CN105405133A - Remote sensing image alteration detection method - Google Patents

Abstract

The invention discloses a remote sensing image alteration detection method comprising the steps that two time phase high-resolution optical remote sensing images X1 and X2 are obtained; image registration is performed on X1 and X2; radiation normalization correction is performed on X1 and X2 by utilizing a multivariate alteration detection method; alteration vector amplitude XM and spectral angle information XSA are respectively calculated according to X1 and X2 after radiation normalization correction; an optimal segmentation threshold T is calculated by utilizing the Bayes theory and an expectation maximization algorithm according to the XM; a pseudo training sample area is selected according to T and XM; XM and XSA are combined to act as input of a core FCM, and optimal model parameter value selection is performed on a core FCM combining spatial neighborhood information model according to the pseudo training sample area; and the alteration area and the non-alteration area of the optical remote sensing images are determined by adopting the method of core FCM combining spatial neighborhood information according to the selected optimal model parameter values. The remote sensing image alteration detection method is more robust and higher in precision.

Description

Remote sensing image change detection method

Technical Field

The invention relates to the technical field of remote sensing image change detection, in particular to a remote sensing image change detection method.

Background

With the continuous accumulation of multi-temporal high-resolution remote sensing data and the successive establishment of spatial databases, how to extract and detect change information from the remote sensing data has become an important research topic of remote sensing science and geographic information science. According to the remote sensing images of the same region in different time phases, the information of dynamic changes of cities, environments and the like can be extracted, and scientific decision-making basis is provided for departments of resource management and planning, environmental protection and the like. The 'twelve five' in China will enlarge and extend the high-resolution earth observation engineering which is started and implemented by 'eleven five', pay attention to basic theories and key technical researches including high-resolution remote sensing target and space environment characteristic analysis and high-reliability automatic interpretation, and become research focuses on solving the major requirements of national security and social and economic development.

The change detection of the remote sensing image is to quantitatively analyze and determine the characteristics and the process of the surface change from the remote sensing data in different periods. From different angles and application researches, various researchers have proposed a plurality of effective detection algorithms, such as a Change Vector Analysis (CVA), a fuzzy c-means (fcm) based clustering method, and the like. In the traditional FCM-based multi-temporal optical remote sensing change detection, CVA (continuously variable amplitude) conversion is carried out firstly, and then FCM clustering is carried out on the amplitude of a change vector, so that a change detection result is obtained. In this type of technology, the disadvantage of using FCM is that it is only suitable for spherical or ellipsoidal clustering and is very sensitive to noise and its outliers (Outlier). In addition, only the amplitude of the variation vector is used, so that the original multispectral information is not fully mined, is not robust enough and has low precision.

Aiming at the problems, many scholars try to solve the problems by adding different spatial neighborhood constraints to an FCM target function, but due to the complexity of a high-resolution image detection environment, the lack of target prior information and the like, the algorithms have certain limitations and are not high in accuracy. Therefore, it is necessary to research a new high-resolution visible light remote sensing image change detection technology to effectively overcome the above difficulties.

Disclosure of Invention

The invention aims to solve the technical problem of providing a remote sensing image change detection method, which is a multi-temporal remote sensing image change detection method of adaptive kernel FCM combining CVA and SAM, and has more stable change detection result and higher precision.

In order to solve the above technical problem, the present invention provides a method for detecting a change in a remote sensing image, comprising:

obtaining two-time-phase high-resolution optical remote sensing image X₁And X₂；

To optical remote sensing image X₁And X₂Performing image matchingPreparing;

optical remote sensing image X by using multivariate change detection method₁And X₂Carrying out radiation normalization correction;

optical remote sensing image X after normalization correction according to radiation₁And X₂Respectively calculating the variation vector amplitude X_MAnd spectral angular information X_SA；

According to the magnitude X of the variation vector_MCalculating to obtain an optimal segmentation threshold value T by using a Bayes principle and a maximum expectation algorithm;

according to the optimal segmentation threshold value T and the variation vector amplitude value X_MSelecting a pseudo-training sample region;

mixing X_MAnd X_SACombining the input data as the kernel FCM, and selecting the optimal model parameter value of the kernel FCM combined with the spatial neighborhood information model according to the pseudo training sample region;

and determining a change region and a non-change region of the optical remote sensing image by adopting a method of combining kernel FCM with space neighborhood information according to the selected optimal model parameter value.

