and join or neighbouring relations are write correctly, otherwise clerical error; If t i=t 1'=4 set up, and think and write correctly from relation, otherwise clerical error;
(2) comparison of the relative orientation relation between stroke in parts: according to stroke s in standard Chinese character parts hwith respect to stroke s kposition relation, judge stroke s in user's Hanzi component h' with respect to stroke s k' position relation, traversal stroke s h' unique point set s h'={ p 1 h, p 2 h..., p m h, relatively stroke s h' in the size of X Y coordinate figure of each point, obtain stroke point and concentrate maximal value and the minimum value x of x and y coordinate jmin', x jmax', y jmin', y jmax' and stroke s jthe coordinate figure x of the x of ' central point and y jcenter', y jcenter', wherein x jcenter &prime; = x j min &prime; + x j max &prime; 2 , y jcenter &prime; = y j min &prime; + y j max &prime; 2 ; In like manner can obtain stroke s imaximal value and the minimum value x of ' a concentrated x and y coordinate imin', x imax', y imin', y imax' and stroke s i' x and the y coordinate figure x of central point icenter', y icenter', wherein x icenter &prime; = x i min &prime; + x i max &prime; 2 , y icenter &prime; = y i min &prime; + y i max &prime; 2 ; Between stroke, the method for position relation judgement is as follows:
1) left and right position relation: suppose stroke s j' at stroke s i' the right, judge x jmin'-x imax' >0 and ((ε 1for a certain threshold value) whether set up simultaneously, set up simultaneously and think stroke s j' at stroke s i' position relation write correctly, otherwise clerical error;
2) upper and lower position relation: suppose stroke s j' at stroke s i' top, judge y jmin'-y imax' >0 and ((ε 1for a certain threshold value) whether set up simultaneously, set up simultaneously and think stroke s j' at stroke s i' position relation write correctly, otherwise clerical error.Can obtain the relative orientation relationship characteristic collection between stroke in user's Hanzi component by said method, whether comparison standard Chinese character is identical with the relative orientation relation between corresponding stroke in user's Hanzi component, if the same think that the relative orientation relation between two strokes is write correctly in user's Chinese character, otherwise think clerical error;
(3) comparison of the distance relation between stroke in parts: calculate stroke s in standard Chinese character hwith stroke s kdistance, computing method are: establish stroke S in standard Chinese character S hpoint set be s h={ p 1 h, p 2 h..., p m h, stroke S kpoint set be s k={ p 1 k, p 2 k..., p n k.First look for S h, S kthe immediate point of original coordinates, method is: recycle ratio is value, gets its reckling, now corresponding coordinate with be stroke S hwith stroke S kin immediate point, minimum value is two distances between stroke in like manner can obtain corresponding stroke s in user's Chinese character k' and stroke s hcoordinate (the p in ' middle closest approach i h '(x), p i h ') and (p (y) j k '(x), p j k '(y)), the distance of two points of proximity d &prime; = ( p h h &prime; ( x ) - p j k &prime; ( x ) ) 2 + ( p i h &prime; ( y ) - p j k &prime; ( y ) ) 2 , Judgement | d-d ' | < ε 22for a certain threshold value) whether set up, if set up, think user's Chinese-character stroke s j' and stroke s i' distance relation write correctly, otherwise think clerical error; According to the topological relation feature set between stroke in standard Chinese character parts, circulation carry out above-mentioned steps judge in user's Hanzi component, write without the distance relation between the stroke of overlapping relation whether correct.
Relative orientation relation between the parts that 5, define in traversal standard Chinese character, the distance relation feature set between parts, set up user part feature set and the processing of comparing, whether write correct specification according to the position relationship between comparison result decision means, method is as follows:
(1) comparison of the relative orientation relation between parts: according to parts c in standard Chinese character jwith respect to parts c iposition relation, judge parts c in user's Chinese character j' with respect to parts c i' position relation, traversal parts c j' unique point set c j'={ s 1' (p 1' ..., p n') ..., s k' (p 1' ..., p m'), comparing unit point is concentrated the size of the XY coordinate figure of each point, obtains maximal value and the minimum value x of the concentrated x of parts point and y coordinate jmin', x jmax', y jmin', y jmax' and parts c jthe coordinate x of the x of ' central point and y jcenter', y jcenter', wherein x jcenter &prime; = x j min &prime; + x j max &prime; 2 , y jcenter &prime; = y j min &prime; + y j max &prime; 2 ; In like manner can obtain parts c imaximal value and the minimum value x of ' a concentrated x and y coordinate imin', x imax', y imin', y imax' and parts c i' x and the y coordinate x of central point icenter', y icenter', wherein x icenter &prime; = x i min &prime; + x i max &prime; 2 , y icenter &prime; = y i min &prime; + y i max &prime; 2 ; Each side is that the determination methods of relation is as follows:
