InSAR subsidence result image processing method and device, readable medium and electronic equipment

Abstract

The application belongs to the technical field of image processing, and discloses an InSAR subsidence result image processing method and device, a readable medium and an electronic equipment. The method can effectively improve subsidence monitoring accuracy, retain more detail information, guarantee data continuity, improve processing efficiency and applicability, and provide more reliable technical support for geological disaster monitoring and sustainable development of cities.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of image processing, in particular to an InSAR subsidence result image processing method and device, readable medium and electronic equipment. BACKGROUND

[0002] With the development of technology, InSAR (Interferometric Synthetic Aperture Radar) technology has been widely used in the field of ground subsidence monitoring. However, the existing InSAR subsidence result data processing method has some limitations. The traditional global detrending method is difficult to effectively eliminate the trend item in the non-uniform subsidence area, and is easy to cause the loss of local detail information. At the same time, the manual selection of the detrending area not only has low efficiency, but also has strong subjectivity, which affects the accuracy and reliability of subsidence monitoring. Therefore, there is an urgent need for a new InSAR subsidence result image processing method to improve the accuracy and efficiency of subsidence monitoring, and better meet the needs of geological disaster monitoring and urban sustainable development. SUMMARY

[0003] The present application provides an InSAR subsidence result image processing method, device, readable medium and electronic equipment to improve the accuracy and efficiency of InSAR subsidence monitoring.

[0004] In a first aspect, the present application provides an InSAR subsidence result image processing method, which comprises:

[0005] calculating a first subsidence gradient and a second subsidence gradient of an InSAR subsidence result image; wherein the first subsidence gradient is the subsidence gradient in the horizontal direction, and the second subsidence gradient is the subsidence gradient in the vertical direction;

[0006] dividing the InSAR subsidence result image into a plurality of image regions according to the first subsidence gradient and the second subsidence gradient of the InSAR subsidence result image;

[0007] for each image region, determining the trend information of the image region by using a preset linear model, and performing detrending information processing on the image region by using the trend information of the image region to obtain a processed image region;

[0008] splicing each processed image region, and performing fusion processing on the boundaries between each processed image region to obtain a processed InSAR subsidence result image.

[0009] In a second aspect, the present application provides an InSAR subsidence result image processing device, which comprises:

[0010] The first unit is configured to calculate a first subsidence gradient and a second subsidence gradient of an InSAR subsidence result image, wherein the first subsidence gradient is a subsidence gradient in a horizontal direction, and the second subsidence gradient is a subsidence gradient in a vertical direction.

[0011] The second unit is configured to divide the InSAR subsidence result image into a plurality of image regions according to the first subsidence gradient and the second subsidence gradient of the InSAR subsidence result image.

[0012] The third unit is configured to, for each image region, determine trend information of the image region by using a preset linear model, and perform trend information removal processing on the image region by using the trend information of the image region to obtain a processed image region.

[0013] The fourth unit is configured to stitch the processed image regions, and perform fusion processing on boundaries between the processed image regions to obtain a processed InSAR subsidence result image.

[0014] In a third aspect, the present application provides a readable medium comprising execution instructions, when a processor of an electronic device executes the execution instructions, the electronic device executes the method according to any one of the first aspect.

[0015] In a fourth aspect, the present application provides an electronic device comprising a processor and a memory storing execution instructions, when the processor executes the execution instructions stored in the memory, the processor executes the method according to any one of the first aspect.

[0016] As can be seen from the above technical solutions, the present application has the following beneficial effects compared with the prior art:

[0017] 1. Improved subsidence monitoring precision: By dividing the InSAR image into image regions according to the first subsidence gradient and the second subsidence gradient of the InSAR subsidence result image, the subsidence characteristics of different image regions can be targeted for trend removal processing. Compared with the traditional global trend removal method, the trend item can be more effectively eliminated, the influence of system errors such as atmospheric delay and orbit error on the subsidence result can be reduced, and thus the subsidence monitoring precision is improved, providing more reliable data support for geological disaster monitoring and the like.

[0018] 2. More details are retained: The problem of local detail information loss caused by global trend removal is avoided, the local subsidence characteristics of each image region are better retained, which is helpful for more comprehensively and accurately understanding the subtle changes of the surface subsidence, and is of great significance for analyzing the subsidence reasons and predicting the subsidence development trend.

