CN107292213B - Handwriting quantitative inspection and identification method - Google Patents
Handwriting quantitative inspection and identification method Download PDFInfo
- Publication number
- CN107292213B CN107292213B CN201610191639.1A CN201610191639A CN107292213B CN 107292213 B CN107292213 B CN 107292213B CN 201610191639 A CN201610191639 A CN 201610191639A CN 107292213 B CN107292213 B CN 107292213B
- Authority
- CN
- China
- Prior art keywords
- handwriting
- features
- value
- level
- quantitative
- 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.)
- Expired - Fee Related
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/30—Writer recognition; Reading and verifying signatures
- G06V40/33—Writer recognition; Reading and verifying signatures based only on signature image, e.g. static signature recognition
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Collating Specific Patterns (AREA)
- Character Discrimination (AREA)
Abstract
A handwriting quantitative inspection and identification method belongs to a handwriting quantitative inspection and identification method in the court science field, and particularly relates to a handwriting quantitative inspection and identification method. The invention provides an objective and standard handwriting quantitative inspection and identification method. The invention comprises the following steps: 1) constructing a handwriting characteristic hierarchical structure and a value evaluation analysis model; 2) determining the weight value of each handwriting characteristic; 3) establishing a corresponding relation between the handwriting characteristic occurrence rate and the coincidence value; 4) making a hierarchical feature comparison table, and calculating the correlation value of the handwriting feature occurrence rate and the coincidence value; 5) and establishing the corresponding relation between the correlation value and the inspection appraisal opinion.
Description
Technical Field
The invention belongs to a handwriting quantitative inspection and identification method in the field of court science, and particularly relates to a handwriting quantitative inspection and identification method.
Background
Handwriting is a system of characters and symbols with personal characteristics formed by writing activities. Handwriting testing is a specialized test performed by an authenticating person to determine whether a handwritten script on a physical evidence as evidence was written by a particular person.
At present, handwriting inspection and identification are mainly based on the special knowledge and experience of handwriting identification technicians, and whether the physical evidence handwriting and the sample handwriting are the same person handwriting is determined by comparing the specificity, the conformity and the stability of various characteristics of the physical evidence handwriting and the sample handwriting and comprehensively evaluating the physical evidence handwriting and the sample handwriting. The handwriting inspection and identification process is mainly based on expert experience qualitative judgment, and a quantitative evaluation and quantitative comparison method based on objective statistical data is lacked, so that inspection and identification opinions lack convincing objective bases, or when the identification opinions are diverged, an objective quantitative standard is not available for unified recognition, the inspection and judgment work is often difficult, the technical development of handwriting inspection and identification is limited, and the scientificity of handwriting inspection and identification is widely questioned.
Disclosure of Invention
Aiming at the problems, the invention provides an objective and standard handwriting quantitative inspection and identification method.
In order to achieve the purpose, the invention adopts the following technical scheme, and the invention comprises the following steps:
1) constructing a handwriting characteristic hierarchical structure and a value evaluation analysis model;
2) determining the weight value of each handwriting characteristic;
3) establishing a corresponding relation between the handwriting characteristic occurrence rate and the coincidence value;
4) making a hierarchical feature comparison table, and calculating the correlation value of the handwriting feature occurrence rate and the coincidence value;
5) and establishing the corresponding relation between the correlation value and the inspection appraisal opinion.
As another preferable scheme, the step 1) of the present invention includes: setting a target layer-handwriting characteristic hierarchical structure and value; determining a middle layer, namely macroscopic level features, mesoscopic level features and microscopic level features; and a refinement scheme layer, wherein the macro level features comprise features, the meso level features comprise features, and the micro level features comprise features.
As another preferred scheme, the macro-level features are overall features, the middle-level features are appearance, content, space and time features, and the micro-level features are two-dimensional and three-dimensional features.
As another preferred solution, the macro-level features in the solution layer of the present invention include: h is1-features are overall layout, line direction and ruled line relationships; h is2-overall style, writing level; h is3-overall composition and inter-word relationships;
the mesoscopic hierarchical features in the solution layer include: z is a radical of1-shape and size; z is a radical of2-structural content; z is a radical of3Writing time sequence of strokes and components; z is a radical of4-stroke and radical collocation proportion space;
the micro-level features in the recipe layer include: w is a1-dot, horizontal, vertical, left-falling, right-falling, bending and hooking; w is a2The corresponding connection relation, trend and connection form among strokes; w is a3-writing speed and distribution rule; w is a4The weight and distribution rule of writing pressure.
