CN1920856A - Computer assisted calligraphic works distinguishing method between true and false - Google Patents

Abstract

The invention relates to a method for checking the truth-false of calligraphy article by computer, wherein it comprises: (1) extracting the strike character of calligraphy word; (2) stating and analyzing the characters of structure and strike of different calligraphy workers; (3) building the personal vector of single calligraphy worker; 94), based on the calligraphy worker of suspected article, building the true mode of said calligraphy worker; (5), based on the personal character vector, comparing the characters of suspected article, checking the suspected point and calculating the false probability; (6) giving out check result that based on the false article, giving out the false probability and relative evidence. The invention has the advantages that: it can use computer to check the calligraphy articles with objective evidence.

Description

A kind of computer assisted calligraphic works distinguishing method between true and false

Technical field

The present invention relates to the digital picture identification in the computer technology, relate in particular to a kind of computer assisted calligraphic works distinguishing method between true and false.The performance computing machine is than storage, statistics and the search function of human brain advantage, for the calligraphy connoisseur provides quantification, objective ' irrefutable evidence '.

Background technology

Because calligraphy work has higher artistic value in history, and is expensive, so counterfeit continues to bring out in commodity market.By means of modern high technology especially computer, to fake to come easilier, the number that makes fakement is far more than authentic work.Buy if a width of cloth fakement is worked as authentic work, obviously loss is heavy.The calligraphy authenticity is finished by experienced calligraphy art man all the time, evaluation is subjective, is main flow with " prestige style " mainly, judges with personal experience's intuition, also have in addition by seeing money seal whether reliably " money seal group " and have with history that not record be foundation " record and send ".Be subjected to the limitation of human brain, any connoisseur is general only to be familiar with several or several calligraphists that sends, can only remember a part of data, by rule of thumb with the sensation judgement true and false, and lacks " real example ".This makes calligraphy identify " mysterious " that " understand by thinking and cannot be explained in words " often occur, allows people's indigestion.

The present invention utilizes computing machine more than the powerful storage memory function of human brain, retrieval and analysis comparing function, and corresponding authentic work style in all style personal characteristics in the suspicious works and the database is compared.For example certain width of cloth authentic work works has 20 writing brush word, and each writing brush word has 10 unique points, supposes each potential feature forgery probability of successful of adulterator up to 80%, and then the view picture works are forged to such an extent that with the living probability of authentic work be $\frac{1}{20}\×{(80\%)}^{10}=0.0053687,$ It is very little promptly to forge probability of successful, and always having suspicious points can be detected.The present invention detects all suspicious points and calculates the possibility size that it is the authentic work feature, and providing the view picture works at last is possibility sizes of authentic work.If judging suspicious works is that the possibility of counterfeit is bigger, it is suspicious then providing which point simultaneously.

Summary of the invention

The objective of the invention is to overcome the weak point of prior art, propose a kind of computer assisted calligraphic works distinguishing method between true and false.In storage, calculating and advantage analytically, auxiliary calligraphy work real and fake discrimination is for the calligraphy real and fake discrimination provides objectively a kind of method of " evidence " by means of computing machine.

Comprise the steps:

(1) by following the tracks of the method for writing brush word skeleton and profile simultaneously, extracts the writing brush word stroke feature;

(2) the different calligraphists' of statistics and analysis writing brush word is in knot volume morphing and the modal feature of stroke;

(3) utilize Gaussian distribution to choose the characteristic feature feature of single calligraphist's writing brush word, and calculate the weight of each feature, structure characteristic feature vector;

(4),, construct this calligraphist's authentic work model according to the authentic work works of this calligraphist in the database at the calligraphist that suspicious works are declared according to gauss of distribution function;

(5) according to each feature in calligraphist's the characteristic feature proper vector, corresponding feature detects suspicious points in the suspicious works of comparative analysis; Calculate this suspicious probability that is characterized as counterfeit;

(6) provide qualification result, so provide counterfeit " evidence " when being the probability of counterfeit to being judged to be the suspicious works of counterfeit, providing.

Described a kind of computer assisted calligraphic works distinguishing method between true and false is characterized in that describedly by following the tracks of the method for writing brush word skeleton and profile simultaneously, and the step of extracting the writing brush word stroke feature is as follows:

(1) follows the tracks of the writing brush word skeleton, detect intersection point bunch;

(2) follow the tracks of the writing brush word profile, detect flex point;

(3) extract correct stroke profile according to the flex point of the intersection point of skeleton bunch and profile.