[0019] 3. Ensuring data continuity: after the trend information of the image area is removed by using the trend information of the image area, the boundaries between each processed image area are fused, which can effectively eliminate the traces of the boundaries of the spliced image areas, ensure the continuity and integrity of the entire settlement result data, make the final settlement result more smooth and natural, and conform to the actual situation, which is convenient for subsequent data analysis and application.

[0020] Further effects of the above-described non-conventional preferred modes will be described below in conjunction with the specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present application or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings described below are only some embodiments described in the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0022] Figure 1 A flowchart of an InSAR settlement result image processing method provided by the present application is shown in the figure;

[0023] Figure 2 A flowchart of a first settlement gradient and a second settlement gradient calculation provided by the present application is shown in the figure;

[0024] Figure 3 A flowchart of a first settlement gradient and a second settlement gradient calculation provided by the present application is shown in the figure;

[0025] Figure 4 A flowchart of a first settlement gradient and a second settlement gradient calculation provided by the present application is shown in the figure;

[0026] Figure 5 A flowchart of a first settlement gradient and a second settlement gradient calculation provided by the present application is shown in the figure;

[0027] Figure 6 A flowchart of a first settlement gradient and a second settlement gradient calculation provided by the present application is shown in the figure;

[0028] Figure 7 A structural diagram of an InSAR settlement result image processing device provided by the present application is shown in the figure;

[0029] Figure 8 A structural diagram of an electronic device provided by the present application is shown in the figure. DETAILED DESCRIPTION

[0030] In order to make the purposes, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described below in a clear and complete manner with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0031] Various non-limiting embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0032] Referring to Figure 1 , a method for processing an InSAR subsidence result image is shown in an embodiment of the present application. In the present embodiment, the method may, for example, include the following steps:

[0033] S101: Calculate a first subsidence gradient and a second subsidence gradient of the InSAR subsidence result image.

[0034] The first subsidence gradient is a subsidence gradient in a horizontal direction, i.e., a subsidence gradient of the subsidence result in an x horizontal direction; and the second subsidence gradient is a subsidence gradient in a vertical direction, i.e., a subsidence gradient of the subsidence result in a y vertical direction. It can be understood that the first subsidence gradient and the second subsidence gradient of the InSAR subsidence result image include the first subsidence gradient and the second subsidence gradient of each pixel point in the InSAR subsidence result image.

[0035] In an implementation manner, the first subsidence gradient and the second subsidence gradient of each pixel point in the InSAR subsidence result image can be calculated by using a Sobel operator or a Prewitt operator. That is, the Sobel operator or the Prewitt operator is used to calculate the subsidence gradient of the InSAR image in the horizontal and vertical directions. As an example, assuming that the Sobel operator is used, each operator is a 3x3 window, and the window is slid in the InSAR subsidence result image, from left to right and from top to bottom, moving one pixel at a time, the value of the window is multiplied with the pixel value of the InSAR subsidence result image, and all the products are added again, and the sum is taken as the pixel value at the center of the window.

[0036] The gradient in the X direction in the Sobel operator is:

[0037] -1 0 1 -2 0 2 -1 0 1

[0038] The gradient in the Y direction in the Sobel operator is:

[0039] -1 -2 -1 0 0 0 1 2 1

[0040] The gradient in the X direction in the Prewitt operator is:

[0041] -1 0 1 -1 0 1 -1 0 1

[0042] The gradient in the Y direction in the Prewitt operator is:

[0043] -1 -1 -1 0 0 0 1 1 1

[0044] Two values are obtained for each pixel point on the post-sedimentation result by calculation, which are the gradient values in the X direction and the Y direction.

[0045] Next, for example, it is assumed that the pixel value of the InSAR sedimentation result image is as shown in Figure 1 , the sobel operator is calculated to obtain the result as shown in Figure 3 . Then, the window is slid one pixel point to the right to obtain the result as shown in Figure 4 . Next, the window is slid one pixel point downward from the position of Figure 3 to obtain the result as shown in Figure 5 . Then, the window is slid one pixel point to the right from the position of Figure 5 to obtain the result as shown in Figure 6 . Thus, the first sedimentation gradient and the second sedimentation gradient of the InSAR sedimentation result image are obtained (as shown in Table 1 below), two values exist at each position, which are the gradient in the horizontal direction and the gradient in the vertical direction, and the edge position of the InSAR sedimentation result image can be set as (0, 0) or NaN.

[0046] Table 1

[0047] NaN NaN NaN NaN NaN （8，32） （8，32） NaN NaN （8，32） （12，26） NaN NaN NaN NaN NaN

[0048] S102: According to the first sedimentation gradient and the second sedimentation gradient of the InSAR sedimentation result image, the InSAR sedimentation result image is divided into a plurality of image regions.

[0049] In this embodiment, the InSAR sedimentation result image can be divided into a plurality of image regions according to the first sedimentation gradient and the second sedimentation gradient of the InSAR sedimentation result image.

[0050] Specifically, for each pixel point in the InSAR sedimentation result image, a feature vector of the pixel point can be generated according to the first sedimentation gradient and the second sedimentation gradient of the pixel point. That is, for each pixel point (i, j), the horizontal direction gradient gx[i][j] and the vertical direction gradient gy[i][j] thereof are combined into a feature vector: features[i][j]= [gx[i][j], gy[i][j]].

[0051] Then, the feature vectors of all pixel points in the InSAR subsidence result image can be clustered based on the K-means clustering algorithm to obtain a clustering result; wherein the clustering result is K clusters. K-means is an iterative clustering algorithm, and its goal is to divide a dataset into K clusters, so that each data point belongs to the cluster closest to it, and the centroid of each cluster is the mean of all data points in the cluster. The specific algorithm steps are as follows:

[0052] (1) Select K value: determine how many clusters the data will be divided into;

[0053] (2) Select K value: determine how many clusters the data will be divided into;

[0054] (3) For each data point in the dataset, calculate its distance from the K centroids;

[0055] (4) Assign each data point to the cluster whose centroid is closest to it;

[0056] (5) For each cluster, recalculate the centroid, which is the mean of all data points in the cluster;

[0057] (6) Repeat steps (3), (4), and (5) until the movement of the centroids is less than a certain threshold.

[0058] Next, according to the clustering result, the clustering effect index corresponding to the clustering result can be determined, and the clustering result can be optimized according to the clustering effect index corresponding to the clustering result to obtain an optimized clustering result. The clustering effect index is one of the following: Silhouette Coefficient, Calinski-Harabasz index.

[0059] Silhouette Coefficient is an index for evaluating the clustering effect, which considers the cohesion and separation of clusters. For each data point i, the calculation formula of its Silhouette Coefficient si is as follows:

[0060] si = (bi - ai) / max(ai, bi)

[0061] Where ai is the average distance of data point i from other data points in its own cluster, and bi is the minimum average distance of data point i from all data points in other clusters.

[0062] The silhouette coefficient ranges from -1 to 1. The closer the silhouette coefficient is to 1, the better the clustering effect is. The closer the silhouette coefficient is to -1, the worse the clustering effect is. The closer the silhouette coefficient is to 0, the general clustering effect is.

[0063] The Calinski-Harabasz index, also known as the variance ratio criterion, evaluates the clustering effect by calculating the ratio of inter-cluster dispersion and intra-cluster dispersion. The calculation formula of the Calinski-Harabasz index is as follows:

[0064] CH = (SSB / (K - 1)) / (SSW / (N - K))

[0065] SSB is the inter-cluster dispersion, representing the weighted sum of squared distances between the center points of all clusters and the center point of the entire data set; SSW is the intra-cluster dispersion, representing the sum of squared distances between each data point and the center point of its own cluster; K is the number of clusters; N is the number of data points, and the larger the CH (Calinski-Harabasz) index, the better the clustering effect is.

[0066] Finally, the InSAR subsidence result image can be divided into multiple image regions according to the optimized clustering result. For example, if the optimized clustering result is N clusters, the InSAR subsidence result image can be divided into N image regions, and N is greater than 1.

[0067] That is, the K-means clustering algorithm can be used to cluster the feature vectors and divide the InSAR image into K regions. The selection of K value can be optimized based on indicators such as the silhouette coefficient or the Calinski-Harabasz index (CH index), such as setting a range of K values such as [5, 20]. After the K-means clustering algorithm is used to divide the regions, the silhouette coefficient or CH index of each classification is calculated, and the K value used when the silhouette coefficient or CH index is the largest is selected.

[0068] S103: For each image region, a predetermined linear model is used to determine the trend information of the image region, and the image region is subjected to trend information removal processing using the trend information of the image region to obtain a processed image region.

[0069] In the embodiment, the trend information of the image region can be determined by using a preset linear model. It should be noted that the atmospheric data and the orbit data are both data that need to be input in the InSAR processing. If the input data is inaccurate, the obtained result is inaccurate. If the input data has errors, the obtained result also has errors. The trend information can be understood as an error term of the InSAR result caused by inaccurate input orbit data, that is, the trend information is error information of the InSAR subsidence result image caused by inaccurate data.

[0070] Specifically, the pixel change trend type of the image region can be determined first. If the pixel change trend type of the image region is a first change trend type, a quadratic linear model is obtained by using a least square method fitting, and the quadratic linear model is used as the preset linear model. The first change trend type is a change in which the pixel value of a pixel point of the image region increases or decreases from one direction. The quadratic linear model is z = a * x^2 + b * y^2 + c * x * y + d * x + e * y + f, where z represents a fitting value of the quadratic linear model, that is, the trend information, a represents a quadratic term coefficient of x, that is, the influence degree of the square term of x on z, x represents a row number, b represents a quadratic term coefficient of y, that is, the influence degree of the square term of y on z, c represents a coefficient of x * y, that is, the influence degree of the product term of x and y on z, y represents a column number, d represents a linear term coefficient of x, that is, the influence degree of x on z, e represents a linear term coefficient of y, that is, the influence degree of y on z, and f represents a constant term, that is, an intercept.

[0071] If the pixel change trend type of the image area is the second change trend type, the least squares method is used to fit a cubic linear model, and the cubic linear model is used as the preset linear model. The second change trend type is that the pixel values ​​of the pixel points in the image area increase or decrease in multiple directions. The cubic linear model is z = a * x^3 + b * y^3 + c * x^2 * y + d * x * y^2 + e * x^2 + f * y^2 + g * x * y+ h * x + i * y + j; wherein z represents the fitting value of the cubic linear model, that is, the trend information, a represents the cubic term coefficient of x, that is, the degree of influence of the cubic term of x on z, x represents the row number, b represents the cubic term coefficient of y, that is, the degree of influence of the cubic term of y on z, c represents the influence of the first product term of y and the square of x on z, y represents the column number, and d represents the influence of the first product term of x and the square of y on z. , e represents the quadratic term coefficient of x, that is, the influence of the square term of x on z, f represents the quadratic term coefficient of y, that is, the influence of the square term of y on z, g represents the influence of the product term of x and y on z, h represents the linear term coefficient of x, that is, the influence of x on z, i represents the linear term coefficient of y, that is, the influence of y on z, and j represents the constant term, that is, the intercept.

[0072] Least squares fitting can be understood as using the least squares method to estimate the model parameters of a selected linear model, minimizing the sum of squared residuals. Taking a quadratic polynomial model as an example, the goal is to find parameters a, b, c, d, e, and f that minimize the following expression: sum((z_i - (a * x_i^2 + b * y_i^2 + c * x_i * y_i + d* x_i + e * y_i + f))^2), where (x_i, y_i) are the coordinates of the pixel and z_i is the amount of sedimentation at that pixel.

[0073] Finally, the row and column numbers of the image region may be input into the preset linear model to obtain trend information of the image region.

[0074] That is, the application eliminates the trend item after each partition by K-means clustering algorithm on the settlement results, and the specific method is to use least squares to fit a linear model, and subtract the linear model from the settlement amount. The linear model can be set as a quadratic or cubic linear model. If the trend is complex, it is set as cubic, otherwise it is set as quadratic. The linear model can also be set as a cubic linear model. It can be understood that if the result is an increasing or decreasing change from one direction, it is a simple trend, which can be set as 2 times. If it increases or decreases from multiple directions, it is complex, which can be set as 3 times. It can also be set as 3 times uniformly, because if it is a simple trend, the cubic term of the linear model can be 0, which is the same as the quadratic linear model.

[0075] In this way, a linear model is obtained by least squares fitting, and the input of the linear model is the row and column number, and the output is the trend information. The row and column numbers of the entire image area are input, and the trend information of the entire image area is obtained by the linear model. The image area is subtracted to obtain the image area after removing the trend information.

[0076] After determining the trend information of the image area, the trend information of the image area can be used to remove the trend information of the image area to obtain a processed image area. For example, the pixel value of the image area can be subtracted from the trend information of the image area to obtain a processed image area.

[0077] S104: Splicing each processed image area, and performing fusion processing on the boundaries between each processed image area to obtain a processed InSAR settlement result image.

[0078] In this embodiment, the boundaries between each processed image area can be fused to obtain a processed InSAR settlement result image.

[0079] As an example, for each image region, a pixel buffer of the image region is determined according to a boundary between the image region and a neighboring image region. A first weight and a second weight respectively corresponding to each pixel point in the pixel buffer are determined according to a distance between each pixel point and the boundary, i.e. W1=1-d, W2=d. The pixel value of each pixel point is adjusted according to the first weight and the second weight respectively corresponding to each pixel point, the pixel value of each pixel point in the InSAR subsidence result image and the pixel value of each pixel point in the processed image region, to obtain an adjusted image region. The pixel buffer can be understood as a region of a pixel point set within a preset number of pixel points from the boundary. It can be understood that the first weight and the second weight respectively corresponding to each pixel point are preset, for example, as shown in Table 2.

[0080] For example, after the detrending information operation for each image region, all the image regions need to be mosaicked together. Direct mosaicking will cause obvious traces of mosaicking boundaries. In order to eliminate the traces of mosaicking boundaries, the pixel values of n pixel points (i.e. pixel buffer) inward from the boundary of each image region are recalculated, and an increasing and decreasing weight is assigned to the original InSAR result and the InSAR result after removing the trend, and the specific weight is calculated as follows.

[0081] If a pixel buffer (i.e. pixel buffer) of n is established, the step of increasing or decreasing the weight is 1 / n. Table 2 below is the weight of establishing a 20-pixel buffer:

[0082] Table 2

[0083] Distance from border (pixels) 1 2 3 ... 19 20 Weight of original result 0 0.05 0.1 ... 0.95 1 Weight of result with trend removed 1 0.95 0.9 ... 0.05 0

[0084] The pixel value of each pixel point on the 20 pixels is equal to the value of the original result (i.e. the pixel value of the pixel point in the InSAR subsidence result image) x the weight of the original result (i.e. the first weight) + the value of the result after removing the trend (i.e. the pixel value of the pixel point in the processed image region) x the weight of the result after removing the trend (i.e. the second weight).

[0085] Finally, the processed InSAR subsidence result image can be obtained according to all the adjusted image regions. That is, the partition after removing the trend and the boundary processing is mosaicked together to obtain the final subsidence result.

[0086] From the above technical solutions, the present application has the following beneficial effects compared with the prior art:

[0087] 1. Improve the accuracy of subsidence monitoring: By dividing the InSAR image into different regions according to the first and second subsidence gradients of the InSAR subsidence result image, the subsidence characteristics of different image regions can be processed specifically. Compared with the traditional global detrending method, the influence of system errors such as atmospheric delay and orbit error on the subsidence result can be effectively eliminated, thereby improving the accuracy of subsidence monitoring and providing more reliable data support for geological disaster monitoring.

[0088] 2. Retain more detailed information: The problem of losing local detailed information caused by global detrending is avoided, and the local subsidence characteristics of each image region are better preserved, which helps to better understand the subtle changes of surface subsidence and is of great significance for analyzing the causes of subsidence and predicting the development trend of subsidence.

[0089] 3. Ensure data continuity: After removing the trend information of the image region using the trend information of the image region, the boundaries between the processed image regions are fused, which can effectively eliminate the traces of the boundaries of the spliced image regions, ensure the continuity and integrity of the entire subsidence result data, and make the final subsidence result more smooth and natural, which conforms to the actual situation and is convenient for subsequent data analysis and application.

[0090] 4. Improve processing efficiency and applicability: The K-means clustering algorithm is used to realize adaptive region division and detrending, which has high automation degree and reduces manual intervention, thereby improving the efficiency of data processing. At the same time, this method is suitable for different types and characteristics of InSAR subsidence data, has strong universality and applicability, and can be widely applied in geological disaster monitoring, urban sustainable development and other fields, thereby providing strong technical support for related research and decision-making.

[0091] That is, adaptive region division: compared with the traditional global detrending method, the InSAR image can be divided into multiple regions with different subsidence characteristics according to the subsidence characteristics, thereby better adapting to the non-uniform subsidence region. Adaptive model selection: for different regions, linear model, quadratic polynomial model or cubic polynomial model can be selected for fitting, thereby better eliminating the trend term and extracting local subsidence information. Local detrending and inlaying: local detrending and inlaying technology are combined to eliminate the trend term while ensuring data continuity.

[0092] Obviously, the InSAR subsidence result image processing method proposed in the present application can effectively improve the accuracy of subsidence monitoring, retain more detailed information, ensure data continuity, improve processing efficiency and applicability, and provide more reliable technical support for geological disaster monitoring and urban sustainable development.

[0093] As Figure 7 shown, a specific embodiment of an InSAR subsidence result image processing device provided by the present application is described. The device described in this embodiment, i.e. the entity device for executing the method described in the above embodiment. The technical solution is essentially consistent with the above embodiment, and the corresponding description in the above embodiment is also applicable to this embodiment. The device comprises:

[0094] The first unit 701 is configured to calculate a first subsidence gradient and a second subsidence gradient of an InSAR subsidence result image; wherein the first subsidence gradient is a subsidence gradient in a horizontal direction, and the second subsidence gradient is a subsidence gradient in a vertical direction;

[0095] The second unit 702 is configured to divide the InSAR subsidence result image into a plurality of image regions according to the first subsidence gradient and the second subsidence gradient of the InSAR subsidence result image;

[0096] The third unit 703 is configured to, for each image region, determine trend information of the image region by using a preset linear model, and perform trend information removal processing on the image region by using the trend information of the image region, to obtain a processed image region;

[0097] The fourth unit 704 is configured to stitch the processed image regions, and perform fusion processing on boundaries between the processed image regions, to obtain a processed InSAR subsidence result image.

[0098] Optionally, the first unit 701 is configured to:

[0099] The first subsidence gradient and the second subsidence gradient of each pixel point in the InSAR subsidence result image are calculated by using a Sobel operator or a Prewitt operator.

[0100] Optionally, the second unit 702 is configured to:

[0101] For each pixel point in the InSAR subsidence result image, a feature vector of the pixel point is generated according to the first subsidence gradient and the second subsidence gradient of the pixel point;

[0102] The feature vectors of all pixel points in the InSAR subsidence result image are clustered based on a K-means clustering algorithm, to obtain a clustering result; wherein the clustering result is K clusters.

[0103] According to the clustering result, a clustering effect index corresponding to the clustering result is determined.

[0104] According to the clustering effect index corresponding to the clustering result, the clustering result is optimized to obtain an optimized clustering result;

[0105] According to the optimized clustering result, the InSAR subsidence result image is divided into a plurality of image regions.

[0106] Optionally, the clustering effect index is one of the following: contour coefficient, Calinski-Harabasz index.

[0107] Optionally, the third unit 703 is configured to:

[0108] determine a pixel change trend type of the image region;

[0109] if the pixel change trend type of the image region is a first change trend type, a quadratic linear model is fitted using a least square method, and the quadratic linear model is taken as the preset linear model;

[0110] if the pixel change trend type of the image region is a second change trend type, a cubic linear model is fitted using a least square method, and the cubic linear model is taken as the preset linear model;

[0111] the row and column numbers of the image region are input into the preset linear model to obtain trend information of the image region.

[0112] Optionally, the third unit 703 is configured to:

[0113] the pixel value of the image region is subtracted from the trend information of the image region to obtain a processed image region.

[0114] Optionally, the fourth unit 704 is configured to:

[0115] for each image region, a pixel buffer area of the image region is determined according to a boundary between the image region and an adjacent image region; first weights and second weights respectively corresponding to each pixel point are determined according to distances between the pixel point and the boundary; pixel values of the pixel points are respectively adjusted according to the first weights and the second weights respectively corresponding to the pixel points, pixel values of the pixel points in the InSAR subsidence result image, and pixel values of the pixel points in the processed image region, to obtain an adjusted image region;

[0116] a processed InSAR subsidence result image is obtained according to all the adjusted image regions.

[0117] Therefore, the device can effectively improve the monitoring precision of subsidence, retain more detailed information, ensure data continuity, improve processing efficiency and applicability, and provide more reliable technical support for geological disaster monitoring and sustainable development of cities.

[0118] Figure 8 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. At the hardware level, the electronic device includes a processor and can optionally further include an internal bus, a network interface, and a memory. The memory can include a memory such as a random-access memory (RAM) and can further include a non-volatile memory such as at least one disk memory. Of course, the electronic device can further include other hardware required by a business.

[0119] The processor, the network interface, and the memory can be connected to each other through the internal bus, which can be an industry standard architecture (ISA) bus, a peripheral component interconnect (PCI) bus, or an extended industry standard architecture (EISA) bus. The bus can be divided into an address bus, a data bus, and a control bus. For ease of representation, Figure 8 Only one bidirectional arrow is used in the figure to represent the bus, but it does not mean that there is only one bus or only one type of bus.

[0120] The memory is configured to store execution instructions. Specifically, the execution instructions are computer programs that can be executed. The memory can include a memory and a non-volatile memory and provide the processor with execution instructions and data.

[0121] In a possible implementation manner, the processor reads corresponding execution instructions from the non-volatile memory into the memory and then runs, and can also obtain corresponding execution instructions from other devices to form an InSAR subsidence result image processing device at a logical level. The processor executes the execution instructions stored in the memory to implement the InSAR subsidence result image processing method provided in any embodiment of the present application through the executed execution instructions.

[0122] The above as the present application Figure 1The method executed by the InSAR subsidence result image processing apparatus provided by the embodiment shown can be applied in a processor or implemented by the processor. The processor can be an integrated circuit chip with signal processing capability. In the implementation process, each step of the above method can be completed by integrated logic circuits of hardware in the processor or instructions in the form of software. The processor mentioned above can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. Each method, step and logic block disclosed in the embodiment of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0123] The steps of the method disclosed in the embodiment of the present application can be directly embodied as hardware decoding processor execution completion, or executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a random memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register, etc. The storage medium in the art. The storage medium is located in the memory, and the processor reads the information in the memory, and combines the hardware to complete the steps of the above method.

[0124] The embodiment of the present application also proposes a readable medium, and the readable storage medium stores execution instructions. When the stored execution instructions are executed by the processor of the electronic device, the electronic device can execute the InSAR subsidence result image processing method provided in any embodiment of the present application, and is specifically used for executing the above evaluation method.

[0125] The electronic device described in each of the foregoing embodiments can be a computer.

[0126] Those skilled in the art should understand that the embodiments of the present application can be provided as a method or a computer program product. Therefore, the present application can adopt a complete hardware embodiment, a complete software embodiment, or a combination of software and hardware.

[0127] Each of the embodiments in the present application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each of the embodiments focuses on the difference from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiments.

[0128] It should also be noted that the terms "comprising", "containing", or any other variant 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 can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0129] The above only describes the embodiments of the present application and is not intended to limit the present application. The present application can have various modifications and changes for those skilled in the art. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the scope of the claims of the present application.

Claims

1. A method for processing an InSAR subsidence result image, characterized in that, The method comprises: calculating a first subsidence gradient and a second subsidence gradient of each pixel point in an InSAR subsidence result image by using a Sobel operator or a Prewitt operator; wherein the first subsidence gradient is a subsidence gradient in a horizontal direction, and the second subsidence gradient is a subsidence gradient in a vertical direction; generating a feature vector of each pixel point in the InSAR subsidence result image according to the first subsidence gradient and the second subsidence gradient of the pixel point; performing clustering on the feature vectors of all pixel points in the InSAR subsidence result image based on a K-means clustering algorithm to obtain a clustering result; wherein the clustering result is K clusters; determining a clustering effect index corresponding to the clustering result according to the clustering result; optimizing the clustering result according to the clustering effect index corresponding to the clustering result to obtain an optimized clustering result; dividing the InSAR subsidence result image into a plurality of image regions according to the optimized clustering result; for each image region, determining a pixel change trend type of the image region by using a preset linear model; if the pixel change trend type of the image region is a first change trend type, fitting a quadratic linear model by using a least square method, and taking the quadratic linear model as the preset linear model; if the pixel change trend type of the image region is a second change trend type, fitting a cubic linear model by using a least square method, and taking the cubic linear model as the preset linear model; inputting row and column numbers of the image region into the preset linear model to obtain trend information of the image region; and performing trend information removal processing on the image region by using the trend information of the image region to obtain a processed image region; wherein the first change trend type is that pixel values of pixel points of an image region increase or decrease from one direction, and the second change trend type is that pixel values of pixel points of an image region increase or decrease from multiple directions; splicing the processed image regions, and performing fusion processing on boundaries between the processed image regions to obtain a processed InSAR subsidence result image.

2. The method of claim 1, wherein, The clustering effect index is one of the following: a contour coefficient, and a Calinski-Harabasz index.

3. The method of claim 1, wherein, The processing of the image region by using the trend information of the image region to remove the trend information to obtain a processed image region comprises: subtracting the trend information of the image region from pixel values of the image region to obtain the processed image region.

4. The method of claim 1, wherein, The fusion processing of the boundaries between the processed image regions to obtain the processed InSAR subsidence result image comprises: For each image region, a pixel buffer of the image region is determined according to a boundary between the image region and a neighboring image region; first weights and second weights respectively corresponding to each pixel point in the pixel buffer are determined according to distances between the pixel point and the boundary; pixel values of the pixel points are respectively adjusted according to the first weights and the second weights respectively corresponding to the pixel points, pixel values of the pixel points in the InSAR subsidence result image, and pixel values of the pixel points in the processed image region, to obtain an adjusted image region; An InSAR subsidence result image after processing is obtained according to all the adjusted image regions.

5. An apparatus for processing an InSAR subsidence result image, characterized by comprising: The apparatus comprises: A first unit configured to calculate first subsidence gradients and second subsidence gradients of each pixel point in an InSAR subsidence result image by using a Sobel operator or a Prewitt operator; wherein the first subsidence gradient is a subsidence gradient in a horizontal direction, and the second subsidence gradient is a subsidence gradient in a vertical direction; A second unit configured to, for each pixel point in the InSAR subsidence result image, generate a feature vector of the pixel point according to the first subsidence gradient and the second subsidence gradient of the pixel point; cluster the feature vectors of all the pixel points in the InSAR subsidence result image based on a K-means clustering algorithm to obtain a clustering result; wherein the clustering result comprises K clusters; determine a clustering effect index corresponding to the clustering result according to the clustering result; optimize the clustering result according to the clustering effect index corresponding to the clustering result to obtain an optimized clustering result; and divide the InSAR subsidence result image into a plurality of image regions according to the optimized clustering result; A third unit configured to, for each image region, determine a pixel change trend type of the image region by using a preset linear model; if the pixel change trend type of the image region is a first change trend type, fit a quadratic linear model by using a least square method, and use the quadratic linear model as the preset linear model; if the pixel change trend type of the image region is a second change trend type, fit a cubic linear model by using a least square method, and use the cubic linear model as the preset linear model; input row and column numbers of the image region into the preset linear model to obtain trend information of the image region; and perform trend information removal processing on the image region by using the trend information of the image region to obtain a processed image region; wherein the first change trend type is a change in which pixel values of pixel points in an image region increase or decrease from one direction, and the second change trend type is a change in which pixel values of pixel points in an image region increase or decrease from multiple directions; A fourth unit configured to splice the processed image regions, and perform fusion processing on boundaries between the processed image regions to obtain an InSAR subsidence result image after processing.

6. A readable medium characterized by The readable medium includes execution instructions, when a processor of an electronic device executes the execution instructions, the electronic device executes the method as claimed in any one of claims 1-4.

7. An electronic device, comprising: The electronic device includes a processor and a memory storing execution instructions, when the processor executes the execution instructions stored in the memory, the processor executes the method as claimed in any one of claims 1-4.

Images