As another preferred scheme, the step 2) of the invention is determined by a 1-9-level scaling method and a root method.
As another preferred scheme, the occurrence rate of the handwriting characteristics in the step 3) is divided into: the characteristic occurrence rate of writing of few people is 0.1%; the occurrence rate of writing features of few people is 10%; the occurrence rate of the characters written by a few people is 30%; the occurrence rate of the characters which respectively account for half of people for writing is 50 percent; most people have 70% incidence of writing features;
the coincidence degree value of the handwriting characteristics is divided into: completely meets 100 percent, extremely meets 97.5 percent, very meets 92.5 percent and better meets 87.5 percent; the total difference is 0%, the difference is extremely large 2.5%, the difference is greatly large 7.5%, and the difference is large 12.5%.
As another preferable scheme, the step 4) of creating the hierarchical feature comparison table includes creating a macro, meso and micro hierarchical feature comparison table.
As another preferred scheme, the following formula is adopted for calculating the correlation value between the handwriting characteristic occurrence rate and the coincidence value in the step 4) of the invention:
virepresents: a correlation value; ciRepresents: a conformity value; piRepresents: the occurrence rate of the characteristics, K1 and K2 respectively represent: two constants; n represents: the number of features.
Secondly, the invention calculates the correlation value v of the handwriting characteristicsiThen, respectively calculating handwriting macroscopical VHMiddle view VZMicro VWAnd (3) calculating the total quantitative association value V of handwriting characteristics at each level of characteristic association value:
macro level feature correlation value: vH＝(vh1×h1+vh2×h2+vh3×h3+vh4×h4)×H
Mesoscopic hierarchical feature correlation value: vz＝（vz1×z1+vz2×z2+vz3×z3+vz4×z4)×Z
Microscopic level feature correlation value: vW＝(vw1×w1+vw2×w2+vw3×w3+vw4×w4)×W
The formula for V is: v is VH+VZ+VW
In addition, the corresponding relation between the handwriting characteristic quantitative total correlation value V and the inspection and identification opinions is as follows: 100 > v.gtoreq.95 is definitely identical, 95 > v.gtoreq.85 is very likely to be identical, 85 > v.gtoreq.70 is very likely to be identical, 70 > v.gtoreq.55 is likely to be identical, 55 > v.gtoreq.45 cannot draw a conclusion, 45 > v.gtoreq.30 is likely to be non-identical, 30 > v.gtoreq.15 is very likely to be non-identical, 15 > v.gtoreq.5 is very likely to be non-identical, and 5.
The invention has the beneficial effects.
The invention uses modern system scientific theory and related mathematical methods to establish a multivariate regression mathematical model which is related by the occurrence probability and the conformity of handwriting characteristics, uses an analytic hierarchy process to calculate and determine the weighted value of various characteristics at each level, combines basic statistical data to determine the interval value of the occurrence probability and the conformity of the handwriting characteristics, then uses the internal related mathematical formula of the handwriting characteristics to calculate the quantitative related value of a specific case, and finally evaluates the quantitative related value according to the stability of the handwriting characteristics to provide the corresponding quantitative inspection and identification method for identifying opinions.
The method converts the qualitative judgment of expert experience into a handwriting characteristic weight value, and carries out inspection and identification on an interval value of probability based on objective statistical data; the handwriting detection and identification quantitative method has the advantages of operation process modeling of detection and identification, mathematization of the detection and identification process and objectification of detection and identification opinions.
Drawings
The invention is further described with reference to the following figures and detailed description. The scope of the invention is not limited to the following expressions.
Fig. 1 is a schematic block diagram of the present invention.
FIG. 2 is a diagram of a handwriting macro-level feature comparison table according to the present invention.
FIG. 3 is a table showing the comparison of the features of the handwriting in the invention.
FIG. 4 is a schematic diagram of a handwriting micro-level feature comparison table according to the present invention.
FIG. 5 is a three-dimensional model diagram of handwriting verification opinion and feature occurrence and conformity correlation values according to the present invention.
Detailed Description
As shown in the figure, the present invention comprises the following steps:
1) constructing a handwriting characteristic hierarchical structure and a value evaluation analysis model;
2) determining the weight value of each handwriting characteristic;
3) establishing a corresponding relation between the handwriting characteristic occurrence rate and the coincidence value;
4) making a hierarchical feature comparison table, and calculating the correlation value of the handwriting feature occurrence rate and the coincidence value;
5) and establishing the corresponding relation between the correlation value and the inspection appraisal opinion.
