CN105069766A - Inscription restoration method based on contour feature description of Chinese character image - Google Patents
Inscription restoration method based on contour feature description of Chinese character image Download PDFInfo
- Publication number
- CN105069766A CN105069766A CN201510441241.4A CN201510441241A CN105069766A CN 105069766 A CN105069766 A CN 105069766A CN 201510441241 A CN201510441241 A CN 201510441241A CN 105069766 A CN105069766 A CN 105069766A
- Authority
- CN
- China
- Prior art keywords
- stroke
- point
- chinese character
- inscription
- contour
- 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.)
- Granted
Links
Landscapes
- Controls And Circuits For Display Device (AREA)
- Image Analysis (AREA)
Abstract
The invention provides an inscription restoration method based on the contour feature description of a Chinese character image. The method is applied in the fields of computer virtual reality technology, machine vision, image processing and inscription restoration, wherein the information is limited in a homologous inscription or the calligraphy works of a single writer. During the pre-treatment process, after the inscription is divided into single Chinese characters, the component decomposition and the stroke decomposition are conducted on each extracted single Chinese character. The acquired components and the acquired strokes of Chinese characters together form a component-stroke template set. During the restoration process of damaged Chinese characters, firstly, the structural information and the contour information of the credible component/stroke part of the damaged Chinese characters are extracted. Based on the partial correspondence relationship of the structure matching and contour segments, a template that is highest in similarity is found out of the component-stroke template set to be adopted as a signal source for restoring the damaged part of the damaged Chinese characters. Compared with other image restoration methods, the structural information and the contour detail information of the damaged part of Chinese characters can be restored on the premise that the original style of the inscription is well maintained.
Description
Technical field
The invention belongs to virtual reality technology, computer vision and field of Computer Graphics, specifically for the Chinese character unit stroke decomposition method of the image repair of an inscription on a tablet.
Background technology
An inscription on a tablet Digital repair is a branch of image repair.At present in image repair, forefathers have done a large amount of work.In the development in image repair field, mainly contain both direction according to the implication of image mending and the mode of repairing.One is the implication itself from image mending, and namely to the reparation of noise tiny on image, irregular cut, what this situation generally adopted is solve based on partial differential equation or variational method.Restorative procedure based on variational problem has more sufficient theoretical foundation, but the shortcoming of the method repairs image not nature, and restoring area profile is obvious.Can blooming be produced when repairing large-area defect area on picture by the method based on partial differential equation, not be suitable in this way and repair large-area texture region.Two disappearances being aimed at large regions on picture are damaged, and what these class methods adopted is image mending method based on sample texture and texture generation technique, effectively avoids the generation of blooming compared with direction one.Utilize the texture information of image to guide the process of textures synthesis, effectively improve the repairing quality of image, maintain picture structure texture.And the subject matter of this method is mainly recovered in the structural information of failure area.Many image repair methods improved are proposed in order to address this problem forefathers, as the image repair method based on sample, it utilizes the isophote direction of affected area peripheral information and hair line direction to arrange preferred value for area to be repaired block, the part that priority is large will preferentially be repaired, the pixel that priority is little will be repaired after giving as security, and these priority are recorded in a table.In reparation, guided the priority ranking of reparation by each pixel priority in continuous updating form, keep image in this part-structure continuous print object to reach.The shortcoming of this method to keep linear (as straight line) structure, is difficult to people's satisfaction for curvilinear structures effect.The image repair method that structure based propagates (structurepropagation) is the another kind of image repair method focusing on structure repair.In this approach, first specify by the mode of adding boost line the mode that in image, failure area surrounding structure extends at failure area by user, namely draw the structure lines of failure area according to the experience hobby etc. of people.Repaired in failure area by structure-borne under boost line instructs.Setting up image set is the image repair method that another can be repaired structure, by matching the image similar with damaging image possibility in millions of image, then utilizes this image corresponding part to fill up image failure area.This method can have efficient recovery original image lost part, but repair structure depends on image library.
Be directed to Chinese character image (as calligraphy, an inscription on a tablet etc.), its reparation feature is, focuses on the reparation of structural information and profile details, and does not emphasize the reparation of texture information.That repairs the image of Chinese character focuses on Recover from damaging partial structural information, the method that image structure based after Chinese character is through binaryzation is propagated is not suitable for bianry image, and be that an inscription on a tablet that calligraphy work finite sum is handed down only has a small amount of usable samples based on the difficult point of Chinese Character Set method, limited sample size, differs when repairing and finds suitable Chinese character as template surely.