The implementation of the invention has the following beneficial effects: the method combines the change vector amplitude of the multi-temporal remote sensing image and a multi-temporal spectrum angle mapping map (SAM) as the input of a kernel FCM, and then obtains the final change detection result based on the kernel FCM and a method of combining spatial neighborhood information. The kernel parameters and the like in the kernel FCM target function are selected through the pseudo training samples acquired based on the CVA technology, and the change detection result is more stable and has higher precision.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a schematic flow chart diagram illustrating an embodiment of a method for detecting changes in remote sensing images according to the present invention;

FIG. 2 is an original high resolution optical remote sensing image map;

FIG. 3 is a graph comparing the results of experiments conducted by the method of the present invention with other methods

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Fig. 1 is a schematic flow chart of an embodiment of a method for detecting changes in remote sensing images provided by the present invention, and the present invention is a method for detecting changes in remote sensing images in multiple time phases, which is mainly suitable for high-resolution optical remote sensing images, and as shown in fig. 1, the present invention includes the steps of:

s101, acquiring two-time-phase high-resolution optical remote sensing image X₁And X₂。

Wherein, X₁、X₂The two high-resolution optical remote sensing images in the same region and different time phases.

S102, aiming at optical remote sensing image X₁And X₂And performing image registration.

Specifically, step S102 specifically includes the steps of:

and S1021, adopting ENVI14.8 remote sensing software for optical remote sensing image X₁And X₂And carrying out geometric coarse correction.

The geometric rough correction comprises the following specific operation steps: (1) displaying a reference image and an image to be corrected; (2) collecting ground control points GCPs; GCPs are uniformly distributed in the whole image, and the number of the GCPs is at least more than or equal to 9; (3) calculating an error; (4) selecting a polynomial model; (5) and (5) resampling and outputting by using bilinear interpolation. The bilinear difference method is as follows: if the unknown function f is evaluated at point P as (x, y), we assume that the function f is known to be at Q₁₁＝(x₁,y₁),Q₁₂＝(x₁,y₂),Q₂₁＝(x₂,y₁) And Q₂₂＝(x₂,y₂) Values of four points. If a coordinate system is chosen such that the coordinates of these four points are (0,0), (0,1), (1,0), and (1,1), respectively, then the bilinear interpolation formula can be expressed as:

f(x,y)≈f(0,0)(1-x)(1-y)+f(1,0)x(1-y)+f(0,1)(1-x)y+f(1,1)xy。

s1022, roughly correcting the geometry of the X by utilizing an automatic matching and triangulation method₁And X₂And carrying out geometric fine correction.

The triangulation method comprises the steps of constructing a Delaunay triangulation network by adopting a point-by-point insertion method, determining affine transformation model parameters inside each triangle by utilizing the row and column numbers of three vertexes of each triangle and the geographic coordinates of the same-name points of the corresponding reference image, and correcting an image to be corrected to obtain a corrected remote sensing image.

S103, utilizing a multivariate change detection Method (MAD) to perform optical remote sensing image X₁And X₂And carrying out radiation normalization correction.

Specifically, step S103 specifically includes the steps of:

s1031, obtaining optical remote sensing image X₁And X₂The linear combination of the brightness values of all wave bands obtains the difference shadow enhanced by the change informationAn image;

s1032, determining a changed area and an unchanged area according to the difference image passing threshold;

s1033, completing relative radiation correction through a mapping equation of two time phase pixel pairs corresponding to the unchanged area.

