CN102254319B - Method for carrying out change detection on multi-level segmented remote sensing image - Google Patents

Method for carrying out change detection on multi-level segmented remote sensing image Download PDF

Info

Publication number
CN102254319B
CN102254319B CN2011100978078A CN201110097807A CN102254319B CN 102254319 B CN102254319 B CN 102254319B CN 2011100978078 A CN2011100978078 A CN 2011100978078A CN 201110097807 A CN201110097807 A CN 201110097807A CN 102254319 B CN102254319 B CN 102254319B
Authority
CN
China
Prior art keywords
image
remote sensing
spot
images
phase
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.)
Active
Application number
CN2011100978078A
Other languages
Chinese (zh)
Other versions
CN102254319A (en
Inventor
徐成华
杨金锋
马鹏飞
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhongke Star Map Co., Ltd.
Original Assignee
ZHONGKE JIUDU (BEIJING) SPATIAL INFORMATION TECHNOLOGY Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by ZHONGKE JIUDU (BEIJING) SPATIAL INFORMATION TECHNOLOGY Co Ltd filed Critical ZHONGKE JIUDU (BEIJING) SPATIAL INFORMATION TECHNOLOGY Co Ltd
Priority to CN2011100978078A priority Critical patent/CN102254319B/en
Publication of CN102254319A publication Critical patent/CN102254319A/en
Application granted granted Critical
Publication of CN102254319B publication Critical patent/CN102254319B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention discloses a method for carrying out change detection on a multi-level segmented remote sensing image, and is characterized in that the method comprises nine steps, and in the method, after carrying out denoising and smoothing and the like on the image, through merging patch objects, a change detection result of the remote sensing image is obtained finally. In the method, the patch objects subjected to segmentation are compared by using a direct comparison method firstly, if the patch objects are changed, the patch objects are subjected to secondary segmentation by using an algorithm, then characteristic values are extracted and compared. By using the method disclosed by the invention, the pseudo change information can be removed effectively, and the change detection accuracy can be improved; and compared with the existing methods, by using the method disclosed by the invention, a more detailed and complex change result can be obtained, thereby realizing the change detection on sub-patch objects.

