CN106485693A - Card side converts the multi-temporal remote sensing image change detecting method with reference to MRF model - Google Patents

Abstract

The invention discloses a multi-temporal remote sensing image change detection method based on chi-square transformation combined with an MRF model. The steps are as follows: inputting multi-temporal remote sensing images, image registration, radiation normalization correction, calculating multi-temporal difference images, iterative CST transformation detection, Input the MRF model, and obtain the final change detection result based on the modulus value of the difference image. The invention overcomes the problems that the prior art is difficult to solve the problems of complex background information and serious noise interference of high spatial resolution remote sensing images.

Description

Multi-temporal remote sensing image change detection method based on chi-square transform combined with MRF model

technical field

The invention belongs to the technical field of remote sensing image processing, and in particular relates to a multi-temporal remote sensing image change detection method based on a chi-square transformation combined with an MRF model.

Background technique

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 these remote sensing data has become an important research topic in remote sensing science and geographic information science. According to remote sensing images of different time phases in the same area, dynamic information such as cities and environments can be extracted to provide scientific decision-making basis for resource management and planning, environmental protection and other departments. my country's "Twelfth Five-Year Plan" will increase and expand the implementation of high-resolution earth observation projects that have been launched during the "Eleventh Five-Year Plan", focusing on basic theories including high-resolution remote sensing target and space environment feature analysis and high-reliability automatic interpretation And key technology research is becoming the focus of research to solve the major needs of national security and social and economic development.

Change detection of remote sensing images is to quantitatively analyze and determine the characteristics and process of surface changes from remote sensing data of different periods. Scholars from various countries have proposed many effective detection algorithms from different angles and applied research, such as Change Vector Analysis (CVA), clustering methods based on Fuzzy C-means (FCM), and so on. Among them, the traditional multi-temporal optical remote sensing change detection based on Chi-Squared Transform (CST), first calculates the mean value and variance matrix of the difference image, and then determines the threshold of change detection based on the confidence level, and then obtains the change detection result. In this type of technology, the disadvantage of using CST is that only the spectral information of the multi-temporal high-resolution difference image is used, and the spatial information is not used. Also, when calculating the mean and variance matrices of the difference images, the estimation accuracy is not high.

In view of the above problems, it is necessary to study new high-resolution visible light remote sensing image change detection technology to effectively overcome the above difficulties.

Contents of the invention

In order to solve the technical problems raised by the above-mentioned background technology, the present invention aims to provide a multi-temporal remote sensing image change detection method based on chi-square transform combined with MRF model, which overcomes the difficulty in solving the complex background information and serious noise interference of high-spatial-resolution remote sensing images in the prior art. The problem.

In order to realize above-mentioned technical purpose, technical scheme of the present invention is:

The chi-square transform combined with the MRF model multi-temporal remote sensing image change detection method includes the following steps:

(1) Input high-resolution optical remote sensing images of two temporal phases, denoted as X ₁ and X ₂ respectively;

(2) Carry out image registration on X ₁ and X ₂ ;

( ₃ ) Carry out radiation normalization correction _on X1 and X2 using multivariate change detection method;

(4) Calculate the multi-temporal difference image D _X =X ₁ -X ₂ ;

(5) Initialize the non-changing area in the multi-temporal difference image D _X , calculate the mean vector and variance matrix of the non-changing area, and calculate the chi-square value of each point on the multi-temporal difference image;

(6) On the basis of a given confidence level, calculate the detection threshold, and detect according to the detection threshold, and determine the changing area and non-changing area in the multi-temporal difference image;

(7) Compare the non-changing area determined in step (6) with the non-changing area determined in step (5), if the two are consistent, use the detection result obtained in step (6) as the change detection result, if the two are different , then use the non-changing region determined in step (6) as a new non-changing region, return to step (5), and loop iterations;

(8) The change detection result obtained in step (7) is used as the input of the MRF model, and the final change detection result is obtained based on the modulus of the multi-temporal difference image D _X .

Further, the specific process of step (8) is as follows:

(a) Calculate the modulus of the multi-temporal difference image D _X :

In the above formula, X _1b and X _2b represent the images of the b-th band of X ₁ and X ₂ respectively, B represents the number of bands of each time-phase remote sensing image, and (i, j) are the coordinates of the image;

(b) Construct the energy function of the MRF model:

In the above formula, U _data represents the data item, U _context represents the spatial local energy item, M ₁ and M ₂ represent the height and width of the image respectively, Y(i,j) represents the value of the coordinate (i,j) of the change detection result, Y ^S (i,j) is the neighborhood system of coordinates (i,j);

(c) The ICM optimization algorithm is used to solve the U minimization to obtain the final change detection result.

