Big data application information cloud platform based on Beidou

Abstract

The invention discloses a big data application information cloud platform based on Beidou, which relates to the technical field of big data application and comprises an image analysis unit, an image registration unit, an interferogram generation unit and a deformation output unit. Satellite image data of a designated area is cut through the image analysis unit, the time difference value and the spatial difference value between each piece of image data and other image data are calculated, it is ensured that construction of a connection diagram can be smoother and more efficient, the interference diagram generation unit analyzes a difference interference diagram, and the accuracy of the connection diagram is improved. According to the method, interference information contained in the earth surface deformation result map is removed, meanwhile, orbit parameters are updated, the deformation rate calculated subsequently can be more accurate, detail features are added into the earth surface deformation result map through Beidou positioning data, and it is guaranteed that the feedback deformation result map can help operators to conduct data analysis to the maximum extent.

Description

Technical Field

[0001] The present invention relates to the technical field of big data applications, and particularly to a big data application information cloud platform based on Beidou. Background Technique

[0002] The big data application information cloud platform based on Beidou is a comprehensive platform integrating Beidou satellites and big data technologies. Through high-precision positioning, real-time data collection, intelligent analysis, and visual display, it provides users with comprehensive location services and decision-making support. In the invention patent with the application number 202010639228.0, it is disclosed that "a big data application information cloud platform based on Beidou, belonging to the technical field of big data information technology, solves the problems of poor management ability and low efficiency of the existing information processing platform, and includes a service presentation layer, a data application layer, a data service layer, a data calculation layer, a data collection layer, and a Beidou information application module. The service presentation layer includes a report display module, a data visualization module, and a security warning module. The data application layer includes a security warning service module, a Jnetcms content management system, and a data service open module. The data service layer includes a data service tool and a data service storage module. The data calculation layer includes a mining analysis platform, a data integration platform, and a tool platform. The data collection layer includes a big data platform and a data exchange platform. The Beidou information application module includes Beidou satellite navigation system equipment positioning, trajectory tracking, SOS positioning, and Beidou short message. The present invention improves the product management ability and after-sales service efficiency."

[0003] The above-mentioned existing technology solves problems such as low information processing and feedback efficiency, which cannot meet the actual needs of users. However, when the system is running, since it does not obtain satellite image data of a specified area, the platform cannot analyze the surface deformation result of the specified area based on the existing satellite image data, and at the same time, it is also unable to adjust the initial orbit parameters included in the image data, resulting in a large deviation in the orbit parameters corresponding to each image data, affecting the process of the operator's analysis of the surface deformation data. Summary of the Invention

[0004] The purpose of the present invention is to provide a big data application information cloud platform based on Beidou to solve the problems raised in the above background technique.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A big data application information cloud platform based on Beidou, including an interferogram generation unit and a deformation output unit.

[0006] An image analysis unit, after obtaining satellite image data at multiple different time points, calculates the time difference according to the recording time, calculates the spatial difference according to the sensor position, sets a threshold, and if both the time difference and the spatial difference are lower than the threshold, adds connection edges between the images to form a connection graph, selects the image data within the time range, counts the number of connection edges, determines the image data with the number of connection edges greater than the preset value, and then analyzes the spatial difference to determine multiple central image data from all the image data;

[0007] An image registration unit, the image registration unit counts the time difference between each central image data and other image data, takes the image with the smallest difference as the adjacent image data, calculates the geometric transformation parameters between the central image data and the adjacent image data, performs an initial registration operation, sets the number of loops, performs Gaussian blur and downsampling on the image data to generate multiple blurred image data, selects key points in the image data, counts the gradient magnitude and direction of the pixels in the neighborhood of the key points, determines the main direction and constructs a feature representation, calculates the similarity between the key points of the adjacent image data and the central image data, selects the best matching points, calculates the geometric parameters of the image data and completes the registration.

[0008] Preferably, the image analysis unit includes a time difference calculation module and a spatial difference calculation module. After the time difference calculation module obtains satellite image data corresponding to a specified area at multiple different time points, it clips the satellite image data and calculates the time difference between each image data and other image data according to the recording time. The spatial difference calculation module determines the sensor position when collecting images according to the recording time of each image data, counts the sensor positions corresponding to each image data, and calculates the spatial difference between each image data and other image data according to the sensor positions.

[0009] Preferably, the image analysis unit further includes a connection graph generation module and a central image determination module. After the connection graph generation module counts the time difference and spatial difference between the image data and other image data, it sets a threshold. If both the time difference and the spatial difference are lower than the threshold, it adds connection edges between the images, otherwise it does not add connection edges. After adding connection edges to all the images, a connection graph is obtained. The central image determination module determines the recording time of each image data in the connection graph, sets a time range, selects all the image data with the recording time within this range, counts the number of connection edges corresponding to the selected image data, determines the image data with the number of connection edges greater than the preset value, and analyzes multiple central image data according to the spatial difference between the image data and other image data.

[0010] Preferably, the image registration unit includes an adjacent image screening module and an initial registration determination module. After the adjacent image screening module calculates the time difference between each central image data and other image data, it takes the image data with the smallest time difference as the adjacent image data of the central image data. After the initial registration determination module determines the adjacent image data of each central image data, it calculates the geometric transformation parameters between the images according to the sensor positions corresponding to the central image data and the adjacent image data, and completes the initial registration operation of the adjacent image data and the central image data according to the geometric transformation parameters.

[0011] Preferably, the image registration unit further includes a blurred image analysis module and a main direction calculation module. The blurred image analysis module sets the number of loops, performs Gaussian blur operations on the central image data and the adjacent image data respectively to obtain multiple blurred image data, downsamples them to determine new blurred image data, and repeats the operation until the loop ends, and then outputs the finally obtained multiple blurred image data. After the main direction calculation module selects key points from the multiple blurred image data, it sets the neighborhood size, calculates the gradient magnitude and direction of all pixel points in the neighborhood corresponding to each key point, sets different gradient direction ranges, determines the number of pixel points corresponding to each gradient direction range, and takes the median value corresponding to the gradient direction range with the largest number as the main direction of the key point.

[0012] Preferably, the image registration unit further includes a feature representation statistics module, a matching point selection module and a secondary registration execution module. The feature representation statistics module constructs the feature representation corresponding to each key point according to the gradient magnitude and direction of all pixel points in the neighborhood corresponding to each key point, so as to statistically obtain the main direction and feature representation of each key point in the adjacent image data and the central image data. The matching point selection module determines the similarity between the key points in the adjacent image data and the key points in the central image data, and selects the best matching points of the key points in the adjacent image data in the central image data according to the similarity. The secondary registration execution module calculates the corresponding geometric parameters according to the feature representation and relative positions of the key points and the best matching points, and completes the registration operation of the adjacent image data and the central image data according to the geometric parameters.

[0013] Preferably, the interferogram generation unit includes an interferogram analysis module and a noise removal module. After the interferogram analysis module determines the initial interferogram and the corresponding phase based on the registered central image data and adjacent image data, it extracts the initial orbit parameters of the central image data, calculates the flat-earth phase and the topographic phase according to the initial orbit parameters and the DEM data, subtracts the original phase from the flat-earth phase and the topographic phase to obtain a new phase, and analyzes the differential interferogram corresponding to the central image data and the adjacent image data based on the new phase. After the noise removal module filters the differential interferogram to remove noise, it analyzes the filtered image using the branch cut method to obtain the phase unwrapping image.

[0014] Preferably, the interferogram generation unit further includes an orbit parameter determination module and a flat-earth phase calculation module. After the orbit parameter determination module counts the differential interferogram and the phase unwrapping image of each central image data and adjacent image data, it sets the initial orbit parameters and the control step corresponding to the central image data, determines the deviation function corresponding to the orbit parameters of each cycle, and calculates the orbit parameters under different cycle numbers using the orbit parameter update algorithm. The orbit parameters of the last cycle are used as the final orbit parameters. The flat-earth phase calculation module recalculates the flat-earth phase according to the orbit parameters, subtracts the current flat-earth phase from the phase values in the differential interferogram and the phase unwrapping image to obtain a new differential interferogram.

[0015] Preferably, the deformation output unit includes a deformation analysis module and an average value calculation module. The deformation analysis module extracts the deformation phase and the residual phase from the phase data of all differential interferograms, calculates the displacement rate and the residual topography using the deformation phase, and analyzes the deformation rate values of each pixel point in different time series according to the displacement rate. The average value calculation module analyzes the deformation result image of the specified area according to the deformation rate values of each pixel point in different time series. After setting the window size, it traverses all pixel points in the deformation result image, takes each pixel point as the center point of the window, counts the numerical sizes of the other pixel points except the center point, and calculates the average value corresponding to the current window center point according to the other pixel values.

[0016] Preferably, the deformation output unit further includes a pixel value replacement module and a deformation result determination module. If the difference between the pixel value of the window center point and the average value is greater than the threshold, the pixel value replacement module replaces the pixel value of the window center point with the average value; otherwise, it does not perform any operation. After the deformation result determination module determines the numerical value of each pixel point in the deformation result image, it adds detailed features to the deformation result image according to the Beidou positioning data and outputs it through the visualization interface.

[0017] Compared with the prior art, the beneficial effects of the present invention are:

[0018] The present invention shears the satellite image data of a specified area through an image analysis unit, and calculates the time difference and spatial difference between each image data and other image data, ensuring that the construction of the connection map can be more smooth and efficient. At the same time, the image registration unit matches the central image data and adjacent image data multiple times, effectively improving the registration effect. And the interferogram generation unit also analyzes the differential interferogram, removes the interference information contained therein, and updates the orbital parameters, so that the subsequent calculated deformation rate can be more accurate. The Beidou positioning data is used to add detailed features to the surface deformation result map, ensuring that the feedback deformation result map can help the operator perform data analysis to the greatest extent. Brief Description of the Drawings

[0019] Figure 1 It is a schematic diagram of the overall system process provided by an embodiment of the present invention;

[0020] Figure 2 It is a block diagram of the internal modules of the image analysis unit provided by an embodiment of the present invention;

[0021] Figure 3 It is a block diagram of the internal modules of the image registration unit provided by an embodiment of the present invention;

[0022] Figure 4 It is a block diagram of the internal modules of the interferogram generation unit provided by an embodiment of the present invention;

[0023] Figure 5 It is a block diagram of the internal modules of the deformation output unit provided by an embodiment of the present invention.

[0024] In the figure: 1. Image analysis unit; 101. Time difference calculation module; 102. Spatial difference calculation module; 103. Connection map generation module; 104. Central image determination module; 2. Image registration unit; 201. Adjacent image screening module; 202. Initial registration determination module; 203. Blurred image analysis module; 204. Main direction calculation module; 205. Feature representation statistics module; 206. Matching point selection module; 207. Secondary registration execution module; 3. Interferogram generation unit; 301. Interferogram analysis module; 302. Noise removal module; 303. Orbital parameter determination module; 304. Flat earth phase calculation module; 4. Deformation output unit; 401. Deformation analysis module; 402. Average value calculation module; 403. Pixel value replacement module; 404. Deformation result determination module. Detailed Embodiment

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] Please refer to Figures 1 - 5 , the present invention provides a technical solution: a Beidou-based big data application information cloud platform, including an interference pattern generation unit 3 and a deformation output unit 4;

[0027] An image analysis unit 1. After the image analysis unit 1 obtains satellite image data at multiple different time points, it calculates the time difference according to the recording time, calculates the spatial difference according to the sensor position, sets a threshold. If both the time difference and the spatial difference are lower than the threshold, connection edges are added between the images to form a connection graph. The image data within the time range is selected, the number of connection edges is counted, the image data with the number of connection edges greater than a preset value is determined, and then the spatial difference is analyzed to determine multiple central image data from all the image data;

[0028] An image registration unit 2. The image registration unit 2 counts the time differences between each central image data and other image data, takes the image with the smallest difference as the adjacent image data, calculates the geometric transformation parameters between the central image data and the adjacent image data, performs an initial registration operation, sets the number of loops, performs Gaussian blurring and downsampling on the image data to generate multiple blurred image data, selects important points in the image data, counts the gradient amplitude and direction of the pixels in the neighborhood of the important points, determines the main direction and constructs a feature representation, calculates the similarity between the important points of the adjacent image data and the central image data, selects the best matching points, calculates the geometric parameters of the image data and completes the registration.

[0029] The image analysis unit 1 includes a time difference calculation module 101 and a spatial difference calculation module 102. After the time difference calculation module 101 obtains satellite image data corresponding to a specified area at multiple different time points, it clips the satellite image data and calculates the time difference between each image data and other image data according to the recording time. The spatial difference calculation module 102 determines the sensor position when collecting images according to the recording time of each image data, counts the sensor positions corresponding to each image data, and calculates the spatial difference between each image data and other image data according to the sensor position;

[0030] The image analysis unit 1 further includes a connection graph generation module 103 and a central image determination module 104. After the connection graph generation module 103 calculates the time difference and spatial difference between the image data and other image data, it sets a threshold. If both the time difference and spatial difference are lower than the threshold, connection edges are added between the images; otherwise, no connection edges are added. After adding connection edges to all images, a connection graph is obtained. The central image determination module 104 determines the recording time of each piece of image data in the connection graph, sets a time range, selects all the image data whose recording time is within this range, then calculates the number of connection edges corresponding to the selected image data, determines the image data with the number of connection edges greater than a preset value, and analyzes multiple central image data according to the spatial difference between the image data and other image data;

[0031] The image registration unit 2 includes an adjacent image screening module 201 and an initial registration determination module 202. After the adjacent image screening module 201 calculates the time difference between each piece of central image data and other image data, it takes the image data with the smallest time difference as the adjacent image data of the central image data. After the initial registration determination module 202 determines the adjacent image data of each piece of central image data, it calculates the geometric transformation parameters between the images according to the sensor positions corresponding to the central image data and the adjacent image data, and completes the initial registration operation of the adjacent image data and the central image data according to the geometric transformation parameters;

[0032] The image registration unit 2 further includes a blurred image analysis module 203 and a main direction calculation module 204. The blurred image analysis module 203 sets the number of loops, performs Gaussian blur operations on the central image data and the adjacent image data respectively to obtain multiple blurred image data, downsamples them to determine new blurred image data, and repeats the operation until the loop ends, and then outputs the finally obtained multiple blurred image data. After the main direction calculation module 204 selects key points from the multiple blurred image data, it sets the neighborhood size, calculates the gradient amplitude and direction of all pixel points in the neighborhood corresponding to each key point, sets different gradient direction ranges, determines the number of pixel points corresponding to each gradient direction range, and takes the median value corresponding to the gradient direction range with the largest number as the main direction of the key point;

[0033] The image registration unit 2 further includes a feature representation statistics module 205, a matching point selection module 206, and a secondary registration execution module 207. The feature representation statistics module 205 constructs a feature representation corresponding to each key point based on the gradient magnitude and direction of all pixel points within the neighborhood corresponding to each key point, thereby statistically obtaining the main direction and feature representation of each key point in the adjacent image data and the central image data. The matching point selection module 206 determines the similarity between the key points in the adjacent image data and the key points in the central image data, and selects the best matching points of the key points in the adjacent image data within the central image data according to the similarity. The secondary registration execution module 207 calculates the corresponding geometric parameters based on the feature representations and relative positions of the key points and the best matching points, and completes the registration operation of the adjacent image data and the central image data according to the geometric parameters;

[0034] The interferogram generation unit 3 includes an interferogram analysis module 301 and a noise removal module 302. After the interferogram analysis module 301 determines the initial interferogram and the corresponding phase based on the registered central image data and adjacent image data, it extracts the initial orbit parameters of the central image data, calculates the flat earth phase and the terrain phase according to the initial orbit parameters and the DEM data, subtracts the original phase from the flat earth phase and the terrain phase to obtain a new phase, and analyzes the differential interferogram corresponding to the central image data and the adjacent image data according to the new phase. After the noise removal module 302 removes the noise from the differential interferogram through filtering, it analyzes the filtered image using the branch cut method to obtain the phase unwrapping image;

[0035] The interferogram generation unit 3 further includes an orbit parameter determination module 303 and a flat earth phase calculation module 304. After the orbit parameter determination module 303 statistically analyzes the differential interferogram and the phase unwrapping image of each central image data and adjacent image data, it sets the initial orbit parameters corresponding to the central image data and the control step size, determines the deviation function corresponding to the orbit parameters for each cycle, calculates the initial orbit parameters and the corresponding deviation function using the orbit parameter update algorithm to obtain the orbit parameters under different cycle numbers, and takes the orbit parameters of the last round as the final orbit parameters. The flat earth phase calculation module 304 recalculates the flat earth phase according to the orbit parameters, subtracts the current flat earth phase from the phase values in the differential interferogram and the phase unwrapping image to obtain a new differential interferogram. The orbit parameter update algorithm is specifically:

[0036]

[0037] ρ h+1 = ρ h -(D Τ D + SE) -1 ·D T U h (ρ)

[0038] Among them, ρ h+1 represents the orbital parameter after the (h + 1)-th round of iteration, ρ h represents the orbital parameter after the h-th round of iteration, U h (ρ) represents the deviation function of the orbital parameter after the h-th round of iteration, U1(ρ) represents the deviation function of the orbital parameter after the 1st round of iteration, U2(ρ) represents the deviation function of the orbital parameter after the 2nd round of iteration, U x (ρ) represents the deviation function of the orbital parameter after the x-th round of iteration, ρ1 represents the orbital parameter after the 1st round of iteration, ρ2 represents the orbital parameter after the 2nd round of iteration, ρ y represents the orbital parameter after the y-th round of iteration, represents the partial derivative of the ratio between the deviation function of the orbital parameter after the 1st round of iteration and the orbital parameter after the 1st round of iteration, represents the partial derivative of the ratio between the deviation function of the orbital parameter after the 2nd round of iteration and the orbital parameter after the 1st round of iteration, represents the partial derivative of the ratio between the deviation function of the orbital parameter after the x-th round of iteration and the orbital parameter after the 1st round of iteration, represents the partial derivative of the ratio between the deviation function of the orbital parameter after the 1st round of iteration and the orbital parameter after the 2nd round of iteration, represents the partial derivative of the ratio between the deviation function of the orbital parameter after the 2nd round of iteration and the orbital parameter after the 2nd round of iteration, represents the partial derivative of the ratio between the deviation function of the orbital parameter after the x-th round of iteration and the orbital parameter after the 2nd round of iteration, represents the partial derivative of the ratio between the deviation function of the orbital parameter after the 1st round of iteration and the orbital parameter after the y-th round of iteration, represents the partial derivative of the ratio between the deviation function of the orbital parameter after the 2nd round of iteration and the orbital parameter after the y-th round of iteration, represents the partial derivative of the ratio between the deviation function of the orbital parameter after the x-th round of iteration and the orbital parameter after the y-th round of iteration, where x, y, h represent parameters, S represents the control step size, E represents the identity matrix, D represents the orbital parameter combination matrix, D Τ represents the transposed matrix corresponding to the combination matrix;

[0039] The deformation output unit 4 includes a deformation analysis module 401 and an average value calculation module 402. The deformation analysis module 401 extracts the deformation phase and the residual phase according to the phase data of all differential interferograms. After calculating the displacement rate and the residual terrain using the deformation phase, the deformation rate values of each pixel point on different time series are analyzed according to the displacement rate. The average value calculation module 402 analyzes the deformation result map of a specified area according to the deformation rate values of each pixel point on different time series. After setting the window size, all pixel points in the deformation result map are traversed, and each pixel point is used as the center point of the window. The numerical sizes corresponding to the other pixel points except the center point are statistically counted, and the average value corresponding to the current window center point is calculated according to the other pixel values;

[0040] The deformation output unit 4 further includes a pixel value replacement module 403 and a deformation result determination module 404. If the difference between the pixel value of the window center point and the average value is greater than the threshold, the pixel value replacement module 403 uses the average value as the pixel value of the window center point, otherwise it does not perform any operation. After the deformation result determination module 404 determines the numerical value of each pixel point in the deformation result map, it adds detailed features to the deformation result map according to the Beidou positioning data and outputs it through the visualization interface.

[0041] Working principle: In the present invention, the time difference calculation module 101 in the image analysis unit 1 calculates the time difference between each image data and other image data according to the recording time, the space difference calculation module 102 calculates the space difference between each image data and other image data according to the sensor position, and after the connection graph generation module 103 adds connection edges to all images to obtain a connection graph, the adjacent image screening module 201 in the image registration unit 2 determines the adjacent image data of each central image data, the initial registration determination module 202 completes the initial registration operation of the adjacent image data and the central image data according to the geometric transformation parameters, and multiple blurred image data are obtained through the blurred image analysis module 203. The main direction calculation module 204 determines the key points and the corresponding main directions in the image data, the feature representation statistics module 205 statistically obtains the feature representations of each key point, the best matching points of each key point are screened out through the matching point selection module 206, and the secondary registration execution module 207 completes the registration operation of the adjacent image data and the central image data. According to the new phase, the interference pattern analysis module 301 in the interference pattern generation unit 3 analyzes the differential interference patterns corresponding to the central image data and the adjacent image data, the noise removal module 302 determines the phase unwrapping map, the orbit parameter determination module 303 calculates the final orbit parameters, a new differential interference pattern is obtained through the flat earth phase calculation module 304, the deformation analysis module 401 in the deformation output unit 4 analyzes the deformation rate values of each pixel point in different time series, the average value calculation module 402 outputs the average value corresponding to the center point of the current window, the average value is used as the pixel value of the window center point through the pixel value replacement module 403, and the deformation result determination module 404 adds detailed features to the deformation result graph and outputs it through the visualization interface.

[0042] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover a non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device.

[0043] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A Beidou-based big data application information cloud platform, comprising an interference pattern generation unit (3) and a deformation output unit (4), characterized in that: An image analysis unit (1), wherein after acquiring a plurality of satellite image data at different time points, the image analysis unit (1) calculates a time difference according to the recording time, and calculates a spatial difference according to the sensor position, sets a threshold, and if both the time difference and the spatial difference are lower than the threshold, adds connecting edges between the images to form a connection graph, selects image data within a time range, counts the number of connecting edges, determines image data with a number of connecting edges greater than a preset value, and then analyzes the spatial difference to determine a plurality of central image data from all the image data; An image registration unit (2) is provided, wherein the image registration unit (2) counts the time difference between each central image data and other image data, takes the image with the minimum difference as the adjacent image data, calculates the geometric transformation parameters of the central image data and the adjacent image data, sets the number of cycles after performing the initial registration operation, performs Gaussian blurring and downsampling on the image data, generates multiple blurred image data, selects important points in the image data, counts the gradient amplitude and direction of pixels in the neighborhood of the important points, determines the main direction and constructs a feature representation, calculates the similarity between the adjacent image data and the important points of the central image data, selects the best matching point, calculates the geometric parameters of the image data and completes the registration.

2. The Beidou-based big data application information cloud platform according to claim 1, characterized in that: The image analysis unit (1) comprises a time difference calculation module (101) and a space difference calculation module (102). After acquiring a plurality of satellite image data corresponding to a specified area at different time points, the time difference calculation module (101) cuts the satellite image data and calculates the time difference between each image data and other image data according to the recording time. The space difference calculation module (102) determines the sensor position when the image is collected according to the recording time of each image data, counts the sensor position corresponding to each image data, and calculates the space difference between each image data and other image data according to the sensor position.

3. The Beidou-based big data application information cloud platform according to claim 2, characterized in that: The image analysis unit (1) also includes a connection diagram generation module (103) and a central image determination module (104). The connection diagram generation module (103) counts the time difference and space difference between the image data and other image data, and then sets a threshold. If the time difference and space difference are both lower than the threshold, a connection edge is added between the images. Otherwise, no connection edge is added. After adding connection edges to all images, a connection diagram is obtained. The central image determination module (104) determines the recording time of each image data in the connection diagram, sets a time range, selects all image data whose recording time is within the range, counts the number of connection edges corresponding to the selected image data, determines image data whose number of connection edges is greater than a preset value, and analyzes multiple central image data according to the space difference between the image data and other image data.

4. The Beidou-based big data application information cloud platform according to claim 1, characterized in that: The image registration unit (2) comprises an adjacent image screening module (201) and an initial registration determination module (202). The adjacent image screening module (201) counts the time difference between each central image data and other image data, and then takes the image data with the smallest time difference as the adjacent image data of the central image data. The initial registration determination module (202) determines the adjacent image data of each central image data, and then calculates the geometric transformation parameters between the images according to the sensor positions corresponding to the central image data and the adjacent image data, and completes the initial registration operation of the adjacent image data and the central image data according to the geometric transformation parameters.

5. The Beidou-based big data application information cloud platform according to claim 4, characterized in that: The image registration unit (2) further comprises a blurred image analysis module (203) and a main direction calculation module (204). The blurred image analysis module (203) sets the number of cycles, performs Gaussian blur operations on the central image data and the adjacent image data respectively, obtains a plurality of blurred image data, performs downsampling on the data, thereby determining new blurred image data, repeats the operation until the cycle ends, and outputs the plurality of blurred image data finally obtained. The main direction calculation module (204) selects important points from the plurality of blurred image data, sets the size of the neighborhood, counts the gradient amplitude and direction of all pixel points in the neighborhood corresponding to each important point, sets different gradient direction ranges, determines the number of pixel points corresponding to each gradient direction range, and takes the median value corresponding to the gradient direction range with the largest number as the main direction of the important point.

6. The Beidou-based big data application information cloud platform according to claim 5, characterized in that: The image registration unit (2) further comprises a feature representation statistics module (205), a matching point selection module (206) and a secondary registration execution module (207). The feature representation statistics module (205) constructs a feature representation of the corresponding important point according to the gradient amplitude and direction of all pixel points in the neighborhood corresponding to each important point, thereby counting the main direction and feature representation of each important point in the adjacent image data and the central image data. The matching point selection module (206) determines the similarity between the important points in the adjacent image data and the important points in the central image data, and selects the best matching point of the important point in the adjacent image data in the central image data according to the similarity. The secondary registration execution module (207) calculates the corresponding geometric parameters according to the feature representation and relative position of the important point and the best matching point, and completes the registration operation of the adjacent image data and the central image data according to the geometric parameters.

7. The Beidou-based big data application information cloud platform according to claim 1, characterized in that: The interference pattern generation unit (3) comprises an interference pattern analysis module (301) and a noise point removal module (302). The interference pattern analysis module (301) determines the initial interference pattern and the corresponding phase according to the aligned central image data and the adjacent image data, extracts the initial orbit parameters of the central image data, calculates the flat ground phase and the terrain phase according to the initial orbit parameters and the DEM data, subtracts the original phase from the flat ground phase and the terrain phase to obtain a new phase, and analyzes the differential interference pattern corresponding to the central image data and the adjacent image data according to the new phase. The noise point removal module (302) removes the noise points by filtering the differential interference pattern, and then analyzes the filtered pattern using a branch cutting method to obtain a phase unwrapping pattern.

8. The Beidou-based big data application information cloud platform according to claim 7, characterized in that: The interference pattern generation unit (3) also includes an orbit parameter determination module (303) and a flat ground phase calculation module (304). After the orbit parameter determination module (303) counts the differential interference pattern and phase unwrapping pattern of each central image data and adjacent image data, it sets the initial orbit parameter and control step size corresponding to the central image data, determines the deviation function corresponding to each cycle orbit parameter, calculates the initial orbit parameter and the corresponding deviation function using an orbit parameter update algorithm, obtains the orbit parameter under different cycle times, and uses the last round of orbit parameters as the final orbit parameter. The flat ground phase calculation module (304) recalculates the flat ground phase according to the orbit parameter, subtracts the current flat ground phase from the phase value in the differential interference pattern and the phase unwrapping pattern, and obtains a new differential interference pattern.

9. The Beidou-based big data application information cloud platform according to claim 1, characterized in that: The deformation output unit (4) comprises a deformation analysis module (401) and an average value calculation module (402). The deformation analysis module (401) extracts deformation phase and residual phase according to phase data of all differential interference patterns, calculates displacement rate and residual topography using the deformation phase, and then analyzes deformation rate values ​​of each pixel point on different time series according to the displacement rate. The average value calculation module (402) analyzes a deformation result map of a specified area according to the deformation rate values ​​of each pixel point on different time series, and after setting the window size, traverses all pixel points in the deformation result map, takes each pixel point as the window center point, counts the numerical values ​​corresponding to other pixel points except the center point, and calculates the average value corresponding to the current window center point according to the other pixel values.

10. The Beidou-based big data application information cloud platform according to claim 9, characterized in that: The deformation output unit (4) also includes a pixel value replacement module (403) and a deformation result determination module (404). If the difference between the pixel value of the window center point and the average value is greater than a threshold value, the pixel value replacement module (403) uses the average value as the pixel value of the window center point, otherwise no operation is performed. After the deformation result determination module (404) determines the value of each pixel point in the deformation result image, it adds detail features to the deformation result image according to Beidou positioning data and outputs it through a visual interface.

Images