InSAR hidden danger point automatic identification method based on hotspot analysis

Abstract

The invention relates to the technical field of geological disaster monitoring, and discloses an InSAR hidden danger point automatic identification method based on hotspot analysis, and the method comprises the steps: S1, carrying out the mask and format conversion of an unreliable region of an InSAR deformation field; due to geometric distortion caused by imaging geometry of satellites, areas with low vegetation coherence may mislead subsequent hotspot analysis. According to the invention, through combination of remote sensing and GIS technologies, accurate screening models and algorithms are respectively established, and automatic and efficient elimination of unreliable areas is realized; a hotspot analysis technology is introduced into the field of automatic extraction of an InSAR deformation field, and element points are innovatively set according to characteristics of InSAR deformation data and actual requirements of geological hidden danger detection and serve as extraction standards of reliable deformation points with spatial clustering properties when specific conditions are met. The method can accurately extract the points really having the possibility of geological hidden troubles from the complex deformation data.

Description

Technical Field

[0001] The invention belongs to the technical field of geological disaster monitoring, and in particular relates to an InSAR (Infrared SAR) hidden danger point automatic identification method based on hotspot analysis. Background Art

[0002] Geological disasters, such as landslides, ground subsidence, and earthquakes, pose a serious threat to human life, property, and the stable operation of infrastructure. Effective geological disaster monitoring can provide early warnings, buying critical time for implementing protective measures and evacuating people, thereby minimizing losses caused by disasters and effectively ensuring social stability and sustainable development. Therefore, geological disaster monitoring is crucial.

[0003] The problem with these technologies is that traditional geological disaster monitoring methods are mostly based on ground-based measurement methods, such as total station measurement and level measurement. While these traditional methods can achieve high measurement accuracy in local areas, they have many drawbacks. Their monitoring coverage is extremely limited, making it difficult to fully cover areas prone to geological disasters, such as large mountainous areas and urban fringes. Monitoring efficiency is quite low, requiring a large investment of manpower, material resources, and time, often struggling to cope with large-scale monitoring tasks. Furthermore, they are easily significantly restricted by natural conditions such as topography, making implementation extremely difficult in complex terrain environments and unable to achieve comprehensive and rapid monitoring goals. Summary of the Invention

[0004] In view of the problems existing in the prior art, the present invention provides an InSAR hidden danger point automatic identification method based on hot spot analysis, which can overcome the above problems or at least partially solve the above problems.

[0005] The present invention is implemented as follows: an InSAR hidden danger point automatic identification method based on hot spot analysis, comprising: S1. Masking and format conversion of unreliable areas of InSAR deformation field; Geometric distortion caused by satellite imaging geometry and areas with low vegetation coherence may mislead subsequent hotspot analysis, leading to phase gradient discontinuity and analytical errors. Therefore, it is necessary to combine remote sensing and GIS technology to analyze and eliminate them to further obtain more reliable deformation areas. After removing the unreliable areas, the large-scale data in raster form is converted into point data based on the above results. Then the coordinate system is converted from WGS-84 to UTM projection coordinate system. Finally, in order to reduce the processing pressure of the computer, a large number of stable points in the deformation field are removed and do not participate in subsequent calculations. S2. Hotspot Analysis (Getis−Ord ) technology performs spatial statistics on the deformation values obtained over a large range to obtain high-value and low-value points with obvious clustering properties in space; hotspot analysis can be generally divided into three steps: Step 1: Spatial incremental autocorrelation analysis; the whole process is based on GlobalMoran's The purpose of the statistics is to determine the appropriate distance threshold for spatial analysis for each element; the z-score of the element is calculated at different distances, and finally the peak value of the z-score is used as the distance threshold parameter for spatial analysis; Step 2: Use the calculated distance threshold statistics, Statistics is a method of monitoring whether the feature points have significant spatial clustering properties by looking at each feature in the environment of neighboring features within a certain distance. The feature points with clustering properties obtained are not only high values themselves, but also surrounded by the same high and low value points. Step 3: Extract hot and cold values. According to the characteristics of the InSAR deformation field, this experiment sets the key points and When is a reliable deformation point with spatial clustering properties, the elements are classified and output to obtain the deformation potential points; S3. The deformation areas identified by hotspot analysis still have deformation errors caused by factors such as satellite imaging itself. Mean filtering is used to smooth the results to remove the influence of isolated noise points. A binary image is generated using the correlation coefficient threshold method, and the internal filling is performed to extract the boundaries of the phase discontinuity area. Considering that the area of the hidden danger point is within a certain range, and the larger area may be due to the discontinuity of the interference phase caused by the atmospheric effect, and the smaller area may be due to the discontinuity of the interference phase caused by the area with poor coherence such as vegetation, the appropriate threshold is selected to use the number of pixels in the formed closed area for screening, remove the identified larger and smaller areas, and retain the real deformation area.

[0006] Compared with the prior art, the present invention has the following beneficial effects: 1. Improve the efficiency of geological hazard detection: by introducing hotspot analysis (Getis-Ord ) technology, which automatically extracts high-value and low-value points with obvious clustering properties, changes the previous time-consuming and labor-intensive manual interpretation, greatly improves the speed of identifying potential geological hazard points in large-scale deformation monitoring data, and can conduct preliminary screening of large areas more quickly, buying precious time for timely response measures, and effectively improving the overall efficiency of geological hazard detection.

[0007] 2. Enhanced Data Processing Accuracy: This technology addresses the numerous discrete error deformation points found in InSAR data by focusing on spatially clustered deformation information, preventing these discrete error points from interfering with analysis results. This allows the extracted high- and low-value points to more accurately reflect the actual deformation trends of the geological body. Combined with geological factor screening, this technology further eliminates irrelevant or interfering information, allowing accurate identification of potential hazards, significantly improving the accuracy and reliability of geological hazard detection results.

[0008] 3. Reduce dependence on sample libraries: Different from existing automatic extraction methods that rely on large sample libraries, the technical solution of the present invention reduces problems such as missed judgments and misjudgments caused by incomplete or mismatched sample libraries, broadens the adaptability to different geological conditions and deformation characteristics, and can better cope with complex and changeable geological environments. Whether in common geological structure areas or in special geological and landform areas, it can effectively detect geological hazards, reduce the initial preparation cost and complexity of data processing, and improve the versatility and practicality of the technology.

[0009] 4. Filling the gap in technological application: The introduction of hotspot analysis technology in the field of automatic extraction of InSAR deformation fields has opened up a new direction for technological application, provided reference and reference for subsequent related research and technological development, promoted the further development of InSAR data processing technology in the field of geological hazard detection, and helped to improve the entire geological disaster monitoring and early warning system. It has important academic value and practical guiding significance, and is expected to be more widely promoted and applied in future geological research and engineering applications, generating greater economic and social benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 The embodiment of the present invention provides a geometric relationship table between satellite incident angle and different slope deformation observation values; Figure 2 The embodiment of the present invention provides a geometric relationship diagram between satellite incidence angle and deformation observation values of different slope aspects; Figure 3 This is a flowchart of a hotspot analysis technique provided by an embodiment of the present invention; Figure 4 This is a diagram showing the details of the hotspot analysis technology process provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0011] In order to further understand the content, features and effects of the present invention, the following embodiments are given as examples and described in detail with reference to the accompanying drawings.

[0012] The structure of the present invention is described in detail below with reference to the accompanying drawings.

[0013] like Figures 1 to 4As shown in the example, an embodiment of the present invention provides an automatic InSAR risk point identification method based on hotspot analysis. However, due to geometric distortion caused by satellite imaging geometry and areas with low vegetation coherence, subsequent hotspot analysis may be misleading, resulting in phase gradient discontinuities and analytical errors. Therefore, it is necessary to combine remote sensing and GIS technology to analyze and eliminate these areas to further obtain more reliable deformation areas.

[0014] 1. Slope. Based on remote sensing and GIS technology, slope data is automatically generated from DEM data. Deformed areas with a slope of less than 5% are eliminated based on the slope conditions that could cause disasters.

[0015] 2. Water bodies. Research on automatic water body identification technology based on SAR data. SAR images are used to identify water bodies and generate water body range vector data, and deformation areas within the water body range are eliminated.

[0016] 3. Geometric distortion: Based on the satellite's flight parameters and DEM files, the system automatically calculates the unreliable areas of the image that are geometrically distorted and removes these deformed areas.

[0017] After removing the unreliable areas, the large-scale data in raster format is converted into point data based on the above results. Then the coordinate system is converted from WGS-84 to the UTM projection coordinate system (projected according to the 3° zone or 6° zone according to actual needs). Finally, in order to reduce the computer processing pressure, a large number of stable points in the deformation field are removed and do not participate in subsequent calculations.

[0018] Hot spot analysis (Getis−Ord Hotspot analysis uses spatial statistics on large-scale deformation values to identify high- and low-value points with distinct spatial clustering. Hotspot analysis can be generally divided into three steps.

[0019] 1. Spatial incremental autocorrelation analysis. The whole process is based on Global Moran's The statistics are calculated, and the formula is as follows (1 As shown, its purpose is to determine the appropriate distance threshold for spatial analysis for each feature. The z-score of the feature is calculated at different distances, and the peak z-score found is used as the distance threshold parameter for spatial analysis.

[0020] (1) in is the total number of features when performing spatial clustering, is the average value of all elements, is the spatial weight of the feature, is the sum of all spatial weights and is given by: (2) From GlobalMoran's According to the results of the index, the z value is used to determine the degree of spatial autocorrelation. The calculation formula is as follows: the spatial autocorrelation of the elements at each distance is calculated step by step, and finally the distance corresponding to the peak is proposed for hot spot analysis.

[0021] (3) Among them, calculation and The formula is as follows: (4) (5) 2. Use the calculated distance threshold to statistics, Statistics is a method of monitoring whether a feature point has significant spatial clustering properties by examining each feature in the surrounding environment of neighboring features within a certain distance. The feature points with clustering properties obtained are not only high-value (low-value) in themselves, but also surrounded by the same high-value and low-value points. The formula is as follows: In the above formula is the total number of elements, is the total number of features within the distance threshold, is the deformation value represented by the element, is the average value of the element. Statistics can determine whether each feature is spatially clustered or uniformly randomly distributed. The Getis−Ord The statistical results are given by its z value and The value here indicates that The values are the significance levels in hypothesis testing. The z values represent the clustering at 90% ( value is 0.10) 95% ( value is 0.05) and 99% ( The corresponding z-scores are ±1.65, ±1.96, and ±2.58, respectively. The positive and negative z-scores represent the positive and negative deformation values of the features, respectively.

[0022] Extract the hot and cold values. According to the characteristics of the InSAR deformation field, this experiment sets the key points (99% confidence level) and When is a reliable deformation point with spatial clustering properties, the elements are classified and output to obtain the deformation potential points.

[0023] The deformation areas identified by hotspot analysis still contain deformation errors caused by factors such as the satellite's own imaging. Mean filtering is used to smooth the results and remove the influence of isolated noise points. A binary image is generated using the correlation coefficient threshold method, and the internal filling is performed to extract the boundaries of the phase discontinuity areas.

[0024] Considering that the area of the hidden danger point is within a certain range, and the larger area may be due to the discontinuity of the interference phase caused by the atmospheric effect, and the smaller area may be due to the discontinuity of the interference phase caused by the area with poor coherence such as vegetation, the appropriate threshold is selected to use the number of pixels in the formed closed area for screening, remove the identified larger and smaller areas, and retain the real deformation area.

[0025] It should be noted that, in this document, relational terms such as first and second, etc., are used only 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 terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0026] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any technician familiar with this patent will not depart from the scope of the technical solution of the present invention.

Claims

1. An automatic identification method of InSAR potential danger points based on hotspot analysis, characterized by: include, S1. Masking and format conversion of unreliable regions of InSAR deformation field; Geometric distortion caused by satellite imaging geometry and areas with low vegetation coherence may mislead subsequent hotspot analysis, leading to phase gradient discontinuity and analytical errors. Therefore, it is necessary to combine remote sensing and GIS technology to analyze and eliminate them to further obtain more reliable deformation areas. After removing the unreliable areas, the large-scale data in raster form is converted into point data based on the above results. Then the coordinate system is converted from WGS-84 to UTM projection coordinate system. Finally, in order to reduce the processing pressure of the computer, a large number of stable points in the deformation field are removed and do not participate in subsequent calculations. S2. Hotspot Analysis (Getis−Ord ) technology performs spatial statistics on the deformation values obtained over a large range to obtain high-value and low-value points with obvious clustering properties in space; hotspot analysis can be generally divided into three steps: Step 1: Spatial incremental autocorrelation analysis; the whole process is based on GlobalMoran's The purpose of the statistics is to determine the appropriate distance threshold for spatial analysis for each element; the z-score of the element is calculated at different distances, and finally the peak value of the z-score is used as the distance threshold parameter for spatial analysis; Step 2: Use the calculated distance threshold statistics, Statistics is a method of monitoring whether the feature points have significant spatial clustering properties by looking at each feature in the environment of neighboring features within a certain distance. The feature points with clustering properties obtained are not only high values themselves, but also surrounded by the same high and low value points. Step 3: Extract hot and cold values. According to the characteristics of the InSAR deformation field, this experiment sets the key points and When is a reliable deformation point with spatial clustering properties, the elements are classified and output to obtain the deformation potential points; S3. The deformation areas identified by hotspot analysis still have deformation errors caused by factors such as satellite imaging itself. Mean filtering is used to smooth the results to remove the influence of isolated noise points. A binary image is generated using the correlation coefficient threshold method, and the internal filling is performed to extract the boundaries of the phase discontinuity area. Considering that the area of the hidden danger point is within a certain range, and the larger area may be due to the discontinuity of the interference phase caused by the atmospheric effect, and the smaller area may be due to the discontinuity of the interference phase caused by the area with poor coherence such as vegetation, the appropriate threshold is selected to use the number of pixels in the formed closed area for screening, remove the identified larger and smaller areas, and retain the real deformation area.

2. The method for automatically identifying potential danger points using InSAR based on hotspot analysis according to claim 1, wherein: The UTM projection coordinate system can be projected according to the 3° zone or 6° zone according to actual needs.

Images