CN106971400A - A kind of method for repairing and mending and its device of image cut-off rule - Google Patents
A kind of method for repairing and mending and its device of image cut-off rule Download PDFInfo
- Publication number
- CN106971400A CN106971400A CN201710149441.1A CN201710149441A CN106971400A CN 106971400 A CN106971400 A CN 106971400A CN 201710149441 A CN201710149441 A CN 201710149441A CN 106971400 A CN106971400 A CN 106971400A
- Authority
- CN
- China
- Prior art keywords
- line segment
- cut
- segmentation line
- rule
- length threshold
- 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
Abstract
The present invention relates to a kind of method for repairing and mending and its device of image cut-off rule.The cut-off rule that image is obtained after over-segmentation is made up of discrete cut-point, to described cut-point by Continuity Analysis, makes the cut-point connect to form no less than one segmentation line segment；Length threshold or length threshold scope are set according to the distance of cut-off rule, to the segmentation line segment more than length threshold or length threshold scope as effective segmentation line segment, invalid segmentation line segment is used as the segmentation line segment below length threshold or length threshold scope；Then, above-mentioned effective segmentation line segment is retained, and the discontinuous point between the segmentation line segment effective to this is repaired.The present invention is repaired on the basis of the correct cut-off rule of image is retained to the discontiguous area of cut-off rule, overcomes the defect of prior art algorithm, method is simply easily realized, improves the accuracy repaired to cut-off rule.
Description
Technical field
The present invention relates to a kind of method for repairing and mending and its device of image cut-off rule, belong to the technology neck of Digital Image Segmentation
Domain.
Background technology
Image segmentation is exactly to divide the image into several regions specific, with unique properties and propose interesting target
Technology and process.The more classical algorithm of the natural scene image ration of division includes：Segmentation, feature based based on border
Mean-shift methods, the segmentation optimized based on energy function and dividing method based on graph theory etc..These algorithm researches are in recent years
To achieve suitable progress and be applied in numerous areas.
Scene image is typically some discrete cut-points according to the segmentation result of above-mentioned dividing method, continuous using some
Property analysis method be attached the cut-off rule formed afterwards also likely to be present interruption and mistake situation, this be accomplished by some segmentation
Post-processing algorithm is repaired and error correction to existing cut-off rule.Conventional method has low pass filtering method, function-fitting method, minimum
Square law etc., but these methods have the limitation of itself.LPF method can only be repaired or a small amount of point of error correction, and function
Fitting process needs to estimate the mathematical function that edge meets, and least square fitting often have modified original accurate detection cut-off rule,
It can not all solve the problems, such as to split post processing well.
The content of the invention
It is an object of the invention to overcome the deficiencies in the prior art, it is proposed that a kind of method for repairing and mending of image cut-off rule, solution
Determine due to producing the segmentation line segment of mistake when being connected to separate division point, so as to influence to ask cut-off rule repairing accuracy
Topic, the invention also provides a kind of repair apparatus of image cut-off rule.
The present invention is achieved by following scheme：
1. a kind of method for repairing and mending of image cut-off rule, step is as follows：
Step 1, the cut-off rule that image is obtained after over-segmentation is made up of discrete cut-point, and described cut-point is passed through
Continuity Analysis, makes the cut-point connect to form no less than one segmentation line segment；
Step 2, length threshold or length threshold scope are set according to the distance of cut-off rule, for judging having for segmentation line segment
Effect property；To the segmentation line segment more than length threshold or length threshold scope as effective segmentation line segment, in length threshold
Segmentation line segment below value or length threshold scope is used as invalid segmentation line segment；
Step 3, above-mentioned effective segmentation line segment is retained, and the discontinuous point between the segmentation line segment effective to this is repaiied
Mend.
Further, judge that the continuity between cut-point is analyzed using the method based on distance in step 1.
Further, the method for method or Weighted Segments length that line segment length is taken based in step 2 determines length threshold
Value or length threshold scope.
Further, to the segmentation line segment between the length threshold scope in step 2, according to adjacent effective point
Secant section determined by effective range come judge between length threshold scope segmentation line segment validity.
Further, it is to the mode that the discontinuous point between effective segmentation line segment is repaired in step 3：Set adjacent
The line of demarcation repairing area scope of Effective line, the gradient between the search intersegmental breakpoint of adjacent virtual lines in the range of the repairing area
The path of maximum, and the path is repaired；Or using linear interpolation method, polynomial fitting method between segmentation line segment
Discontinuous point repaired.
2. a kind of repair apparatus of image cut-off rule, the device includes cut-point link block, extracts effectively segmentation line segment
Module and repairing cut-off rule root module composition, it is specific as follows：
1) cut-point link block：The cut-off rule that image is obtained after over-segmentation is made up of discrete cut-point, to described
Cut-point by Continuity Analysis, make the cut-point connect to be formed no less than one segmentation line segment；
2) effective cut-off rule root module is extracted:Length threshold or length threshold scope are set according to the distance of cut-off rule, used
In the validity for judging segmentation line segment；Effective segmentation is used as to the segmentation line segment more than length threshold or length threshold scope
Line segment, invalid segmentation line segment is used as the segmentation line segment below length threshold or length threshold scope；
3) cut-off rule root module is repaired:Retain above-mentioned effective segmentation line segment, and between the segmentation line segment effective to this
Discontinuous point repaired.
Further, judge that the continuity between cut-point is carried out in cut-point link block using the method based on distance
Analysis.
Further, the method that line segment length is taken based in effective cut-off rule root module or Weighted Segments length are extracted
Method determines length threshold or length threshold scope.
Further, extract in effective cut-off rule root module to the segmentation line segment between the length threshold scope,
The effective range according to determined by the adjacent effective segmentation line segment judges the segmentation line segment between the length threshold scope
Validity.
Further, the mode repaired in repairing cut-off rule root module to the discontinuous point between effective segmentation line segment
For：The line of demarcation repairing area scope of adjacent virtual lines section is set, adjacent virtual lines are searched in the range of the repairing area intersegmental
The path of maximum of gradients between breakpoint, and the path is repaired；Or using linear interpolation method, polynomial fitting method pair
Discontinuous point between segmentation line segment is repaired.
Present invention beneficial effect compared to the prior art is：
The present invention proposes a kind of method for repairing and mending and its device of image cut-off rule, passes through what image was formed due to segmentation
Separate division clicks through Line Continuity analysis, forms it into some segmentation line segments, and set length threshold scope to segmentation line segment
Validity judged, retain its effective segmentation line segment；Then the discontinuous point between effective segmentation line segment is repaiied
Mend.Wrong part is rejected in the segmentation line segment that the present invention is formed to reconnecting, it is to avoid cause to mislead when repairing cut-off rule.
The discontiguous area of cut-off rule is repaired on the basis of the correct cut-off rule of image is retained, and overcomes lacking for prior art algorithm
Fall into, method is simply easily realized, improve the accuracy repaired to cut-off rule.
Brief description of the drawings
Fig. 1 is the overview flow chart of the embodiment of the present invention；
Fig. 2 is the cut-off rule Effective judgement flow chart based on high-low threshold value of the embodiment of the present invention；
Fig. 3 is the image cut-off rule breakpoint repairing flow chart of the embodiment of the present invention；
Fig. 4 is the intersegmental breakpoint repairing principle schematic diagram of cut-off rule of the embodiment of the present invention.
Embodiment
The present invention will be further described in detail with reference to the accompanying drawings and examples.
A kind of method for repairing and mending of image cut-off rule, the key step of this method includes：Cut-off rule Continuity Analysis, cut-off rule
The intersegmental breakpoint repairing of Effective judgement and cut-off rule of section；The step of below to this method, is specifically described successively：
1. cut-off rule Continuity Analysis；
The scene image obtained by partitioning algorithm is typically to be made up of the cut-point of series of discrete, it is therefore desirable to this
A little discrete click-through Line Continuity analyses, i.e., connect cut-point to form some segmentation line segments.Segmentation during scene cut
Point is usually unidirectional, i.e., all cut-points are corresponding or for image seat for image coordinate x one and only one y value
One and only one x value of mark y is corresponding.By cut-point for exemplified by coordinate x one and only one y value, adopting in the present embodiment
Continuity between being judged with the distance between cut-point.And it regard the set of continuity point as segmentation line segment, continuity point
Collection share C language, vector, stack or customized structure and preserved.
Concretely comprise the following steps and preset y directions continuity threshold value, travel through all cut-points, by y-coordinate between cut-point away from
A segmentation line segment is connected into from the cut-point less than continuity threshold value.Wherein, continuity threshold value is chosen for empirical value.
2. split line segment Effective judgement
Effective cut-off rule and the class of invalid cut-off rule two can be divided into according to length by above method formation segmentation line segment；Its
In, effective cut-off rule is the segmentation result for the scene image for needing reservation, and invalid cut-off rule should be removed as noise.
Specific judgment mode is as follows：
As shown in Fig. 2 the method based on line segment length sets a length threshold or threshold range.Then respectively statistically
State the length of gained segmentation line segment.Then, line segment of the line segment more than threshold value or threshold range will be split and is set to effective cut-off rule
Section, will split segmentation line end of the line segment below threshold value or threshold range and is set to invalid segmentation line segment, between threshold range
Between segmentation line segment can be used as segmentation line segment undetermined.Method in the present embodiment using adjacent virtual lines section is undetermined to determine
The validity of line segment, that is, calculate effective model of the adjacent effective segmentation line segment of segmentation line segment undetermined on the y directions of line of demarcation
Enclose, effective range is to constitute this maximum and minimum value for effectively splitting all cut-point y directions on line segment.Composition is calculated to treat
Surely the average value of cut-points all on line segment in y-direction is split, when this average value is in effective range, then this section is split
Line is judged as effectively splitting line segment, otherwise judges that described segmentation line segment undetermined is invalid segmentation line segment.Wherein, threshold value or threshold value
The threshold range in empirical value, the present embodiment that is chosen for of scope is chosen for Low threshold for picture traverse/50, and high threshold is chosen for 3
Low threshold again.
As other embodiment, the method given threshold scope or threshold value of Weighted Segments length can also be used.
3. the repairing of the intersegmental breakpoint of cut-off rule
The invalid segmentation line segment of above-mentioned selection is removed, the discontinuous point (or line) between adjacent Effective line is repaiied
Mend.The mode that selection class figure is cut in the present embodiment is repaired, as shown in Figure 3.First, adjacent virtual lines section is calculated in line of demarcation
Regional extent in repairing area scope, the present embodiment is a rectangle.Then, original scene figure is calculated in the regional extent
Sobel operators are selected to calculate image gradient in the gradient image of picture, the present embodiment.Segmentation line segment breakpoint repairing problem is changed into
Maximum flow problem in graph theory, as shown in figure 4, the region to be repaired of one 4 × 5 is illustrated as, wherein each round dot T represents one
Pixel, shade round spot S represents the connected mode and weights between two end points of the adjacent effective segmentation line segment of two bars, pixel
Represented using line segment R.Algorithm target is the feasible flow for the maximum weight for finding the S that joins two endpoints, and the path is included
Pixel be split line segment between required for repairing point.Weights R is line segment line beginning or end picture in the present embodiment
The Grad of element.Solving maximum flow problem, method has had Goldberg-Tarja methods, Ford-Fulkerson methods etc. at present.Make
For other embodiment, it can equally be calculated when calculating image gradient using Prewitt, Roberts operator.
The intersegmental breakpoint method for repairing and mending of cut-off rule chosen in the present embodiment is class figure segmentation method, and this method is only preferred reality
Apply mode.As other embodiment, repairing point can equally be repaiied using item formula fitting process or linear interpolation method
Mend.
Under the thinking that the present invention is provided, to above-mentioned implementation by the way of being readily apparent that to those skilled in the art
Technological means in example enters line translation, replacement, modification, and plays a part of and the basic phase of relevant art means in the present invention
Goal of the invention that is same, realizing is also essentially identical, and the technical scheme so formed is finely adjusted to be formed to above-described embodiment, this
Technical scheme is planted to still fall within protection scope of the present invention.
Claims (10)
1. a kind of method for repairing and mending of image cut-off rule, it is characterised in that step is as follows：
Step 1, the cut-off rule that image is obtained after over-segmentation is made up of discrete cut-point, described cut-point is passed through continuous
Property analysis, make the cut-point connect to be formed no less than one segmentation line segment；
Step 2, length threshold or length threshold scope are set according to the distance of cut-off rule, for judging the effective of segmentation line segment
Property；To the segmentation line segment more than length threshold or length threshold scope as effective segmentation line segment, in length threshold
Or the segmentation line segment below length threshold scope is used as invalid segmentation line segment；
Step 3, above-mentioned effective segmentation line segment is retained, and the discontinuous point between the segmentation line segment effective to this is repaired.
2. the method for repairing and mending of a kind of image cut-off rule according to claim 1, it is characterised in that use and be based in step 1
The method of distance judges that the continuity between cut-point is analyzed.
3. the method for repairing and mending of a kind of image cut-off rule according to claim 1, it is characterised in that be taken based in step 2
The method of line segment length or the method for Weighted Segments length determine length threshold or length threshold scope.
4. the method for repairing and mending of a kind of image cut-off rule according to claim 1, it is characterised in that between institute in step 2
State the segmentation line segment between length threshold scope, according to it is adjacent it is effective segmentation line segment determined by effective range come judge between
The validity of segmentation line segment between length threshold scope.
5. the method for repairing and mending of a kind of image cut-off rule according to claim 1, it is characterised in that to effective in step 3
The mode repaired of discontinuous point between segmentation line segment is：The line of demarcation repairing area scope of adjacent virtual lines section is set,
The path of maximum of gradients between the intersegmental breakpoint of adjacent virtual lines is searched in the range of the repairing area, and the path is repaiied
Mend；Or the discontinuous point between segmentation line segment is repaired using linear interpolation method, polynomial fitting method.
6. a kind of repair apparatus of image cut-off rule, it is characterised in that the device includes cut-point link block, extracts and effectively divide
Secant root module and repairing cut-off rule root module composition, it is specific as follows：
1) cut-point link block：The cut-off rule that image is obtained after over-segmentation is made up of discrete cut-point, to described point
Cutpoint makes the cut-point connect to form no less than one segmentation line segment by Continuity Analysis；
2) effective cut-off rule root module is extracted:Length threshold or length threshold scope are set according to the distance of cut-off rule, for sentencing
The validity of disconnected segmentation line segment；Effective cut-off rule is used as to the segmentation line segment more than length threshold or length threshold scope
Section, invalid segmentation line segment is used as the segmentation line segment below length threshold or length threshold scope；
3) cut-off rule root module is repaired:Retain above-mentioned effective segmentation line segment, and between the segmentation line segment effective to this between
Breakpoint is repaired.
7. the repair apparatus of a kind of image cut-off rule according to claim 6, it is characterised in that in cut-point link block
Judge that the continuity between cut-point is analyzed using the method based on distance.
8. the repair apparatus of a kind of image cut-off rule according to claim 6, it is characterised in that extract effectively segmentation line segment
The method of method or Weighted Segments length that line segment length is taken based in module determines length threshold or length threshold scope.
9. the repair apparatus of a kind of image cut-off rule according to claim 6, it is characterised in that extract effectively segmentation line segment
To the segmentation line segment between the length threshold scope in module, according to determined by adjacent effective segmentation line segment effectively
Scope come judge between length threshold scope segmentation line segment validity.
10. a kind of repair apparatus of image cut-off rule according to claim 6, it is characterised in that repairing segmentation line segment mould
It is to the mode that the discontinuous point between effective segmentation line segment is repaired in block：The line of demarcation repairing of adjacent virtual lines section is set
Regional extent, the path of maximum of gradients between the search intersegmental breakpoint of adjacent virtual lines in the range of the repairing area, and to this
Repaired in path；Or the discontinuous point between segmentation line segment is repaired using linear interpolation method, polynomial fitting method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710149441.1A CN106971400B (en) | 2017-03-10 | 2017-03-10 | Method and device for repairing image dividing line |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710149441.1A CN106971400B (en) | 2017-03-10 | 2017-03-10 | Method and device for repairing image dividing line |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106971400A true CN106971400A (en) | 2017-07-21 |
| CN106971400B CN106971400B (en) | 2020-11-10 |
Family
ID=59329317
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710149441.1A Active CN106971400B (en) | 2017-03-10 | 2017-03-10 | Method and device for repairing image dividing line |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106971400B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001076154A (en) * | 1999-08-23 | 2001-03-23 | Xerox Corp | Method for segmenting digital image and digital image processor |
| WO2006005728A2 (en) * | 2004-07-10 | 2006-01-19 | Evotec Technologies Gmbh | Method for detecting geometrical structures in images |
| CN102938066A (en) * | 2012-12-07 | 2013-02-20 | 南京大学 | Method for reconstructing outer outline polygon of building based on multivariate data |
| CN104021574A (en) * | 2014-07-04 | 2014-09-03 | 武汉武大卓越科技有限责任公司 | Method for automatically identifying pavement diseases |
| CN104751178A (en) * | 2015-03-31 | 2015-07-01 | 上海理工大学 | Pulmonary nodule detection device and method based on shape template matching and combining classifier |
| CN104933709A (en) * | 2015-06-04 | 2015-09-23 | 西安理工大学 | Automatic random-walk CT lung parenchyma image segmentation method based on prior information |
| CN105930800A (en) * | 2016-04-21 | 2016-09-07 | 北京智芯原动科技有限公司 | Lane line detection method and device |
| CN106373168A (en) * | 2016-11-24 | 2017-02-01 | 北京三体高创科技有限公司 | Medical image based segmentation and 3D reconstruction method and 3D printing system |
-
2017
- 2017-03-10 CN CN201710149441.1A patent/CN106971400B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2001076154A (en) * | 1999-08-23 | 2001-03-23 | Xerox Corp | Method for segmenting digital image and digital image processor |
| WO2006005728A2 (en) * | 2004-07-10 | 2006-01-19 | Evotec Technologies Gmbh | Method for detecting geometrical structures in images |
| CN102938066A (en) * | 2012-12-07 | 2013-02-20 | 南京大学 | Method for reconstructing outer outline polygon of building based on multivariate data |
| CN104021574A (en) * | 2014-07-04 | 2014-09-03 | 武汉武大卓越科技有限责任公司 | Method for automatically identifying pavement diseases |
| CN104751178A (en) * | 2015-03-31 | 2015-07-01 | 上海理工大学 | Pulmonary nodule detection device and method based on shape template matching and combining classifier |
| CN104933709A (en) * | 2015-06-04 | 2015-09-23 | 西安理工大学 | Automatic random-walk CT lung parenchyma image segmentation method based on prior information |
| CN105930800A (en) * | 2016-04-21 | 2016-09-07 | 北京智芯原动科技有限公司 | Lane line detection method and device |
| CN106373168A (en) * | 2016-11-24 | 2017-02-01 | 北京三体高创科技有限公司 | Medical image based segmentation and 3D reconstruction method and 3D printing system |
Non-Patent Citations (1)
| Title |
|---|
| 孙涵等: "一种快速实用的直线检测算法", 《计算机应用研究》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106971400B (en) | 2020-11-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105913415B (en) | A kind of image sub-pixel edge extracting method with extensive adaptability | |
| CN105260713A (en) | Method and device for detecting lane line | |
| CN109285179A (en) | A kind of motion target tracking method based on multi-feature fusion | |
| CN105869174B (en) | A kind of Sky Scene image partition method | |
| CN104408711A (en) | Multi-scale region fusion-based salient region detection method | |
| CN104867145B (en) | IC element welding point defect detection methods based on VIBE models | |
| CN104851085A (en) | Method and system automatically obtaining detection zone in image | |
| CN102903093A (en) | Poisson image fusion method based on chain code mask | |
| CN106156749A (en) | Method for detecting human face based on selective search and device | |
| CN110889328A (en) | Method, device, electronic equipment and storage medium for detecting road traffic condition | |
| CN104463138A (en) | Text positioning method and system based on visual structure attribute | |
| CN105335960B (en) | The image partition method of jointing edge detection and watershed algorithm | |
| Li et al. | Bounding box–based technique for pavement crack classification and measurement using 1 mm 3D laser data | |
| CN106504247B (en) | A kind of quick-fried automatic testing method of the insulator chain based on visible images and device | |
| CN106651792A (en) | Method and device for removing strip noise of satellite image | |
| CN108573238A (en) | A kind of vehicle checking method based on dual network structure | |
| CN110033437A (en) | A kind of cut deal impression intelligent identification Method based on convolutional neural networks | |
| CN113240623A (en) | Pavement disease detection method and device | |
| CN108492306A (en) | A kind of X-type Angular Point Extracting Method based on image outline | |
| CN102663787B (en) | Method and system for image path generation | |
| CN103761520B (en) | File and picture based on stroke width is without ginseng binarization method | |
| CN106971400A (en) | A kind of method for repairing and mending and its device of image cut-off rule | |
| CN116596895A (en) | Substation equipment image defect identification method and system | |
| CN107085855B (en) | A kind of quick circular arc detection method | |
| Danilescu et al. | Road anomalies detection using basic morphological algorithms |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 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 |