CN102156984A - Method for determining optimal mark image by adaptive threshold segmentation - Google Patents
Method for determining optimal mark image by adaptive threshold segmentation Download PDFInfo
- Publication number
- CN102156984A CN102156984A CN2011100849024A CN201110084902A CN102156984A CN 102156984 A CN102156984 A CN 102156984A CN 2011100849024 A CN2011100849024 A CN 2011100849024A CN 201110084902 A CN201110084902 A CN 201110084902A CN 102156984 A CN102156984 A CN 102156984A
- Authority
- CN
- China
- Prior art keywords
- image
- average
- described step
- adaptive threshold
- cut apart
- 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
- 238000000034 method Methods 0.000 title claims abstract description 21
- 230000003044 adaptive effect Effects 0.000 title claims abstract description 12
- 230000011218 segmentation Effects 0.000 title claims abstract description 8
- 238000007789 sealing Methods 0.000 claims description 4
- 238000003709 image segmentation Methods 0.000 abstract description 8
- 230000008859 change Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 5
- VMXUWOKSQNHOCA-UKTHLTGXSA-N ranitidine Chemical compound [O-][N+](=O)\C=C(/NC)NCCSCC1=CC=C(CN(C)C)O1 VMXUWOKSQNHOCA-UKTHLTGXSA-N 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method for determining an optimal mark image by adaptive threshold segmentation, which comprises the following steps of: calculating a mean value E and a variance Sigma of a gradient image f (x, y) to be marked; establishing different combinations H Lambda of the mean value E and the variance Sigma by using the mean value E as the centre and using a certain multiple of the variance Sigma as a step length; carrying out threshold segmentation on the gradient image by the different combinations H Lambda to obtain a corresponding binary image sequence f' (x, y); counting the number C Lambda of closed communicated regions each of which areas are greater than a minimum region area parameter in the binary image sequence f' (x, y); and selecting a binary image which has the maximum number C Lambda as a mark image. In the method, the optimal mark image of watershed change can be generated adaptively according to a principle of maximum region number by simply setting the minimum region area parameter; an algorithm is simple, fast and convenient; the minimum region area parameter is insensitive to an image segmentation result; and the minimum region area parameter is not required to be changed ordinarily after being set.
Description
Technical field
The present invention relates to a kind of image processing method, particularly a kind of adaptive threshold that utilizes is cut apart the method for determining optimum marking image.
Background technology
Along with the development of sensor technology, the spatial resolution of satellite remote sensing images is progressively improving.Image resolution ratio before several years is at ten meter levels, and the mean flow rate response of some atural objects has been represented in the brightness of a pixel, even need carry out pixel and decompose.And present up-to-date commercial satellite can reach the spatial resolution of 0.5m, and the level of detail of image obtains tens of times raising.Pixel correspondence the part of atural object target on high-definition picture, and the heterogeneity of ground object target manifests day by day, so image segmentation becomes an inevitable problem.Image segmentation is based on the process that homogeney or heterogeneous criterion are divided into piece image some significant subregions, this process is converted into cut zone to input picture, to further extraction target signature, carry out target measurement and classification and other high-rise processing all are very important.Image segmentation is the key link of high spatial resolution remote sense image surface in object class and identifying, also is one of the difficult point in remote sensing image processing field.
In order to cut apart all the atural object objects in the remote sensing images, mainly contain three classes and realize approach: the first kind is to connect and region labeling realization image segmentation by rim detection, edge; Second class is to merge by the zone to realize image segmentation, comprises level merging from bottom to top, top-down regional split and merging etc.; The 3rd class is to realize image segmentation according to the watershed transform of image gradient, and has developed the multiple improvement algorithms such as merging after cutting apart preceding mark and cutting apart on this basis.
The watershed transform method has fast operation, can generate advantages such as zone closed and that be communicated with, thereby is widely used in remote sensing images are cut apart.But the shortcoming of watershed algorithm also clearly, and it is apparent in view that is exactly the over-segmentation phenomenon.This mainly is because the influence of picture noise, make the gradient of image have the lower zonule of many gray-scale values, these local minimum area correspondences the bottom, catchment basin of watershed algorithm, and the process of water burst at first is the bottom from the basin, so a large amount of irrelevant local minimum area in small, broken bits is to cause the basic reason of over-segmentation.One of ways of addressing this issue is exactly those significant local minimum area of mark, make follow-up watershed segmentation only at these marks the catchment basin carry out.Conventional mark generating algorithm can tentatively solve the over-segmentation problem, but still has problems such as optimum marking-threshold is difficult to determine.
Summary of the invention
Goal of the invention:, the purpose of this invention is to provide a kind of adaptive threshold that utilizes that is easy to definite optimum marking-threshold and cut apart the method for determining optimum marking image at the problem and shortage of above-mentioned existing existence.
Technical scheme: for achieving the above object, the technical solution used in the present invention is that a kind of adaptive threshold that utilizes is cut apart the method for determining optimum marking image, comprises following steps:
(1) calculates gradient image f to be marked (x, average E y) and variances sigma;
(2) being the center with the average E in the described step (1), is step-length with a σ, and a is the positive number of predefined, sets up the various combination H of average E and variances sigma
λ, H
λ=E+ λ a σ, in the formula, λ is the coefficient scope, is integer;
(3) with the various combination H of average E in the described step (2) and variances sigma
λTo gradient image f (x y) carries out Threshold Segmentation, obtain corresponding bianry image sequence f ' (x, y);
(4) successively the described bianry image sequence of statistic procedure (3) f ' (x, y) each the bianry image area in is greater than the number C of the sealing connected region of Minimum Area area
λ
(5) select number of regions C in the described step (4)
λThe bianry image when maximum image that serves as a mark.
(x y), can be the edge feature image itself that is used to carry out watershed transform to gradient image f to be marked in the described step (1), also can be external label information, as textural characteristics, semantic feature image.
The various combination H of average E and variances sigma in the described step (2)
λIn the swing of average E, step-length can adopt 0.1 times of variance, also can adopt greater or lesser numerical value as required; The scope of coefficient lambda swing is [10,10], also can adopt greater or lesser scope as required.
Threshold Segmentation in the described step (3) is calculated simple and fast, and (x y) can adopt the three-dimensional array storage to the bianry image sequence f ' of acquisition.
Number of regions statistics in the described step (4), must deduct the too small zone of area, can realize that minsize can value be 50 pixels, also can adopt greater or lesser value as required by a Minimum Area area parameters of prior setting minsize.
The marking image of determining in the described step (5) also must be eliminated the too small zone of area to reduce insignificant issue, still can realize by Minimum Area area parameters minsize.
Beneficial effect: the present invention only need set a Minimum Area area parameters simply, can generate the optimum marking image that the watershed divide changes according to the principle of maximum area adaptively, the algorithm simple and fast, and the Minimum Area area parameters generally need not to change after setting to image segmentation result and insensitive.
Description of drawings
Fig. 1 is a process flow diagram of the present invention;
Fig. 2 is a result schematic diagram of the present invention of utilizing computer programming language to realize;
Fig. 3 utilizes mark shown in Figure 2 to carry out the result schematic diagram that watershed transform is cut apart.
Embodiment
Below in conjunction with the drawings and specific embodiments, further illustrate the present invention, should understand these embodiment only is used to the present invention is described and is not used in and limit the scope of the invention, after having read the present invention, those skilled in the art all fall within the application's claims institute restricted portion to the modification of the various equivalent form of values of the present invention.
Basic ideas of the present invention are: design a kind of adaptive mobile threshold segmentation method, gradient image is cut apart, obtain optimum marking image automatically.Its main process is: average and the variance of at first calculating gradient image to be marked, making up average and variance then carries out simple threshold values to gradient image and cuts apart, obtain some bianry images, and the number of the sealing connected region in the statistics bianry image, the bianry image that number of regions is maximum promptly is used as marking image.
As shown in Figure 1, setting Minimum Area area parameters minsize is 50 pixels.
If gradient image to be marked be f (x, y), calculate its average E:
In the formula, N is the pixel number of image.Calculate its variances sigma:
With the average is the center, is step-length with 0.1 times of variance, sets up the various combination H of average and variance
λ:
H
λ=E+λ·0.1σ,λ=-10,-9,-8,...,10
In the formula, λ is a coefficient.