Described a kind of computer assisted calligraphic works distinguishing method between true and false is characterized in that the different calligraphists' of described statistics and analysis writing brush word is as follows in the step of knot volume morphing and the modal feature of stroke:

(1) knot volume morphing feature comprises: the depth-width ratio example of word, centre of gravity place, being boundary's number of side ink dot up and down with the center of gravity, is boundary's inclination gradient up and down with the center of gravity;

(2) the stroke morphological feature comprises: stroke entropy, stroke degree of jitter, degree of crook, thickness change.

Described a kind of computer assisted calligraphic works distinguishing method between true and false is characterized in that described structure characteristic feature to measure feature is: all authentic works and average and the variance of single calligraphist's authentic work on each unique point in the difference computational data storehouse; According to Gaussian distribution, if the distribution of mean value of certain feature of single calligraphist in the small probability time range of this eigenwert Gaussian distribution of all calligraphists, then this feature is considered to a composition of this calligraphist's personalized style and features.

Described a kind of computer assisted calligraphic works distinguishing method between true and false, it is characterized in that the described qualification result that provides, so to being judged to be the suspicious works of counterfeit, providing the method that provides counterfeit " evidence " when being the probability of counterfeit is: it in the suspicious works may be the writing brush word of counterfeit that frame goes out, when clicking this word, provide the contrast of this word and corresponding authentic work; Simultaneously, provide the source of the authentic work writing brush word that is used as contrast.

The useful effect that the present invention has is: the invention solves the calligraphy work real and fake discrimination and finish, differentiate the problem that is subject to human brain subjective experience limitation by the calligraphist all the time, " suspicious evidence " for the calligraphy work real and fake discrimination provides objectively, quantized helps to clarify the painting and calligraphy market that present true and false works fish eyes mixes.

Description of drawings

Fig. 1 is a computer assisted calligraphic works distinguishing method between true and false system framework block diagram;

Fig. 2 is the committed step workflow diagram of computer assisted calligraphic works distinguishing method between true and false;

Fig. 3 is the single writing brush word cutting synoptic diagram of computer aided calligraphy works real and fake discrimination;

Fig. 4 (a) is the chain code direction signal that computer assisted calligraphic works distinguishing method between true and false adopts;

Fig. 4 (b) is that the corresponding image pixel of Fig. 4 (a) is put the signal of 8 neighborhoods coding;

Fig. 5 is the Gaussian distribution synoptic diagram that computer assisted calligraphic works distinguishing method between true and false adopts;

Fig. 6 is the described example of extracting the writing brush word stroke by the method for following the tracks of writing brush word skeleton and profile simultaneously of computer assisted calligraphic works distinguishing method between true and false;

Fig. 7 is horizontal stroke of skeleton, corresponding stroke direction histogram and corresponding stroke entropy contrast thereof in three different writing brush word;

Fig. 8 is that stroke width calculates synoptic diagram, and wherein red pixel is the skeleton point, and d is the corresponding stroke width of skeleton point;

Fig. 9 is the suspicious works true and false estimation synoptic diagram of computer assisted calligraphic works distinguishing method between true and false;

Figure 10 is the real and fake discrimination example of claiming to the large works of Wu Chang, and what blue bounding box frame went out is that suspicious writing brush word is arranged;

Figure 11 is that suspicious " evidence " of the suspicious writing brush word of Figure 10 showed, the horizontal pen that green frame contour goes out is suspicious " evidence ",

Second row, the third line have provided the horizontal stroke literary style reference in the large authentic work works of corresponding Wu Chang;

Figure 12 be first last row of row in Figure 11 authentic work reference result " can " original work of word browse;

Specific implementation method

The present invention proposes a kind of computer assisted calligraphic works distinguishing method between true and false, and in conjunction with the accompanying drawings, its enforcement is described in detail as follows:

System framework of the present invention may further comprise the steps as shown in Figure 1:

1. scan original authentic work calligraphy work, and it is carried out the metadata mark, comprise the key word of doing the name of an article, author, dynasty, comment, set up database table and deposit works metadata and the store path of corresponding original scanning works image in disk.