Summary of the invention
The object of the invention is in order to overcome above-mentioned shortcoming, solve the shortcoming that conventional images treatment technology can not have efficient recovery an inscription on a tablet image broken parts structural information, a kind of an inscription on a tablet restorative procedure described based on Chinese character image contour feature is provided, solve the shortcoming being difficult to recover defect area structure and profile details information for an inscription on a tablet image, better solve the problem of an inscription on a tablet image not program degree defect.
To achieve these goals, the technical solution adopted in the present invention is: a kind of an inscription on a tablet restorative procedure described based on Chinese character image contour feature, is characterized in that following steps:
Step 1, use the method for grey level histogram to carry out the single-row segmentation of an inscription on a tablet according to arrangement feature, single-rowly go out the gray scale regularity of distribution by statistics of histogram to what be partitioned into again and be partitioned into individual Chinese character.
Step 2, unit stroke decomposition is carried out to the individual Chinese character obtained in step one, build template set.
Step 3, use SC descriptors to be described to all templates in template, the coupling for similar unit stroke is retrieved.
When damaging Chinese character in step 4, reparation an inscription on a tablet, the difference for the order of severity of Chinese character defect is divided into based on the reparation of parts and the reparation based on single stroke.For the situation of most of region defect of certain parts in a Chinese character.First obtain the topology information damaging Hanzi component, carry out the description of SC descriptor by topology and parts skeleton sampled point.The component set extracted in step 2 is measured based on the topological resemblance degree of SC algorithm, retrieves similarity the highest, i.e. the most possible parts close with the defect parts information source of repairing as defect area.For the situation of Local Damaged only having certain stroke, first use shape description matrix to be described to the trusted area profile of impaired stroke, also same expression is carried out to each stroke in the stroke set obtained.In stroke set, the information source that the stroke the highest with defect stroke similarity repair as stroke is retrieved through IS-Matching contour segment matching algorithm.Utilize said method find as restoration information source template guided under complete the reparation of broken parts in Chinese character.
Compare relative to other technology existing and the invention has the advantages that: the present invention is directed an inscription on a tablet image repair technology, repair process has suffered Chinese character conformational analysis, by parts, stroke carries out the reparation of an inscription on a tablet Chinese character at many levels.First build template set, according to repairing by different level in various degree of Chinese character defect, relate to structural information and the profile details information of Chinese character simultaneously.
Accompanying drawing explanation
Fig. 1 is the process schematic that the present invention is based on an inscription on a tablet restorative procedure that Chinese character image contour feature describes;
Fig. 2 is an inscription on a tablet column split schematic diagram in Chinese character character extraction of the present invention;
Fig. 3 is the schematic diagram that an inscription on a tablet column split in Chinese character character extraction of the present invention goes out individual character;
Fig. 4 is parts defect repair process flow diagram during this bright Chinese character is repaired;
Fig. 5 is that during Chinese character of the present invention is repaired, stroke level repairs process flow diagram;
Fig. 6 is that result displaying repaired in Chinese character of the present invention.
Embodiment
As shown in Figure 1, the present invention is based on an inscription on a tablet restorative procedure employing following steps that Chinese character image contour feature describes:
Step one, an inscription on a tablet are split:
(1) first carry out binary conversion treatment again to after an inscription on a tablet image smoothing, remove noise on the basis retaining Chinese character marginal information in an inscription on a tablet.
(2) write from top to bottom according to an inscription on a tablet writing process, so in the vertical direction keeps alignment, the histogrammic Wave crest and wave trough pretreated an inscription on a tablet in the vertical direction being carried out to projection formation is obvious, can find column split position accurately, be partitioned into single-row.
(3) be partitioned on single-row basis, to single-row carrying out, horizontal direction projecting, the ratio that the height of structure Chinese character is wide retrains, and adjusts, be correctly partitioned into individual Chinese character to the narrow and wide situation of segmentation.
Step 2, parts and Stroke decomposition:
(1) decomposition of stroke.Use concordance list thinning algorithm to carry out refinement to the Chinese character that step one is partitioned into and obtain skeleton, in size be the concordance list of 256 guidance under complete the thinning process of Chinese character image.The skeleton of single pixel wide that what refinement obtained is.Three constraints must be met: 1. internal point can not be deleted in thinning process.2. straight line end points can not be deleted.3. if fruit dot P is frontier point, after removing P, if connected component does not increase, then P can delete.Obtain Chinese character skeleton through refinement, the point of crossing of these skeleton points is that separation can obtain many stroke skeleton segments, uses merge algorithm to carry out merging can obtain complete stroke skeleton to these skeleton segments.The complete stroke skeleton of stroke generates and meets following three rules: 1. two stroke skeleton segments are connected, and 2. the angle difference of two stroke skeleton segments and target stroke skeleton segment is less than given threshold values.3. the stroke skeleton obtained after merging and the differential seat angle of target stroke skeleton are less than given threshold values.After obtaining complete stroke skeleton, the corresponding relation according to skeleton and profile obtains complete stroke.
