CN102411712B - Handwriting-based method for identity identification and terminal thereof - Google Patents

Abstract

The invention provides a handwriting-based method for identity identification and a terminal thereof. The method comprises the following steps that: a handwriting collection module collects to-be-identified handwriting; a feature extraction module extracts an identification feature of the to-be-identified handwriting, wherein the identification feature is used for characterizing specificity of the to-be-identified handwriting; a determination module determines whether the identification feature is matched with a standard feature of standard handwriting or not, wherein the standard feature is stored in a handwriting sample database; and if it is determined and obtained that the identification feature is matched with the standard feature, the identity is legal; If not, the identity is illegal and the standard feature is used for characterizing specificity of the standard handwriting. According to the method, handwriting is employed to carry out determination on legality of user identity; because handwriting of each person is unique and will not be changed randomly in a long time, stability of identity identification is ensured; in addition, handwriting is not easy to be imitated, so that identity identification employing handwriting has high security.

Description

Method and terminal based on the identification of person's handwriting

Technical field

The present invention relates to identity recognizing technology, relate in particular to a kind of method and terminal of the identification based on person's handwriting.

Background technology

Along with developing rapidly of microelectric technique, volume is little, function is used abundant personal electric product (as smart mobile phone, PDA, panel computer etc.) has become indispensable articles for use in people's routine work life.In above-mentioned personal electric product, generally can store always commercial data, privacy data etc.For preventing that the unprincipled fellow from usurping the personal electric product, and utilize wherein that institute's canned data carries out unlawful activities, and bring loss for owner or the related personnel of personal electric product, therefore, in the personal electric product, generally be provided with the mechanism of identity validation.

At present, most popular identity validation mechanism is fingerprint recognition in the personal electric product.Fingerprint recognition is basis, and under state of nature, everyone fingerprint all is inequality at pattern, breakpoint and the intersection point of skin lines, and the fingerprint characteristic that is to say everyone all is unique, and constant throughout one's life.Relying on this uniqueness and stability, fingerprint and individual are carried out corresponding one by one, when carrying out identification, by individual's fingerprint and the fingerprint of preserving are in advance compared, if the identical legal identity that then is judged as, otherwise is illegal identity.

Though, whether can identify user identity by fingerprint recognition legal, but, fingerprint also can have the wound means to be changed by operation or some, this has just caused fingerprint recognition is not again stable means of identification, for example, when the finger injuries of typing fingerprint, then just can not carry out the identification of identity by this finger of recycling this moment.In addition, on some people's fingerprint characteristic, be difficult to imaging, cause identifying difficulty and misclassification rate height.Also have, when using fingerprint recognition each time, on the fingerprint recognition collection head, all can stay user's finger mark, and these fingerprints that remain on the fingerprint recognition collection head can be used to carry out copying of fingerprint, in case fingerprint is replicated, then any fingerprint that copies that can use per capita carries out the identification of identity, has so just reduced the security of fingerprint recognition.

Summary of the invention

The invention provides a kind of method and terminal of the identification based on person's handwriting, to improve stability and the security of identification.

The invention provides a kind of method of the identification based on person's handwriting, may further comprise the steps:

The person's handwriting acquisition module is gathered person's handwriting to be identified;

Characteristic extracting module is extracted the recognition feature of described person's handwriting to be identified, and described recognition feature is used for characterizing the specificity of person's handwriting to be identified;

Judge module judges whether described recognition feature and the standard feature that is stored in the standard handwriting in the person's handwriting sample library module mate, if judgement knows that described recognition feature and described standard feature are complementary, then identity is legal; Otherwise identity is illegal, and described standard feature is used for characterizing the specificity of described standard handwriting.

The present invention also provides a kind of terminal of the identification based on person's handwriting, comprising:

The person's handwriting acquisition module is used for gathering person's handwriting to be identified;

Characteristic extracting module, for the recognition feature of extracting described person's handwriting to be identified, described recognition feature is used for characterizing the specificity of person's handwriting;

Person's handwriting sample library module is used for the storage standards person's handwriting;

Judge module is used for judging whether described recognition feature and the standard feature that is stored in the standard handwriting of person's handwriting sample library module mate, if judgement knows that described recognition feature and described standard feature are complementary, then identity is legal; Otherwise identity is illegal, and described standard feature is used for characterizing the specificity of described standard handwriting.