1) upper and lower position relation: suppose parts c j' at parts c i' top, judge y jmin'-y imax' >0 and ((ε 3for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (a);
2) left and right position relation: suppose parts c j' at parts c i' the right, judge x jmin'-x imax' >0 and ((ε 3for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (b);
3) upper left position relation: suppose parts c j' at parts c i' lower left, judge y jmin'-y imax' >-ε 4, x imax'-x jmax' > ε 4and x imin'-x jmin' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (c);
4) upper right position relation: suppose parts c j' at parts c i' lower right, judge y jmin'-y imax' >-ε 4, x jmax'-x imax' > ε 4and x jmin'-x imin' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (d);
5) upper left semi-surrounding position relation: suppose parts c j' the left side and top by parts c i' semi-surrounding, judges y imax'-y jcenter' > ε 4, x imax'-x jcenter' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (e);
6) lower-left semi-surrounding position relation: suppose parts c j' the left side with below by parts c i' semi-surrounding, judges y jcenter-' y imin' > ε 4, x imax'-x jcenter' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (f);
7) upper right semi-surrounding position relation: suppose parts c j' the right and top by parts c i' semi-surrounding, judges y imax'-y jcenter' > ε 4, x jcenter'-x imin' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (g);
8) bottom right semi-surrounding position relation: suppose parts c j' the right with below by parts c i' semi-surrounding, judges y jcenter'-y imin' > ε 4, x jcenter'-x imin' > ε 4((ε 4for a certain threshold value) whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (h);
9) the semi-surrounding position relation under upper bag: suppose parts c j' the right and left and top by parts c i' semi-surrounding, judges y imax'-y jcenter' > ε 4((ε 4for a certain threshold value), x imin' <x jmin' and x imax '>x jmax' whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (i);
10) the semi-surrounding position relation of wrapping under: suppose parts c j' the right and left with below by parts c i' semi-surrounding, judges y jcenter'-y imin' > ε 4((ε 4for a certain threshold value), x imin' <x jmin' and x imax' >x jmax' whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (j);
11) the right semi-surrounding position relation of left bag: suppose parts c j' both sides up and down and the left side by parts c i' semi-surrounding, judges x imax'-x jcenter' > ε 4((ε 4for a certain threshold value), y imin' <y jmin' and y imax' >y jmax' whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (k);
12) entirely surround position relation: suppose parts c j' by parts c i' entirely surround, judge x imin' <x jmin' and x imax' >x jmax', y imin' <y jmin' and y imax' >y jmax' whether set up simultaneously, set up simultaneously and think parts c j' with parts c i' position relation write correctly, otherwise clerical error, as shown in Fig. 3 (i);
Can obtain the relative orientation relationship characteristic collection between user's Hanzi component by said method, relative orientation relation in comparison standard Chinese character parts and user's Chinese character between corresponding component, if position relation is identical, thinks that the relative orientation relation between two parts is write correctly in user's Chinese character, otherwise think clerical error;
(2) comparison of the distance relation between parts: calculate parts c in standard Chinese character jwith parts c idistance, computing method are: traversal parts c junique point set c j={ s 1(p 1..., p n) ..., s k(p 1..., p m), the size of the XY coordinate figure of each point in comparing unit, obtains parts point and concentrates maximal value and the minimum value x of x and y coordinate jmin, x jmax, y jmin, y jmax, in like manner can obtain parts c ipoint is concentrated maximal value and the minimum value x of x and y coordinate imin, x imax, y imin, y imax; Judge that according to the relative orientation relation between parts in standard Chinese character whether correct distance relation between user's Hanzi component write, determination methods is:
1) left and right position relation: suppose parts c jat parts c iright side, parts c jwith parts c idistance be d=x imin-x imax;
2) upper and lower position relation: suppose parts c jat parts c itop, parts c jwith parts c idistance be d=y imin-y jmax;
3) upper left position relation: suppose parts c jat parts c ilower left, if x imin>x jmax, parts c jwith parts c idistance be otherwise parts c jwith parts c idistance be d=y jmin-y imax;
4) upper right position relation: suppose parts c jat parts c ilower right, if x jmin>x imax, parts c jwith parts c idistance be otherwise parts c jwith parts c idistance be d=y jmin-y imax; Utilize said method to calculate corresponding component c in user's Chinese character j' with parts c i' distance d ', judgement | d-d ' | < ε 55for a certain threshold value) whether set up, if set up, think user's Hanzi component c jwith respect to parts c ibetween distance relation write correctly, otherwise think clerical error; In like manner can judge whether the distance relation between miscellaneous part is write correctly.

Claims (6)

1. the hand-written Chinese character strokes position relationship correction judgement method based on template matches, is characterized in that, concrete steps are as follows:
(1) whole word feature set and the component feature collection of Criterion Chinese character, component feature collection comprises in parts in the topological relation between stroke, parts distance relation five category features between the distance relation between stroke in the relative orientation relation between stroke, parts, relative orientation relation and parts between parts;
(2) written handwriting of recording user handwritten Chinese character, obtains its stroke feature;
(3) according to whole word feature, user's Chinese character is mated with standard Chinese character, set up one-to-one relationship between user's Chinese-character stroke and standard Chinese character stroke, confirm the attribute of the every unicursal of user's Chinese character, and stroke is assigned in parts;
(4) according to the distance relation between stroke in the relative orientation relation between stroke and parts in the topological relation between stroke in described parts, parts, user's Chinese character and standard Chinese character are compared, whether correct according to the position relationship between stroke in comparison result decision means;
(5) according to the relative orientation relation between described parts and the distance relation between parts, user's Chinese character and standard Chinese character are compared, whether correct according to the position relationship between comparison result decision means.
2. the hand-written Chinese character strokes position relationship correction judgement method based on template matches according to claim 1, it is characterized in that, in described step (1): whole word feature set comprises overlapping relation four category features between type, absolute positional relation, relative position relation and the stroke of each stroke in this Chinese character.
3. the hand-written Chinese character strokes position relationship correction judgement method based on template matches according to claim 1, is characterized in that, the concrete steps of described step (2) are as follows:
The written handwriting of recording user handwritten Chinese character, according to the person's handwriting situation of movement of user's handwritten Chinese character on platform, obtains this handwritten Chinese character point set, and the point set forming from the first stroke of a Chinese character of starting writing in user writing process is defined as to a stroke; Traversal stroke point set, carries out noise reduction process to original point set, and the stroke after noise reduction is carried out to redundant classification.
4. the hand-written Chinese character strokes position relationship correction judgement method based on template matches according to claim 1, is characterized in that, the concrete steps of described step (3) are as follows:
Taking standard Chinese character as benchmark, carry out successively following judgement: the judgement of (1) stroke type uniqueness; (2) absolute position uniqueness judgement; (3) intersect the judgement of feature uniqueness; (4) relative position similarity judgement; (5) remain unfulfilled by step (1)-(4) user's hand-written Chinese character strokes that corresponding relation is set up, calculate respectively and in itself and standard Chinese character, do not complete corresponding relation and set up the stroke type similarity of stroke, absolute positional relation similarity, relative position relation similarity, and three kinds of similarities are carried out to overall treatment, set up the corresponding relation of user's hand-written Chinese character strokes and standard Chinese character stroke according to comprehensive similarity;
According to the division of parts class in standard Chinese character, corresponding user's Chinese character stroke is assigned in different parts, obtain the component feature of user's Chinese character.
5. the hand-written Chinese character strokes position relationship correction judgement method based on template matches according to claim 1, is characterized in that, the concrete steps of described step (4) are as follows:
(1) comparison of the topological relation between stroke in parts: the topological relation feature set in the parts that define taking standard Chinese character between stroke is as benchmark, topological relation in traversal user Chinese character corresponding component between stroke also completes judgement, judge that whether user's Chinese character is identical with the topological relation between stroke in standard Chinese character corresponding component, if the same think that the topological relation between stroke writes correctly, otherwise clerical error;
(2) comparison of the relative orientation relation between stroke in parts: whether comparison standard Chinese character is identical with the relative orientation relation between corresponding stroke in user's Hanzi component, if the same think that the relative orientation relation between two strokes is write correctly in user's Chinese character, otherwise think write wrong;
(3) comparison of the distance relation between stroke in parts: find two immediate points of stroke original coordinates, judge whether their distance is less than a certain threshold value, think that user writing is correct if be less than, otherwise think write wrong; Whether the topological relation feature in traversal standard Chinese character parts between stroke, utilize said method to judge to write without the distance relation between the stroke of overlapping relation in user's Hanzi component correct.
6. the hand-written Chinese character strokes position relationship correction judgement method based on template matches according to claim 1, is characterized in that, the concrete steps of described step (5) are as follows:
(1) comparison of the relative orientation relation between parts: according to the relative orientation relation about parts and parts defining in standard Chinese character, judge the relative orientation relation of parts and parts in user's Chinese character, if position relation is identical, think that the relative orientation relation between parts is write correctly in user's Chinese character, otherwise think write wrong;
(2) comparison of the distance relation between parts: judge according to the distance relation between the parts that define in standard Chinese character whether correct the distance relation between user's Hanzi component is write.Judge that user's Hanzi component spacing deviation is whether within the threshold range of standard Chinese character parts spacing deviation, if thinking that user writing is correct, otherwise think write wrong.
CN201410439007.3A 2014-08-29 2014-08-29 Method for judging correctness of position relation of strokes of handwritten Chinese character based on template matching Pending CN104182766A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410439007.3A CN104182766A (en) 2014-08-29 2014-08-29 Method for judging correctness of position relation of strokes of handwritten Chinese character based on template matching

