CN104537654B - Printed image tampering forensic methods based on half-tone dot location distortion - Google Patents
Printed image tampering forensic methods based on half-tone dot location distortion Download PDFInfo
- Publication number
- CN104537654B CN104537654B CN201410805834.XA CN201410805834A CN104537654B CN 104537654 B CN104537654 B CN 104537654B CN 201410805834 A CN201410805834 A CN 201410805834A CN 104537654 B CN104537654 B CN 104537654B
- Authority
- CN
- China
- Prior art keywords
- half tone
- tone dot
- image
- cluster
- dot
- 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
Landscapes
- Image Processing (AREA)
- Editing Of Facsimile Originals (AREA)
- Image Analysis (AREA)
Abstract
The invention belongs to the field of information security and signal and information processing technologies and relates to a printed image tampering forensic method based on half-tone dot location extraction in a laser printing image and according to location distortion features. The method comprises the steps that the center locations of half-tone dots in a printed image are obtained by means of a Gaussian model matching method, 12-dimensional location features are extracted from all half-tone dot center points, clustering is performed on the extracted features by using a k-means clustering method, and clustering results of all the half-tone dots are marked in the printed image; a tampering reality of the printed image can be identified according to the clustering results, and tampered areas can be determined. As the half-tone dot locations are detected, the detection results are not affected by image content, and the fine tampered image areas can be detected. When the method is used for tampering detection, concision and visualization are achieved. The printed image tampering forensic method is suitable for the field of information security.
Description
Technical field
The invention belongs to information security field, Signal and Information Processing technology, are related to halftoning in Laser Printed Image
Evidence collecting method is distorted in the extraction of point position and the printing carried out according to position distortions feature.
Background technology
Current digital-code printer has become the office equipment of indispensability in people's daily life.And with printer, photocopier
Popularization, forge the case of document, image also in the trend that increases year by year.Offender can be set using cheap printing, duplicating
Standby and image processing software distorts all kinds of vital documents.These criminal offences can not only cause economic dispute and criminal case
Part, and also affect the affairs and nation's security of government.It is in order to take precautions against and hit such criminal activity, judicial and public
Peace department differentiates the verity of mimeograph documents in the urgent need to effectively, easily and efficiently printing forensic technologies.As laser is beaten
Print machine has the advantages that speed fast, high definition, holding time are long thus obtains widely using for company and government department.
Textural characteristics, geometric distortion feature, noise characteristic and sensor class are utilized mainly currently for print image evidence obtaining
Type identification etc. is explored.Seung-Jin Ryu in 2010 etc. carry out print image source using halftoning textural characteristics
Evidence obtaining.The half tone dot arrangement of print image is considered as into a kind of texture, texture is analyzed by analyzing high resolution scanning image
Feature.Different according to the arrangement of the half tone dot of different printers, the arrangement of identical printer half tone dot is identical.Using Hough
The straight line information of lattice arrangement is extracted in conversion, builds rectangular histogram according to the angle character of the straight line for obtaining, by calculating several
The histogrammic meansigma methodss of image angle are representing model printer, then the source for being detected to complete printer using similarity is taken
Card.
Orhan Bulan propose a kind of print image source evidence forensics skill based on half tone dot geometric distortion within 2009
Art.As the deviation of the position in laser printer between photosensitive drums and pentaprism and rotating speed causes half tone dot in print image
Actual arrangement and ideal alignment structure existence position distortion.Extracting half tone dot center carries out frequency domain transform and extracts a little
In the battle array cycle, estimate original point position, and carry out rotation compensation to correct what is introduced in scanning process using fitting a straight line mode
Rotation offset, extracts geometric distortion feature, and the printer source of scanogram is obtained by correlation detection.Above two
Method is all the global feature of texture feature extraction or geometric distortion feature as the page, enters line printer source evidence forensics.