S104, normalizing the corrected optical remote sensing image X according to the radiation₁And X₂Respectively calculating the variation vector amplitude X_MAnd spectral angular information X_SA。

Specifically, step S104 includes the steps of:

s1041, normalizing the corrected optical remote sensing image X according to the radiation₁And X₂Calculating to obtain the amplitude X of the variation vector_M。

Wherein, $X_M(i,j)=\sqrt{\underset{b=1}{\overset{B}{\Σ}}{\left(X_{1b}\right(i,j)-X_{2b}(i,j\left)\right)}^2},$ wherein B represents the number of wave bands of each time phase remote sensing image, (i, j) is the coordinate of the image, X_1bRepresents X₁B band image of (1), X_2bRepresents X₂B-band images of (1);

s1042, normalizing the corrected optical remote sensing image X according to the radiation₁And X₂Calculating to obtain the amplitude X of the variation vector_M，

Wherein, $X_{SA}(i,j)=arccos\left(\underset{b=1}{\overset{B}{\Σ}}\right(X_{1b}\left(i,j\right)X_{2b}\left(i,j\right)\left)\sqrt{\underset{b=1}{\overset{B}{\Σ}}{X}_{1b}^2(i,j)}\sqrt{\underset{b=1}{\overset{B}{\Σ}}{X}_{2b}^2(i,j)}\right).$

s105, according to the change vector amplitude X_MAnd calculating to obtain an optimal segmentation threshold value T by using Bayes principle and maximum Expectation-Maximization (EM).

Specifically, step S105 specifically includes the steps of:

s1051, estimating X by adopting maximum expectation algorithm_MUnchanged class omega on image_nMean value m of_nSum variance σ_nOf the variation class omega_cMean value m of_cThe sum variance is σ_cWherein

$m_n^{t+1}=\left\{\underset{X(i,j)\∈X_M}{\mathrm{\Σ}}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}X(i,j)\right\}/\left\{\underset{X(i,j)\∈X_M}{\mathrm{\Σ}}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}\right\}$

${\left({\mathrm{\σ}}_n^2\right)}^{t+1}=\{\underset{X(i,j)\∈X_M}{\mathrm{\Σ}}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}\[X(i,j)-m_n^t\]\}/\left\{\underset{X(i,j)\∈X_M}{\mathrm{\Σ}}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}\right\}$

$m_c^{t+1}=\left\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j\left|{\mathrm{\ω}}_c\right))}{p^t\left(X\right(i,j))}X(i,j)\right\}/\left\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j\left|{\mathrm{\ω}}_c\right))}{p^t\left(X\right(i,j))}\right\}$

${\left({\mathrm{\σ}}_c^2\right)}^{t+1}=\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j\left|{\mathrm{\ω}}_c\right))}{p^t\left(X\right(i,j))}\[X(i,j)-m_c^t\]\}/\left\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j\left|{\mathrm{\ω}}_c\right))}{p^t\left(X\right(i,j))}\right\}$

where t represents the number of iterations, the t superscript represents the value at the t-th iteration of the current content, e.g.,represents m_nThe value at the t +1 th iteration, the rest of which are similar,to representThe value at the t +1 th iteration, $p^{t+1}\left({\mathrm{\ω}}_n\right)=\frac{\underset{X(i,j)\∈X_M}{\mathrm{\Σ}}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}}{IJ},p^{t+1}\left({\mathrm{\ω}}_c\right)=\frac{\underset{X(i,j)\∈X_M}{\mathrm{\Σ}}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_c\left)\right)}{p^t\left(X\right(i,j\left)\right)}}{IJ},$ i and J represent the number of rows and columns of the image respectively,represents X_MUnchanged class omega on image_nThe gaussian distribution that is obeyed to,represents X_MVariation class omega on image_cA gaussian distribution obeyed;

s1052, solving a formula according to Bayes minimum error criterion $({\mathrm{\σ}}_n^2-{\mathrm{\σ}}_c^2)T^2+2(m_n{\mathrm{\σ}}_c^2-m_c{\mathrm{\σ}}_n^2)T+m_c^2{\mathrm{\σ}}_n^2-m_n^2{\mathrm{\σ}}_c^2-2{\mathrm{\σ}}_n^2{\mathrm{\σ}}_c^{2}ln\[\frac{{\mathrm{\σ}}_{c}p\left({\mathrm{\ω}}_n\right)}{{\mathrm{\σ}}_{n}p\left({\mathrm{\ω}}_c\right)}\]=0,$ And obtaining an optimal segmentation threshold value T.

S106, according to the optimal segmentation threshold T and the change vector amplitude X_MA pseudo-training sample region is selected.

Specifically, step S106 includes the steps of:

s1061, according to the optimal segmentation threshold value T and the change vector amplitude X_MSelecting unchanged pseudo training set samples as

S1062, according to the optimal segmentation threshold value T and the change vector amplitude X_MSelecting the pseudo training set sample of the variation class asWherein is X_M15% of the dynamic range.

