CN103198480B - Based on the method for detecting change of remote sensing image of region and Kmeans cluster - Google Patents
Based on the method for detecting change of remote sensing image of region and Kmeans cluster Download PDFInfo
- Publication number
- CN103198480B CN103198480B CN201310114150.0A CN201310114150A CN103198480B CN 103198480 B CN103198480 B CN 103198480B CN 201310114150 A CN201310114150 A CN 201310114150A CN 103198480 B CN103198480 B CN 103198480B
- Authority
- CN
- China
- Prior art keywords
- region
- image
- area
- gray
- pixel
- 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.)
- Expired - Fee Related
Links
Abstract
The invention discloses a kind of method for detecting change of remote sensing image based on region and Kmeans cluster.Mainly solve existing method testing result and contain in isolated pixel point and region the problem having cavity.Performing step is: phase images X when (1) reads in two width registration different
1and X
2; (2) to X
1and X
2error image is constructed after filtering; (3) maximum entropy threshold segmentation is carried out to differential image and extract area-of-interest and affirmative non-changing region; (4) proper vector in area-of-interest and affirmative non-changing region is calculated; (5) adopt Kmeans clustering algorithm, according to the feature of affirmative non-changing region and all area-of-interests, all regions are divided into two classes; (6) according to cluster result, will correspond to that higher class region of differential image gray-scale value as region of variation, other regions in differential image, as non-changing region, obtain final change testing result.The consistance of testing result energy retaining zone inside of the present invention and removal isolated pixel point, improve accuracy of detection, can be used for monitoring resource and Disaster Assessment.
Description
Technical field
The invention belongs to technical field of image processing, relate to remote sensing image change and detect, specifically a kind of method for detecting change of remote sensing image based on region and Kmeans cluster, is applicable to remote sensing image processing and analysis.
Background technology
Remote Sensing Imagery Change Detection is exactly the region changed in extraction two phase remote sensing images, be widely used in numerous areas that is economic and society, as the field such as even military surveillance and battlefield estimation is planned in land resource monitoring, aquatic monitoring, environmental monitoring, forestry monitoring, agricultural investigation, vegetative coverage and weather monitoring, disaster monitoring and assessment, city management.
In practical application, what the change as the change of land cover pattern, land deterioration and the change of desertification, the change, forest cover change, crops etc. of rivers and lakes presented on remote sensing images is not single or two or three pixels or block change, but the variform regional change of size, even if there is the city management planning, disaster monitoring etc. of the region of variation of more small size, the area of its region of variation is also can to form the zonule of connection.The resolution of remote sensing images improves gradually, and the connected pixel in real change region increases thereupon, and the method for detecting change of remote sensing image based on region is also more aobvious important.Human eye is not read and understand the information of image by pixel, but carry out based on region.Comparatively obtain differential image for two phase image ratio, region of variation is wherein the key area of remote Sensing Image Analysis and decipher, is visual important area.
In order to the change information of two phase remote sensing images and the degree of accuracy improving change detection effectively can be detected, carry out changing the effective way that detection is then head it off from the angle of region and vision attention.Ten thousand red woodss in 2012 etc. are at document " carrying out SAR image change detection research in interested regional level. " (mapping journal, 2012,41 (2): 239-245.) propose one and carry out SAR image change detection in interested regional level.First the method extracts area-of-interest to differential image; Then regard each area-of-interest as a data point and calculate its feature, the threshold technology according to threshold criteria is split all data points, obtains final change testing result.The method requires that the area-of-interest extracted must contain non-changing region, otherwise region of variation mistake can be divided into non-changing region, greatly reduces Detection results; The method FCM cluster is extracted region of interest and is existed more undetected, causes larger transmission error and loses too much change information.Within 2010, Zhang Min proposes a kind of method for detecting change of remote sensing image based on Chi-square Distance geometry significance measure in Xian Electronics Science and Technology University in its patented claim " method for detecting change of remote sensing image based on significance measure " (number of patent application: 201210051159.7, publication number: CN102629377A).First the method is carried out significance measure to differential chart and is obtained difference image, split to difference image with based on the differential image of Chi-square distance respectively, utilize the segmentation result of the segmentation result correction difference image of Chi-square distance difference image, obtain final change testing result.The method can suppress the generation of pseudo-change information, more accurate by revising the location making region of variation, but this method is more undetected owing to existing, and can not keep the edge of region of variation well.