The content of the invention
The technical problem to be solved in the present invention is that invention is a kind of to be lost using half tone dot position for the deficiencies in the prior art
True print image distorts evidence collecting method, is the microscopic feature of the position using half tone dot, analyzes each local feature of the page,
To find local difference, that enters line printer distorts evidence obtaining.Half tone dot feature of the present invention using printer, extracts printed drawings
Half tone dot positional information as in, and judge which whether there is position distortions, so realize print image distort evidence obtaining.This
The testing result of method is not affected by picture material, as the position using half tone dot is detected, the result of this experiment
Each half tone dot is accurate to, trickle tampered image region can be detected, either distorted with figure or usurping of distorting of different figure
Change region can be accurately positioned out.
The technical solution used in the present invention is to distort the side of evidence obtaining using half tone dot position distortions to carry out print image
Method, is characterized in that, the method obtains the center of half tone dot in print image using Gauss model matching process, to each
Half tone dot central point extract 12 tie up position features, carried feature is clustered using k- means clustering methods, and by each
The cluster result of half tone dot is marked in print image；The fact of distorting of print image can be differentiated by cluster result, and really
Determine tampered region；The method is comprised the following steps that：
1) Spot detection of half tone dot
The center of half tone dot in print image is obtained using Gauss model matching process.First, according to halftoning
The shape and its pixel value Changing Pattern of point, builds an ideal Gaussian Matching Model matrix for meeting dimensional gaussian distribution, will
This Gauss Matching Model carries out traversal matching on testing image, that is, take with Gauss model size identical image block and therewith
Correlation coefficient is calculated, its formula is as follows：
Wherein, AmnFor Gauss model matrix, BmnIt is the image block matched with Gauss model, its size is m rows n row, D
(Amn) and D (Bmn) variance of Gauss model matrix and matching image block is represented respectively.Molecule represents the association side of the two
Difference, denominator represent the product of the two standard deviation, then whole formula represents the correlation coefficient between two matrixes.
Gauss model is carried out into traversal calculating on testing image, the correlation matrix of testing image is obtained, in matrix
In both for best match ask the position of local maximum position, and the center of half tone dot, by half tone dot
Heart position is marked with white point, then obtained half tone dot position mark figure as shown in Figure 4.
2) half tone dot position feature is extracted
Its 12 dimension position feature is calculated according to the center of the half tone dot for obtaining, for certain halftoning central point A, meter
6 half tone dots nearest with the half tone dot position in half tone dot around which are calculated, then this 6 half tone dots for obtaining are
Its adjacent half tone dot, after obtaining adjacent half tone dot position, calculates point A and is adjacent distance a little as 6 dimensional features, then
Point A is calculated respectively is adjacent line a little and horizontal line angle as other 6 dimensional feature；
3) tampered region is determined using clustering algorithm
The present invention adopts k- means clustering algorithms, due to needing to distinguish original in the present invention and distorting two class half tone dots,
Therefore, classification k=2；All samples in data set are divided into 2 clusters and need to meet following condition：Among same cluster
Between object, distance is minimum；Between the object in different clusters, distance is maximum；The process of k- mean algorithms is as follows：First, from there is n numbers
According to arbitrarily 2 objects of selection in the data set of object as initial cluster centre；Remaining other objects, according to these point with
They are respectively allocated in the cluster for giving cluster centre place by the distance of cluster centre；Then, calculate that each obtains is new
The average of all objects in cluster centre, i.e. this cluster；Finally, constantly steps be repeated alternatively until that canonical measure function is opened
Begin till convergence, the present invention is using mean square deviation as canonical measure function.
The invention has the advantages that be directed to print image distorts evidence obtaining, using half tone dot in print image
The method of arrangement position distortion carries out print image and distorts evidence obtaining, and determines tampered region.We extract print image first
Half tone dot center in scanning figure, then calculates the positional information feature of each point half tone dot adjacent with surrounding, using k-
Half tone dot is divided into original half tone dot and distorts two class of half tone dot by means clustering algorithm, and using different gray values half-and-half
The cluster result of tone point is marked, and testing staff is by the other tampered image of the labelling illustrated handbook obtained by clustering algorithm.This
It is bright suitable for information security field, by experimental result picture as can be seen that the testing result of the present invention is not affected by picture material,
And as the position using half tone dot is detected, the result of this experiment is accurate to each half tone dot, can detect
Trickle tampered image region, either distorts with figure or the tampered region that different figure is distorted can be by accurate and effective positioning
Go out, and regardless of whether with Professional knowledge, tampering detection is carried out using this method, and succinctly, intuitively judged
Tampered region.