(2) decomposition of component.Reorganized by objective contour sampled point by Edge track, the objective contour obtaining chain decomposes for parts.1. determine starting point.2. detect the next pixel in eight fields of current point counterclockwise and upgrade current point.3. repeat above-mentioned two steps until follow the tracks of end.Utilize the edge point set with chain form obtained, obtain after the input skeleton of Chinese character and the corresponding relation of target component through the optimal principle, find the corresponding relation of point and skeleton, the point corresponding to parts skeleton extracted and connects in the position disconnected, obtaining the parts split.
Step 3, based on the subassembly selection of topological structure coupling and the contour segment registration of semi-match:
(1) Turbo codes and subassembly selection.For the situation of parts defect, need to find close parts, parts are made up of multiple stroke, have respective topological structure.To the parts in template set through concordance list thinning algorithm, obtain the parts skeleton of single pixel wide degree, use SC descriptor to be described skeleton sampling point set, by the shape histogram counted below formula between skeleton point.
001"/>
Wherein,
when representing centered by current skeleton point, the skeleton in a kth region is counted; # operation represent q be fall into a kth region be different from p
ithe quantity of all the other points on the profile of point, k represents the zone number that polar coordinate system divides two dimensional surface, p
ifor skeleton to be described, point, q represents and is different from p
iskeleton point, bin (k) represents the kth region be divided in polar coordinate system;
Same, Turbo codes is carried out to the Hanzi component of defect.Similarity d between each template of formula (6) (7) computing machine.
002"/>
Wherein, p
ithe point on shape sampling point set A, q
jbe the point on form sampling point set B, A and B is two point sets to be compared, and N is the number of regions that polar coordinate space divides.
Formula (7) utilizes the result of formula (6) to give the similarity of two point sets, retrieves the information source that the parts the highest with defect parts similarity are repaired as parts from template set.
003"/>
Wherein, p
ithe point of shape sampling point set A on the h of region, q
jthe point of shape sampling point set B on the h of region; A
hand B
hbe two subsets of point set in polar coordinate space on the h of region to be compared; N is the number of regions that polar coordinate space divides;
(2) profile describes.For the situation that individual character stroke damages, because the discrimination of the topological structure of single stroke is little, the profile by damaging trusted area in stroke is needed to be described the selection of apply to Template stroke.First need to use shape description matrix to represent damage stroke trusted area profile, what obtain is a N*N matrix as formula (8), and N is contour segment sampling number.
Wherein angle α
ija p
iwith a p
jbetween line and some p
iwith reference point p
j3rd some p of position
j-△between the angle of line.
(3) contour segment registration.Be extracted defect stroke reliable part contour segment set Ε={ e
1, e
2... e
k, k represents profile hop count.K point set P is obtained to each section of profile uniform sampling
i={ p
1, p
2..., p
n, n represents the sampling number of each section, and each contour segment n can be different.Similar expression P is carried out to the profile of stroke each in template set
m={ p
1, p
2, p
n, wherein m represents the numbering of each stroke in stroke set.According to describing method above, shape description matrix is set up to each stroke in the credible contour segment of defect stroke and template set.In order to find the process of the optimum profiles matching section in credible contour segment and stroke set between each outline strokes, the process namely finding corresponding contour segment is exactly the process of in compute matrix similar piece.Formula (9) gives concrete method for solving.
004"/>
Wherein r represents that profile similar section is counted, A and B represents the shape of two measuring similarities.S and m represents the sampling number of shape A and B respectively.
Step 4, Chinese character are repaired by different level:
(1) according to the degree difference that Chinese character damages, the reparation of different levels is carried out.The situation of parts defect is carried out to the reparation of hierarchy of components, first according to previously described subassembly selection algorithm, select best parts recovery template, the process flow diagram that Fig. 4 recovers to Hanzi component broken parts.