Further, U _data is further expressed as,

In the above formula, μ _Y(i,j) ∈ {μ _n , μ _c }, and μ _n denote the variance and mean of the non-changing region, respectively, and _μc denote the variance and mean of the changing region, respectively.

Further, it is characterized in that U _context is further expressed as,

In the above formula, (p, q) are the neighborhood coordinates of (i, j), β is a parameter to control the local energy item in the space, and I is an indicator function, which is defined as follows:

Further, in step (2), performing image registration on X ₁ and X ₂ includes geometric coarse correction and geometric fine correction, the process of geometric coarse correction:

(A) Select X ₁ and X ₂ as the reference image and the image to be corrected respectively;

(B) Collect ground control points on the reference image and the image to be corrected respectively, the number of ground control points is greater than or equal to 9, and the ground control points are evenly distributed on the image;

(C) Calculate the mean square error at each ground control point of the reference image and the image to be corrected;

(D) correcting the image to be corrected by polynomial correction method;

(E) resampling the image to be corrected by using bilinear interpolation;

The geometric fine correction is to use the automatic matching and triangulation method to correct the remote sensing image after the geometric rough correction.

Further, in step (5), the detection threshold is calculated using the following formula:

In the above formula, 1-α is the confidence level, C _ij represents the chi-square value of D _X at coordinates (i, j), is the detection threshold.

The beneficial effect brought by adopting the above-mentioned technical scheme:

The invention introduces the MRF model into the CST detection result, so that the change detection based on the CST transformation has the capability of spatial constraints. In the change detection, the iterative method is used to estimate the mean and standard deviation of the non-changing area, which overcomes the shortcomings of only using the entire multi-band difference image to calculate the mean and variance matrix, making the result of change detection more reliable and robust. .

Description of drawings

Fig. 1 is method flowchart of the present invention;

Figure 2(a) and 2(b) are the schematic diagram of the third band of the high-resolution IKONOS image of the Mina area in Saudi Arabia in January 2007, and the schematic diagram of the third band of the high-resolution IKONOS image of the Mina area in Saudi Arabia in December 2007 ;

Figure 3 is a reference image for change detection;

Figures 4(a), 4(b), 4(c), and 4(d) are schematic diagrams of the detection results of the EM-CVA algorithm, the ICST algorithm, the RCST algorithm, and the algorithm of the present invention, respectively.

detailed description

The technical solutions of the present invention will be described in detail below in conjunction with the accompanying drawings.

As shown in Figure 1, the multi-temporal remote sensing image change detection method based on chi-square transformation combined with MRF model, the steps are as follows:

Step 1: Input two high-resolution optical remote sensing images of the same area and different time phases, denoted as X ₁ and X ₂ respectively.

Step 2: Carry out image registration on X ₁ and X ₂ , which is divided into two steps: rough correction and fine correction:

For geometric rough correction, use the relevant functions in ENVI4.8 software to realize, the specific operation steps are:

(1) Display the reference image and the image to be corrected;

(2) Collect ground control points GCPs, GCPs should be evenly distributed in the entire image, and the number of GCPs should be at least greater than or equal to 9;

(3) Calculate the mean square error;

(4) Using polynomial correction method to correct the image to be corrected;

(5) Use bilinear interpolation for resampling output. If you want to find the value of the unknown function f at point P=(x,y), assume that the known function f is at Q ₁₁ =(x ₁ ,y ₁ ), Q ₁₂ = (x ₁ ,y ₂ ), Q ₂₁ ＝(x ₂ ,y ₁ ), and Q ₂₂ ＝(x ₂ ,y ₂ ), if a coordinate system is selected so that the coordinates of these four points are respectively (0,0), (0,1), (1,0) and (1,1), 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 For the precise geometric correction, the multispectral remote sensing image data that has undergone geometric rough correction is used for geometric fine correction by automatic matching and triangulation. Use the point-by-point interpolation method to construct the Delaunay triangulation network. For each triangle, use the row and column numbers of its three vertices and the geographical coordinates of the corresponding point of the same name in the reference image to determine the affine transformation model parameters inside the triangle, and correct the image to be corrected. , to get the corrected remote sensing image.

Step 3: Perform radiation normalization correction on X ₁ and X ₂ by using the multivariate alteration detection (Multivariate Alteration Detection, MAD) method. First find a linear combination of brightness values in each band of X ₁ and X ₂ to obtain a difference image with enhanced change information, determine the changed and unchanged areas through the threshold value, and then use the mapping equation of the two-temporal pixel pairs corresponding to the unchanged area to complete Relative radiometric correction.