Description of the drawings
Fig. 1 is that the print image detected based on half tone dot position distortions distorts forensics process block diagram.
Fig. 2 is print image and its regional enlarged drawing.Figure left-half is the half tone dot arrangement after print image amplification,
Each stain represents a half tone dot；Figure right half part represents the pixel value composition after single half tone dot is amplified.
Fig. 3 is the black white image form of expression of dimensional Gaussian model；Fig. 4 is print image half tone dot and its center
Schematic diagram, wherein white point represent its center；Fig. 5 is to distort front design sketch；Fig. 6 is to distort rear design sketch；Fig. 7 is utilization
Distance, the tampering detection result of angle character.
Specific embodiment
The specific embodiment of the present invention is described in detail below in conjunction with technical scheme and accompanying drawing.
The present invention distorts evidence obtaining for the image that generated by laser printer, by printer print image mechanism
Research, finds printer by the half tone dot of small marshalling to represent the gray value of electronic image to be printed, gray scale
Change by the size of half tone dot or sparse describing.During the splicing of image is distorted, tampered region can usually pass through
Among then rotation or scaling are spliced to background image again.Therefore, the half tone dot arrangement existence position in distorted image region is lost
True phenomenon, can carry out distorting evidence obtaining by the detection of half tone dot position distortions.As print image is advised by a series of arrangements
The halftoning dot matrix of rule is constituted, and whether the algorithm extracts the positional information of half tone dot, and right according to the arrangement architecture of dot matrix
Together judging the tampered region of print image, the concrete block diagram of algorithm is as shown in Figure 1.
In embodiment, using the printer and model EPSON of model HP laser Jet 520Lx
The scanner of perfection 1200PHOTO, rate is 600dpi respectively for scanning.First, by the image scanning of laser printing into number
Word image and amplification obtain a scanogram to be measured, as shown in Figure 2.To be detected in testing image in each half tone dot
Heart position；By positional information calculation position feature, position feature includes half tone dot center with adjacent half tone dot center
The distance between, the angle of inclination of half tone dot center and the adjacent half tone dot line of centres.It is adjacent with each half tone dot
Half tone dot totally 6, arrangement position is respectively in the apex of approximate hexagon, therefore the feature extracted includes 6 dimension distance feature
With 6 dimension angle characters, totally 12 tie up.Then, pass through 12 extracted dimension position features, find the half tone dot of malposition, profit
Half tone dot is divided into apart from the normal class of normal class, exception class and angle, exception class with clustering method, and is entered with different gray values
Line flag；Finally, specifically usurping by the gray scale marker for judgment mimeograph documents scanogram to the clustering algorithm in print image
Change region.That what is implemented comprises the following steps that：
1) Spot detection of half tone dot
The present invention adopts a kind of half tone dot center extraction method matched based on Gauss model, due to half tone dot shape
Circle is similar to, and pixel value outwards gradually increases from center, the center of half tone dot is calculated using dimensional Gaussian model.Will
Gauss Matching Model carries out traversal matching on testing image, and the center of each half tone dot is obtained using correlation coefficient size
Position.As shown in Fig. 2 left-half is represented prints the half tone dot for obtaining, right half part is represented and amplifies this half tone dot
Shape afterwards and pixel composition.Our shapes and its pixel value Changing Pattern according to half tone dot, builds a symbol first
Close dimensional gaussian distribution Gauss Matching Model matrix, by this Gauss Matching Model on testing image successively carry out traversal
Match somebody with somebody, shown in Fig. 3, represent the preferable half tone dot model for meeting dimensional gaussian distribution for obtaining.Gauss model is often moved and is moved a step,
One piece is taken with Gauss model size identical image block and correlation coefficient is calculated therewith.Its formula is as follows：
Wherein, AmnFor Gauss model matrix, BmnIt is the image block matched with Gauss model, its size is m row n
Row, D (Amn) and D (Bmn) variance of Gauss model matrix and matching image block is represented respectively.Molecule represents the association of the two
Variance, denominator represent the product of the two standard deviation, then whole formula represents the correlation coefficient between two matrixes.