S107, mixing X_MAnd X_SAAnd combining the kernel FCM and the space neighborhood information model to be input as kernel FCM, and selecting optimal model parameter values for the kernel FCM and the space neighborhood information model according to the pseudo training sample region.

Specifically, step S107 specifically includes the steps of:

s1071, mixing X_MAnd X_SAThe combination is used as the input of a kernel FCM, and the kernel FCM and space neighborhood information combination model is constructed as follows: $J_m=\underset{j=1}{\overset{C}{\mathrm{\Σ}}}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{u}_{jk}^m(1-K(X_{MS}\left(k\right),v_j\left)\right)+\mathrm{\α}\underset{j=1}{\overset{C}{\mathrm{\Σ}}}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{u}_{jk}^m(1-K({\stackrel{\‾}{X}}_{MS}\left(k\right),v_j\left)\right),$

where C is the number of clusters, N is the total number of samples,representing the fuzzy membership of the kth sample to the jth class center, m being the weighted index of the membership,the parameter α controls the penalty effect,is X_MLocal mean image and X_SAThe local mean value image of (a) is combined, $K(x,y)=\exp \{-\frac{{(x-y)}^2}{g^2}\}.$

s1072, setting the value range of the parameter α and the kernel parameter g, and searching the change index C by using a pseudo training sample set_indexThe values of α and g at the smallest were used as the optimal model parameter values.

Wherein, the index of variation C_index＝D_index/k_T，k_TKappa coefficient, N, representing model parameters on a pseudo-training sample set_n(α, g) shows minimizing the number of unchanged pixels of the entire image acquired using the objective function given α and g, N_c(α, g) shows the number of pixels that change for the entire image given α and g, TN_n(α, g) indicates the number of unchanged pixels in the pseudo-training sample set given α and g, TN_c(α, g) shows the number of changed pixels in the pseudo training sample set given α and g.

And S108, determining a change region and a non-change region of the optical remote sensing image by adopting a method of combining kernel FCM with space neighborhood information according to the selected optimal model parameter value.

Specifically, step S108 specifically includes:

s1081, setting the cluster number C in a kernel FCM combined space neighborhood information model to be 2, taking the cluster number C as the center of an initial unchanged class and an initial changed class, and selecting and changing a vector amplitude X_MSetting a weighting index m of the membership degree to be 2, wherein the weighting index m is a constant greater than 0, and the values of the parameter α and the kernel parameter g are selected as the optimal model parameter values;

s1082, calculating X_M，X_SAThe window size is set to 3 × 3;

s1083, adopt formula $u_{jk}=\frac{{\left(\right(1-K\left(X_{MS}\right(k),v_j))+\mathrm{\α}(1-K\left({\stackrel{\‾}{X}}_{MS}\right(k),v_j)\left)\right)}^{-1/(m-1)}}{\underset{j=1}{\overset{C}{\Σ}}{\left(\right(1-K\left(X_{MS}\right(k),v_j))+\mathrm{\α}(1-K\left({\stackrel{\‾}{X}}_{MS}\right(k),v_j)\left)\right)}^{-1/(m-1)}}$ Updating mouldPasting and dividing the matrix;

s1084, adopt formula $v_j=\frac{\underset{k=1}{\overset{N}{\Σ}}{u}_{jk}^m\left(K\right(X_{MS}\left(k\right),v_j)X_{MS}\left(k\right)+\mathrm{\α}K\left({\stackrel{\‾}{X}}_{MS}\right(k),v_j){\stackrel{\‾}{X}}_{MS}\left(k\right))}{\underset{k=1}{\overset{N}{\Σ}}{u}_{jk}^m\left(K\right(X_{MS}\left(k\right),v_j)+\mathrm{\α}K\left({\stackrel{\‾}{X}}_{MS}\right(k),v_j))}$ Updating the clustering center;

s1085, repeatedly updating the fuzzy partition matrix and the clustering centers until the clustering centers of two adjacent iterations are smaller than the threshold;

s1086, dividing the matrix u according to the fuzzy_jkAnd determining a final change detection image to obtain a change area and a non-change area of the optical remote sensing image.