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410439007.3A CN104182766A (en) 2014-08-29 2014-08-29 Method for judging correctness of position relation of strokes of handwritten Chinese character based on template matching

Publications (1)

Publication Number Publication Date
CN104182766A true CN104182766A (en) 2014-12-03

Family

ID=51963792

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410439007.3A Pending CN104182766A (en) 2014-08-29 2014-08-29 Method for judging correctness of position relation of strokes of handwritten Chinese character based on template matching

Country Status (1)

Country Link
CN (1) CN104182766A (en)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104537343A (en) * 2014-12-22 2015-04-22 中山大学 Multistage precision progressive stroke identification method based on grid
CN104966096A (en) * 2015-06-10 2015-10-07 南京师范大学 Important writing characteristic tagging-based handwritten Chinese character neatness evaluation method
CN105160338A (en) * 2015-08-17 2015-12-16 南京师范大学 Handwritten Chinese character correctness distinguishing method based on labeling relationship
CN105678348A (en) * 2016-01-07 2016-06-15 陕西师范大学 Normative evaluation method and system for handwritten Chinese characters
CN106056049A (en) * 2016-05-20 2016-10-26 广东小天才科技有限公司 Chinese character writing stroke detection method and device
CN108416249A (en) * 2017-02-10 2018-08-17 肖奇 A kind of written handwriting identification system and method
CN109189314A (en) * 2018-08-13 2019-01-11 广东小天才科技有限公司 Writing bootstrap technique, device, equipment and the medium of handwriting equipment
CN109635687A (en) * 2018-11-30 2019-04-16 南京师范大学 The Chinese-character text running hand calculated based on timing point set writes automatic Evaluation of Quality method and system
CN111079483A (en) * 2019-05-15 2020-04-28 广东小天才科技有限公司 Writing standard judgment method and electronic equipment
CN112926419A (en) * 2021-02-08 2021-06-08 北京百度网讯科技有限公司 Character judgment result processing method and device and electronic equipment

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
PAK-KEUNG LAI ET AL.: ""Chinese glyph generation using character composition and beauty evaluation metrics"", 《INTERNATIONAL JOURNAL OF RADIATION BIOLOGY》 *