Gauss model is carried out into traversal calculating on testing image, the correlation matrix of testing image is obtained, is worked as Gauss
When model is moved to the position with a half tone dot aligned in position, it is known that, its correlation coefficient has a local maximum, therefore
The position of Matrix local maximum position both for best match, that is, the center of half tone dot.Experimental result
Figure is as shown in figure 4, white point represents the half tone dot center extracted in Fig. 4.
2) half tone dot position feature is extracted
Its 12 dimension position feature is calculated according to the center of the half tone dot for obtaining.For certain halftoning central point A, meter
6 half tone dots nearest with the half tone dot position in half tone dot around which are calculated, then this 6 half tone dots for obtaining are
Its adjacent half tone dot.After obtaining the position of adjacent half tone dot, the distance between point A 6 points most adjacent with which are calculated
As 6 dimensional features, then point A and the angle between the line and horizontal line of its most consecutive points are calculated respectively as other 6 dimensional feature.
Table 1 is that original half tone dot changes citing with the character numerical value for distorting half tone dot.
1 original half tone dot of table changes citing with the character numerical value for distorting half tone dot
3) tampered region is determined using clustering algorithm
The present invention adopts k- means clustering algorithms, due to needing to distinguish original in the present invention and distorting two class half tone dots,
Therefore, classification k=2.All samples in data set are divided into 2 clusters and need to meet following condition：Among same cluster
Between object, distance is minimum；Between the object in different clusters, distance is maximum.The process of k- mean algorithms is as follows：First, from there is n numbers
According to arbitrarily 2 objects of selection in the data set of object as initial cluster centre.Remaining other objects, according to these point with
They are respectively allocated in the cluster for giving cluster centre place by the distance of cluster centre.Then, calculate that each obtains is new
The average of all objects in the cluster centre of cluster, i.e. this cluster.Finally, constantly steps be repeated alternatively until canonical measure
Till function starts convergence, the present invention is using mean square deviation as canonical measure function.
Eigenvalue is divided into two classes using k- mean clusters by the present invention, and a class is tampered region, another kind of for original area.
The weights of two classes for adopting distance cluster to obtain are respectively labeled as into 0 and 80；The two class weights marks that will be obtained using slope cluster
It is designated as 0 and 50；The weights of each point are summed up, and the gray value weights for obtaining corresponded on gray level image is marked table
Show.Experiment effect is as shown in fig. 7, according to distance and angle of inclination feature, whole data are polymerized to two classes respectively, different with four kinds
Gray value be marked, it is different according to gray value, it may be clearly seen that Detection results.
In embodiment, choose a width and license plate image is printed for distorting evidence obtaining experiment.Fig. 5 is real car plate printed drawings
Picture, Fig. 6 are different figure tampered image, i.e., by rotation, scale the letter " G " that digital " 9 " in image are replaced with other width figure.
Concrete operation method：First, take this algorithm to read the altimetric image to be checked of Fig. 6, mapping is treated using Gauss model
Ask related as carrying out traversal, calculate the center of all half tone dots in image, carry out the extraction of half tone dot position；Secondly,
For each half tone dot center, extract 12 and tie up position feature, including 6 dimension distance feature and 6 dimension angle characters；Finally,
The cluster that classification number is 2 is carried out to put forward two groups of distances and angle character respectively using k- means clustering algorithms, and cluster is tied
Fruit is marked in print image using different gray values, and concrete outcome is as shown in fig. 7, the region that gray value white is concentrated is to usurp
Change region.
The results show, the testing result of the present invention are usurped not by picture material, tampered region area, with figure or different figure
The impact of factor such as change, as long as tampered region can accurately and effectively detect tampered region through scaling or rotation.And
Detection method is simple, and experiment effect is simply easy to identify, for the personnel without specialty evidence obtaining knowledge, it is also possible to using this technology
Carry out print image tampering detection.
Claims (1)
1. a kind of utilization half tone dot position distortions is characterized in that carrying out the method that print image distorts evidence obtaining, the method profit
The center of half tone dot in print image is obtained with Gauss model matching process, 12 are extracted to each half tone dot central point
Dimension position feature, is clustered to carried feature using k- means clustering methods, and by the cluster result mark of each half tone dot
Note is in print image；The fact of distorting of print image can be differentiated by cluster result, and determine tampered region；The method
Comprise the following steps that：
1) Spot detection of half tone dot
The center of half tone dot in print image is obtained using Gauss model matching process；First, according to half tone dot
Shape and its pixel value Changing Pattern, build an ideal Gaussian Matching Model matrix for meeting dimensional gaussian distribution, by this
Gauss Matching Model carries out traversal matching on testing image, that is, take with Gauss model size identical image block and calculate therewith
Correlation coefficient, its formula are as follows：
Wherein, AmnFor Gauss model matrix, BmnIt is the image block matched with Gauss model, its size is m rows n row, D
(Amn) and D (Bmn) variance of Gauss model matrix and matching image block is represented respectively；Molecule represents the association side of the two
Difference, denominator represent the product of the two standard deviation, then whole formula represents the correlation coefficient between two matrixes；
Gauss model is carried out into traversal calculating on testing image, the correlation matrix of testing image is obtained, is asked in a matrix
The position of local maximum is the position of best match, and the center of half tone dot, by the centre bit of half tone dot
Put and be marked with white point, then obtain half tone dot position mark figure；
2) half tone dot position feature is extracted
Its 12 dimension position feature is calculated according to the center of the half tone dot for obtaining, for certain halftoning central point A, which is calculated
6 half tone dots nearest with the half tone dot position in surrounding half tone dot, then this 6 half tone dots for obtaining are its phase
Adjacent half tone dot, after obtaining adjacent half tone dot position, calculates point A and is adjacent distance a little as 6 dimensional features, then distinguish
Calculating point A is adjacent line a little and horizontal line angle as other 6 dimensional feature；
3) tampered region is determined using clustering algorithm
The present invention adopts k- means clustering algorithms, due to needing to distinguish original in the present invention and distorting two class half tone dots, therefore,
Classification k=2；All samples in data set are divided into 2 clusters and need to meet following condition：Between the object among same cluster
Distance is minimum；Between the object in different clusters, distance is maximum；The process of k- mean algorithms is as follows：First, from there is n data object
Data set in arbitrarily select 2 objects as initial cluster centre；Remaining other objects, according in these points and cluster
They are respectively allocated in the cluster for giving cluster centre place by the distance of the heart；Then, calculate in the new cluster that each obtains
The average of all objects in the heart, i.e. this cluster；
Finally, till constantly steps be repeated alternatively until that canonical measure function starts convergence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410805834.XA CN104537654B (en) | 2014-12-19 | 2014-12-19 | Printed image tampering forensic methods based on half-tone dot location distortion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410805834.XA CN104537654B (en) | 2014-12-19 | 2014-12-19 | Printed image tampering forensic methods based on half-tone dot location distortion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104537654A CN104537654A (en) | 2015-04-22 |
| CN104537654B true CN104537654B (en) | 2017-04-12 |
Family
ID=52853173
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410805834.XA Expired - Fee Related CN104537654B (en) | 2014-12-19 | 2014-12-19 | Printed image tampering forensic methods based on half-tone dot location distortion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104537654B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107392949B (en) * | 2017-07-17 | 2019-11-05 | 湖南优象科技有限公司 | Image zone duplicating and altering detecting method and detection device based on local invariant feature |
| CN110533632B (en) * | 2019-07-18 | 2022-05-10 | 数字广东网络建设有限公司 | Image blurring tampering detection method and device, computer equipment and storage medium |