Summary of the invention
The object of the invention is to for the deficiency in above-mentioned method for detecting change of remote sensing image, propose a kind of method for detecting change of remote sensing image based on region and Kmeans cluster, undetected to reduce, accurately region of variation edge, location, improves the precision that change detects.
Implementation of the present invention, comprises the steps:
(1) two width remote sensing images X are read in
1and X
2, and to X
1and X
2carry out neighbour average filtering, obtain filtered image Y
1and Y
2;
(2) according to filtered image Y
1and Y
2, structural differences image: Y=|Y
1-Y
2|;
(3) the maximum entropy threshold T of principle of maximum entropy calculated difference image Y is utilized;
(4) pixel being not less than maximum entropy threshold T by gray-scale value in differential image Y forms several connected regions, and regards each connected region as a region of interest ROI, and all area-of-interests is formed an area-of-interest set { ROI};
(5) gray-scale value in differential image Y is less than the pixel of α * T as affirmative non-changing pixel, and non-changing pixels certainly all in differential image Y are formed a non-changing pixel set NRset certainly, this affirmative non-changing pixel set NRset is regarded as a non-changing region NR certainly, wherein, 0< α <1;
(6) area-of-interest set { area grayscale average G, the area grayscale maximal value M of each area-of-interest and the nucleus gray average K of area-of-interest in ROI} is calculated, and by the vector [G of these three structural feature, M, K] as the proper vector of this area-of-interest;
(7) area grayscale average U, the area grayscale maximal value V of non-changing region the NR certainly and nucleus gray average S in affirmative non-changing region is calculated, and by the vector [U of these three structural feature, V, S] as the proper vector affirming non-changing region;
(8) { in ROI}, each area-of-interest is as regional ensemble R={ROI} ∪ NR to be sorted, and ∪ is for getting union symbol will to affirm non-changing region NR and area-of-interest set;
(9) proper vector adopting Kmeans clustering algorithm to treat all regions in specification area set R carries out cluster, and using that larger for the area grayscale mean value component of cluster centre vector class region as final region of variation, using other regions remaining in differential image Y then as final non-changing region, obtain final change testing result.
The present invention compared with prior art tool has the following advantages:
(1) the present invention is owing to being carry out change to detect on the level in region, extracts area-of-interest, improve segmentation precision, reduce the transmission error of region segmentation to testing result differential image segmentation.
(2) cooperation of affirmative non-changing set of pixels is a non-changing region participation Kmeans cluster certainly by the present invention, ensure that all regions participating in Kmeans cluster are necessarily containing region of variation and non-changing region, avoid in the region participating in Kmeans cluster and only contain region of variation thus be divided into the situation in non-changing region to occur region of variation mistake.
Simulation results show, compared with the existing methods, the present invention can take into account false-alarm and undetected, can keep the edge of region of variation, improve the precision of detection, solve undetected more problem in existing method.
Accompanying drawing explanation
Fig. 1 is process flow diagram of the present invention;
Fig. 2 is that three group of two phase remote sensing images using of the present invention and change thereof detect reference diagram;
Fig. 3 uses the present invention and existing method to the change testing result figure of Fig. 2 remote sensing images.
Embodiment
As follows to performing step of the present invention with reference to Fig. 1:
Step 1, reads in the two width remote sensing images Xs of areal at the registration do not obtained in the same time
1and X
2, this two width image X
1and X
2size be I capable J row.
Step 2, to first phase image X
1carry out Neighborhood Filtering, obtain filtered first phase image Y
1.