Description

A kind of remote sensing image variation detection method of multilayer division
Technical field
The present invention relates to a kind of method of Remote Sensing Imagery Change Detection, relate in particular to a kind of remote sensing image variation detection method of multilayer division, belong to technical field of image processing.
Background technology
Remote Sensing Imagery Change Detection refers to the remote sensing images of areal different times are analyzed, therefrom detect the information that atural object changes in time, these Information Availabilities are in geosystem information updating, Monitoring of Resource and Environment, target dynamic supervision and military attack recruitment evaluation etc.
Present detection technique is divided into:
1. based on the change detecting method of pixel
Variation based on pixel detects, it realizes mainly depending on the change of spectral reflectance value in the remote sensing images that caused by feature changes, causing and these changes may be real change due to atural object, may be also due to sensing station, solar incident angle, cloud, the difference of the non-atural object factor such as mist causes, therefore, need to carry out registration to image relatively, directly carry out pixel after registration and subtract each other, obtain differential image.
Suppose that image is input as the not remote sensing images X (t1) of phase simultaneously of two width, X (t2), and done registration process.
Differential image:
Xchange=|X(t1)-X(t2)|
2. OO change detection techniques
(1) compare after the classification:
The method that compares after classification: at first respectively two phase images are carried out image segmentation, then image segmentation result is combined the classification again (being OO classification) of carrying out image with spectral classification result separately, during then to two, the classified image of phase compares, obtain changing testing result, as shown in Figure 1.
(2) directly relatively:
Method directly relatively: at first select the not remote sensing image of phase simultaneously of two scapes, comprise multispectral data and panchromatic wave-band data, the data of selecting are carried out pre-service, all wave bands of pretreated two phase remote sensing images are formed a scape image, this image is cut apart, set up the not mapping relations one by one of the object of phase remote sensing image simultaneously of two scapes; Based on segmentation result, the feature set that builds each object is described, and utilizes these eigenwerts to carry out algebraic manipulation, sets a change threshold, in conjunction with result of calculation, finally obtains modified-image, and its overall plan as shown in Figure 2.
The prior art shortcoming:
(1) variation based on pixel detects, and a lot of pseudo-change informations are arranged, and changes accuracy of detection not high, is not suitable for high-resolution remote sensing image.
(2) OO change detecting method adopts direct comparison method, namely after image segmentation, directly extracts the variation detection that eigenwert is carried out figure spot object, is difficult to detect the situation of change of inferior figure spot object (cut zone is inner).Relative method after the employing classification namely after image segmentation, is classified to scheme the spot object, thereby obtains the situation of change of all kinds of key elements.The method is subject to the degree of accuracy of classification results, has the phenomenon of error in classification accumulation.
Summary of the invention
The object of the invention is to overcome the deficiency of above-mentioned existing Remote Sensing Imagery Change Detection technology, propose a kind of OO figure spot change detecting method.
The technical solution adopted in the present invention is: a kind of remote sensing image variation detection method of multilayer division, and detection is divided into following steps:
1. choose the not remote sensing image of phase simultaneously of regional two scapes to be detected, 01 phase and 02 phase;
2. choose 01 phase remote sensing image as reference images, 02 phase remote sensing image is as detecting image;
3. the remote sensing image of selecting is carried out the image pre-service, comprise noise reduction process, radiant correction and Image registration;
(1) noise reduction process: the noise of image is present in the HFS of image, and wants to reject the noise in high-frequency signal, the low-and high-frequency of image need to be separated, and utilizes small echo to change and carries out Image Denoising by Use, rejects the irrelevant information that is mingled in image;
(2) radiant correction: adopt the statistical regression method, take reference images as master image, carry out radiant correction to detecting image, the method found same place that atural object Change of types and the stable ground object sample point of spectral quality do not occur in the phase image at 2 o'clock, utilize the linear dependence relation of its gray-scale value to proofread and correct, do not need the radiation calibration of sensor and the clutter reflections rate data that relevant atmospheric parameter just can obtain regularization;
(3) Image registration: adopt affine invariant feature extraction algorithm to realize Image registration, at first method builds the SIFT descriptor with affine unchangeability, and utilizes this descriptor that the reference mark of extracting is mated, and obtains transformation parameter and realizes Image registration;
4. the reference images of step 3 being selected is carried out OO multi-scale division: adopt the partitioning algorithm of mean shift (Mean-shift) that reference images is cut apart for the first time, obtain figure spot object and the boundary information of reference images, make up these boundary informations and be converted to and cut apart polar plot;
5. the polar plot stack cut apart with step 4 gained detects image, and this image is cut apart, and obtains and reference images figure spot object one to one;
6. based on above-mentioned segmentation result, obtain each figure spot object, extract a plurality of eigenwerts of object, mainly comprise gray level co-occurrence matrixes-entropy, average and density;
The present invention comes computation of mean values, wherein C according to formula (1) iThe pixel value of i pixel in the expression imaged object; N represents the pixel number of imaged object;
C ‾ = 1 n Σ i = 1 n C i - - - ( 1 )
Utilize formula (2) to calculate entropy, wherein (i, j) individual element in p (i, j) expression gray level co-occurrence matrixes; G represents the exponent number of gray level co-occurrence matrixes;
f = Σ i = 1 G Σ j = 1 G P ( i , j ) log [ P ( i , j ) ] - - - ( 2 )
Utilize formula (3) to come the bulk density value, wherein s represents the area of imaged object, and Var (x) expression is asked variance to the x coordinate of all pixels in image, and Var (y) expression is asked variance to the y coordinate of all pixels in image;
d = s 1 + Var ( x ) + Var ( y ) - - - ( 3 )
7. based on the resulting eigenwert of step 6, set corresponding threshold value, carry out eigenwert and detect, be weighted fusion for testing result, obtain changing graphic and non-changing graphic;
8. for changing graphic, cut apart for the second time, at first choose the figure spot object that detects image and cut apart as reference images, obtain a plurality of inferior figure spot objects, repeat the 4-7 step;
9. result merges: by denoising, level and smooth aftertreatment, merge each figure spot object, finally obtain the variation testing result of remote sensing image.
At first the present invention adopts cuts apart rear direct comparison method, if figure spot object changes, algorithm carries out secondary splitting to these figure spot objects, and then the extraction eigenwert compares.The present invention can effectively remove pseudo-change information, improves to change accuracy of detection, and compares existing method, and the present invention can obtain complicated more in detail result of variations, realizes that the variation of inferior figure spot object detects.
Description of drawings
The present invention is further detailed explanation below in conjunction with the drawings and specific embodiments.
Fig. 1 is the rear relative method scheme block diagram of classification.
Fig. 2 is direct comparison method scheme block diagram.
Fig. 3 is remote sensing image of many phases of the present invention scenic spot change in resources detection method processing flow chart.
Fig. 4 is that variation of the present invention detects the overall technology process flow diagram.
Embodiment
For the high score remote sensing image after two phase registrations, select first phase wherein as changing the benchmark image that detects, this image is carried out Image Segmentation, form the object polar plot, then utilize the Object Segmentation technology to unify cutting to two phase images, form the object piece of a series of correspondences, utilize change detection techniques to carry out extracting change information for every a pair of object piece, integrate by result at last and obtain last result of variations.Fig. 3 has summed up OO remote sensing image of many phases scenic spot change in resources detection technique treatment scheme.
Compare with the variation detection based on pixel, OO change detection technology is utilized the object overall permanence, and judgement overall variation situation has higher counting yield and accuracy of detection.Can effectively overcome the region of variation erroneous judgement that causes due to season, weather, registration bias disconnected.
In the Resource Access change detection techniques of the present invention proposes OO image of many phases scenic spot, after utilizing the segmentation result of first phase image to carry out Object Segmentation to two phase images, form a plurality of spot of figure one to one objects, for the figure spot object that changes, carry out secondary splitting, compare again based on this segmentation result.Twice partitioning algorithm all adopts the method (mean-shift) of mean shift, and the overall technology flow process as shown in Figure 4.
The 02 phase figure spot object of cutting apart the correspondence of polar plot gained for 01 phase figure spot object of first phase Image Segmentation gained and second phase remote sensing image stack, at first, 02 phase figure spot object is carried out secondary splitting, obtain a plurality of inferior figure spot objects, the vector that forms this figure spot is cut apart figure, to 01 phase figure spot object this polar plot that superposes, obtain corresponding inferior figure spot object.Then, extract respectively the various eigenwerts of the two phases inferior figure spot object of secondary splitting gained, as textural characteristics, spectral signature and shape facility etc., the variation of carrying out object level detects, each is changed testing result merges, then merge inferior figure spot and change testing result, finally obtain the result of variations of two phase figure spot objects by denoising, the aftertreatment such as level and smooth.
Implementation step:
1. choose the not remote sensing image of phase simultaneously of regional two scapes to be detected, 01 phase and 02 phase;
2. choose 01 phase remote sensing image as reference images, 02 phase remote sensing image is as detecting image;
3. the remote sensing image of selecting is carried out the image pre-service.Comprise noise reduction process, radiant correction and Image registration;
(1) noise reduction process: the noise of image generally is present in the HFS of image, and want to reject noise in high-frequency signal, the low-and high-frequency of image need to be separated, the present invention utilizes the small echo variation to carry out Image Denoising by Use, rejects the irrelevant information that is mingled in image;
(2) radiant correction: the present invention adopts the statistical regression method, take reference images as master image, carry out radiant correction to detecting image, the method found same place that atural object Change of types and the stable ground object sample point of spectral quality do not occur in the phase image at 2 o'clock, utilize the linear dependence relation of its gray-scale value to proofread and correct, do not need the radiation calibration of sensor and the clutter reflections rate data that relevant atmospheric parameter just can obtain regularization;
(3) Image registration: the present invention adopts affine invariant feature extraction algorithm to realize Image registration, at first method builds the SIFT descriptor with affine unchangeability, and utilize this descriptor that the reference mark of extracting is mated, obtain transformation parameter and realize Image registration;
4. the reference images of step 3 being selected is carried out OO multi-scale division: the present invention adopts the partitioning algorithm of mean shift (Mean-shift) that reference images is cut apart for the first time, obtains figure spot object and the boundary information of reference images.Make up these boundary informations and be converted to and cut apart polar plot;
5. the polar plot stack cut apart with step 4 gained detects image, and this image is cut apart, and obtains and reference images figure spot object one to one;
6. based on above-mentioned segmentation result, obtain each figure spot object, extract a plurality of eigenwerts of object, mainly comprise textural characteristics (gray level co-occurrence matrixes-entropy), spectral signature (average) and shape facility (density);
The present invention comes computation of mean values according to formula (1), and wherein Ci represents the pixel value of i pixel in imaged object;
C ‾ = 1 n Σ i = 1 n C i - - - ( 1 )
Utilize formula (2) to calculate entropy, wherein (i, j) individual element in p (i, j) expression gray level co-occurrence matrixes;
f = Σ i = 1 G Σ j = 1 G P ( i , j ) log [ P ( i , j ) ] - - - ( 2 )
Utilize formula (3) to come the bulk density value, wherein Var represents to ask variance;
d = s 1 + Var ( x ) + Var ( y ) - - - ( 3 )
7. based on the resulting eigenwert of step 6, set corresponding threshold value, carry out eigenwert and detect, be weighted fusion for testing result, obtain changing graphic and non-changing graphic;
8. for changing graphic, cut apart for the second time, at first choose the figure spot object that detects image and cut apart as reference images, obtain a plurality of inferior figure spot objects, repeat the 4-7 step;
9. result merges: by denoising, level and smooth aftertreatment, merge each figure spot object, finally obtain the variation testing result of remote sensing image.
Characteristics of objects is extracted:
Compare based on the change detection techniques of pixel with tradition, the maximum characteristics of OO change detection techniques are that the object of its consideration is not single pixel, but a cutting object integral body, when extracting feature, be not aimed at some pixels, but for an integral body.The corresponding stack features vector of cutting object.At this, for each cutting object, texture feature extraction (gray level co-occurrence matrixes), spectral signature (indexes such as vegetation, water body) and shape facility (being mainly marginal density) etc., their common composition characteristic vectors.
Object-level change detection:
The present invention extracts to scheme spot as object the proper vector (textural characteristics, spectral signature, shape facility etc.) of cutting apart gained figure spot object, utilizes differential technique directly various characteristic images to be carried out difference and calculates.By to the comparative analysis of various changing features testing result, select the best several eigenwerts of testing result: object average, gray level co-occurrence matrixes entropy, density.
Result merges:
Based on the change detecting method of characteristics of objects, the variation of different size there is good applicability, can suppresses not the puppet that phase remote sensing image tone difference simultaneously causes and change, process rising to object level and process from Pixel-level.The object of the detected variation of different characteristic has certain difference, the variation that has in image can not detect by a certain feature but can be by another kind of feature detection out, the variation testing result that this explanation obtains in a certain feature has only been utilized the advantage of this a kind of feature, its result might not be reliable, according to expertise, the result that different characteristic is obtained merges can obtain better result.Invention has proposed the variation testing result that average, entropy, three kinds of features of density obtain is weighted fusion, finally obtains the variation testing result of Fusion Features.
Topmost characteristics of the present invention are, judgement two phase images are not to change for certain pixel when directly changing, but for the overall variation of a cutting object.Its outstanding advantage is:
(1) by the secondary splitting to figure spot object, can access more detailed result of variations in inferior figure spot object.
(2) effectively eliminate the pixel that causes because registration is inaccurate and change illusion, improve and change the accuracy that detects;
(3) adopt the partitioning algorithm of mean shift, consider based on the integral body of cutting object, be conducive to improve feature extraction speed and matching efficiency;
(4) easily carry out the integral body judgement in conjunction with various features, as texture, shape, spectrum etc., greatly improve accuracy rate.

Claims (1)

1. the remote sensing image variation detection method of a multilayer division is characterized in that: detect and be divided into following steps:
(1), choose the not remote sensing image of phase simultaneously of regional two scapes to be detected, 01 phase and 02 phase;
(2), choose 01 phase remote sensing image as reference images, 02 phase remote sensing image is as detecting image;
(3), the remote sensing image of selecting is carried out the image pre-service, comprise noise reduction process, radiant correction and Image registration;
(A) noise reduction process: the noise of image is present in the HFS of image, and wants to reject the noise in high-frequency signal, the low-and high-frequency of image need to be separated, and utilizes small echo to change and carries out Image Denoising by Use, rejects the irrelevant information that is mingled in image;
(B) radiant correction: adopt the statistical regression method, take reference images as master image, carry out radiant correction to detecting image, this statistical regression method found same place that atural object Change of types and the stable ground object sample point of spectral quality do not occur in the phase image at 2 o'clock, utilize the linear dependence relation of its gray-scale value to proofread and correct, do not need the radiation calibration of sensor and the clutter reflections rate data that relevant atmospheric parameter just can obtain regularization;
(C) Image registration: adopt affine invariant feature extraction algorithm to realize Image registration, at first method builds the SIFT descriptor with affine unchangeability, and utilizes this descriptor that the reference mark of extracting is mated, and obtains transformation parameter and realizes Image registration;
(4), the reference images of step 3 being selected is carried out OO multi-scale division: adopt the partitioning algorithm of mean shift that reference images is cut apart for the first time, obtain figure spot object and the boundary information of reference images, make up these boundary informations and be converted to and cut apart polar plot;
(5), the polar plot stack cut apart of step 4 gained is detected image, this image is cut apart, obtained and reference images figure spot object one to one;
(6), based on above-mentioned segmentation result, obtain each figure spot object, extract a plurality of eigenwerts of object, comprise gray level co-occurrence matrixes-entropy, average and density;
The present invention comes computation of mean values according to formula (1), and wherein Ci represents the pixel value of i pixel in imaged object; N represents the pixel number of imaged object;
C ‾ = 1 n Σ i = 1 n C i - - - ( 1 )
Utilize formula (2) to calculate entropy, wherein (i, j) individual element in p (i, j) expression gray level co-occurrence matrixes; G represents the exponent number of gray level co-occurrence matrixes;
f = Σ i = 1 G Σ j = 1 G P ( i , j ) log [ P ( i , j ) ] - - - ( 2 )
Utilize formula (3) to come the bulk density value, wherein s represents the area of imaged object, and Var (x) expression is asked variance to the x coordinate of all pixels in image, and Var (y) expression is asked variance to the y coordinate of all pixels in image;
d = s 1 + Var ( x ) + Var ( y ) - - - ( 3 )
(7), based on the resulting eigenwert of step 6, set corresponding threshold value, carry out eigenwert and detect, be weighted fusion for testing result, obtain changing graphic and non-changing graphic;
(8), for changing graphic, cut apart for the second time, at first choose the figure spot object that detects image and cut apart as reference images, obtain a plurality of inferior figure spot objects, repeat the 4-7 step;
(9), result merges: by denoising, level and smooth aftertreatment, merge each figure spot object, finally obtain the variation testing result of remote sensing image.
CN2011100978078A 2011-04-19 2011-04-19 Method for carrying out change detection on multi-level segmented remote sensing image Active CN102254319B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2011100978078A CN102254319B (en) 2011-04-19 2011-04-19 Method for carrying out change detection on multi-level segmented remote sensing image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2011100978078A CN102254319B (en) 2011-04-19 2011-04-19 Method for carrying out change detection on multi-level segmented remote sensing image

Publications (2)

Publication Number Publication Date
CN102254319A CN102254319A (en) 2011-11-23
CN102254319B true CN102254319B (en) 2013-06-19

Family

ID=44981561

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2011100978078A Active CN102254319B (en) 2011-04-19 2011-04-19 Method for carrying out change detection on multi-level segmented remote sensing image

Country Status (1)

Country Link
CN (1) CN102254319B (en)

Families Citing this family (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102855487B (en) * 2012-08-27 2015-04-22 南京大学 Method for automatically extracting newly added construction land change image spot of high-resolution remote sensing image
CN103489193B (en) * 2013-09-30 2016-07-06 河海大学 OO high-resolution remote sensing image change detecting method based on convergence strategy
CN103646402B (en) * 2013-12-20 2016-05-18 中国科学院遥感与数字地球研究所 Detect method and the device of remote sensing road image precision
CN104751166B (en) * 2013-12-30 2018-04-13 中国科学院深圳先进技术研究院 Remote Image Classification based on spectral modeling and Euclidean distance
CN103955943A (en) * 2014-05-21 2014-07-30 西安电子科技大学 Non-supervision change detection method based on fuse change detection operators and dimension driving
CN104156955A (en) * 2014-08-04 2014-11-19 华中农业大学 Method for detecting variation of high resolution remote sensing images
CN104268838B (en) * 2014-09-28 2017-02-01 武汉大学 Wavelet denoising algorithm oriented to hyperspectral databases
CN106886988B (en) * 2015-12-11 2020-07-24 中国科学院深圳先进技术研究院 Linear target detection method and system based on unmanned aerial vehicle remote sensing
CN107248172A (en) * 2016-09-27 2017-10-13 中国交通通信信息中心 A kind of remote sensing image variation detection method based on CVA and samples selection
CN107085708B (en) * 2017-04-20 2020-06-09 哈尔滨工业大学 High-resolution remote sensing image change detection method based on multi-scale segmentation and fusion
CN107421514A (en) * 2017-06-21 2017-12-01 广东中冶地理信息股份有限公司 A kind of refuse landfill range monitoring method
CN107480634A (en) * 2017-08-12 2017-12-15 天津市测绘院 A kind of geographical national conditions ground mulching monitoring method based on multistage target classification
CN109409389B (en) * 2017-08-16 2020-01-24 香港理工大学深圳研究院 Multi-feature-fused object-oriented change detection method
CN108053408B (en) * 2017-12-05 2020-05-19 中国资源卫星应用中心 High-automation land use updating method based on remote sensing satellite image
CN109903274B (en) * 2019-01-31 2020-05-15 兰州交通大学 High-resolution remote sensing image change detection method and system
CN110084205A (en) * 2019-04-30 2019-08-02 合肥工业大学 One kind being based on improved object-oriented Classification of High Resolution Satellite Images method
CN110427997A (en) * 2019-07-25 2019-11-08 南京信息工程大学 Improvement CVA change detecting method towards complicated remote sensing image background
CN111680704B (en) * 2020-06-11 2021-01-15 生态环境部卫星环境应用中心 Automatic and rapid extraction method and device for newly-increased human active plaque of ocean red line
CN112101325B (en) * 2020-11-18 2021-02-19 航天宏图信息技术股份有限公司 Method and device for detecting farmland change, electronic equipment and storage medium

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4568845B2 (en) * 2007-04-26 2010-10-27 三菱電機株式会社 Change area recognition device
CN101604445B (en) * 2009-07-24 2012-08-22 武汉大学 Method based on convex module for detecting variation of object level of remote sensing images
CN101937079B (en) * 2010-06-29 2012-07-25 中国农业大学 Remote sensing image variation detection method based on region similarity

Also Published As

Publication number Publication date
CN102254319A (en) 2011-11-23

Similar Documents

Publication Publication Date Title
CN102254319B (en) Method for carrying out change detection on multi-level segmented remote sensing image
CN105574534A (en) Significant object detection method based on sparse subspace clustering and low-order expression
CN103020605B (en) Based on the bridge recognition of Decision-level fusion
CN102855622B (en) A kind of infrared remote sensing image sea ship detection method based on significance analysis
CN103971115A (en) Automatic extraction method for newly-increased construction land image spots in high-resolution remote sensing images based on NDVI and PanTex index
CN103632363A (en) Object-level high-resolution remote sensing image change detection method based on multi-scale fusion
Shaoqing et al. The comparative study of three methods of remote sensing image change detection
CN107330875B (en) Water body surrounding environment change detection method based on forward and reverse heterogeneity of remote sensing image
CN103745468A (en) Significant object detecting method based on graph structure and boundary apriority
CN104200461A (en) Mutual information image selected block and sift (scale-invariant feature transform) characteristic based remote sensing image registration method
CN102110227B (en) Compound method for classifying multiresolution remote sensing images based on context
CN105335965B (en) Multi-scale self-adaptive decision fusion segmentation method for high-resolution remote sensing image
CN103077515B (en) Detection method for building change by multispectral image
CN105956557A (en) Object-oriented timing sequence remote sensing image cloud coverage area automatic detection method
CN103034858A (en) Secondary clustering segmentation method for satellite cloud picture
CN104680536A (en) Method for detecting SAR image change by utilizing improved non-local average algorithm
CN104517286A (en) SAR (synthetic aperture radar) image registration based on self-adaption threshold segmentation and combination optimization
Xu et al. The comparative study of three methods of remote sensing image change detection
CN103500343A (en) Hyperspectral image classification method based on MNF (Minimum Noise Fraction) transform in combination with extended attribute filtering
CN103913166A (en) Star extraction method based on energy distribution
Baraldi et al. Operational performance of an automatic preliminary spectral rule-based decision-tree classifier of spaceborne very high resolution optical images
CN103971364A (en) Remote sensing image variation detecting method on basis of weighted Gabor wavelet characteristics and two-stage clusters
Jamil et al. Tree species extraction and land use/cover classification from high-resolution digital orthophoto maps
Huang et al. Classification of very high spatial resolution imagery based on the fusion of edge and multispectral information
CN104200460A (en) Image registration method based on images characteristics and mutual information

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
C41 Transfer of patent application or patent right or utility model
TR01 Transfer of patent right

Effective date of registration: 20160126

Address after: 100022, No. 19, building 1, No. 22, No. 2 West Fourth Ring Road, Haidian District, Beijing

Patentee after: Space Star Technology (Beijing) Co., Ltd.

Address before: 100022, No. 19, building 1, No. 22, No. 2 West Fourth Ring Road, Haidian District, Beijing

Patentee before: Zhongke Jiudu (Beijing) Spatial Information technology Co., Ltd.

CP03 Change of name, title or address
CP03 Change of name, title or address

Address after: 101399 No. 2 East Airport Road, Shunyi Airport Economic Core Area, Beijing (1st, 5th and 7th floors of Industrial Park 1A-4)

Patentee after: Zhongke Star Map Co., Ltd.

Address before: 100022 Beijing Haidian District North Fourth Ring West Road 19 22 Building 1 Floor 2

Patentee before: Space Star Technology (Beijing) Co., Ltd.