| CN110895811B (en) * | 2019-11-05 | 2023-05-09 | 泰康保险集团股份有限公司 | Image tampering detection method and device |
| CN112116565B (en) * | 2020-09-03 | 2023-12-05 | 深圳大学 | Method, apparatus and storage medium for generating countersamples for falsifying a flip image |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101415062A (en) * | 2007-10-16 | 2009-04-22 | 佳能株式会社 | Information processing apparatus, image processing apparatus and method thereof |
| CN101854461A (en) * | 2010-04-20 | 2010-10-06 | 大连理工大学 | Printed document evidence obtaining method for detecting authenticity of document by using half-tone information |
| CN102739914A (en) * | 2011-03-29 | 2012-10-17 | 富士施乐株式会社 | Image processing apparatus, image processing method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040090654A1 (en) * | 2002-09-17 | 2004-05-13 | Luc Minnebo | FM screening with sub-dot phase modulation |
| JP5381166B2 (en) * | 2009-03-04 | 2014-01-08 | オムロン株式会社 | Model image acquisition support apparatus, model image acquisition support method, and model image acquisition support program |
-
2014
- 2014-12-19 CN CN201410805834.XA patent/CN104537654B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101415062A (en) * | 2007-10-16 | 2009-04-22 | 佳能株式会社 | Information processing apparatus, image processing apparatus and method thereof |
| CN101854461A (en) * | 2010-04-20 | 2010-10-06 | 大连理工大学 | Printed document evidence obtaining method for detecting authenticity of document by using half-tone information |
| CN102739914A (en) * | 2011-03-29 | 2012-10-17 | 富士施乐株式会社 | Image processing apparatus, image processing method |
Non-Patent Citations (2)
| Title |
|---|
| Detecting Halftone Dots for Offset Print Quality Assessment Using Soft Computing;Jen Lundstrom et al;《2011 IEEE International conference on Fuzzy Systems (FUZZ)》;20100923;第978-984页 * |
| Geometric distortion signatures for printer identification;Orhan Bulan et al;《Acoustics speech and signal processing 2009》;20090430;第1401-1404页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104537654A (en) | 2015-04-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110598699B (en) | Anti-counterfeiting bill authenticity distinguishing system and method based on multispectral image | |
| CN101894260B (en) | Method for identifying forgery seal based on feature line randomly generated by matching feature points | |
| CN104537654B (en) | Printed image tampering forensic methods based on half-tone dot location distortion | |
| US7587066B2 (en) | Method for detecting fraud in a value document such as a check | |
| CN101894268B (en) | Seal authenticity identification method capable of eliminating receipt interference | |
| Forczmański et al. | Robust stamps detection and classification by means of general shape analysis | |
| US8903155B2 (en) | Optical waveform generation and use based on print characteristics for MICR data of paper documents | |
| CN101587540B (en) | Printer verification method for detecting document source by means of geometric distortion of page document | |
| CN112016563B (en) | Method for identifying authenticity of circular seal | |
| CN101118592B (en) | Printers evidence obtaining method based on character printing feature | |
| CN103748866B (en) | Forensic authentication system and method | |
| Ryu et al. | Document forgery detection with SVM classifier and image quality measures | |
| Wu et al. | A printer forensics method using halftone dot arrangement model | |
| CN102646194B (en) | Method for performing printer type evidence obtainment by utilizing character edge features | |
| Benhamza et al. | Image forgery detection review | |
| Khan et al. | Urdu optical character recognition technique using point feature matching; a generic approach | |
| Chhabra et al. | Detecting fraudulent bank checks | |
| CN101854461B (en) | Printed document evidence obtaining method for detecting authenticity of document by using half-tone information | |
| Pfeifer et al. | Detection of commercial offset printing in counterfeited banknotes | |
| Dhapare et al. | Detection of Counterfeit Currency using Image Processing Techniques | |
| KR101030548B1 (en) | Color laser printer identification system and method by analyzing halftone texture of cyan color image | |
| Danti et al. | Grid based feature extraction for the recognition of Indian currency notes | |
| ALAMERİ et al. | Computational Methods for Forgery Detection in Printed Official Documents | |
| CN106408021B (en) | Handwriting and print form identification method based on stroke thickness | |
| CN113177510B (en) | Inkjet printing file tracing method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| C10 | Entry into 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 |
Granted publication date: 20170412 Termination date: 20191219 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |