CN106778774A - A kind of high-resolution remote sensing image man-made features profile testing method - Google Patents
A kind of high-resolution remote sensing image man-made features profile testing method Download PDFInfo
- Publication number
- CN106778774A CN106778774A CN201611052418.2A CN201611052418A CN106778774A CN 106778774 A CN106778774 A CN 106778774A CN 201611052418 A CN201611052418 A CN 201611052418A CN 106778774 A CN106778774 A CN 106778774A
- Authority
- CN
- China
- Prior art keywords
- remote sensing
- sensing image
- man
- template
- curve
- 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
- 238000012360 testing method Methods 0.000 title claims abstract description 13
- 238000000034 method Methods 0.000 claims abstract description 15
- 238000012795 verification Methods 0.000 claims abstract description 7
- 238000001514 detection method Methods 0.000 claims abstract description 5
- 238000005070 sampling Methods 0.000 claims abstract description 5
- 238000009499 grossing Methods 0.000 claims description 5
- 238000009826 distribution Methods 0.000 claims description 4
- 230000000750 progressive effect Effects 0.000 claims description 3
- 238000003708 edge detection Methods 0.000 claims description 2
- 238000003672 processing method Methods 0.000 claims description 2
- 238000007670 refining Methods 0.000 claims description 2
- AWSBQWZZLBPUQH-UHFFFAOYSA-N mdat Chemical compound C1=C2CC(N)CCC2=CC2=C1OCO2 AWSBQWZZLBPUQH-UHFFFAOYSA-N 0.000 claims 1
- 238000012216 screening Methods 0.000 claims 1
- 238000000605 extraction Methods 0.000 abstract description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 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 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of high-resolution remote sensing image man-made features profile testing method.Comprise the following steps:Step 1, first to input remote sensing imageIDown-sampling is carried out, is obtainedI s , willI s Convolution is carried out with gaussian kernel function, image is obtainedI sc , willI s SubtractI sc , obtain differential imageD;Step 2, with gray-scale statistical comparison window to differential imageDRim detection is carried out, candidate's collection of curves is obtained;Step 3, is verified and is screened using the verification method based on standard error function to candidate's collection of curves, obtains selected collection of curves;Step 4, is smoothed to selected collection of curves, exports high-resolution remote sensing image man-made features contour detecting result.The profile information of the man-made features in remote sensing image is excavated to greatest extent, can apply to the accurate extraction of the man-made features such as building, road.
Description
Technical field
The present invention relates to a kind of remote sensing image process field, specifically a kind of high-resolution remote sensing image man-made features wheel
Wide detection method.
Background technology
Remote sensing image is the concentrated expression of the spectrum and geometric properties of ground target and phenomenon on image, not exclusively body
The pixel unit set of existing brightness and chromaticity, and with complicated spectral signature and architectural feature.High-resolution
Remote sensing image is a kind of very special digital picture, and its complexity is far above normal image.High-resolution remote sensing image
Contour detecting is Analysis of Remote Sensing Information and the basis for understanding, is a challenge of digital image processing field.At present, many scholars
Constantly propose from the point of view of relative profile etection theory and method, but the achievement that oneself is delivered, these methods there is problems:
(1) because remote sensing image is influenceed by sensor, position of sun etc. are multifactor in imaging process, showed in image
Target scalar information is not only incomplete, and contains much noise.Edge and noise all show as the larger of gray scale in spatial domain
Mutation, is then reflected as being both high fdrequency component in frequency domain, and the result of contour detecting is performed thereon usually noise as edge
Put and detect, make real profile due to not being detected by noise jamming.
(2) remote sensing image information amount is enriched, and compared with general pattern, the content that it is included is more more than common image, no
Influenced each other and interference with information between atural object, not notable or fuzzy border makes that objective contour interested must be extracted and becomes
Obtain extremely difficult.
The content of the invention
The invention provides a kind of high-resolution remote sensing image man-made features profile testing method, with high-definition remote sensing shadow
As data source, differential image being done on metric space, it is possible to reduce interference of the noise to extraction profile, protect to greatest extent
The marginal information of picture picture, the amount of calculation of method is small, and adaptive ability is strong, output result reliability.
To realize that the technical scheme that target of the invention is used is:Method is comprised the following steps:
Step 1:Down-sampling first is carried out to input remote sensing image I, I is obtaineds, by IsConvolution is carried out with gaussian kernel function, shadow is obtained
As Isc, by IsSubtract Isc, obtain differential image D;
Step 2:Rim detection is carried out to the differential image D of step 1 with gray-scale statistical comparison window GSC, candidate's curve set is obtained
Close Scont;
Step 3:In view of the rule of the gray-scale statistical comparison window GSC Normal Distributions of step 2, using based on standard error
Candidate collection of curves S of the verification method of function to step 2contIn all curves verified and screened, obtain selected song
Line set Fcont;
Step 4:To the selected collection of curves F of step 3contIn all curves be smoothed respectively, export high-resolution
Remote sensing image man-made features contour detecting result.
The standard deviation of described gaussian kernel functionWherein k is balance coefficient, for balance outline
Sawtooth and fog-level, s are the ratio of adjacent yardstick in metric space.
Described gray-scale statistical comparison window GSC is moveable, and move mode is in the form of progressive scan.
Described gray-scale statistical comparison window GSC is by bright template MLWith dark template MDAdjacent window, it is wide by side
Degree w determines the area size that window is included, and calculates taking for candidate's curve in gray-scale statistical comparison window GSC according to below equation
Value v:
In formula (1), M and N is respectively bright template MLWith dark template MDIn the pixel number that includes,
LmAnd DnRespectively bright template MLIn m-th pixel and dark template MDIn nth pixel, as gray-scale statistical comparison window GSC
Without boundary information, homogenous region, L are coverednIt is constantly equal to Dn, i.e. δ (Lm,DnDuring)=0.5,When gray-scale statistical compares
Window GSC includes significant boundary information, LnPerseverance is less than Dn, i.e. δ (Lm,DnDuring)=1, v=M × N;As the value v of candidate's curve
More than threshold value T1When, using candidate's curve as refinement curve.
Described lateral width w refers to bright template MLWith dark template MDMinimum enclosed rectangle short side, using two valuesWherein,Ensure bright template MLWith dark template MDThe pixel of a line is at least covered, is contrasted for detecting
Degree profile high.Ensure bright template MLWith dark template MDThe pixel of three rows is at least covered, noise, side are contained for detecting
The fuzzy profile of edge.
The described verification method based on standard error function is using the standard error value of below equation calculated curve
SER:
In formula (2), upper limit of integralIn h expression formulasWhen carrying
The standard error value SER for refining curve is less than threshold value T2When, as selected curve.
Described smoothing processing method uses Bezier exponential smoothing.
Described candidate's curve refers to the result for directly being detected with edge detection algorithm and being obtained, wherein comprising too short or invalid
Edge, it is necessary to subsequent step further checking and screen.
The beneficial effects of the invention are as follows:The profile information of the man-made features in remote sensing image is excavated to greatest extent, can be with
It is applied to the accurate extraction of the man-made features such as building, road.
Brief description of the drawings
Fig. 1 is overall process flow figure of the invention.
Fig. 2 is gray-scale statistical comparison window schematic diagram of the invention.
Specific embodiment
Describe specific embodiment of the invention in detail below in conjunction with the accompanying drawings.
Fig. 1 is overall process flow figure of the invention:
In step 101, it is Quickbird panchromatic images to be input into pending high-resolution remote sensing image I, and spatial resolution is
0.81 meter.
In step 102, down-sampling is carried out to high-resolution remote sensing image I and obtains Is, wherein, the sampling interval is 2 pixels.
In step 103, by IsConvolution is carried out with gaussian kernel function, image I is obtainedsc, by IsSubtract Isc, obtain differential image
D, wherein, the standard deviation of gaussian kernel functionWherein k is balance coefficient, for the sawtooth and mould of balance outline
Paste degree, s is the ratio of adjacent yardstick in metric space.Through repetition test, k is set as 1.5, s is set as 0.8.
In step 104, rim detection is carried out to the differential image D in step 103 with gray-scale statistical comparison window GSC, obtained
To candidate's collection of curves Scont, wherein, GSC be it is moveable, move mode using progressive scan in the form of, based on high-resolution
The man-made features of remote sensing image follow the principles distribution feature and consider subsequent step amount of calculation and profile relevant information retain
Trade-off problem between degree, is tested with different sensors and various sizes of remote sensing image, and discovery sets moving step pitch P
5 best results are put, gray-scale statistical comparison window GSC is by bright template MLWith dark template MDAdjacent window, it is wide by side
Degree w determines the area size that window is included, and calculates taking for candidate's curve in gray-scale statistical comparison window GSC according to below equation
Value v:
Wherein, M and N are respectively bright template MLWith dark template MDIn the pixel number that includes,Lm
And DnRespectively bright template MLIn m-th pixel and dark template MDIn nth pixel, when gray-scale statistical comparison window GSC not
Containing boundary information, homogenous region is covered, without boundary information, LnIt is constantly equal to Dn, i.e. δ (Lm,DnDuring)=0.5),When
Gray-scale statistical comparison window GSC includes significant boundary information, LnPerseverance is less than Dn, i.e. δ (Lm,DnDuring)=1), v=M × N;When
The value v of candidate's curve is more than threshold value T1When, using candidate's curve as refinement curve.Lateral width w uses two valuesWherein,Ensure bright template MLWith dark template MDThe pixel of a line is at least covered, for detecting contrast
Profile high;Ensure bright template MLWith dark template MDThe pixel of three rows is at least covered, noise, edge are contained for detecting
Fuzzy profile.
In step 105, it is contemplated that the rule of the gray-scale statistical comparison window GSC Normal Distributions of step 104, using base
In standard error function verification method to candidate's collection of curves S of step 4contIn all curves verified and screened,
Obtain selected collection of curves Fcont, wherein, the verification method based on standard error function is using below equation calculated curve
Standard error value SER:
Wherein, upper limit of integralIn h expression formulasWork as refinement
The standard error value SER of curve is less than threshold value T2When, as selected curve.
In step 106, to the selected collection of curves F of step 105contIn all curves utilize Bezier exponential smoothing
It is smoothed respectively.
In step 107, profile is exported.
Through experiment, the threshold value T in step 1041With threshold value T in step 1062It is respectively set to 0.75 and 0.6.
Fig. 2 is gray-scale statistical comparison window schematic diagram of the invention.
200 is the part of differential image D, and 201 is one section of contour curve, and 202 is bright template ML, 203 is dark template MD, 204
It is lateral width w.
Claims (8)
1. a kind of high-resolution remote sensing image man-made features profile testing method, it is characterised in that comprise the following steps:
Step 1:Down-sampling first is carried out to input remote sensing image I, I is obtaineds, by IsConvolution is carried out with gaussian kernel function, image is obtained
Isc, by IsSubtract Isc, obtain differential image D;
Step 2:Rim detection is carried out to the differential image D of step 1 with gray-scale statistical comparison window GSC, candidate's curve set is obtained
Close Scont;
Step 3:In view of the rule of the gray-scale statistical comparison window GSC Normal Distributions of step 2, using based on standard error
Candidate collection of curves S of the verification method of function to step 2contIn all curves verified and screened, obtain selected song
Line set Fcont;
Step 4:To the selected collection of curves F of step 3contIn all curves be smoothed respectively, output high-resolution it is distant
Sense image man-made features contour detecting result.
2. a kind of high-resolution remote sensing image man-made features profile testing method according to claim 1, it is characterised in that
Described gaussian kernel function, its standard deviationWherein k is balance coefficient, for the sawtooth and mould of balance outline
Paste degree, s is the ratio of adjacent yardstick in metric space.
3. a kind of high-resolution remote sensing image man-made features profile testing method according to claim 1, it is characterised in that
Gray-scale statistical comparison window GSC is moveable, and move mode is in the form of progressive scan.
4. a kind of high-resolution remote sensing image man-made features profile testing method according to claim 1, it is characterised in that
Gray-scale statistical comparison window GSC is by bright template MLWith dark template MDAdjacent window, window bag is determined by lateral width w
The area size for containing, and the value v of candidate's curve in gray-scale statistical comparison window GSC is calculated according to below equation:
In formula (1), M and N is respectively bright template MLWith dark template MDIn the pixel number that includes,
LmAnd DnRespectively bright template MLIn m-th pixel and dark template MDIn nth pixel;As gray-scale statistical comparison window GSC
Without boundary information, homogenous region, L are coverednIt is constantly equal to Dn, i.e. δ (Lm,DnDuring)=0.5),When gray-scale statistical ratio
Significant boundary information, L are included compared with window GSCnPerseverance is less than Dn, i.e. δ (Lm,DnDuring)=1, v=M × N;When taking for candidate's curve
Value v is more than threshold value T1When, using candidate's curve as refinement curve.
5. a kind of high-resolution remote sensing image man-made features profile testing method according to claim 4, it is characterised in that
Lateral width w refers to bright template MLWith dark template MDMinimum enclosed rectangle short side, using two valuesIts
In,Ensure bright template MLWith dark template MDAt least cover the pixel of a line, the profile high for detecting contrast;Ensure bright template MLWith dark template MDThe pixel of three rows is at least covered, for detecting the wheel containing noise, edge blurry
It is wide.
6. a kind of high-resolution remote sensing image man-made features profile testing method according to claim 1, it is characterised in that
Verification method based on standard error function is using the standard error value SER of below equation calculated curve:
In formula (2), upper limit of integralIn h expression formulasWhen carrying
The standard error value SER for refining curve is less than threshold value T2When, as selected curve.
7. a kind of high-resolution remote sensing image man-made features profile testing method according to claim 1, it is characterised in that
Smoothing processing method uses Bezier exponential smoothing.
8. a kind of high-resolution remote sensing image man-made features profile testing method according to claim 1, it is characterised in that
Candidate's curve refers to directly to detect the result that obtains with edge detection algorithm, wherein comprising too short or invalid edge, it is necessary to after
Further checking and the screening of continuous step.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611052418.2A CN106778774B (en) | 2016-11-25 | 2016-11-25 | High-resolution remote sensing image artificial ground feature contour detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611052418.2A CN106778774B (en) | 2016-11-25 | 2016-11-25 | High-resolution remote sensing image artificial ground feature contour detection method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106778774A true CN106778774A (en) | 2017-05-31 |
CN106778774B CN106778774B (en) | 2020-04-03 |
Family
ID=58912363
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611052418.2A Expired - Fee Related CN106778774B (en) | 2016-11-25 | 2016-11-25 | High-resolution remote sensing image artificial ground feature contour detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106778774B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111047616A (en) * | 2019-12-10 | 2020-04-21 | 中国人民解放军陆军勤务学院 | Remote sensing image landslide target constraint active contour feature extraction method |
CN113313724A (en) * | 2021-05-27 | 2021-08-27 | 深圳企业云科技股份有限公司 | Line detection processing method for resisting resampling of mobile phone camera |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101777181A (en) * | 2010-01-15 | 2010-07-14 | 西安电子科技大学 | Ridgelet bi-frame system-based SAR image airfield runway extraction method |
CN102253184A (en) * | 2011-06-29 | 2011-11-23 | 南京信息工程大学 | Remote sensing inversion method for land surface evapotranspiration of arid and semi-arid regions |
CN103208011A (en) * | 2013-05-05 | 2013-07-17 | 西安电子科技大学 | Hyperspectral image space-spectral domain classification method based on mean value drifting and group sparse coding |
CN103559493A (en) * | 2013-10-21 | 2014-02-05 | 中国农业大学 | Method for extracting linear ground objects |
CN103839265A (en) * | 2014-02-26 | 2014-06-04 | 西安电子科技大学 | SAR image registration method based on SIFT and normalized mutual information |
CN103886597A (en) * | 2014-03-24 | 2014-06-25 | 武汉力成伟业科技有限公司 | Circle detection method based on edge detection and fitted curve clustering |
CN104794495A (en) * | 2015-05-04 | 2015-07-22 | 福建师范大学 | Large-format remote-sensing image region classifying method based on straight line statistical characteristics |
CN105740873A (en) * | 2016-02-01 | 2016-07-06 | 福建师范大学 | Artificial feature straight line contour detection method of remote-sensing image |
-
2016
- 2016-11-25 CN CN201611052418.2A patent/CN106778774B/en not_active Expired - Fee Related
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101777181A (en) * | 2010-01-15 | 2010-07-14 | 西安电子科技大学 | Ridgelet bi-frame system-based SAR image airfield runway extraction method |
CN102253184A (en) * | 2011-06-29 | 2011-11-23 | 南京信息工程大学 | Remote sensing inversion method for land surface evapotranspiration of arid and semi-arid regions |
CN103208011A (en) * | 2013-05-05 | 2013-07-17 | 西安电子科技大学 | Hyperspectral image space-spectral domain classification method based on mean value drifting and group sparse coding |
CN103559493A (en) * | 2013-10-21 | 2014-02-05 | 中国农业大学 | Method for extracting linear ground objects |
CN103839265A (en) * | 2014-02-26 | 2014-06-04 | 西安电子科技大学 | SAR image registration method based on SIFT and normalized mutual information |
CN103886597A (en) * | 2014-03-24 | 2014-06-25 | 武汉力成伟业科技有限公司 | Circle detection method based on edge detection and fitted curve clustering |
CN104794495A (en) * | 2015-05-04 | 2015-07-22 | 福建师范大学 | Large-format remote-sensing image region classifying method based on straight line statistical characteristics |
CN105740873A (en) * | 2016-02-01 | 2016-07-06 | 福建师范大学 | Artificial feature straight line contour detection method of remote-sensing image |
Non-Patent Citations (7)
Title |
---|
FENGHUA HUANG等: "ICA-ASIFT-BASED multi-temporal matching of high-resolution remote sensing urban images", 《CYBERNETICS AND INFORMATION TECHNOLOGIES》 * |
H.BANCH等: "methods and example for remote sensing data assimilation in land surface process modeling", 《IEEE TRANSACTIONS ON GEOSCIENCE AND REOMOTE SENSING》 * |
傅立光等: "基于灰度的图像边缘检测与匹配算法的研究", 《电脑知识与技术》 * |
徐平等: "基于边缘像元投影的微小轴承亚像素边缘检测", 《仪器仪表学报》 * |
景雨: "海上溢油遥感图像的边缘检测算法研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
曹风云等: "自适应多方向灰度形态学图像边缘检测算法", 《光学技术》 * |
殷润民等: "一种基于灰度差统计的边缘检测方法", 《计算机工程》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111047616A (en) * | 2019-12-10 | 2020-04-21 | 中国人民解放军陆军勤务学院 | Remote sensing image landslide target constraint active contour feature extraction method |
CN113313724A (en) * | 2021-05-27 | 2021-08-27 | 深圳企业云科技股份有限公司 | Line detection processing method for resisting resampling of mobile phone camera |
CN113313724B (en) * | 2021-05-27 | 2022-04-08 | 深圳企业云科技股份有限公司 | Line detection processing method for resisting resampling of mobile phone camera |
Also Published As
Publication number | Publication date |
---|---|
CN106778774B (en) | 2020-04-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101727662B (en) | SAR image nonlocal mean value speckle filtering method | |
CN101539629B (en) | Remote sensing image change detection method based on multi-feature evidence integration and structure similarity | |
CN102096921B (en) | SAR (Synthetic Aperture Radar) image change detection method based on neighborhood logarithm specific value and anisotropic diffusion | |
CN107301661A (en) | High-resolution remote sensing image method for registering based on edge point feature | |
CN101561932B (en) | Method and device for detecting real-time movement target under dynamic and complicated background | |
CN112307901B (en) | SAR and optical image fusion method and system for landslide detection | |
CN103759676A (en) | Non-contact type workpiece surface roughness detecting method | |
CN102819740B (en) | A kind of Single Infrared Image Frame Dim targets detection and localization method | |
CN102360503B (en) | SAR (Specific Absorption Rate) image change detection method based on space approach degree and pixel similarity | |
CN104867150A (en) | Wave band correction change detection method of remote sensing image fuzzy clustering and system thereof | |
CN102842120B (en) | Image blurring degree detection method based on supercomplex wavelet phase measurement | |
Raffei et al. | A low lighting or contrast ratio visible iris recognition using iso-contrast limited adaptive histogram equalization | |
CN107590816B (en) | Water body information fitting method based on remote sensing image | |
CN107341790A (en) | A kind of image processing method of environment cleanliness detection | |
CN107403433A (en) | A kind of complicated cloud infrared small target in background detection method | |
CN107133922A (en) | A kind of silicon chip method of counting based on machine vision and image procossing | |
CN106023134A (en) | Automatic grain boundary extraction method for steel grain | |
CN111007039A (en) | Automatic extraction method and system for sub-pixel level water body of medium-low resolution remote sensing image | |
CN109978940A (en) | A kind of SAB air bag size vision measuring method | |
CN103914829B (en) | Method for detecting edge of noisy image | |
CN107229910A (en) | A kind of remote sensing images icing lake detection method and its system | |
CN108335310B (en) | Portable grain shape and granularity detection method and system | |
CN104933719B (en) | One kind integration segment spacing is from detection image edge method | |
CN106778774A (en) | A kind of high-resolution remote sensing image man-made features profile testing method | |
CN103035185A (en) | Method for brightness correction of defective pixels of digital monochrome image |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200403 |