Step 4: For the input multi-temporal high-resolution images X ₁ and X ₂ , calculate the multi-temporal difference image D _X :

D _X ＝X ₁ -X ₂

Step 5: Treat the entire D _X as a non-changing area, and calculate its mean vector m and variance matrix Σ, and calculate the chi-square value of each point of the difference image:

C _ij ＝(x _ij -m) ^T Σ ^-1 (x _ij -m)～χ ² (b)

In the above formula, C _ij represents the chi-square value of the (i, j) coordinate point of the difference image, which obeys the chi-square distribution with the degree of freedom b; x _ij represents the vector value of the difference image at the (i, j) coordinate point; Σ ^-1 indicates the inverse matrix of the variance matrix; b indicates the number of bands of the difference image.

Step 6: Given a confidence level 1-α, use the following formula to calculate the detection threshold:

When the confidence level is 1-α, the value of C _ij is greater than The probability is α. If the value of α is small, it is greater than The C _ij of can be regarded as an outlier or a change point, so the threshold is determined as And obtain preliminary detection results according to the threshold.

Step 7: Compare the non-changing area determined in the detection result of step 6 with the non-changing area (whole D _X ) determined in step 5, if both are consistent, use the detection result obtained in step 6 as the change detection result, if both or different, then use the non-changing area determined in step 6 as the new non-changing area, return to step 5, and loop iteratively.

Step 8: Use the change detection result obtained in step 7 as the input of the MRF model, and obtain the final change detection result based on the modulus of the difference image. The specific implementation steps are:

(1) First calculate the modulus X _M of D _X :

In the above formula, X _1b and X _2b represent the b-band images of the first time phase and the second time phase multispectral images X ₁ and X ₂ respectively, B represents the number of bands of each time phase remote sensing image, (i,j) are the coordinates of the image;

(2) The energy function for constructing the MRF model is as follows:

In the above formula, U _data represents the data item, U _context represents the spatial local energy item, M ₁ and M ₂ represent the height and width of the image respectively, Y(i,j) represents the value of the coordinate (i,j) of the change detection result, Y ^S (i,j) is the neighborhood system of coordinates (i,j).

Among them, the data item U _data can be further expressed as:

In the above formula, μ _Y(i,j) ∈ {μ _n ,μ _c }, and μ _n denote the variance and mean of the non-changing region, respectively, and _μc denote the variance and mean of the changing region, respectively.

Among them, the local local space energy item U _context can be further expressed as:

In the above formula, (p,q) are the neighborhood coordinates. Adopt 8 neighborhood systems in the present invention, β is the parameter of control space local energy item, and I is indicator function, is defined as follows:

(3) The ICM optimization algorithm is used to solve U minimization, and the final change detection result is obtained.

Effect of the present invention can be further illustrated by following experimental results and analysis:

The experimental data of the present invention is the multi-temporal IKNOS high-resolution image data of the Mina area of Saudi Arabia, and the image size is 700 × 950, using three bands of B1, B2 and B3, and Fig. 2 (a) and Fig. 2 (b) are two Temporal remote sensing images in the B3 band.

In order to verify the effectiveness of the present invention, the change detection method of the present invention is compared with the following change detection methods:

(1) CVA-based EM method (CVA-EM) [Italy's Bruzzone L. et al. in the article "Automatic analysis of difference image for unsupervised change detection" (IEEE Transactions on Geoscience and Remote Sensing, 2000, 38 (3): 1171-1182 .) in the proposed detection method].

(2) Based on the iterative CST detection (ICST) method [B.Desclée, P.Bogaert, and P.Defourny in the article "Forest change detection by statistical object-based method" (Remote Sensingof Environment, 2006, 102 (1-2 ): the method mentioned in 1-12.).

(3) CST detection based on robust estimation (RCST) method [Aiye Shi et al. in the article "Unsupervised change detection based on robust chi-squared transform for bittemporal remotely sensed images" (International of Remote Sensing, 2006, 102 (1-2): 1-12.) mentioned method].

(4) The method of the present invention.

Figure 3 is a reference image for change detection. Under the condition of a confidence level of 0.99, the detection results of the above four methods are shown in Figure 4(a), 4(b), 4(c), and 4(d). FN, the total number of errors OE and Kappa coefficient four indicators to measure. The closer FP, FN and OE are to 0, and the closer the Kappa coefficient is to 1, it indicates that the performance of the change detection method is better. The test results are shown in Table 1. It can be seen from Table 1 that the performance of the detection method proposed by the present invention is better than the other three detection methods, which indicates that the change detection method proposed by the present invention is effective.

Table 1

The above embodiments are only to illustrate the technical ideas of the present invention, and can not limit the protection scope of the present invention with this. All technical ideas proposed in accordance with the present invention, any changes made on the basis of technical solutions, all fall within the protection scope of the present invention. Inside.

Claims (6)

1. The multi-temporal remote sensing image change detection method combined with the chi-square transformation of the MRF model, is characterized in that, comprising the following steps: (1) Input high-resolution optical remote sensing images of two temporal phases, denoted as X ₁ and X ₂ respectively; (2) Carry out image registration on X ₁ and X ₂ ; (3) Carry out radiation normalization correction on X ₁ and X ₂ respectively by using multivariate change detection method; (4) Calculate the multi-temporal difference image D _X =X ₁ -X ₂ ; (5) Initialize the non-changing area in the multi-temporal difference image D _X , calculate the mean vector and variance matrix of the non-changing area, and calculate the chi-square value of each point on the multi-temporal difference image; (6) On the basis of a given confidence level, calculate the detection threshold, and detect according to the detection threshold, and determine the changing area and non-changing area in the multi-temporal difference image; (7) Compare the non-changing area determined in step (6) with the non-changing area determined in step (5), if the two are consistent, use the detection result obtained in step (6) as the change detection result, if the two are different , then use the non-changing region determined in step (6) as a new non-changing region, return to step (5), and loop iterations; (8) The change detection result obtained in step (7) is used as the input of the MRF model, and the final change detection result is obtained based on the modulus of the multi-temporal difference image D _X . 2. according to the described Chi-square transformation of claim 1 in conjunction with the multi-temporal remote sensing image change detection method of MRF model, the concrete process of step (8) is as follows: (a) Calculate the modulus of the multi-temporal difference image D _X : ${\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\sqrt{\underset{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; B</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}$ In the above formula, X _1b and X _2b represent the images of the b-th band of X ₁ and X ₂ respectively, B represents the number of bands of each time-phase remote sensing image, and (i, j) are the coordinates of the image; (b) Construct the energy function of the MRF model: $\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\underset{\mathrm{<span\; class="notranslate">\; j</span>}}{\overset{{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; Y</span>}}^{\mathrm{<span\; class="notranslate">\; S</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>$ In the above formula, U _data represents the data item, U _context represents the spatial local energy item, M ₁ and M ₂ represent the height and width of the image respectively, Y(i,j) represents the value of the coordinate (i,j) of the change detection result, Y ^S (i,j) is the neighborhood system of coordinates (i,j); (c) The ICM optimization algorithm is used to solve the U minimization to obtain the final change detection result. 3. according to the described Chi-square transformation of claim 2 in conjunction with the multi-temporal remote sensing image change detection method of MRF model, it is characterized in that: U _data is further expressed as, ${\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&pi;\&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}{<span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; \&rsqb;</span>}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}$ In the above formula, μ _Y(i,j) ∈ {μ _n , μ _c }, and μ _n denote the variance and mean of the non-changing region, respectively, and _μc denote the variance and mean of the changing region, respectively. 4. according to claim 2 said chi-square transformation combined with the multi-temporal remote sensing image change detection method of MRF model, it is characterized in that: U _context is further expressed as, ${\mathrm{<span\; class="notranslate">\; u</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; no</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; x</span>}\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; Y</span>}}^{\mathrm{<span\; class="notranslate">\; S</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; q</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Element;</span>{\mathrm{<span\; class="notranslate">\; Y</span>}}^{\mathrm{<span\; class="notranslate">\; S</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>}{<span\; class="notranslate">\; \&Sigma;</span>}\mathrm{<span\; class="notranslate">\; \&beta;</span>}\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; q</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>$ In the above formula, (p, q) are the neighborhood coordinates of (i, j), β is the parameter to control the local energy item in the space, and I is the indicator function, which is defined as follows: $\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; q</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}& \mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; q</span>}<span\; class="notranslate">\; )</span>\\ \mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; h</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; w</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; the\; s</span>}\mathrm{<span\; class="notranslate">\; e</span>}\end{array}\right.<span\; class="notranslate">\; .</span>$ 5. according to claim 1 said chi-square transformation combined with the multi-temporal remote sensing image change detection method of MRF model, it is characterized in that: in step (2), carrying out image registration to X ₁ and X ₂ includes geometric rough correction and geometric Fine correction, the process of geometric coarse correction: (A) Select X ₁ and X ₂ as the reference image and the image to be corrected respectively; (B) Collect ground control points on the reference image and the image to be corrected respectively, the number of ground control points is greater than or equal to 9, and the ground control points are evenly distributed on the image; (C) Calculate the mean square error at each ground control point of the reference image and the image to be corrected; (D) correcting the image to be corrected by polynomial correction method; (E) resampling the image to be corrected by using bilinear interpolation; The geometric fine correction is to use the automatic matching and triangulation method to correct the remote sensing image after the geometric rough correction. 6. according to the described Chi-square transformation of claim 1 in conjunction with the multi-temporal remote sensing image change detection method of MRF model, it is characterized in that, in step (5), adopt following formula to calculate detection threshold: $\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; \&chi;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}$ In the above formula, 1-α is the confidence level, C _ij represents the chi-square value of D _X at coordinates (i, j), is the detection threshold.