Method and the terminal of the identification based on person's handwriting provided by the invention adopt the recognition feature of the hand-written person's handwriting of user and the standard feature that is stored in the person's handwriting sample library module to mate, if coupling is by then being judged as validated user, otherwise are the disabled user.Because everyone person's handwriting has uniqueness, and in long one period (such as several years), be can arbitrarily not change, therefore guaranteed the stability of identification.Therefore in addition, because that person's handwriting is difficult for is imitated, adopts person's handwriting to carry out identification and have higher security.

Description of drawings

Accompanying drawing is used to provide further understanding of the present invention, and constitutes the part of instructions, is used from explanation the present invention with embodiments of the invention one, is not construed as limiting the invention.In the accompanying drawings:

Fig. 1 is the process flow diagram of method embodiment 1 that the present invention is based on the identification of person's handwriting;

Fig. 2 is the process flow diagram of method embodiment 2 that the present invention is based on the identification of person's handwriting;

Fig. 3 is the process flow diagram of method embodiment 3 that the present invention is based on the identification of person's handwriting;

Fig. 4 is the structural representation of terminal embodiment 1 that the present invention is based on the identification of person's handwriting;

Fig. 5 is the structural representation of terminal embodiment 2 that the present invention is based on the identification of person's handwriting.

Embodiment

For the purpose, technical scheme and the advantage that make the embodiment of the invention clearer, below in conjunction with the accompanying drawing in the embodiment of the invention, technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that obtains under the creative work prerequisite.

As shown in Figure 1, the present invention is based on the process flow diagram of method embodiment of the identification of person's handwriting, may further comprise the steps:

S100: the person's handwriting acquisition module is gathered person's handwriting to be identified;

The present personal electric product of main flow, (Personal Digital Assistant, PDA), panel computer etc., their display screen is touching display screen as smart mobile phone, personal digital assistant.The touching display screen here can be resistive touch screen or capacitive touch screen.

Because the operation of all kinds of personal electric products aspect identification is similar, so, be that example is set forth to the identification that is applied on the smart mobile phone only hereinafter.

When needs carry out identification, for example start shooting, during operation such as screen release or application program unlatching, smart mobile phone can point out the user to carry out identification, at this moment, the user just can writing at the enterprising style of writing word of touching display screen, in the user writing literal, touching display screen is just real-time to have got access to the handwriting information corresponding with writing words.

S200: characteristic extracting module is extracted the recognition feature of described person's handwriting to be identified, and described recognition feature is used for characterizing the specificity of person's handwriting;

Characteristic extracting module is handled the handwriting information that touching-type monitor collects, to obtain the recognition feature of person's handwriting to be identified.Because the literal that everyone writes all has the uniqueness of self, therefore identify this uniqueness with recognition feature, also be the specificity of person's handwriting.

S300: judge module judges whether described recognition feature and the standard feature that is stored in the person's handwriting sample library module mate, if judgement knows that described recognition feature and described standard feature are complementary, then identity is legal; Otherwise identity is illegal, and described standard feature is used for characterizing the specificity of described standard handwriting.

Before carrying out identification, at first in person's handwriting sample library module, make up a standard feature sample storehouse, standard feature sample storehouse can adopt the mode of features training to make up.The handwriting characteristic of a plurality of literal is generally contained in standard feature sample storehouse, also can only contain the handwriting characteristic of a literal certainly.

The structure in standard feature sample storehouse can be finished by once making up, and also can be divided into repeatedly making up.When repeatedly making up, whether have the person's handwriting in the back input in the criterion feature samples storehouse, there has been above-mentioned person's handwriting if judge standard feature sample storehouse, then it being abandoned need not, do not store above-mentioned person's handwriting if judge standard feature sample storehouse, then will extract the feature of this person's handwriting, and this person's handwriting will be deposited in the standard feature sample storehouse.