The method of Neighborhood Filtering is a lot, such as, have four neighbour average filterings, four neighborhood medium filterings, eight neighborhood mean filter, eight neighborhood medium filtering and N
3neighbour average filterings etc., the present embodiment adopts N
3neighbour average filtering, its filter step is as follows:
2a) at first phase image X
1in, centered by pixel (i, j), choose the neighborhood territory pixel gained set N of this pixel
ij={ (i, j), (i-2, j), (i+2, j), (i-1, j-1), (i-1, j), (i-1, j+1), (i, j-2), (i, j-1), (i, j+1), (i, j+2), (i+1, j-1), (i+1, j), (i+1, j+1) }, wherein, i and j is respectively the row and column sequence number in image, their limited range is i=3 ..., I-2, j=3 ..., J-2;
2b) calculation procedure 2a) middle set N
ijin the average of whole grey scale pixel values, and using this average as first phase image Y after filtering
1at the gray-scale value Y of pixel (i, j)
1(i, j);
2c) for first phase image X
1middle i=3 ..., I-2 and j=3 ..., the boundary pixel point (i, j) beyond J-2, by the gray-scale value X of this pixel
1(i, j) is as first phase image Y after filtering
1at the gray-scale value Y of pixel (i, j)
1(i, j);
2d) repeat step 2a) to 2c), until process first phase image X
1in all pixel, obtain filtered first phase image Y
1.
Step 3, to second time phase images X
2carry out the Neighborhood Filtering same with step 2, phase images Y when obtaining filtered second
2.
Step 4, to filtered image Y
1and Y
2the gray-scale value of corresponding pixel points carry out subtracting each other and taking absolute value, obtain a width differential image Y, Y=|Y
1-Y
2|, the size of differential image Y is the capable J row of I.
Step 5, the maximum entropy threshold T by following formulae discovery differential image Y:
Wherein, gray-scale value th is the arbitrary gray level between the minimal gray level and maximum gray scale of image,
the value of gray-scale value th when getting maximal value for making objective function, p
q=n
q/ (I × J) is the probability that in image, gray level q occurs, n
qthat in image, gray level is the number of the pixel of q,
be the probability sum that in image, all gray levels being not more than threshold value th occur, Gray is the maximum gray scale of image.
Step 6, the pixel being not less than maximum entropy threshold T by gray-scale value in differential image Y forms several connected regions, each connected region is regarded as a region of interest ROI, and all area-of-interests is formed an area-of-interest set { ROI}.
Step 7, obtains the area-of-interest set { proper vector of regional in ROI}.
7a) calculate the nucleus gray average K of the area grayscale average G of region of interest ROI, area grayscale maximal value M and area-of-interest according to the following formula:
Wherein, ROI be area-of-interest set an area-of-interest in ROI}, Nroi is the number of pixel in region of interest ROI, Y (i, j) be the gray-scale value of pixel (i, j) in differential image Y, KP is the nucleus of region of interest ROI, KP={ (i, j) | Y (i, j) >G, (i, j) ∈ ROI}, Nkp are the number of pixel in nucleus KP;
7b) by step 7a) in area grayscale average G, area grayscale maximal value M and nucleus gray average K form vector [G, M, K], and using the proper vector of this vector as region of interest ROI;
7c) repeat step 7a) to 7b), until process area-of-interest set, { all regions in ROI} obtain the area-of-interest set { proper vector of regional in ROI}.
Step 8, gray-scale value in differential image Y is less than the pixel of α * T as affirmative non-changing pixel, and non-changing pixels certainly all in differential image Y are formed a non-changing pixel set NRset certainly, this affirmative non-changing pixel set NRset is regarded as a non-changing region NR certainly, wherein, T is the maximum entropy threshold of differential image Y, and contraction coefficient 0< α <1, chooses α=0.5 in embodiments of the present invention.
Step 9, calculates the proper vector of non-changing region NR certainly.
9a) be calculated as follows area grayscale average U, the area grayscale maximal value V of non-changing region the NR certainly and nucleus gray average S in affirmative non-changing region:
Wherein, NR is affirmative non-changing region, and Nnr is the number of pixel in affirmative non-changing region NR, Y (i, j) be the gray-scale value of pixel (i, j) in differential image Y, KPNR is the nucleus of non-changing region NR certainly, KPNR={ (i, j) | Y (i, j) >U, (i, j) ∈ NR}, Nnrkp are the number of pixel in nucleus KPNR;
9b) by step 9a) in area grayscale average U, area grayscale maximal value V and nucleus gray average S form vector [U, V, S], and using this vector as affirming the proper vector of non-changing region NR.
Step 10, will affirm that { ROI} merges into a regional ensemble to be sorted: R={ROI} ∪ NR, and wherein, ∪ is for getting union symbol for non-changing region NR and area-of-interest set.
Step 11, the proper vector adopting clustering algorithm to treat all regions in specification area set R carries out cluster, and using that larger for the area grayscale mean value component of cluster centre vector class region as final region of variation, using other regions remaining in differential image Y then as final non-changing region, obtain final change testing result.
Clustering algorithm has a lot, such as there are Kmeans clustering algorithm and FCM clustering algorithm etc., the present embodiment adopts document " Some Methods for Classification and Analysis of Multivariate Observations. " (InL.M.LeCam and J.Neyman, editors, Proceedings of Fifth Berkeley Symposium onMathematical Statistics and Probability, volume1, pages281-297, Berkeley, CA, 1967.University of California Press.) in Kmeans clustering algorithm carry out cluster.
Effect of the present invention further illustrates by following experimental result and analysis:
1. experimental data and evaluation index
Emulation experiment of the present invention data used are three groups of true remotely-sensed data collection.First group of true remotely-sensed data collection is two width Landsat-5 satellite TM the 4th band spectrum images of Italian Sardinia, two width image sizes are 300 × 412 pixels, the change occurred between them is caused by lake middle water level rises, and comprises 7626 change pixels and 115974 non-changing pixels; Its two phases original image and change detect reference diagram respectively as shown in Fig. 2 (a1), Fig. 2 (a2), Fig. 2 (a3).Two width Landsat-7EM+ the 4th band spectrum images in second group of true remotely-sensed data Ji Shi Mexico countryside, the size of two width images is 512 × 512 pixels, the change occurred between them is destroyed caused by large-area local vegetation by fire, comprises 25589 change pixels and 236555 non-changing pixels; Its two phases original image and change detect reference diagram respectively as shown in Fig. 2 (b1), Fig. 2 (b2), Fig. 2 (b3).3rd group of true remote sensing image data collection original image is two width Landsat-5 satellite TM the 4th wave band multispectral image compositions in west area, Italian Elba island respectively in August, 1994 and in September, 1994, two width image sizes are 326 × 414 pixels, change between them is because forest fire destroys caused by a large amount of vegetation, 2415 change pixels and 132549 non-changing pixels are comprised in change reference diagram, its two phases original image and change detect reference diagram respectively as Fig. 2 (c1), Fig. 2 (c2), shown in Fig. 2 (c3), (change detects the white pixel region representation region of variation in reference diagram, black-pixel region represents non-changing region).
The objective evaluation index weighing change detection algorithm in the present invention adopts false-alarm number, undetected number, total error number and accuracy.
2. the method for contrast of the present invention's use is as described below:
Control methods 1 is Wan Honglin etc. " to be carried out SAR image change detection research " and proposes a kind ofly in interested regional level, carrying out SAR image change detection at document in interested regional level, is designated as ROI_FCM_THR method.
Control methods 2, that Xian Electronics Science and Technology University is at its patented claim " method for detecting change of remote sensing image based on significance measure " (number of patent application: 201210051159.7, publication number: CN102629377A) in propose a kind of method for detecting change of remote sensing image based on Chi-square Distance geometry significance measure, be designated as SMCD method.
3. experiment content and analysis
In order to verify that the inventive method is carried out change detection and can be improved region segmentation precision and accuracy of detection on regional level, the inventive method and ROI_FCM_THR method are contrasted; In order to verify the region of variation of differentiation effectively and the non-changing region of three kinds of provincial characteristicss that the inventive method adopts, the inventive method and SMCD method are contrasted.
Emulation 1, to three groups of true remotely-sensed data collection in accompanying drawing 2, carry out change by existing ROI_FCM_THR method to detect, the evaluation index of change testing result is as the first row of table 1, shown in fourth line and the 7th row, its change testing result figure as shown in Figure 3, wherein Fig. 3 (a1) is the change testing result figure of ROI_FCM_THR method to first group of true remotely-sensed data collection, Fig. 3 (b1) is the change testing result figure of ROI_FCM_THR method to second group of true remotely-sensed data collection, Fig. 3 (c1) is the change testing result figure of ROI_FCM_THR method to the 3rd group of true remotely-sensed data collection.
Emulation 2, to three groups of true remotely-sensed data collection in accompanying drawing 2, carry out change by existing SMCD method to detect, change shown in second row of evaluation index as table 1 of testing result, fifth line and the 8th row, its change testing result figure as shown in Figure 3, wherein Fig. 3 (a2) is the change testing result figure of SMCD method to first group of true remotely-sensed data collection, Fig. 3 (b2) is the change testing result figure of SMCD method to second group of true remotely-sensed data collection, Fig. 3 (c2) is the change testing result figure of SMCD method to the 3rd group of true remotely-sensed data collection.
Emulation 3, to three groups of true remotely-sensed data collection in accompanying drawing 2, carry out change by the inventive method to detect, change shown in evaluation index the third line as table 1 of testing result, the 6th row and the 9th, its change testing result figure as shown in Figure 3, wherein Fig. 3 (a3) is the change testing result figure of the inventive method to first group of true remotely-sensed data collection, Fig. 3 (b3) is the change testing result figure of the inventive method to second group of true remotely-sensed data collection, Fig. 3 (c3) is the change testing result figure of the inventive method to the 3rd group of true remotely-sensed data collection.
As can be seen from Table 1, in the change testing result of three kinds of methods, the overall evaluation of the inventive method to three groups of true remotely-sensed data collection is best.The inventive method to total error number of first group of true remotely-sensed data collection 310 pixels fewer than total error number of ROI_FCM_THR method, 97 pixels fewer than total error number of SMCD method; To total error number of second group of true remotely-sensed data collection 4034 pixels fewer than total error number of ROI_FCM_THR method, 870 pixels fewer than total error number of SMCD method; To total error number of the 3rd group of true remotely-sensed data collection 10341 pixels fewer than total error number of ROI_FCM_THR method, 170 pixels fewer than total error number of SMCD method; Total error number of ROI_FCM_THR method is too much, and total error number of the inventive method is minimum in existing method.To the undetected number of first group of remotely-sensed data collection undetected number few 669 pixels respectively than ROI_FCM_THR method, 337 pixels fewer than the undetected number of SMCD method; To the undetected number of second group of remotely-sensed data collection undetected number few 5612 pixels respectively than ROI_FCM_THR method, 1205 pixels fewer than the undetected number of SMCD method, to the undetected number of the 3rd group of remotely-sensed data collection 138 pixels fewer than the undetected number of SMCD method; ROI_FCM_THR method is too much to the undetected number of first and second groups of remotely-sensed data collection, is at least 2 times of the inventive method; The undetected number of the inventive method is minimum in existing method.This shows, the inventive method can detect change information comparatively comprehensively, exactly, reduces undetected number, takes into account undetected and false-alarm preferably, has higher accuracy of detection.As can be seen from the design sketch 3 of three groups of experimental data collection, compare with SMCD method with ROI_FCM_THR method, the inventive method can keep the marginal information of region of variation preferably, and the isolated pixel point of flase drop is also less.
The true remote sensing image data centralized procurement of three groups, table 1 differently changes the performance evaluation of testing result
Claims (4)
1., based on a method for detecting change of remote sensing image for region and Kmeans cluster, comprise the steps:
(1) two width remote sensing images X are read in
1and X
2, and to X
1and X
2carry out neighbour average filtering, obtain filtered image Y
1and Y
2;
(2) according to filtered image Y
1and Y
2, structural differences image: Y=|Y
1-Y
2|;
(3) the maximum entropy threshold T of principle of maximum entropy calculated difference image Y is utilized:
Wherein, gray-scale value th is the arbitrary gray level between the minimal gray level and maximum gray scale of image,
the value of gray-scale value th when getting maximal value for making objective function, p
q=n
q/ (I × J) is the probability that in image, gray level q occurs, n
qthat in image, gray level is the number of the pixel of q,
p
qbe the probability sum that in image, all gray levels being not more than threshold value th occur, Gray is the maximum gray scale of image;
(4) pixel being not less than maximum entropy threshold T by gray-scale value in differential image Y forms several connected regions, and regards each connected region as a region of interest ROI, and all area-of-interests is formed an area-of-interest set { ROI};
(5) gray-scale value in differential image Y is less than the pixel of α * T as affirmative non-changing pixel, a non-changing pixel set NRset is certainly formed by non-changing pixels certainly all in differential image Y, regard affirmative non-changing pixel set NRset as a non-changing region NR certainly, wherein, 0< α <1;
(6) area-of-interest set { area grayscale average G, the area grayscale maximal value M of each area-of-interest and the nucleus gray average K of area-of-interest in ROI} is calculated, and by the vector [G of these three structural feature, M, K] as the proper vector of this area-of-interest;
(7) area grayscale average U, the area grayscale maximal value V of non-changing region the NR certainly and nucleus gray average S in affirmative non-changing region is calculated, and by the vector [U of these three structural feature, V, S] as the proper vector affirming non-changing region;
(8) { in ROI}, each area-of-interest is as regional ensemble R={ROI} ∪ NR to be sorted, and ∪ is for getting union symbol will to affirm non-changing region NR and area-of-interest set;
(9) proper vector adopting Kmeans clustering algorithm to treat all regions in specification area set R carries out cluster, and using that larger for the area grayscale mean value component of cluster centre vector class region as final region of variation, using other regions remaining in differential image Y then as final non-changing region, obtain final change testing result.
2. the method for detecting change of remote sensing image based on region and Kmeans cluster according to claim 1, wherein described in step (1) to X
1and X
2carry out neighbour average filtering, adopt N
3neighbour average filtering, its filter step is as follows:
1a) at image X
1in, centered by pixel (i, j), choose the neighborhood territory pixel gained set of this pixel: N
ij={ (i, j), (i-2, j), (i+2, j), (i-1, j-1), (i-1, j), (i-1, j+1), (i, j-2), (i, j-1), (i, j+1), (i, j+2), (i+1, j-1), (i+1, j), (i+1, j+1) }, wherein, i and j is respectively the row and column sequence number in image, and their limited range is i=3,, I-2, j=3,, J-2, I are image X
1maximum row sequence number, J is image X
1maximum column sequence number;
1b) calculation procedure 1a) middle set N
ijin the average of whole grey scale pixel values, and using this average as filtered image Y
1at the gray-scale value Y of this point
1(i, j);
1c) for image X
1middle i=3 ..., I-2 and j=3 ..., the boundary pixel point (i, j) beyond J-2, using the gray-scale value of this pixel as filtered image Y
1at the gray-scale value Y of this point
1(i, j);
1d) repeat step 1a) to 1c), until process image X
1whole pixels, obtains filtered image Y
1;
1e) according to step 1a) to 1d) process image X
2in all pixel, obtain filtered image Y
2.
3. the method for detecting change of remote sensing image based on region and Kmeans cluster according to claim 1, calculating area-of-interest set wherein described in step (6) area grayscale average G, the area grayscale maximal value M of each area-of-interest and the nucleus gray average K of area-of-interest in ROI}, be calculated as follows:
Wherein, ROI be area-of-interest set an area-of-interest in ROI}, Nroi is the number of pixel in region of interest ROI, Y (i, j) be the gray-scale value of pixel (i, j) in differential image Y, KP is the nucleus of region of interest ROI, KP={ (i, j) | Y (i, j) >G, (i, j) ∈ ROI}, Nkp are the number of pixel in nucleus KP.
4. the method for detecting change of remote sensing image based on region and Kmeans cluster according to claim 1, calculating wherein described in step (7) is area grayscale average U, the area grayscale maximal value V of non-changing region NR and the nucleus gray average S in affirmative non-changing region certainly, is calculated as follows:
Wherein, NR is affirmative non-changing region, Nnr is the number of pixel in affirmative non-changing region NR, KPNR is the nucleus of non-changing region NR certainly, KPNR={ (i, j) | Y (i, j) >U, (i, j) ∈ NR}, Nnrkp are the number of pixel in nucleus KPNR.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310114150.0A CN103198480B (en) | 2013-04-02 | 2013-04-02 | Based on the method for detecting change of remote sensing image of region and Kmeans cluster |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310114150.0A CN103198480B (en) | 2013-04-02 | 2013-04-02 | Based on the method for detecting change of remote sensing image of region and Kmeans cluster |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103198480A CN103198480A (en) | 2013-07-10 |
CN103198480B true CN103198480B (en) | 2015-09-02 |
Family
ID=48720986
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310114150.0A Expired - Fee Related CN103198480B (en) | 2013-04-02 | 2013-04-02 | Based on the method for detecting change of remote sensing image of region and Kmeans cluster |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103198480B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017066353A1 (en) * | 2015-10-12 | 2017-04-20 | Drone Seed Co. | Forestry information management systems and methods streamlined by automatic biometric data prioritization |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103400383A (en) * | 2013-07-22 | 2013-11-20 | 西安电子科技大学 | SAR (synthetic aperture radar) image change detection method based on NSCT (non-subsampled contourlet transform) and compressed projection |
CN104881677B (en) * | 2015-05-08 | 2017-12-05 | 北京师范大学 | Method is determined for the optimum segmentation yardstick of remote sensing image ground mulching classification |
CN106557767B (en) * | 2016-11-15 | 2019-04-09 | 北京唯迈医疗设备有限公司 | A kind of method of ROI region in determining interventional imaging |
CN106803245B (en) * | 2016-11-29 | 2020-07-03 | 中国铁道科学研究院集团有限公司铁道建筑研究所 | Railway roadbed state evaluation method based on ground penetrating radar periodic detection |
CN106780471B (en) * | 2016-12-23 | 2020-05-12 | 贵州电网有限责任公司电力科学研究院 | Transformer substation equipment infrared image change detection method based on Markov random field |
CN107358625B (en) * | 2017-06-12 | 2021-07-06 | 西安电子科技大学 | SAR image change detection method based on SPP Net and region-of-interest detection |
CN109242889B (en) * | 2018-08-27 | 2020-06-16 | 大连理工大学 | SAR image change detection method based on context significance detection and SAE |
CN109255388B (en) * | 2018-09-28 | 2021-12-31 | 西北工业大学 | Unsupervised heterogeneous remote sensing image change detection method |
CN110415208B (en) * | 2019-06-10 | 2023-10-17 | 西安电子科技大学 | Self-adaptive target detection method and device, equipment and storage medium thereof |
CN111160977A (en) * | 2019-12-31 | 2020-05-15 | 中国移动通信集团黑龙江有限公司 | Method, device, equipment and medium for acquiring user relation interest characteristic graph |
CN111476813B (en) * | 2020-04-28 | 2023-04-07 | 兰州交通大学 | Image change detection method, image change detection device, electronic equipment and storage medium |
CN114324185A (en) * | 2022-01-04 | 2022-04-12 | 浙江大学 | Underwater polarization detection device based on Stokes vector |
-
2013
- 2013-04-02 CN CN201310114150.0A patent/CN103198480B/en not_active Expired - Fee Related
Non-Patent Citations (5)
Title |
---|
Fuzzy clustering algorithms for unsupervised change detection in remote sensing images;Ashish Ghosh et al;《Information Sciences》;20111231;第181卷;全文 * |
Multiscale Unsupervised Change Detection on Optical Images by Markov Random Fields and Wavelets;Gabriele Moser et al;《IEEE GEOSCIENCE AND REMOTE SENSING LETTERS》;20110731;第8卷(第4期);全文 * |
The attentional field has a Mexican hat distribution;Notger G. Muller et al;《Vision Research》;20051231;第45卷;全文 * |
万红林 等.在感兴趣的区域层面上进行SAR图像变化检测的方法研究.《测绘学报》.2012,第41卷(第2期),全文. * |
辛芳芳 等.利用改进Fisher分类器进行遥感图像变化检测.《西安电子科技大学学报(自然科学版)》.2012,第39卷(第5期),全文. * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017066353A1 (en) * | 2015-10-12 | 2017-04-20 | Drone Seed Co. | Forestry information management systems and methods streamlined by automatic biometric data prioritization |
EP3361853A4 (en) * | 2015-10-12 | 2019-06-19 | Drone Seed Co. | Forestry information management systems and methods streamlined by automatic biometric data prioritization |
Also Published As
Publication number | Publication date |
---|---|
CN103198480A (en) | 2013-07-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103198480B (en) | Based on the method for detecting change of remote sensing image of region and Kmeans cluster | |
Sahebjalal et al. | Analysis of land use-land covers changes using normalized difference vegetation index (NDVI) differencing and classification methods | |
CN101650439B (en) | Method for detecting change of remote sensing image based on difference edge and joint probability consistency | |
CN103236063B (en) | Based on the SAR image oil spilling detection method of multiple dimensioned spectral clustering and decision level fusion | |
CN100507603C (en) | Hyperspectral image abnormal point detection method based on selective kernel principal component analysis | |
CN103473764B (en) | A kind of remote sensing image object variations detection method | |
CN104299232B (en) | SAR image segmentation method based on self-adaptive window directionlet domain and improved FCM | |
CN102063720B (en) | Treelets-based method for detecting remote sensing image changes | |
CN102629380B (en) | Remote sensing image change detection method based on multi-group filtering and dimension reduction | |
CN105549009B (en) | A kind of SAR image CFAR object detection methods based on super-pixel | |
CN103258324B (en) | Based on the method for detecting change of remote sensing image that controlled kernel regression and super-pixel are split | |
CN104217436B (en) | SAR image segmentation method based on multiple features combining sparse graph | |
CN102542295A (en) | Method for detecting landslip from remotely sensed image by adopting image classification technology | |
CN107247927B (en) | Method and system for extracting coastline information of remote sensing image based on tassel cap transformation | |
CN103106658A (en) | Island or reef coastline rapid obtaining method | |
CN102831598A (en) | Remote sensing image change detecting method with combination of multi-resolution NMF (non-negative matrix factorization) and Treelet | |
CN107392887B (en) | Heterogeneous remote sensing image change detection method based on homogeneous pixel point conversion | |
CN103020649A (en) | Forest type identification method based on texture information | |
CN103198482B (en) | Based on the method for detecting change of remote sensing image that disparity map fuzzy membership merges | |
CN104680536A (en) | Method for detecting SAR image change by utilizing improved non-local average algorithm | |
CN104268574A (en) | SAR image change detecting method based on genetic kernel fuzzy clustering | |
Hadeel et al. | Environmental change monitoring in the arid and semi-arid regions: a case study Al-Basrah Province, Iraq | |
CN102567726A (en) | Technology for automatically extracting floating ice in polar ice sheet edge regions | |
Shang et al. | PolSAR land classification by using quaternion-valued neural networks | |
CN103093472B (en) | Based on the remote sensing image change detecting method of doubledictionary intersection rarefaction representation |
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 |
Granted publication date: 20150902 Termination date: 20200402 |
|
CF01 | Termination of patent right due to non-payment of annual fee |