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104537343A (en) * 2014-12-22 2015-04-22 中山大学 Multistage precision progressive stroke identification method based on grid
CN104537343B (en) * 2014-12-22 2018-11-16 中山大学 One kind gesticulating recognition methods based on grid multistage accuracy is progressive
CN104966096A (en) * 2015-06-10 2015-10-07 南京师范大学 Important writing characteristic tagging-based handwritten Chinese character neatness evaluation method
CN105160338B (en) * 2015-08-17 2018-11-16 南京师范大学 Handwritten Chinese character correctness method of discrimination based on mark relationship
CN105160338A (en) * 2015-08-17 2015-12-16 南京师范大学 Handwritten Chinese character correctness distinguishing method based on labeling relationship
CN105678348A (en) * 2016-01-07 2016-06-15 陕西师范大学 Normative evaluation method and system for handwritten Chinese characters
CN105678348B (en) * 2016-01-07 2019-01-11 陕西师范大学 A kind of handwritten Chinese character normalization evaluation method and system
CN106056049A (en) * 2016-05-20 2016-10-26 广东小天才科技有限公司 Chinese character writing stroke detection method and device
CN108416249A (en) * 2017-02-10 2018-08-17 肖奇 A kind of written handwriting identification system and method
CN109189314A (en) * 2018-08-13 2019-01-11 广东小天才科技有限公司 Writing bootstrap technique, device, equipment and the medium of handwriting equipment
CN109189314B (en) * 2018-08-13 2022-01-21 广东小天才科技有限公司 Writing guide method, device, equipment and medium for handwriting equipment
CN109635687A (en) * 2018-11-30 2019-04-16 南京师范大学 The Chinese-character text running hand calculated based on timing point set writes automatic Evaluation of Quality method and system
CN109635687B (en) * 2018-11-30 2022-07-01 南京师范大学 Chinese character text line writing quality automatic evaluation method and system based on time sequence point set calculation
CN111079483A (en) * 2019-05-15 2020-04-28 广东小天才科技有限公司 Writing standard judgment method and electronic equipment
CN112926419A (en) * 2021-02-08 2021-06-08 北京百度网讯科技有限公司 Character judgment result processing method and device and electronic equipment
CN112926419B (en) * 2021-02-08 2023-10-27 北京百度网讯科技有限公司 Character judgment result processing method and device and electronic equipment

Similar Documents

Publication Publication Date Title
CN104182766A (en) Method for judging correctness of position relation of strokes of handwritten Chinese character based on template matching
CN103810506B (en) A kind of hand-written Chinese character strokes recognition methods
CN104182732B (en) Feature based matrix carries out the hand-written Chinese character strokes confirmation method of similarity mode
Zitnick Handwriting beautification using token means
McLure et al. Extending analogical generalization with near-misses
CN110555403A (en) handwritten character evaluation method and system
CN101377854B (en) Method for simulating Chinese characters hand-written handwriting by a computer
CN105678297A (en) Portrait semantic analysis method and system based on label transfer and LSTM model
CN104966096A (en) Important writing characteristic tagging-based handwritten Chinese character neatness evaluation method
CN104268526B (en) A kind of Chinese character picture match and deformation method
CN106156082A (en) A kind of body alignment schemes and device
CN101299236A (en) Method for recognizing Chinese hand-written phrase
CN103093196A (en) Character interactive input and recognition method based on gestures
CN104063723A (en) Stroke reduction method of offline handwritten Chinese character and device thereof
CN106598920A (en) Similar Chinese character classification method combining stroke codes with Chinese character dot matrixes
CN101393693B (en) Computer educating method for Chinese character writing
CN111552457A (en) Statement identification-based front-end development page construction method and device and storage medium
CN107729316A (en) The identification of wrong word and the method and device of error correction in the interactive question and answer text of Chinese
Potrus et al. An evolutionary harmony search algorithm with dominant point detection for recognition-based segmentation of online Arabic text recognition
CN104050483A (en) Feature dimension reduction method based on local orthogonal alignment
CN104239910B (en) Stroke addition recognition method for online handwritten Chinese characters
Hu et al. A Chinese handwriting education system with automatic error detection.
CN107463799A (en) Interaction fusion feature represents the DBP recognition methods with selective ensemble
CN102375994B (en) Method and device for detecting and reducing correctness of order of strokes of written Chinese character
CN113191309A (en) Method and system for recognizing, scoring and correcting handwritten Chinese characters

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20141203

WD01 Invention patent application deemed withdrawn after publication