The effects of the present invention can be further illustrated by the following experimental results and analyses:

the experimental data of the invention is multi-temporal SPOT high-resolution image data in Littoral region of France, the image size is 400 x 400, and three wave bands of B1, B2 and B3 are used. To verify the effectiveness of the present invention, the change detection method of the present invention was compared to the following change detection methods:

(1) CVA-based EM method (CVA-EM) [ detection methods mentioned in the article "Automatics in genetic engineering for Experimental detection and removal detection", 2000,38(3):1171-1182 ], by Bruzzonen, et al, Italy ].

(2) FCM-based spatial neighborhood information classification method (FCM-S) [ Chensongchan et al (IEEETransactionsonSystems, Man, and dCybernetics-PartB: Cybernetics,2004,34(4): 1907. cndot. 1916) ] in the article

(3) The method of the invention.

The detection performance is measured by four indexes of error detection number FP, missing detection number FN, total error number OE and Kappa coefficient. The closer FP, FN and OE are to 0 and the closer Kappa coefficient is to 1, indicating the better performance of the change detection method. The results are shown in Table 1. As can be seen from fig. 2, fig. 3 and table 1, the performance of the detection method of the present invention is superior to the other two detection methods, which indicates that the change detection method of the present invention is effective.

TABLE 1 comparison of multi-temporal SPOT5 image change detection results in Littoral area

Method of producing a composite material	FP	FN	OE	k
					CVA-EM	7919	3882	11801	0.705
FCM-S	1822	6928	8750	0.737
					The method of the invention	2511	4689	7200	0.797
Ideal for	0	0	0	1

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

In the several embodiments provided in the present application, it should be understood that the disclosed method may be implemented in other ways. For example, the above-described system embodiments are merely illustrative, and those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, the components and steps of which have been generally described in terms of their functionality in the foregoing description for clarity of explanation of interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (8)

1. A method for detecting remote sensing image change is characterized by comprising the following steps:

obtaining two-time-phase high-resolution optical remote sensing image X₁And X₂；

To optical remote sensing image X₁And X₂Carrying out image registration;

optical remote sensing image X by using multivariate change detection method₁And X₂Carrying out radiation normalization correction;

optical remote sensing image X after normalization correction according to radiation₁And X₂Are respectively provided withCalculating the magnitude X of the variation vector_MAnd spectral angular information X_SA；

According to the magnitude X of the variation vector_MCalculating to obtain an optimal segmentation threshold value T by using a Bayes principle and a maximum expectation algorithm;

according to the optimal segmentation threshold value T and the variation vector amplitude value X_MSelecting a pseudo-training sample region;

mixing X_MAnd X_SACombining the input data as the kernel FCM, and selecting the optimal model parameter value of the kernel FCM combined with the spatial neighborhood information model according to the pseudo training sample region;

and determining a change region and a non-change region of the optical remote sensing image by adopting a method of combining kernel FCM with space neighborhood information according to the selected optimal model parameter value.

2. The method for detecting changes in remote-sensing images of claim 1, wherein the pair of optical remote-sensing images X₁And X₂Performing image registration, specifically comprising:

optical remote sensing image X by adopting ENVI14.8 remote sensing software₁And X₂Carrying out geometric rough correction;

roughly correcting the geometry of the X by automatic matching and triangulation₁And X₂And carrying out geometric fine correction.

3. A method for detecting changes in remote-sensing images as claimed in claim 1, characterized in that said optical remote-sensing image X is detected by a multivariate change detection method₁And X₂Performing radiation normalization correction, specifically comprising:

obtaining optical remote sensing image X₁And X₂Linearly combining the brightness values of all the wave bands to obtain a difference image with enhanced change information;

determining a changed area and an unchanged area according to the difference image through a threshold;

and finishing the relative radiation correction through a mapping equation of the two-time phase pixel corresponding to the unchanged area.

4. The method for detecting changes in remote-sensing images of claim 1, wherein the optical remote-sensing image X corrected by radiation normalization is used₁And X₂Respectively calculating the variation vector amplitude X_MAnd spectral angular information X_SAThe method specifically comprises the following steps:

optical remote sensing image X after normalization correction according to radiation₁And X₂Calculating to obtain the amplitude X of the variation vector_MWherein $X_M(i,j)=\sqrt{\underset{b=1}{\overset{B}{\Σ}}{\left(X_{1b}\right(i,j)-X_{2b}(i,j\left)\right)}^2},$ wherein B represents the number of wave bands of each time phase remote sensing image, (i, j) is the coordinate of the image, X_1bRepresents X₁B band image of (1), X_2bRepresents X₂B-band images of (1);

optical remote sensing image X after normalization correction according to radiation₁And X₂Calculating to obtain the amplitude X of the variation vector_MWherein $X_{SA}(i,j)=arccos\left(\underset{b=1}{\overset{B}{\Σ}}\right(X_{1b}\left(i,j\right)X_{2b}\left(i,j\right))/\sqrt{\underset{b=1}{\overset{B}{\Σ}}{X}_{1b}^2(i,j)}\sqrt{\underset{b=1}{\overset{B}{\Σ}}{X}_{2b}^2(i,j)}).$

5. the method for detecting changes in remote-sensing images of claim 1, wherein said vector magnitude X varies according to changes_MCalculating to obtain an optimal segmentation threshold value T by using a Bayes principle and a maximum expectation algorithm, and specifically comprising the following steps of:

estimating X using a maximum expectation algorithm_MUnchanged class omega on image_nMean value m of_nSum variance σ_nOf the variation class omega_cMean value m of_cThe sum variance is σ_cWherein

$m_n^{t+1}=\left\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}X(i,j)\right\}/\left\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}\right\}$

${\left({\mathrm{\σ}}_n^2\right)}^{t+1}=\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}\[X(i,j)-m_n^t\]\}/\left\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}\right\}$

$m_c^{t+1}=\left\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_c\left)\right)}{p^t\left(X\right(i,j\left)\right)}X(i,j)\right\}/\left\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_c\left)\right)}{p^t\left(X\right(i,j\left)\right)}\right\}$

${\left({\mathrm{\σ}}_c^2\right)}^{t+1}=\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_c\left)\right)}{p^t\left(X\right(i,j\left)\right)}\[X(i,j)-m_c^t\]\}/\left\{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_c\left)\right)}{p^t\left(X\right(i,j\left)\right)}\right\}$

wherein t represents the number of iterations, t superscript represents the value at the t-th iteration of the current content,

$p^{t+1}\left({\mathrm{\ω}}_n\right)=\frac{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_n\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_n\left)\right)}{p^t\left(X\right(i,j\left)\right)}}{IJ},p^{t+1}\left({\mathrm{\ω}}_c\right)=\frac{\underset{X(i,j)\∈X_M}{\Σ}\frac{p^t\left({\mathrm{\ω}}_c\right)p^t\left(X\right(i,j|{\mathrm{\ω}}_c\left)\right)}{p^t\left(X\right(i,j\left)\right)}}{IJ},$

i and J represent the number of rows and columns of the image respectively,represents X_MUnchanged class omega on image_nThe gaussian distribution that is obeyed to,represents X_MVariation class omega on image_cA gaussian distribution obeyed;

solving a formula according to Bayes minimum error criterion $({\mathrm{\σ}}_n^2-{\mathrm{\σ}}_c^2)T^2+2(m_n{\mathrm{\σ}}_c^2-m_c{\mathrm{\σ}}_n^2)T+m_c^2{\mathrm{\σ}}_n^2-m_n^2{\mathrm{\σ}}_c^2-2{\mathrm{\σ}}_n^2{\mathrm{\σ}}_c^{2}ln\[\frac{{\mathrm{\σ}}_{c}p\left({\mathrm{\ω}}_n\right)}{{\mathrm{\σ}}_{n}p\left({\mathrm{\ω}}_c\right)}\]=0,$ And obtaining an optimal segmentation threshold value T.

6. The method for detecting changes in remote-sensing images of claim 1, wherein said determining is based on an optimal segmentation threshold T and a change vector magnitude X_MSelecting a pseudo-training sample region, specifically comprising:

according to the optimal segmentation threshold value T and the variation vector amplitude value X_MSelecting unchanged pseudo training set samples as

According to the optimal segmentation threshold value T and the variation vector amplitude value X_MSelecting a change-like pseudo-training setThe sample is

Wherein is X_M15% of the dynamic range.

7. The method for detecting changes in remote-sensing images of claim 1, wherein said detecting X is performed by a computer_MAnd X_SACombining the kernel FCM and the pseudo training sample region to be used as input of the kernel FCM, and performing optimal model parameter selection on the kernel FCM and the spatial neighborhood information model according to the pseudo training sample region, wherein the optimal model parameter selection specifically comprises the following steps:

mixing X_MAnd X_SAThe combination is used as the input of a kernel FCM, and the kernel FCM and space neighborhood information combination model is constructed as follows:

$J_m=\underset{j=1}{\overset{C}{\Σ}}\underset{k=1}{\overset{N}{\Σ}}{u}_{jk}^m(1-K(X_{MS}\left(k\right),v_j\left)\right)+\mathrm{\α}\underset{j=1}{\overset{C}{\Σ}}\underset{k=1}{\overset{N}{\Σ}}{u}_{jk}^m(1-K({\stackrel{\‾}{X}}_{MS}\left(k\right),v_j\left)\right),$ where C is the number of clusters, N is the total number of samples,representing the fuzzy membership of the kth sample to the jth class center, m being the weighted index of the membership, u_jk∈[0,1]And isThe parameter α controls the penalty effect,is X_MLocal mean image and X_SAThe local mean value image of (a) is combined,

setting the value ranges of the parameter α and the kernel parameter g, and searching the change index C by using a pseudo training sample set_indexα and g as the minimum value as the optimal model parameter value, wherein the change index C_index＝D_index/k_T，k_TKappa coefficient, N, representing model parameters on a pseudo-training sample set_n(α, g) shows minimizing the number of unchanged pixels of the entire image acquired using the objective function given α and g, N_c(α, g) shows the number of pixels that change for the entire image given α and g, TN_n(α, g) indicates the number of unchanged pixels in the pseudo-training sample set given α and g, TN_c(α, g) shows the number of changed pixels in the pseudo training sample set given α and g.

8. The method for detecting changes in remote-sensing images according to claim 7, wherein the determining of the changed area and the unchanged area of the high-resolution optical remote-sensing image by using a method of combining kernel FCM with spatial neighborhood information according to the selected optimal model parameter values specifically comprises:

setting the clustering number C of the kernel FCM in combination with the spatial neighborhood information model to be 2, taking the clustering number C as the center of the initial unchanged class and the initial changed class, and selecting and changing the vector amplitude X_MSetting a weighting index m of the membership degree to be 2, wherein the weighting index m is a constant greater than 0, and the values of the parameter α and the kernel parameter g are selected as the optimal model parameter values;

calculating X_M，X_SAThe window size is set to 3 × 3;

adopt the formula $u_{jk}=\frac{{\left(\right(1-K\left(X_{MS}\right(k),v_j))+\mathrm{\α}(1-K\left({\stackrel{\‾}{X}}_{MS}\right(k),v_j)\left)\right)}^{-1/(m-1)}}{\underset{j=1}{\overset{C}{\Σ}}{\left(\right(1-K\left(X_{MS}\right(k),v_j))+\mathrm{\α}(1-K\left({\stackrel{\‾}{X}}_{MS}\right(k),v_j)\left)\right)}^{-1/(m-1)}}$ Updating the fuzzy partition matrix;

adopt the formula $v_j=\frac{\underset{k=1}{\overset{N}{\Σ}}{u}_{jk}^m\left(K\right(X_{MS}\left(k\right),v_j)X_{MS}\left(k\right)+\mathrm{\α}K\left({\stackrel{\‾}{X}}_{MS}\right(k),v_j){\stackrel{\‾}{X}}_{MS}\left(k\right))}{\underset{k=1}{\overset{N}{\Σ}}{u}_{jk}^m\left(K\right(X_{MS}\left(k\right),v_j)+\mathrm{\α}K({\stackrel{\‾}{X}}_{MS}\left(k\right),v_j\left)\right)}$ Updating the clustering center;

repeatedly updating the fuzzy partition matrix and the clustering center until the clustering center of two adjacent iterations is smaller than the threshold;

partition matrix u from the blur_jkAnd determining a final change detection image to obtain a change area and a non-change area of the optical remote sensing image.