(2) for the situation that unit stroke is impaired, the reparation of stroke level is carried out.Adopting above-mentioned profile to describe algorithm first uses shape description matrix to be described to the trusted area profile of impaired stroke, also same expression is carried out to each stroke in the stroke set obtained, in stroke set, retrieves the stroke the highest with defect stroke similarity through IS-Matching contour segment matching algorithm and as the information source of stroke reparation, defect stroke is repaired.Fig. 5 gives the process flow diagram that Chinese-character stroke level is repaired.
Can find out that the present invention is in the reparation taking homology unit stroke as template guided damage Chinese character from reparation result Fig. 6, can be good at the undamaged portion recovering Chinese character, remain the writing style of original Chinese character simultaneously.
Non-elaborated part of the present invention belongs to techniques well known; The above; be only part embodiment of the present invention, but protection scope of the present invention is not limited thereto, any those skilled in the art are in the technical scope that the present invention discloses; the change that can expect easily or replacement, all should be encompassed within protection scope of the present invention.
Claims (4)
1., based on an inscription on a tablet restorative procedure that Chinese character image contour feature describes, its feature comprises the following steps:
Step 1, according to an inscription on a tablet arrangement feature, by grey level histogram by the writing brush word in an inscription on a tablet image respectively procession segmentation extract individual Chinese character, to the individual character normalized extracted;
Step 2, in the single character extracted not damage Chinese character decomposition be Chinese character basic composition element parts and stroke, these parts strokes formed a set, be referred to as template set, in template set, preserve profile information and the framework information of each unit stroke;
Step 3, use Shape context SC descriptors to be described to all templates in template, the coupling for similar unit stroke is retrieved;
The reparation of step 4, an inscription on a tablet is divided into two points, to the serious Chinese character of defect repair damage Chinese character time, first given its damages the topology information of parts, uses SC to describe given topology information; The source of the template the highest with this topological resemblance degree as restoration information is found out in template set; Select template to carry out contour segment with defect parts reliable part mate and repair; For the Chinese character that individual character stroke damages, shape description matrix is used to carry out shape description to these contour segments, namely the matrix of line angle between point for representing relation between profile each point and reference point is set up, then adopt partial contour matching process (IS-Matching) in the set of stroke template, find the corresponding contour segment on template stroke of contour segment, determine possible similar stroke alternatively stroke by these corresponding contour segments; Utilize said method to find the most close with defect area template guided under complete the reparation of broken parts in Chinese character.
2. an inscription on a tablet restorative procedure based on Chinese character image contour feature according to claim 1, is characterized in that: in described step 1, the detailed process of Chinese character character extraction is as follows:
Input an inscription on a tablet image, according to being written as ancient times from top to bottom, vertical direction keeps the feature of alignment, to calculating its grey level histogram of vertical direction after its gray processing, utilize histogrammic crest to carry out column split for row boundary position, in like manner split in the horizontal direction, obtain individual character Chinese character.
3. an inscription on a tablet restorative procedure based on Chinese character image contour feature according to claim 1, is characterized in that: the detailed process that in described step 2, Chinese character unit stroke decomposes is as follows:
(3.1) to Chinese character contour uniform sampling, contour following algorithm is utilized to find the point chain organization formation of corresponding profile, concrete steps are divided into: determine starting point, search for next target domain point successively and upgrade starting point, like this until follow the tracks of end, object edge is reorganized, obtains the objective contour information of chain; The chain objective contour information after reorganizing is utilized to carry out part segmentation;
(3.2) Stroke decomposition is carried out to the individual character extracted, use image processing techniques obtain the skeleton of the single pixel wide degree of Chinese character and remove burr, obtain first skeleton segment in small, broken bits according to the flex point of skeleton and point of crossing; Under user interactions, complete the merging of stroke skeleton segment, generate complete stroke skeleton, this process must meet following constraint: two stroke section skeletons are connected, the angle difference of two stroke section skeletons is less than a given threshold values; Under the prerequisite meeting above-mentioned constraint, initial pen section skeleton is merged one by one; Utilize the stroke that the corresponding relation of profile and skeleton decomposites;
(3.3) Hanzi component obtained above and stroke are formed a template set being used for repairing.