When carrying out identification, judge module mates with the recognition feature of extracting among the step S200 and the standard feature that is stored in the person's handwriting sample library module, if recognition feature and standard feature are complementary, then judge and know that user's identity is legal, allow the user to use this smart mobile phone, if recognition feature and standard feature do not match, then judge and know that user's identity is illegal, do not allow the user to use this smart mobile phone.

The present invention is based among the method embodiment of identification of person's handwriting, adopt the hand-written person's handwriting of user to carry out the whether judgement of user identity legitimacy, because, everyone person's handwriting has uniqueness, and in long one period (such as several years), be can arbitrarily not change, therefore guaranteed the stability of identification.Therefore in addition, because that person's handwriting is difficult for is imitated, adopts person's handwriting to carry out identification and have higher security.

Further, based on above-described embodiment 1, as shown in Figure 2, except comprising above-mentioned steps S100～S200, before execution in step S200, also comprise:

S101: pretreatment module is carried out medium filtering to described person's handwriting to be identified.

After the person's handwriting acquisition module collected person's handwriting to be identified, pretreatment module was carried out pretreatment operation to described person's handwriting to be identified.Pretreatment operation includes but are not limited to carries out medium filtering to person's handwriting to be identified.Can eliminate information such as noise, redundancy by person's handwriting to be identified is carried out medium filtering, be conducive to improve the accuracy of person's handwriting coupling identification.

Further, based on above-described embodiment 1, as shown in Figure 3, except comprising above-mentioned steps S100～S200, before execution in step S200, also comprise:

S102: pretreatment module is carried out normalized to described person's handwriting to be identified.

After the person's handwriting acquisition module collected person's handwriting to be identified, pretreatment module was carried out pretreatment operation to described person's handwriting to be identified.Pretreatment operation includes but are not limited to carries out normalized to person's handwriting to be identified.Be conducive to improve the judging efficiency whether recognition feature and standard feature mate by normalized.

Based on above-described embodiment, specifically, recognition feature comprises architectural feature and behavioral characteristics, and described architectural feature is used for characterizing the discrete pen section sequence signature that forms of person's handwriting to be identified, and described behavioral characteristics is used for characterizing the velocity series feature of pen trace motion.Comprised behavioral characteristics in recognition feature, behavioral characteristics can reflect the velocity characteristic that individual person writing is, it is difficult to imitated, therefore, has improved this security based on the method for the identification of person's handwriting greatly.

To the person's handwriting to be identified method that forms pen section sequence signature that disperses be, the person's handwriting that the person's handwriting acquisition module is collected is expressed as the spatial spreading coordinate point sequence of constant duration, i.e. and { x (1), y (1) }, { x (2), y (2) } ..., { x (n), y (n) }.Get the curvature at each spatial spreading coordinate points place:

$c\left(i\right)={\left[\frac{\mathrm{dq}}{\mathrm{ds}}\right]}_i,i=\mathrm{1,2,3}...n;$

Wherein,

Carry out smoothing processing to calculating the curvature value that obtains, namely curvature value removes the weighted mean value of three continuity points.After obtaining curvature value, each curvature value and threshold value are compared, greater than threshold value, then the corresponding spatial spreading coordinate points of this curvature value is flex point as if curvature value, serves as a pen section sequence with described flex point with person's handwriting cutting to be identified (dispersing).The threshold value here is a definite value that compares with curvature value, and its size can specifically be determined according to the accuracy requirement of concrete actual needs and identification.After person's handwriting cutting to be identified is pen section sequence, extract following three features than each section in the section sequence at least: pen section starting point coordinate (x _i, y _i), a segment endpoint coordinate (x _e, y _e) and a segment length L _s

Owing to have noise in above-mentioned spatial spreading coordinate point sequence, therefore, the velocity series feature of pen trace motion adopts the flat little method of second order:

$z^{\′}(n+a)=\underset{r=1}{\overset{N}{\mathrm{\Σ}}}{w}_r{z}_r;$

$W_r=\frac{1}{h}\{\frac{12(r-\stackrel{\‾}{r})}{n(n^2-1)}+\frac{180[{(r-\stackrel{\‾}{r})}^2-(n^2-1)/12]2(r_1-\stackrel{\‾}{r})}{n(n^2-1)(n^2-4)}\};$

$\stackrel{\‾}{r}=(n+1)/2;$

$\left\{\begin{array}{c}V\left(t\right)=\sqrt{V_X^2\left(t\right)+V_y^2\left(t\right)}\\ \mathrm{\θ}\left(t\right)\text{arctan[}\frac{V_y\left(t\right)}{V_X\left(t\right)}\text{]}\end{array}\right.;$

Wherein,

N is flat little node number;

N+a is the flat little position of speed point in velocity series of asking;

r ₁Position for flat differential point;

H is the differential interval;

W _rBe the power sequence;

V _x(t) speed component on the rectangular coordinate system x axle in the handwriting;

V _y(t) speed component on the rectangular coordinate system y axle in the handwriting;

V (t) is the mould value of velocity series;

θ (t) is deflection.

In above-mentioned algorithm, the data inlet side point at the velocity series feature two ends of the motion of identifying the handwriting is level and smooth, and center section carries out central smoothing, has improved the precision that data are handled like this, is conducive to improve the accuracy of identification.When each section is level and smooth, then,

At this moment:

$W_r=\frac{1}{h}\frac{12(r-\stackrel{\‾}{r})}{n(n^2-1)}.$

Judge module judges whether described recognition feature and the standard feature that is stored in the person's handwriting sample library module mate the following algorithm of employing:

G=g ₁g ₂…g _m；

X=x ₁x ₂…x _n；

$\left\{\begin{array}{c}D\left(\mathrm{1,1}\right)=d\left(\mathrm{1,1}\right)\\ D(i,j)=d(i,j)+\min \{D(i-j),D(i-1,j-1),D(i,j-1)\};\\ \mathrm{Dm}(G,X)=D(m,n)\end{array}\right.$

$d(i,j)=\sqrt{{(x_i^{g_i}-x_i^{x_j})}^2+{(y_i^{g_i}-y_i^{x_j})}^2}+\sqrt{{(x_e^{g_i}-x_e^{x_j})}^2+{(y_e^{g_i}-y_e^{x_j})}^2}+C_L|L_s^{g_i}-L_s^{x_j}|;$

D _T=λ ₁D _S+λ ₂D _d；

If D _T≤ TH, it is legal then to be judged as identity, otherwise identity is illegal;

Wherein,

g _mBe m pen section in the pen section sequence of person's handwriting to be identified;

x _nBe n pen section in the pen section sequence of standard handwriting;

G is the pen section sequence of person's handwriting to be identified;

X is the pen section sequence of standard handwriting;

D _m(G X) is total matching distance between the pen section sequence of pen section sequence and standard handwriting of person's handwriting to be identified;

D (i, j) be in the pen section sequence of person's handwriting to be identified before in the pen section sequence of i pen section and standard handwriting before matching distance between j section;

(i j) is the matching distance between j section in the pen section sequence of i pen section and standard handwriting in the pen section sequence of person's handwriting to be identified to d;

(i j) is the matching distance between j section behavioral characteristics in the pen section sequence of i pen section behavioral characteristics and standard handwriting in the pen section sequence of person's handwriting to be identified to d ';

Be velocity vector;

X _s, Y _sBe pen section starting point coordinate;

X _e, Y _eBe a segment endpoint coordinate;

L _SLength for the pen section;

C _LBe weight coefficient;

i=1,2,...,m；

j=1,2,...,n；

λ ₁Weight coefficient for the matching distance of standard handwriting;

λ ₂Weight coefficient for the matching distance of person's handwriting to be identified;

D _sBe the total matching distance of standard handwriting;

D _dBe the total matching distance of person's handwriting to be identified;

D _TBe matching distance total between standard handwriting and the person's handwriting to be identified;

TH is identity legitimacy criterion value.

TH is a definite value, and it is fixed to be come by the similarity of concrete identification rank and identification.If it is more little that the grade of the identification that needs, the value of the more high then TH of similarity obtain, otherwise more big.

As shown in Figure 4, the present invention is based on the structural representation of terminal embodiment of the identification of person's handwriting, comprise person's handwriting acquisition module 1, characteristic extracting module 2, person's handwriting sample library module 3 and judge module 4.

Particularly, person's handwriting acquisition module 1 is used for gathering person's handwriting to be identified; Characteristic extracting module 2 is used for extracting the recognition feature of described person's handwriting to be identified, and described recognition feature is used for characterizing the specificity of person's handwriting; Person's handwriting sample library module 3 is used for the storage standards person's handwriting; Judge module 4 is used for judging whether described recognition feature and the standard feature that is stored in the standard handwriting of person's handwriting sample library module mate, if judgement knows that described recognition feature and described standard feature are complementary, then identity is legal; Otherwise identity is illegal, and described standard feature is used for characterizing the specificity of described standard handwriting.

Function and treatment scheme based on the terminal of the identification of person's handwriting that the embodiment of the invention provides can repeat no more referring to said method embodiment herein.

The present invention is based among the terminal embodiment of identification of person's handwriting, carry out the whether judgement of user identity legitimacy with the hand-written person's handwriting of user, because, everyone person's handwriting has uniqueness, and in long one period (such as several years), be can arbitrarily not change, therefore guaranteed the stability of identification.Therefore in addition, because that person's handwriting is difficult for is imitated, adopts person's handwriting to carry out identification and have higher security.

Based on above-described embodiment, as shown in Figure 5, the present invention is based on another embodiment of terminal of the identification of person's handwriting, person's handwriting acquisition module 1, characteristic extracting module 2, person's handwriting sample library module 3 and judge module 4 except comprising above-described embodiment also comprise pretreatment module 5.

Pretreatment module 5 is used for described person's handwriting to be identified is carried out medium filtering and/or normalized.

Can eliminate information such as noise, redundancy by person's handwriting to be identified is carried out medium filtering, be conducive to improve the accuracy of person's handwriting coupling identification.Be conducive to improve the judging efficiency whether recognition feature and standard feature mate by normalized.

It should be noted that at last: above embodiment only in order to technical scheme of the present invention to be described, is not intended to limit; Although with reference to previous embodiment the present invention is had been described in detail, those of ordinary skill in the art is to be understood that: it still can be made amendment to the technical scheme that aforementioned each embodiment puts down in writing, and perhaps part technical characterictic wherein is equal to replacement; And these modifications or replacement do not make the essence of appropriate technical solution break away from the spirit and scope of various embodiments of the present invention technical scheme.

Claims (7)

1. the method based on the identification of person's handwriting is characterized in that, may further comprise the steps:

The person's handwriting acquisition module is gathered person's handwriting to be identified;

Characteristic extracting module is extracted the recognition feature of described person's handwriting to be identified, and described recognition feature is used for characterizing the specificity of person's handwriting to be identified;

Judge module judges whether described recognition feature and the standard feature that is stored in the standard handwriting in the person's handwriting sample library module mate, if judgement knows that described recognition feature and described standard feature are complementary, then identity is legal; Otherwise identity is illegal, and described standard feature is used for characterizing the specificity of described standard handwriting;

Described recognition feature comprises architectural feature and behavioral characteristics, and described architectural feature is used for characterizing the discrete pen section sequence signature that forms of person's handwriting to be identified, and described behavioral characteristics is used for characterizing the velocity series feature of pen trace motion;

The algorithm of described behavioral characteristics is:

$z^{\′}(n+a)=\underset{r=1}{\overset{N}{\mathrm{\Σ}}}{w}_r{z}_r;$

$W_r=\frac{1}{h}\{\frac{12(r-\stackrel{\‾}{r})}{n(n^2-1)}+\frac{180[{(r-\stackrel{\‾}{r})}^2-(n^2-1)/12]2(r_1-\stackrel{\‾}{r})}{n(n^2-1)(n^2-4)}\};$

$\stackrel{\‾}{r}=(n+1)/2;$

$\left\{\begin{array}{c}V\left(t\right)=\sqrt{V_X^2\left(t\right)+V_y^2\left(t\right)}\\ \mathrm{\θ}\left(t\right)\text{arctan[}\frac{V_y\left(t\right)}{V_X\left(t\right)}\text{]}\end{array}\right.;$

Wherein,

N is flat little node number;

N+a is the flat little position of speed point in velocity series of asking;

W _rBe the power sequence;

r ₁Position for flat differential point;

H is the differential interval;

V _x(t) speed component on the rectangular coordinate system x axle in the handwriting;

V _y(t) speed component on the rectangular coordinate system y axle in the handwriting;

V (t) is the mould value of velocity series;

θ (t) is deflection.

2. the method for the identification based on person's handwriting according to claim 1 is characterized in that, before described characteristic extracting module is extracted described recognition feature, also comprises:

Pretreatment module is carried out medium filtering to described person's handwriting to be identified.

3. the method for the identification based on person's handwriting according to claim 1 is characterized in that, before described characteristic extracting module is extracted described recognition feature, also comprises:

Pretreatment module is carried out normalized to described person's handwriting to be identified.

4. the method for the identification based on person's handwriting according to claim 1 is characterized in that, the discrete method that forms pen section sequence signature of person's handwriting to be identified is:

Described characteristic extracting module disperses person's handwriting constant duration to be identified and is the spatial spreading coordinate point sequence, and whether the curvature of judging each spatial spreading coordinate points place is greater than the cutting threshold value, if judge and know that described curvature is greater than the cutting threshold value, then be flex point greater than the corresponding spatial spreading coordinate points of the described curvature of described cutting threshold value, will described person's handwriting cutting to be identified be a section sequence according to described flex point.

5. the method for the identification based on person's handwriting according to claim 4 is characterized in that, each section in the described pen section sequence comprises following three features at least: pen section starting point coordinate, a segment endpoint coordinate and a segment length.

6. the terminal based on the identification of person's handwriting is characterized in that, comprising:

The person's handwriting acquisition module is used for gathering person's handwriting to be identified;

Characteristic extracting module, for the recognition feature of extracting described person's handwriting to be identified, described recognition feature is used for characterizing the specificity of person's handwriting;

Person's handwriting sample library module is used for the storage standards person's handwriting;

Judge module is used for judging whether described recognition feature and the standard feature that is stored in the standard handwriting of person's handwriting sample library module mate, if judgement knows that described recognition feature and described standard feature are complementary, then identity is legal; Otherwise identity is illegal, and described standard feature is used for characterizing the specificity of described standard handwriting;

Described recognition feature comprises architectural feature and behavioral characteristics, and described architectural feature is used for characterizing the discrete pen section sequence signature that forms of person's handwriting to be identified, and described behavioral characteristics is used for characterizing the velocity series feature of pen trace motion;

The algorithm that described characteristic extracting module is extracted described behavioral characteristics is:

$z^{\′}(n+a)=\underset{r=1}{\overset{N}{\mathrm{\Σ}}}{w}_r{z}_r;$

$W_r=\frac{1}{h}\{\frac{12(r-\stackrel{\‾}{r})}{n(n^2-1)}+\frac{180[{(r-\stackrel{\‾}{r})}^2-(n^2-1)/12]2(r_1-\stackrel{\‾}{r})}{n(n^2-1)(n^2-4)}\};$

$\stackrel{\‾}{r}=(n+1)/2;$

$\left\{\begin{array}{c}V\left(t\right)=\sqrt{V_X^2\left(t\right)+V_y^2\left(t\right)}\\ \mathrm{\θ}\left(t\right)\text{arctan[}\frac{V_y\left(t\right)}{V_X\left(t\right)}\text{]}\end{array}\right.;$

Wherein,

N is flat little node number;

N+a is the flat little position of speed point in velocity series of asking;

W _rBe the power sequence;

r ₁Position for flat differential point;

H is the differential interval;

V _x(t) speed component on the rectangular coordinate system x axle in the handwriting;

V _y(t) speed component on the rectangular coordinate system y axle in the handwriting;

V (t) is the mould value of velocity series;

θ (t) is deflection.

7. the terminal of the identification based on person's handwriting according to claim 6 is characterized in that, also comprises:

Pretreatment module is used for described person's handwriting to be identified is carried out medium filtering and/or normalized.

Images