The step 1) comprises the following steps: setting a target layer-handwriting characteristic hierarchical structure and value; determining a middle layer, namely macroscopic level features, mesoscopic level features and microscopic level features; and a refinement scheme layer, wherein the macro level features comprise features, the meso level features comprise features, and the micro level features comprise features.
The macro-level features are overall features, the middle-level features are appearance, content, space and time features, and the micro-level features are two-dimensional and three-dimensional features.
The macro-level features in the solution layer include: h is1-features are overall layout, line direction and ruled line relationships; h is2-overall style, writing level; h is3-overall composition and inter-word relationships;
the mesoscopic hierarchical features in the solution layer include: z is a radical of1-shape and size; z is a radical of2-structural content; z is a radical of3Writing time sequence of strokes and components; z is a radical of4-stroke and radical collocation proportion space;
the micro-level features in the recipe layer include: w is a1-dot, horizontal, vertical, left-falling, right-falling, bending and hooking; w is a2The corresponding connection relation, trend and connection form among strokes; w is a3-writing speed and distribution rule; w is a4The weight and distribution rule of writing pressure.
The handwriting characteristic hierarchical structure and value evaluation analysis model is shown in the following table 1:
table 1: handwriting characteristic hierarchical structure and value evaluation analysis model
The step 2) is determined by a 1-9-level scaling method and a root method.
On the basis of constructing a handwriting characteristic hierarchical structure and a value evaluation analysis model, four handwriting characteristic value judgment matrixes are respectively constructed, and the calculation of handwriting characteristic weight values is completed by utilizing a 1-9-level scaling method and a root method, wherein the specific results are shown in a table 2.
Table 2: weighted value of handwriting characteristics
The handwriting characteristic occurrence rate in the step 3) is divided into: the characteristic occurrence rate of writing of few people is 0.1%; the occurrence rate of writing features of few people is 10%; the occurrence rate of the characters written by a few people is 30%; the occurrence rate of the characters which respectively account for half of people for writing is 50 percent; most people have 70% incidence of writing features;
the coincidence degree value of the handwriting characteristics is divided into: completely meets 100 percent, extremely meets 97.5 percent, very meets 92.5 percent and better meets 87.5 percent; the total difference is 0%, the difference is extremely large 2.5%, the difference is greatly large 7.5%, and the difference is large 12.5%.
And 4) making a hierarchical feature comparison table in the step 4) comprises making a macro, meso and micro hierarchical feature comparison table.
The step 4) of calculating the correlation value of the handwriting characteristic occurrence rate and the coincidence value adopts the following formula:
virepresents: a correlation value; ciRepresents: a conformity value; piRepresents: the occurrence rate of the characteristics, K1 and K2 respectively represent: two constants (K1, K2 are both 50, making the value of vi fall in the interval 0 to 100); n represents: the number of features.
The invention calculates the correlation value v of the handwriting characteristicsiThen, respectively calculating handwriting macroscopical VHMiddle view VZMicro VWAnd (3) calculating the total quantitative association value V of handwriting characteristics at each level of characteristic association value:
macro level feature correlation value: vH＝(vh1×h1+vh2×h2+vh3×h3+vh4×h4)×H
Mesoscopic hierarchical feature correlation value: vz＝(vz1×z1+vz2×z2+vz3×z3+vz4×z4)×Z
Microscopic level feature correlation value: vw＝(vw1×w1+vw2×w2+vw3×w3+vw4×w4)×W
The formula for V is: v is VH+Vz+VW
The corresponding relation between the handwriting characteristic quantitative total correlation value V and the inspection and identification opinions is as follows: 100 > v.gtoreq.95 is definitely identical, 95 > v.gtoreq.85 is very likely to be identical, 85 > v.gtoreq.70 is very likely to be identical, 70 > v.gtoreq.55 is likely to be identical, 55 > v.gtoreq.45 cannot draw a conclusion, 45 > v.gtoreq.30 is likely to be non-identical, 30 > v.gtoreq.15 is very likely to be non-identical, 15 > v.gtoreq.5 is very likely to be non-identical, and 5.
Table 3: corresponding relation between total quantitative association value of handwriting characteristics and inspection and identification opinions
It should be understood that the detailed description of the present invention is only for illustrating the present invention and is not limited by the technical solutions described in the embodiments of the present invention, and those skilled in the art should understand that the present invention can be modified or substituted equally to achieve the same technical effects; as long as the use requirements are met, the method is within the protection scope of the invention.
Claims (9)
1. The handwriting quantitative inspection and identification method is characterized by comprising the following steps of:
1) constructing a handwriting characteristic hierarchical structure and a value evaluation analysis model;
2) determining the weight value of each handwriting characteristic;
3) establishing a corresponding relation between the handwriting characteristic occurrence rate and the coincidence value;
4) making a hierarchical feature comparison table, and calculating the correlation value of the handwriting feature occurrence rate and the coincidence value;
5) establishing a corresponding relation between the correlation value and the inspection and identification opinions;
and 4) calculating the correlation value of the handwriting characteristic occurrence rate and the coincidence value by adopting the following formula:
virepresents: a correlation value; ciRepresents: a conformity value; piRepresents: the occurrence rate of the characteristics, K1 and K2 respectively represent: two constants; n represents: the number of features.
2. The handwriting quantitative verification authentication method according to claim 1, wherein said step 1) comprises: setting a target layer-handwriting characteristic hierarchical structure and value; determining a middle layer, namely macroscopic level features, mesoscopic level features and microscopic level features; and a refinement scheme layer, wherein the macro level features comprise features, the meso level features comprise features, and the micro level features comprise features.
3. The handwriting quantitative inspection authentication method according to claim 2, wherein the macro-level features are general features, the middle-level features are appearance, content, space and time features, and the micro-level features are two-dimensional and three-dimensional features.
4. The handwriting quantitative verification authentication method according to claim 2, wherein the macro-level features in the scheme layer comprise: h is1-features are overall layout, line direction and ruled line relationships; h is2-overall style, writing level; h is3-overall composition and inter-word relationships;
the mesoscopic hierarchical features in the solution layer include: z is a radical of1-shape and size; z is a radical of2-structural content; z is a radical of3-stroke and radical writing chronological order; z is a radical of4-stroke and radical collocation proportion space;
the micro-level features in the recipe layer include: w is a1-dot, horizontal, vertical, left-falling, right-falling, bending, hooking; w is a2-the corresponding connection relations and trends between strokes, connection modalities; w is a3-how fast and how regularly the writing speed is distributed; w is a4The weight and distribution of the writing pressure.
5. The handwriting quantitative verification authentication method according to claim 1, wherein said step 2) is determined by using a 1 to 9-level scaling method and a root method.
6. The handwriting quantitative inspection appraisal method according to claim 1, characterized in that the step 3) handwriting characteristic occurrence rate is divided into: the characteristic occurrence rate of writing of few people is 0.1%; the occurrence rate of writing features of few people is 10%; the occurrence rate of the characters written by a few people is 30%; the occurrence rate of the characters which respectively account for half of people for writing is 50 percent; most people have 70% incidence of writing features;
the coincidence degree value of the handwriting characteristics is divided into: completely meets 100 percent, extremely meets 97.5 percent, very meets 92.5 percent and better meets 87.5 percent; the total difference is 0%, the difference is extremely large 2.5%, the difference is greatly large 7.5%, and the difference is large 12.5%.
7. The handwriting quantitative verification authentication method according to claim 1, wherein said step 4) of making a hierarchical feature comparison table comprises making a macro, meso and micro hierarchical feature comparison table.
8. A handwriting quantitative verification authentication method according to claim 1 and characterised by calculating the associated value v of handwriting featuresiThen, respectively calculating handwriting macroscopical VHMiddle view VZMicro VWAnd (3) calculating the total quantitative association value V of handwriting characteristics at each level of characteristic association value:
macro level feature correlation value: vH＝(vh1×h1+vh2×h2+vh3×h3+vh4×h4)×H
Mesoscopic hierarchical feature correlation value: vz＝(vz1×z1+vz2×z2+vz3×z3+vz4×z4)×Z
Microscopic level feature correlation value: vw＝(vw1×w1+vw2×w2+vw3×w3+vw4×w4)×W
The formula for V is: v is VH+VZ+VW。
9. The method for quantitative verification and appraisal of handwriting according to claim 8, wherein the correspondence between the total quantitative association value V of handwriting characteristics and the appraisal opinions of verification is as follows: 100 > v.gtoreq.95 is definitely identical, 95 > v.gtoreq.85 is very likely to be identical, 85 > v.gtoreq.70 is very likely to be identical, 70 > v.gtoreq.55 is likely to be identical, 55 > v.gtoreq.45 cannot draw a conclusion, 45 > v.gtoreq.30 is likely to be non-identical, 30 > v.gtoreq.15 is very likely to be non-identical, 15 > v.gtoreq.5 is very likely to be non-identical, and 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610191639.1A CN107292213B (en) | 2016-03-30 | 2016-03-30 | Handwriting quantitative inspection and identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610191639.1A CN107292213B (en) | 2016-03-30 | 2016-03-30 | Handwriting quantitative inspection and identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107292213A CN107292213A (en) | 2017-10-24 |
| CN107292213B true CN107292213B (en) | 2020-04-14 |
Family
ID=60087825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610191639.1A Expired - Fee Related CN107292213B (en) | 2016-03-30 | 2016-03-30 | Handwriting quantitative inspection and identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107292213B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112990000A (en) * | 2021-03-11 | 2021-06-18 | 中国刑事警察学院 | Electronic signature handwriting characteristic typing and characteristic chain comparison method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104809451A (en) * | 2015-05-15 | 2015-07-29 | 河海大学常州校区 | Handwriting authentication system based on stroke curvature detection |
| CN104951502A (en) * | 2014-03-10 | 2015-09-30 | Ims保健公司 | Handwriting recognition tool |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AUPR824301A0 (en) * | 2001-10-15 | 2001-11-08 | Silverbrook Research Pty. Ltd. | Methods and systems (npw001) |
| CN100347723C (en) * | 2005-07-15 | 2007-11-07 | 清华大学 | Off-line hand writing Chinese character segmentation method with compromised geomotric cast and sematic discrimination cost |
| CN102324048B (en) * | 2011-09-07 | 2013-07-31 | 中国刑事警察学院 | Pen mark feature computer quantification and comparison assisted analysis system |
| CN102521858B (en) * | 2011-11-28 | 2014-04-02 | 北京盛世宣合信息科技有限公司 | Handwriting generating method for electronic writing brush |
| CN104281831B (en) * | 2013-07-02 | 2017-10-13 | 阿里巴巴集团控股有限公司 | A kind of method and apparatus of person's handwriting checking |
-
2016
- 2016-03-30 CN CN201610191639.1A patent/CN107292213B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104951502A (en) * | 2014-03-10 | 2015-09-30 | Ims保健公司 | Handwriting recognition tool |
| CN104809451A (en) * | 2015-05-15 | 2015-07-29 | 河海大学常州校区 | Handwriting authentication system based on stroke curvature detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107292213A (en) | 2017-10-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105512799B (en) | Power system transient stability evaluation method based on mass online historical data | |
| CN110458187B (en) | Malicious code family clustering method and system | |
| CN103745200B (en) | A kind of facial image recognition method based on word bag model | |
| US11003738B2 (en) | Dynamically non-gaussian anomaly identification method for structural monitoring data | |
| CN111046341A (en) | Unconventional natural gas fracturing effect evaluation and capacity prediction method based on principal component analysis | |
| CN103226595B (en) | The clustering method of the high dimensional data of common factor analyzer is mixed based on Bayes | |
| CN110045227B (en) | power distribution network fault diagnosis method based on random matrix and deep learning | |
| CN104954185A (en) | Cloud computing load prediction method based on deep belief network | |
| CN107038432B (en) | Fingerprint image direction field extraction method based on frequency information | |
| CN111709454A (en) | Multi-wind-field output clustering evaluation method based on optimal copula model | |
| TWI477987B (en) | Methods for sentimental analysis of news text | |
| CN112417893A (en) | Software function demand classification method and system based on semantic hierarchical clustering | |
| CN107292213B (en) | Handwriting quantitative inspection and identification method | |
| CN103970651A (en) | Software architecture safety assessment method based on module safety attributes | |
| TW201508696A (en) | Analytic system of wafer bin map and analytic method thereof | |
| CN108255807A (en) | A kind of method for identifying R＆D direction | |
| Yu et al. | Mental workload classification via online writing features | |
| CN110008972A (en) | Method and apparatus for data enhancing | |
| CN106709598B (en) | Voltage stability prediction and judgment method based on single-class samples | |
| CN110516750B (en) | Steel plate shape quality abnormity detection method based on CART decision tree | |
| Ngabu et al. | CLUSTER FAST DOUBLE BOOTSTRAP APPROACH WITH RANDOM EFFECT SPATIAL MODELING | |
| CN113239730A (en) | Method for automatically eliminating structural false modal parameters based on computer vision | |
| CN112307673A (en) | Grid surface quality detection method based on deep learning | |
| CN105809100A (en) | Handwriting and signature quantization inspection and identification characteristic product method | |
| CN107563029B (en) | Support vector machine approximate model optimization method based on K-fold cross verification method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200414 Termination date: 20210330 |