4. an inscription on a tablet restorative procedure based on Chinese character image contour feature according to claim 1, is characterized in that: in described step 4, the performing step of partial contour matching process is as follows:
(4.1) situation for parts defect first obtains the framework information damaging Chinese character, uses SC descriptor to carry out skeleton description, as formula (1):
Wherein,
when representing centered by current skeleton point, the skeleton in a kth region is counted; # operation represent q be fall into a kth region be different from p
ithe quantity of all the other points on the profile of point, k represents the zone number that polar coordinate system divides two dimensional surface, p
ifor skeleton to be described, point, q represents and is different from p
iskeleton point, bin (k) represents the kth region be divided in polar coordinate system;
(4.2) matching primitives and defect parts similarity in a template; Formula (2) calculates similarity d between two similar shapes corresponding each point;
Wherein, p
ithe point on shape sampling point set A, q
ibe the point on shape sampling point set B, A and B is two point sets to be compared, and N is the number of regions that polar coordinate space divides;
Formula (3) utilizes the result of formula (2) to give the similarity D of two point sets
h;
Wherein, p
ithe point of shape sampling point set A on the h of region, q
jthe point of shape sampling point set B on the h of region; A
hand B
hbe two subsets of point set in polar coordinate space on the h of region to be compared; N is the number of regions that polar coordinate space divides;
(4.3) for the situation of stroke defect, adopt shape description matrix to be described stroke to be repaired, each matrix-block represents corresponding contour segment, and each stroke in template set also carries out same expression;
(4.4) utilize formula (4) to calculate the corresponding point of two contour segments, carry out Image matching, under the guidance of template stroke, complete the reparation of defect stroke;
Wherein, s and m represents the sampling number of shape A and B respectively; R represents that profile similar section is counted, A and B is two point sets to be compared.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510441241.4A CN105069766B (en) | 2015-07-24 | 2015-07-24 | A kind of an inscription on a tablet restorative procedure based on the description of Chinese character image contour feature |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510441241.4A CN105069766B (en) | 2015-07-24 | 2015-07-24 | A kind of an inscription on a tablet restorative procedure based on the description of Chinese character image contour feature |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105069766A true CN105069766A (en) | 2015-11-18 |
| CN105069766B CN105069766B (en) | 2017-12-08 |
Family
ID=54499124
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510441241.4A Expired - Fee Related CN105069766B (en) | 2015-07-24 | 2015-07-24 | A kind of an inscription on a tablet restorative procedure based on the description of Chinese character image contour feature |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105069766B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109410291A (en) * | 2018-09-11 | 2019-03-01 | 北京语言大学 | The treating method and apparatus of burr type pen section |
| CN111067669A (en) * | 2019-12-24 | 2020-04-28 | 北京爱康宜诚医疗器材有限公司 | Method and device for determining acetabular bone defect treatment mode |
| CN112435196A (en) * | 2020-12-02 | 2021-03-02 | 中国人民解放军战略支援部队信息工程大学 | Character repairing method and system based on deep learning |
| CN112837334A (en) * | 2021-04-02 | 2021-05-25 | 河南大学 | Automatic conjugation method of Chinese character and sketch image |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070142952A1 (en) * | 2005-12-19 | 2007-06-21 | Nitto Denko Corporation | Error-correction assistance system |
| CN101458770A (en) * | 2008-12-24 | 2009-06-17 | 北京文通科技有限公司 | Character recognition method and system |
| CN101859382A (en) * | 2010-06-03 | 2010-10-13 | 复旦大学 | License plate detection and identification method based on maximum stable extremal region |
| CN102521805A (en) * | 2011-11-30 | 2012-06-27 | 华中科技大学 | Video word processing method based on interframe information |
| CN102968764A (en) * | 2012-10-26 | 2013-03-13 | 北京航空航天大学 | Chinese character image inpainting method based on strokes |
| CN103208107A (en) * | 2012-01-17 | 2013-07-17 | 展讯通信（上海）有限公司 | Terminal and method and device for repairing image |
| CN104376300A (en) * | 2014-11-03 | 2015-02-25 | 电子科技大学 | Identification method used for intelligent matching of incomplete Chinese characters on basis of grid characteristics |
-
2015
- 2015-07-24 CN CN201510441241.4A patent/CN105069766B/en not_active Expired - Fee Related
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070142952A1 (en) * | 2005-12-19 | 2007-06-21 | Nitto Denko Corporation | Error-correction assistance system |