With H
λFor threshold value gradient image to be marked is carried out simple threshold values and cut apart, acquisition bianry image sequence f ' (x, y), value is 1 expressive notation zone, value is 0 expression background:
Area is greater than the number C of the sealing connected region of Minimum Area area parameters minsize in the statistics bianry image sequence
λ, select the number of regions C in the described step (4)
λThe bianry image when maximum image that serves as a mark.
The present invention can be applicable to that remote sensing images are handled automatically, the sensor information Intelligent Recognition.An example of the present invention realizes that on the PC platform through experimental verification, this adaptive threshold dividing method can obtain comparatively ideal marking image, and the accuracy that watershed transform is cut apart after the mark is higher.As shown in drawings, Fig. 2 is the result of the present invention who utilizes computer programming language to realize, the result shows main atural object such as house, vegetation, road all by mark exactly, and Fig. 3 utilizes this mark to carry out the result that watershed transform is cut apart, and the result shows that main atural object is all cut apart exactly.
Claims (4)
1. one kind is utilized adaptive threshold to cut apart the method for determining optimum marking image, it is characterized in that comprising following steps:
(1) calculates gradient image f to be marked (x, average E y) and variances sigma;
(2) being the center with the average E in the described step (1), is step-length with a σ, and a is the positive number of predefined, sets up the various combination H of average E and variances sigma
λ, H
λ=E+ λ a σ, in the formula, λ is the coefficient scope, is integer;
(3) with the various combination H of average E in the described step (2) and variances sigma
λTo gradient image f (x y) carries out Threshold Segmentation, obtain corresponding bianry image sequence f ' (x, y);
(4) successively the described bianry image sequence of statistic procedure (3) f ' (x, y) each the bianry image area in is greater than the number C of the sealing connected region of Minimum Area area
λ
(5) select number of regions C in the described step (4)
λThe bianry image when maximum image that serves as a mark.
2. cut apart the method for determining optimum marking image according to the described a kind of adaptive threshold that utilizes of claim 1, it is characterized in that: (x y) is edge feature image or the external label information that is used to carry out watershed transform to the gradient image f to be marked in the described step (1).
3. cut apart the method for determining optimum marking image according to the described a kind of adaptive threshold that utilizes of claim 1, it is characterized in that: the various combination H of average E and variances sigma in the described step (2)
λSwing at average E.
4. cut apart the method for determining optimum marking image according to the described a kind of adaptive threshold that utilizes of claim 1, it is characterized in that: (x y) is stored in the three-dimensional array bianry image sequence f ' in the described step (3).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2011100849024A CN102156984B (en) | 2011-04-06 | 2011-04-06 | Method for determining optimal mark image by adaptive threshold segmentation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2011100849024A CN102156984B (en) | 2011-04-06 | 2011-04-06 | Method for determining optimal mark image by adaptive threshold segmentation |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102156984A true CN102156984A (en) | 2011-08-17 |
CN102156984B CN102156984B (en) | 2013-03-06 |
Family
ID=44438465
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2011100849024A Expired - Fee Related CN102156984B (en) | 2011-04-06 | 2011-04-06 | Method for determining optimal mark image by adaptive threshold segmentation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102156984B (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102622598A (en) * | 2012-01-13 | 2012-08-01 | 西安电子科技大学 | SAR (Synthesized Aperture Radar) image target detection method based on zone markers and grey statistics |
CN102799183A (en) * | 2012-08-21 | 2012-11-28 | 上海港吉电气有限公司 | Mobile machinery vision anti-collision protection system for bulk yard and anti-collision method |
CN103778624A (en) * | 2013-12-20 | 2014-05-07 | 中原工学院 | Fabric defect detection method based on optical threshold segmentation |
CN104899853A (en) * | 2014-03-04 | 2015-09-09 | 腾讯科技(深圳)有限公司 | Image region dividing method and device |
CN104050670B (en) * | 2014-06-24 | 2016-08-17 | 广州中医药大学 | In conjunction with simple mutual and the complex background leaf image dividing method in labelling watershed |
CN107808385A (en) * | 2017-11-22 | 2018-03-16 | 新疆大学 | Coloured image watershed segmentation methods based on power-law distribution |
CN107991532A (en) * | 2017-11-21 | 2018-05-04 | 厦门理工学院 | A kind of harmonic data thresholds merging method based under a variety of methods of operation |
CN109191456A (en) * | 2018-09-19 | 2019-01-11 | 电子科技大学 | Lung CT image processing method and system based on two-dimentional S-transformation |
CN109377507A (en) * | 2018-09-19 | 2019-02-22 | 河海大学 | A method of the high-spectrum remote sensing segmentation based on curve of spectrum spectral distance |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050129274A1 (en) * | 2001-05-30 | 2005-06-16 | Farmer Michael E. | Motion-based segmentor detecting vehicle occupants using optical flow method to remove effects of illumination |
CN1885346A (en) * | 2006-06-01 | 2006-12-27 | 电子科技大学 | Detection method for moving target in infrared image sequence under complex background |
CN101170641A (en) * | 2007-12-05 | 2008-04-30 | 北京航空航天大学 | A method for image edge detection based on threshold sectioning |
CN101556693A (en) * | 2009-03-30 | 2009-10-14 | 西安电子科技大学 | Division method for extracted watershed SAR image with threshold method and marking |
CN101789080A (en) * | 2010-01-21 | 2010-07-28 | 上海交通大学 | Detection method for vehicle license plate real-time positioning character segmentation |
CN101853333A (en) * | 2010-05-26 | 2010-10-06 | 中国科学院遥感应用研究所 | Method for picking marks in medical robot navigation positioning images |
CN101923707A (en) * | 2009-07-23 | 2010-12-22 | 北京师范大学 | Watershed algorithm-based high spatial resolution multi-spectral remote sensing image segmentation method |
-
2011
- 2011-04-06 CN CN2011100849024A patent/CN102156984B/en not_active Expired - Fee Related
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050129274A1 (en) * | 2001-05-30 | 2005-06-16 | Farmer Michael E. | Motion-based segmentor detecting vehicle occupants using optical flow method to remove effects of illumination |
CN1885346A (en) * | 2006-06-01 | 2006-12-27 | 电子科技大学 | Detection method for moving target in infrared image sequence under complex background |
CN101170641A (en) * | 2007-12-05 | 2008-04-30 | 北京航空航天大学 | A method for image edge detection based on threshold sectioning |
CN101556693A (en) * | 2009-03-30 | 2009-10-14 | 西安电子科技大学 | Division method for extracted watershed SAR image with threshold method and marking |
CN101923707A (en) * | 2009-07-23 | 2010-12-22 | 北京师范大学 | Watershed algorithm-based high spatial resolution multi-spectral remote sensing image segmentation method |
CN101789080A (en) * | 2010-01-21 | 2010-07-28 | 上海交通大学 | Detection method for vehicle license plate real-time positioning character segmentation |
CN101853333A (en) * | 2010-05-26 | 2010-10-06 | 中国科学院遥感应用研究所 | Method for picking marks in medical robot navigation positioning images |
Non-Patent Citations (4)
Title |
---|
《Urban Remote Sensing Event, 2009 Joint》 20090520 Pengfeng Xiao等 Multispectral remotely sensed imagery segmentation based on first fundamental form , * |
《中国优秀硕士学位论文全文数据库 》 20060914 孔凡壮 遥感图像分割算法的研究与应用 , * |
《中国期刊全文数据库》 20100228 尹高飞等 基于改进标记的高分辨率遥感图像分水岭分割方法 , * |
《计算机应用》 20090830 秦剑等 基于梯度图像分割新方法 , * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102622598A (en) * | 2012-01-13 | 2012-08-01 | 西安电子科技大学 | SAR (Synthesized Aperture Radar) image target detection method based on zone markers and grey statistics |
CN102799183A (en) * | 2012-08-21 | 2012-11-28 | 上海港吉电气有限公司 | Mobile machinery vision anti-collision protection system for bulk yard and anti-collision method |
CN102799183B (en) * | 2012-08-21 | 2015-03-25 | 上海港吉电气有限公司 | Mobile machinery vision anti-collision protection system for bulk yard and anti-collision method |
CN103778624A (en) * | 2013-12-20 | 2014-05-07 | 中原工学院 | Fabric defect detection method based on optical threshold segmentation |
CN104899853B (en) * | 2014-03-04 | 2019-12-06 | 腾讯科技(深圳)有限公司 | Image area dividing method and device |
CN104899853A (en) * | 2014-03-04 | 2015-09-09 | 腾讯科技(深圳)有限公司 | Image region dividing method and device |
CN104050670B (en) * | 2014-06-24 | 2016-08-17 | 广州中医药大学 | In conjunction with simple mutual and the complex background leaf image dividing method in labelling watershed |
CN107991532A (en) * | 2017-11-21 | 2018-05-04 | 厦门理工学院 | A kind of harmonic data thresholds merging method based under a variety of methods of operation |
CN107808385A (en) * | 2017-11-22 | 2018-03-16 | 新疆大学 | Coloured image watershed segmentation methods based on power-law distribution |
CN107808385B (en) * | 2017-11-22 | 2021-05-25 | 新疆大学 | Color image watershed segmentation method based on power law distribution |
CN109377507A (en) * | 2018-09-19 | 2019-02-22 | 河海大学 | A method of the high-spectrum remote sensing segmentation based on curve of spectrum spectral distance |
CN109191456A (en) * | 2018-09-19 | 2019-01-11 | 电子科技大学 | Lung CT image processing method and system based on two-dimentional S-transformation |
CN109377507B (en) * | 2018-09-19 | 2022-04-08 | 河海大学 | Hyperspectral remote sensing image segmentation method based on spectral curve spectral distance |
Also Published As
Publication number | Publication date |
---|---|
CN102156984B (en) | 2013-03-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102156984B (en) | Method for determining optimal mark image by adaptive threshold segmentation | |
CN103049763B (en) | Context-constraint-based target identification method | |
CN102855487B (en) | Method for automatically extracting newly added construction land change image spot of high-resolution remote sensing image | |
CN103793907B (en) | Water body information extracting method and device | |
CN103383733A (en) | Lane video detection method based on half-machine study | |
CN102622598B (en) | SAR (Synthesized Aperture Radar) image target detection method based on zone markers and grey statistics | |
CN104123561A (en) | Spatial gravity model based fuzzy c-means remote sensing image automatic classification method | |
CN101661497A (en) | Remote sensing land use change detection method and system thereof | |
CN102842130B (en) | Method for detecting buildings and extracting number information from synthetic aperture radar image | |
CN106683112A (en) | High-resolution image-based road region building change extraction method | |
CN102867183B (en) | Method and device for detecting littered objects of vehicle and intelligent traffic monitoring system | |
CN103279951A (en) | Object-oriented remote sensing image building and shade extraction method of remote sensing image building | |
CN107945200A (en) | Image binaryzation dividing method | |
CN104361351A (en) | Synthetic aperture radar (SAR) image classification method on basis of range statistics similarity | |
CN107330422A (en) | A kind of method for carrying out mima type microrelief classification to semiarid zone based on high accuracy number elevation model | |
CN103761522B (en) | SAR image river channel extracting method based on minimum circumscribed rectangle window river channel segmentation model | |
CN104951799A (en) | SAR remote-sensing image oil spilling detection and identification method | |
McCormack et al. | A methodology for mapping annual flood extent using multi-temporal Sentinel-1 imagery | |
CN107301649B (en) | Regional merged SAR image coastline detection algorithm based on superpixels | |
Wan et al. | Automatic extraction of flood inundation areas from SAR images: A case study of Jilin, China during the 2017 flood disaster | |
CN105160355A (en) | Remote sensing image change detection method based on region correlation and visual words | |
Zhang et al. | Impervious surface extraction from high-resolution satellite image using pixel-and object-based hybrid analysis | |
CN104077609A (en) | Saliency detection method based on conditional random field | |
CN107507202A (en) | A kind of vegetation rotary island towards high-resolution remote sensing image automates extracting method | |
CN102938069A (en) | Pure and mixed pixel automatic classification method based on information entropy |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130306 |