2. the cutting authentic work calligraphy work page obtains single writing brush word.Detect intersection point bunch by following the tracks of the writing brush word skeleton: for a skeleton point, if the skeleton that is adjacent in its 8 neighborhood o'clock greater than 2, then this point is the point of crossing.The set of adjacent a plurality of point of crossing is called intersection point bunch, is the mark that two pen sections are intersected.Detect flex point by following the tracks of the writing brush word point: a point p _i(x _i, y _i) need satisfy following two conditions for flex point:

(1) the line segment slope that is end point and starting point with this point satisfies: $\frac{y_i-y_{i-k}}{x_i-x_{i-k}}-\frac{y_{i+k}-y_i}{x_{i+k}-x_i}>\mathrm{\θ},$ Wherein θ=0.6 is a threshold value, k=3;

(2) corner angle that is the center with this point satisfies: $c_i=\frac{{\mathrm{pre}}_{\mathrm{ik}}+{\mathrm{next}}_{\mathrm{ik}}}{\left|p_{i+k}{p}_{i-k}\right|}>T,$ Wherein T=1.25 is a threshold value, k=5.

3. the different calligraphist's authentic work of statistics and analysis writing brush word comprises at the modal style and features of stroke:

(1) the stroke entropy of all types of strokes of the different calligraphist's authentic work of statistics and analysis writing brush word:

The stroke entropy of a stroke $E_s=\underset{i=1}{\overset{8}{\mathrm{\Σ}}}\mathrm{sd}\left[i\right]\×\log \left(\mathrm{sd}\right[i\left]\right),$ Wherein $\mathrm{sd}\left[i\right]=\frac{\#\{{\mathrm{code}}_j=i\}}{\mathrm{length}},$ Code _jBe j encoded radio in these stroke 8 direction chain codes, length is the chain code length.

(2) the stroke degree of jitter of the different calligraphist's authentic work of statistics and analysis writing brush word:

Make length _OriginalBe the point number of the stroke that will calculate the shake degree, length _{Zoom out}Dwindle point number after 0.5 times for this stroke, then degree of shake Jitterness is:

$\mathrm{Jitteriness}=\frac{{\mathrm{length}}_{\mathrm{original}}}{{\mathrm{length}}_{0.5}}$

(3) stroke weight of the different calligraphist's authentic work of statistics and analysis writing brush word changes, and to a certain specified type stroke, its thickness changes available mean square deviation square value and is expressed as:

$\mathrm{tv}=\frac{1}{n}\underset{i=1}{\overset{i=n}{\mathrm{\Σ}}}{(w_i-\stackrel{\‾}{w})}^2,$ Wherein $\stackrel{\‾}{w}=\frac{1}{n}\underset{i=1}{\overset{i=n}{\mathrm{\Σ}}}{w}_i$

(4) distortion of the facts by an official historian flexibility tendency of the different calligraphist's authentic work of statistics and analysis writing brush word:

With 8 chain representation stroke skeletons trend, its direction numbering is shown in Fig. 4 (a), and corresponding pictorial element 8 neighborhoods coding is shown in Fig. 4 (b).The stroke tendency changes to the number of an opposite direction from a direction in the statistics chain code sequence, and note is made t.Make lb _iAnd la _iBe respectively peak dot p _iThe length of stroke different directions tendency before and afterwards, tendency length is defined as: l _i=min{lb _i, la _i, the value of flexibility can be calculated by following formula:

$\mathrm{rt}=1+\underset{i=1}{\overset{i=t}{\mathrm{\Σ}}}\frac{t\×l_i+{\mathrm{dif}}_i}{\mathrm{length\; of\; curve}}$

4. the style and features on the knot volume morphing of the different calligraphist's authentic work of statistics and analysis writing brush word comprises:

(1) calculate the height and width ratio of different calligraphist's authentic work writing brush word, as following formula:

${\mathrm{ratio}}_{h/w}=\frac{\mathrm{height}}{\mathrm{width}}$

(2) make the writing brush word image function be f (x, y), calculate the X-axle of writing brush word of authentic work and counterfeit and the centre of gravity place of Y-axle: the p+q rank square of a writing brush word can be write: $m_{\mathrm{pq}}=\underset{x=0}{\overset{x=M-1}{\mathrm{\Σ}}}\underset{y=0}{\overset{y=N-1}{\mathrm{\Σ}}}{x}^p{y}^{q}f(x,y),$ Then its X-direction of principal axis center of gravity is:

$\stackrel{\‾}{x}=\frac{m_{10}}{m_{00}},$ Y-direction of principal axis center of gravity is: $\stackrel{\‾}{y}=\frac{m_{01}}{m_{00}}$

(3) the different calligraphist's authentic work writing brush word of calculating is the boundary with the center of gravity, and reach the degree of tilt ratio of left and right sides pen and ink number and word up and down: the p+q rank central moment of writing brush word can be write:

$u_{\mathrm{pq}}=\underset{x=0}{\overset{x=M-1}{\mathrm{\Σ}}}\underset{y=0}{\overset{y=N-1}{\mathrm{\Σ}}}{(x-\stackrel{\‾}{x})}^p{(y-\stackrel{\‾}{y})}^{q}f(x,y),$

Writing brush word is that boundary's left and right sides pen and ink number changes and can write with X-axle center of gravity: $s_h=\frac{{u_{30}}^{+}}{{u_{30}}^{+}+{u_{30}}^{-}};$

Writing brush word is that boundary's upper and lower sides pen and ink number changes and can write with Y-axle center of gravity: $s_v=\frac{{u_{03}}^{+}}{{u_{03}}^{+}+{u_{03}}^{-}};$

Writing brush word is that boundary's left and right sides degree of tilt ratio is with Y-axle center of gravity: $b_h=\frac{{u_{21}}^{+}}{{u_{21}}^{+}+{u_{21}}^{-}};$

Writing brush word is that boundary's left and right sides pen and ink number changes and can write with X-axle center of gravity: $b_v=\frac{{u_{12}}^{+}}{{u_{12}}^{+}+{u_{12}}^{-}}.$

5. be not the personal characteristics that each feature all is meant the agreement Legalists, therefore above-mentioned style and features carried out Feature Selection, choose personal characteristics and be combined into the proper vector that characterizes appointment calligraphist style:

According to Gaussian statistics model shown in Figure 5, each feature f of all calligraphist's authentic work works in the computational data _iAverage u _iAnd variances sigma _i, and specify feature f in calligraphist's works _iAverage Make λ=1.5 be threshold value, if this feature satisfies following formula:

$|\stackrel{\‾}{x_i}-u|>\mathrm{\λ}\×{\mathrm{\σ}}_i$

Then this feature is that this calligraphist's personalized style is various, adds in the proper vector that characterizes its style.

6. according to each vector in the personalized style and features vector of specifying calligraphist's authentic work works, corresponding feature in the suspicious works of comparative analysis detects suspicious points: order

For specifying the vector f in calligraphist's proper vector _iVariance, u is an average,

Average for the suspicious works of the need submitted to checking.Make λ=1.5 be threshold value, then according to Gaussian statistics model shown in Figure 5, satisfy as this feature following formula, then this suspicious feature may be the counterfeit feature:

$|\stackrel{\‾}{x_i}-u|>\mathrm{\λ}\×{\mathrm{\σ}}_i$

Then this suspicious points is that the probability of counterfeit is:

$p\left(x_i\right)=2\×{\∫}_{l=x_i}^{t=\∞}\frac{1}{\sqrt{2{\mathrm{\π\σ}}_i}}\exp \left(\frac{{(t-\stackrel{\‾}{x_i})}^2}{2\stackrel{\‾}{{\mathrm{\σ}}_i^2}}\right)\mathrm{dt}$

But the probability that this suspicious writing brush word is a counterfeit is:

$p\left(c\right)=\frac{1}{m}\underset{i=1}{\overset{i=m}{\mathrm{\Σ}}}p\left(x_i\right)$

The suspicious works of view picture be that to equal all suspicious individual characters are the average of counterfeit probability for the probability of counterfeit.

7. the result shows:

The user submits a width of cloth works to, computing machine output identification result:

(1) at blue bounding box of suspicious writing brush word upper ledge, represent that this writing brush word is suspicious, may be counterfeit;

(2) when user's click box has the suspicious writing brush word of blue bounding box, eject a new page, show why this word may be " evidence " of counterfeit: the suspicious stroke of suspicious book word is gone out with green profile wire frame, retrieve corresponding authentic work stroke in this calligraphist's authentic work database simultaneously, go out with green profile wire frame equally, so that user's contrast, and, what realizes it is to be counterfeit according to prompting;

(3) wonder whether authentic work is reliable as the user, when coming wherefrom, can click this authentic work writing brush word, eject a new page, show the original authentic work works of this word, and frame goes out this word position.

Embodiment:

(1) the cutting authentic work calligraphy work page obtains single writing brush word, as shown in Figure 3, and structure authentic work database;

(2) extract writing brush word stroke profile, example as shown in Figure 6;

(3) the different calligraphist's authentic work of statistics and analysis writing brush word is at the modal style and features of stroke:

Calculate horizontal the stroke entropy that the green profile of three writing brush word as shown in Figure 7 marks:

$E_s=\underset{i=1}{\overset{8}{\mathrm{\Σ}}}\mathrm{sd}\left[i\right]\×\log \left(\mathrm{sd}\right[i\left]\right),$ Wherein $\mathrm{sd}\left[i\right]=\frac{\#\{{\mathrm{code}}_j=i\}}{\mathrm{length}},$ It is worth shown in Fig. 7 statistical graph, corresponding E _sMark as Fig. 7, know that easily first calligraphy " equals " the stroke entropy maximum of horizontal stroke of word, that is that horizontal stroke of flexibility of this word is big because compare with horizontal stroke of all the other two writing brush word, and it is many to comprise information; With the stroke that straight line draws, the stroke entropy reaches minimum value E _s=0; When stroke was a circle, each direction proportion equated in the 8 direction chain codes of this stroke, and at this moment, the stroke entropy reaches maximal value.Stroke twisting degree is big more, and its stroke entropy is just big more;

Calculate stroke shake degree: the shake degree formula of stroke $\mathrm{Jitteriness}=\frac{{\mathrm{length}}_{\mathrm{original}}}{{\mathrm{length}}_{0.5}}$ Calculate;

Calculate the degree of crook of stroke: the stroke flexibility is by formula $\mathrm{rt}=1+\underset{i=1}{\overset{i=t}{\mathrm{\Σ}}}\frac{t\×l_i+{\mathrm{dif}}_i}{\mathrm{length\; of\; curve}}$ Calculate, wherein t is that the stroke tendency changes to the number of an opposite direction, lb from a direction in the chain code sequence _iAnd la _iBe respectively before the peak dot and the length of stroke different directions tendency afterwards, l _i=min{lb _i, la _i;

The calculating stroke weight changes: stroke weight changes by formula ${\mathrm{\σ}}^2=\frac{1}{t}\underset{i=1}{\overset{i=t}{\mathrm{\Σ}}}{(d_i-w)}^2$ Calculate, wherein $w=\frac{1}{t}\underset{i=1}{\overset{t}{\mathrm{\Σ}}}{d}_i,d_i$ Be the width of i skeleton point, t is the number of skeleton point, shown in Fig. 8 sectional drawing.

(4) style and features of the different calligraphist's authentic work of statistics and analysis writing brush word on the knot volume morphing,

Calculate the depth-width ratio example of writing brush word: ${\mathrm{ratio}}_{h/w}=\frac{\mathrm{height}}{\mathrm{width}};$

Calculate the centre of gravity place of writing brush word: (x, y), wherein $\stackrel{\‾}{x}=,\frac{m_{10}}{m_{00}},\stackrel{\‾}{y}{=\frac{m_{01}}{m_{00}},m}_{\mathrm{pq}}=\underset{x=0}{\overset{x=M-1}{\mathrm{\Σ}}}\underset{y=0}{\overset{y=N-1}{\mathrm{\Σ}}}{x}^p{y}^{q}f(x,y),$

Calculating writing brush word is that boundary's left and right sides pen and ink number changes with X-axle center of gravity: $s_h=\frac{{u_{30}}^{+}}{{u_{30}}^{+}+{u_{30}}^{-}};$

Calculating writing brush word is that boundary's upper and lower sides pen and ink number changes and can write with Y-axle center of gravity: $s_v=\frac{{u_{03}}^{+}}{{u_{03}}^{+}+{u_{03}}^{-}};$

Calculating writing brush word is that boundary's left and right sides degree of tilt ratio is with Y-axle center of gravity: $b_h=\frac{{u_{21}}^{+}}{{u_{21}}^{+}+{u_{21}}^{-}};$

Calculating writing brush word is that boundary's left and right sides pen and ink number changes and can write with X-axle center of gravity: $b_v=\frac{{u_{12}}^{+}}{{u_{12}}^{+}+{u_{12}}^{-}};$

(5) according to each feature in calligraphist's the characteristic feature proper vector, corresponding feature detects suspicious points in the suspicious works of comparative analysis; Calculate this suspicious probability that is characterized as counterfeit, as shown in Figure 9;

(6) user submits suspicious works to, and computing machine output identification result: at blue bounding box of suspicious writing brush word upper ledge of works, representing that this writing brush word is suspicious, may be " evidence " of counterfeit as these works, as shown in figure 10.

(7) when the user clicks suspicious works acceptance of the bid frame the writing brush word of blue bounding box is arranged, eject a new page, as shown in figure 11, show why this word may be " evidence " of counterfeit: the suspicious stroke of suspicious book word is gone out with green profile wire frame, retrieve corresponding authentic work stroke in this calligraphist's authentic work database simultaneously, go out with green profile wire frame equally, so that user's contrast, and, what realizes it is to be counterfeit according to prompting.

(8) with reference to whether reliable, when coming wherefrom, can click this authentic work writing brush word when the user wonders the authentic work writing brush word, eject a new page, as shown in figure 12, show the original authentic work works of this word, and frame go out this word position.

Claims (5)

1. a computer assisted calligraphic works distinguishing method between true and false is characterized in that comprising the steps:

(1) by following the tracks of the method for writing brush word skeleton and profile simultaneously, extracts the writing brush word stroke feature;

(2) the different calligraphists' of statistics and analysis writing brush word is in knot volume morphing and the modal feature of stroke;

(3) utilize Gaussian distribution to choose the characteristic feature feature of single calligraphist's writing brush word, and calculate the weight of each feature, structure characteristic feature vector;

(4),, construct this calligraphist's authentic work model according to the authentic work works of this calligraphist in the database at the calligraphist that suspicious works are declared according to gauss of distribution function;

(5) according to each feature in calligraphist's the characteristic feature proper vector, corresponding feature detects suspicious points in the suspicious works of comparative analysis; Calculate this suspicious probability that is characterized as counterfeit;

(6) provide qualification result, so provide counterfeit " evidence " when being the probability of counterfeit to being judged to be the suspicious works of counterfeit, providing.

2. a kind of computer assisted calligraphic works distinguishing method between true and false according to claim 1 is characterized in that describedly by following the tracks of the method for writing brush word skeleton and profile simultaneously, and the step of extracting the writing brush word stroke feature is as follows:

(1) follows the tracks of the writing brush word skeleton, detect intersection point bunch;

(2) follow the tracks of the writing brush word profile, detect flex point;

(3) extract correct stroke profile according to the flex point of the intersection point of skeleton bunch and profile.

3. a kind of computer assisted calligraphic works distinguishing method between true and false according to claim 1 is characterized in that the different calligraphists' of described statistics and analysis writing brush word is as follows in the step of knot volume morphing and the modal feature of stroke:

(1) knot volume morphing feature comprises: the depth-width ratio example of word, centre of gravity place, being boundary's number of side ink dot up and down with the center of gravity, is boundary's inclination gradient up and down with the center of gravity;

(2) the stroke morphological feature comprises: stroke entropy, stroke degree of jitter, degree of crook, thickness change.

4. a kind of computer assisted calligraphic works distinguishing method between true and false according to claim 1 is characterized in that described structure characteristic feature to measure feature is: all authentic works and average and the variance of single calligraphist's authentic work on each unique point in the difference computational data storehouse; According to Gaussian distribution, if the distribution of mean value of certain feature of single calligraphist in the small probability time range of this eigenwert Gaussian distribution of all calligraphists, then this feature is considered to a composition of this calligraphist's personalized style and features.

5. a kind of computer assisted calligraphic works distinguishing method between true and false according to claim 1, it is characterized in that the described qualification result that provides, so to being judged to be the suspicious works of counterfeit, providing the method that provides counterfeit " evidence " when being the probability of counterfeit is: it in the suspicious works may be the writing brush word of counterfeit that frame goes out, when clicking this word, provide the contrast of this word and corresponding authentic work; Simultaneously, provide the source of the authentic work writing brush word that is used as contrast.

Images