| CN101458770A (en) * | 2008-12-24 | 2009-06-17 | 北京文通科技有限公司 | Character recognition method and system |
| CN101859382A (en) * | 2010-06-03 | 2010-10-13 | 复旦大学 | License plate detection and identification method based on maximum stable extremal region |
| CN102521805A (en) * | 2011-11-30 | 2012-06-27 | 华中科技大学 | Video word processing method based on interframe information |
| CN103208107A (en) * | 2012-01-17 | 2013-07-17 | 展讯通信（上海）有限公司 | Terminal and method and device for repairing image |
| CN102968764A (en) * | 2012-10-26 | 2013-03-13 | 北京航空航天大学 | Chinese character image inpainting method based on strokes |
| CN104376300A (en) * | 2014-11-03 | 2015-02-25 | 电子科技大学 | Identification method used for intelligent matching of incomplete Chinese characters on basis of grid characteristics |
Non-Patent Citations (5)
| Title |
|---|
| JIE CHEN,FUXI ZHU: "HierarchicalMatching for Chinese Calligraphic RetrievalBased on Skeleton Similarity", 《EDUCATION TECHNOLOGY AND COMPUTER SCIENCE》 * |
| YU QIAO 等: "Recovering Dynamic Information from Static Handwritten Images", 《 FRONTIERS IN HANDWRITING RECOGNITION》 * |
| 刘敏 等: "基于骨架图匹配的汉字变形技术", 《北京航空航天大学学报》 * |
| 郑丹晨 等: "基于改进典型形状上下文特征的形状识别方法", 《计算机辅助设计与图形学学报》 * |
| 郭晨 等: "基于计算机图像处理的书法笔画特征提取", 《天津科技大学学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109410291A (en) * | 2018-09-11 | 2019-03-01 | 北京语言大学 | The treating method and apparatus of burr type pen section |
| CN109410291B (en) * | 2018-09-11 | 2023-03-07 | 北京语言大学 | Processing method and device for burr type pen segments |
| CN111067669A (en) * | 2019-12-24 | 2020-04-28 | 北京爱康宜诚医疗器材有限公司 | Method and device for determining acetabular bone defect treatment mode |
| CN112435196A (en) * | 2020-12-02 | 2021-03-02 | 中国人民解放军战略支援部队信息工程大学 | Character repairing method and system based on deep learning |
| CN112435196B (en) * | 2020-12-02 | 2022-11-29 | 中国人民解放军战略支援部队信息工程大学 | Character repairing method and system based on deep learning |
| CN112837334A (en) * | 2021-04-02 | 2021-05-25 | 河南大学 | Automatic conjugation method of Chinese character and sketch image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105069766B (en) | 2017-12-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103941866B (en) | Three-dimensional gesture recognizing method based on Kinect depth image | |
| CN104809446B (en) | Palmmprint area-of-interest rapid extracting method based on correction volar direction | |
| CN105550691B (en) | Adaptive mountain valley ridge line extracting method and system based on scale space | |
| CN105069766A (en) | Inscription restoration method based on contour feature description of Chinese character image | |
| CN104834922A (en) | Hybrid neural network-based gesture recognition method | |
| CN104700092A (en) | Small-character number identification method based on template and feature matching | |
| CN110659644B (en) | Automatic extraction method for strokes of calligraphy single characters | |
| CN107085726A (en) | Oracle bone rubbing individual character localization method based on multi-method denoising and connected component analysis | |
| CN101697229A (en) | Method for extracting region of interest of medical image | |
| CN104392229A (en) | Handwriting identification method based on stroke fragment direction distribution characteristics | |
| CN112699845A (en) | Online non-contact palm vein region-of-interest extraction method | |
| CN105138983A (en) | Pedestrian detection method based on weighted part model and selective search segmentation | |
| CN107944451A (en) | The row cutting method and system of a kind of ancient Tibetan books document | |
| CN104134195A (en) | Image restoration method based on geometric sparsity of blocks | |
| WO2023065397A1 (en) | Stroke order recognition method and system for handwritten chinese character image | |
| CN114494306A (en) | Oracle bone thesaurus character contour repairing method guided by edge gradient covariance | |
| CN107945190A (en) | Bianry image high-speed communication region computational methods | |
| CN114511853B (en) | Character image writing track recovery effect discrimination method | |
| CN111612907A (en) | Multidirectional repairing system and method for damaged ancient building column | |
| Wu et al. | Web based chinese calligraphy learning with 3-d visualization method | |
| CN113362296B (en) | Tunnel crack extraction method and system | |
| CN114332874A (en) | Calligraphy repairing method based on crowdsourcing | |
| Yuan et al. | A detection method of palmprint principal lines based on local minimum gray value and line following | |
| Min et al. | Improved range image segmentation by analyzing surface fit patterns | |
| CN109934837B (en) | 3D plant leaf contour extraction method, device and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| 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: 20171208 Termination date: 20200724 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |