A geological disaster discrimination method and system fusing remote sensing and big data analysis

Abstract

The application provides a geological disaster discrimination method and system fusing remote sensing and big data analysis, comprising obtaining multi-source remote sensing data and historical geological disaster point data of a region to be discriminated, and extracting multiple geological disaster evaluation factors; based on time series of remote sensing data, a time domain stability coefficient of each factor pixel is calculated, representing the fluctuation degree of the factor in the time dimension; the neighborhood range for spatial autocorrelation calculation is determined according to the time domain stability coefficient of each evaluation factor, the spatial correction amount is calculated and the factor information amount is corrected; a trigger factor set and a response factor set are divided; when calculating the information amount of a unit, if the sum of the trigger factor information amount exceeds a preset threshold, the weight of the response factor is increased, and the unit comprehensive information amount is obtained; the time domain stability coefficients of each factor in the unit are aggregated to generate a comprehensive time domain stability index, the index and the comprehensive information amount are weighted and fused, and a geological disaster spatial discrimination result map of the region to be discriminated is drawn.

Description

Technical Field

[0001] This application belongs to the field of disaster identification, and in particular relates to a geological disaster identification method and system that integrates remote sensing and big data analysis. Background Technology

[0002] In the field of disaster prevention and mitigation, multi-source remote sensing data is used for early identification and spatial susceptibility assessment of geological hazards. Information volume models, through statistical analysis of the spatial overlay relationships between historical hazard points and various geological environmental factors, assess the contribution of different states of each factor to the occurrence of geological hazards. By overlaying the information volume of each factor, regional geological hazard zoning maps are generated, providing a scientific basis for land spatial planning and disaster prevention. However, information volume models are usually based on assumptions about the spatial distribution of evaluation factors, and linear overlay of information volume may lead to biased evaluation results. Information volume summation fails to distinguish the differentiated roles of different factors in the incubation and triggering processes of geological hazards. The occurrence of geological hazards is a process of interaction between background conditions and triggering factors; treating all factors equally makes it difficult to represent the trigger-response mechanism. Moreover, most evaluation methods are static assessments, based on remote sensing images and data from a specific period, neglecting the characteristics of geological environment changes over time. Existing evaluation systems often ignore the stability or volatility of surface factors over time, thus affecting the reliability of geological hazard identification results. Summary of the Invention

[0003] This invention proposes a geological hazard identification method integrating remote sensing and big data analysis to address the problems of existing methods neglecting the spatial autocorrelation of geological elements and ignoring the characteristics of geological environment evolution over time. The method includes:

[0004] Acquire multi-source remote sensing data and historical geological hazard point data of the area to be judged, and extract multiple geological hazard evaluation factors from them; calculate the temporal stability coefficient of each evaluation factor at the pixel level based on the time series of the remote sensing data;

[0005] For any evaluation factor, the neighborhood range for spatial autocorrelation calculation is determined based on the time-domain stability coefficient of the evaluation factor, and a spatial correction amount is calculated based on the neighborhood range. The information content of the evaluation factor is corrected using the spatial correction amount to obtain the corrected information content of the evaluation factor.

[0006] Based on the absolute value of the difference in information content before and after correction of each evaluation factor, the evaluation factors are divided into a set of trigger factors and a set of response factors. When superimposing information content for each calculation unit, if the sum of information content of each trigger factor in the unit exceeds a preset trigger threshold, the weight of information content of each factor in the response factor set is increased in the superposition operation to obtain the comprehensive information content of the unit.

[0007] For each calculation unit, the temporal stability coefficients of each evaluation factor within the unit are aggregated to obtain the comprehensive temporal stability index of the unit; the comprehensive information quantity and the comprehensive temporal stability index are weighted and fused to generate the spatial discrimination result map of geological hazards in the area to be judged.

[0008] Furthermore, this invention also relates to a geological disaster identification system that integrates remote sensing and big data analysis, comprising the following modules:

[0009] The calculation module is used to acquire multi-source remote sensing data and historical geological disaster point data of the area to be judged, and extract multiple geological disaster evaluation factors from them; based on the time series of the remote sensing data, the temporal stability coefficient of each evaluation factor at the pixel level is calculated;

[0010] The correction module is used to determine the neighborhood range for spatial autocorrelation calculation based on the temporal stability coefficient of any evaluation factor, calculate the spatial correction amount based on the neighborhood range, and use the spatial correction amount to correct the information content of the evaluation factor to obtain the corrected information content of the evaluation factor.

[0011] The overlay module is used to divide the evaluation factors into a set of trigger factors and a set of response factors based on the absolute value of the difference in information content before and after the correction of each evaluation factor. When overlaying information content for each calculation unit, if the sum of information content of each trigger factor in the unit exceeds a preset trigger threshold, the weight of information content of each factor in the response factor set is increased in the overlay operation to obtain the comprehensive information content of the unit.

[0012] The generation module is used to aggregate the temporal stability coefficients of each evaluation factor within each calculation unit to obtain the comprehensive temporal stability index of the unit; and to perform weighted fusion of the comprehensive information and the comprehensive temporal stability index to generate a spatial discrimination result map of geological hazards in the area to be judged.

[0013] This invention utilizes the temporal stability of evaluation factors to determine the calculation range of spatial autocorrelation, and based on this, spatially corrects the information content of each evaluation factor. Factors are categorized into triggering and response types according to the changes in information content before and after correction. By setting trigger thresholds and adjusting response weights, the invention simulates the disaster-causing mechanism of geological hazards caused by the combined effects of background conditions and triggering factors, thus more closely resembling the actual process. By integrating comprehensive information content with comprehensive temporal stability indicators, the discrimination results not only reflect the susceptibility of the regional geological environment but also the stable state at the current timescale, thereby improving the overall reliability of geological hazard discrimination results. Attached Figure Description

[0014] Figure 1 A flowchart of the first embodiment;

[0015] Figure 2 This is a schematic diagram comparing the temporal stability of pixels;

[0016] Figure 3 A diagram illustrating the division of evaluation factors into trigger and response factor sets;

[0017] Figure 4 This is a diagram illustrating the weighting of response factors. Detailed Implementation

[0018] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0019] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0020] In the first embodiment, the present invention proposes a geological disaster identification method that integrates remote sensing and big data analysis, such as... Figure 1 ,include:

[0021] S1. Obtain multi-source remote sensing data and historical geological disaster point data of the area to be judged, and extract multiple geological disaster evaluation factors from them; based on the time series of the remote sensing data, calculate the temporal stability coefficient of each evaluation factor at the pixel level, which represents the degree of fluctuation of each evaluation factor in the time dimension;

[0022] Multi-source remote sensing data includes, but is not limited to, Landsat series imagery, Sentinel series imagery, Gaofen series imagery, and Digital Elevation Model (DEM) data; historical geological hazard point data are obtained through field surveys and compilation of historical data. The extraction process for evaluation factors is as follows: Based on DEM data, slope, aspect, elevation, topographic relief, and curvature are extracted as topographic factors; the Normalized Difference Vegetation Index (NDVI) is calculated based on multi-temporal Landsat or Sentinel imagery to represent vegetation cover; engineering geological factors such as rock groups and distance from faults are extracted using regional geological maps; and hydrometeorological factors such as multi-year average rainfall and rainfall during heavy rain periods are obtained through spatial interpolation or extraction using meteorological station data or remote sensing rainfall products. All factor data are unified in coordinate system and spatial resolution to form a raster data layer.

[0023] Taking the NDVI factor as an example, remote sensing images of the study area for each month over five consecutive years were selected to form an image time series containing 60 time phases. For any pixel within the study area, there are 60 corresponding NDVI values. The standard deviation and mean of these 60 NDVI values ​​were calculated. Dividing the standard deviation by the mean yielded the CV of that pixel, which served as the temporal stability coefficient of the NDVI factor for that pixel. A larger coefficient indicates more drastic fluctuations in vegetation cover over the five years, resulting in poorer temporal stability; conversely, a smaller coefficient indicates better stability. Figure 2 The same method is used to calculate the time-domain stability coefficient for other factors such as rainfall and soil moisture. For factors such as slope and lithology that change slowly, the time-domain stability coefficient can be set to a constant minimum value.

[0024] In an optional embodiment, calculating the temporal stability coefficient of each evaluation factor at the pixel level based on the time series of the remote sensing data includes:

[0025] For remote sensing data with a time series of length n, for any pixel, calculate the standard deviation of the time series of each evaluation factor for that pixel. and mean Through formula The calculation result is used as the temporal stability coefficient of the pixel.

[0026] Specifically, taking the Normalized Difference Vegetation Index (NDVI) as an example, assume that monthly NDVI images of a certain region for 12 consecutive months have been acquired, i.e., the time series length n is 12. For any pixel P in the image, the NDVI values ​​of the 12 months constitute a time series, for example, the series is 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3.

[0027] Calculate the arithmetic mean In this example, Calculate the standard deviation. According to the standard deviation calculation method, By applying the formula and substituting the calculated mean and standard deviation, the temporal stability coefficient of pixel P on the NDVI factor is obtained as CV = 0.366. This process will be repeated for each pixel in the image and for each evaluation factor to obtain the temporal stability coefficient of each evaluation factor.

[0028] S2, For any evaluation factor, determine the neighborhood range for spatial autocorrelation calculation based on the time-domain stability coefficient of the evaluation factor, and calculate the spatial correction amount based on the neighborhood range. Use the spatial correction amount to correct the information content of the evaluation factor to obtain the corrected information content of the evaluation factor.

[0029] The original information content value of each evaluation factor under each graded state is calculated using the standard information content formula. This formula is the natural logarithm of the ratio of the probability of a disaster occurring under a certain combination of factor states to the overall probability of a disaster occurring in the region. Correction is then performed. Taking the slope factor as an example, for a central pixel, the temporal stability coefficient of that pixel is first obtained. The temporal stability coefficient is proportional to the neighborhood range; that is, pixels with poor stability have more information noise and require a larger neighborhood to capture spatial correlation and smooth the noise. A larger analysis neighborhood, such as a 5×5 pixel window, is set. Pixels with good stability have reliable information, and a smaller analysis neighborhood, such as a 3×3 pixel window, is set. After the neighborhood range is determined, the arithmetic mean of the information content of all pixels within that neighborhood, excluding the central pixel, is calculated as the spatial correction amount.

[0030] In an optional embodiment, determining the neighborhood range for spatial autocorrelation calculation based on the temporal stability coefficient of the evaluation factor includes:

[0031] For any evaluation factor, the temporal stability coefficient at each pixel is determined to be less than the stability threshold. When the temporal stability coefficient at the pixel is less than the stability threshold, the neighborhood range for spatial autocorrelation calculation centered at the pixel is determined to be a 3×3 pixel window.

[0032] When the temporal stability coefficient of the pixel is greater than or equal to the stability threshold, the neighborhood range is determined to be a 5×5 pixel window.

[0033] Set a stability threshold, for example, 0.2. This threshold is used to distinguish the temporal stability of a cell, that is, whether the factor value on the cell changes drastically or gradually over time.

[0034] Traverse each pixel on the temporal stability coefficient map of the evaluation factors. Assume that for evaluation factor A, the temporal stability coefficient of pixel P1 is calculated to be 0.15. Since 0.15 < 0.2, this indicates that the pixel exhibits high stability over time, and the information is less affected by short-term disturbances. Therefore, a small 3×3 pixel window is determined for pixel P1 as the neighborhood range for spatial autocorrelation calculation. Conversely, if the temporal stability coefficient of another pixel P2 is 0.3, since 0.3 ≥ 0.2, this indicates that the pixel varies significantly over time, has poor stability, and the information may contain more noise or short-term fluctuations. Therefore, a large neighborhood is needed to smooth out the influence; a 5×5 pixel window is determined for this pixel as the neighborhood range.

[0035] In an optional embodiment, taking slope factor as an example, a mountainous area with a total area of ​​80 km² is selected for assessment. Through field surveys and historical record review, 40 historical geological hazard points are identified in the area. First, the overall probability of hazard occurrence in the area is calculated as P0 = total number of historical hazard points / total area of ​​the area = 40 / 80 = 0.5. Then, according to the industry standard for geological hazard assessment, the slope factor is divided into four levels: 0-5°, 5-15°, 15-30°, and ≥30°. Through GIS spatial overlay analysis, the area of ​​each level and the number of hazard points within the corresponding range are statistically analyzed: the 0-5° level has an area of ​​20 km² and corresponds to 4 hazard points, with a hazard probability P1 = 4 / 20 = 0.2; the 5-15° level has an area of ​​32 km². 2 The corresponding number of disaster points is 12, with a probability P² = 12 / 3² = 0.375; the area affected by the 15-30° grading is 20 km². 2 There are 16 corresponding disaster points, with a probability P3 = 16 / 20 = 0.8; the area classified as ≥30° is 8 km². 2 There are 8 corresponding disaster sites, and the probability P4 = 8 / 8 = 1.0; finally, substituting into the standard information content formula... The original information values ​​for each grade are calculated sequentially: 0-5° grade I1≈-0.916, 5-15° grade I2≈-0.288, 15-30° grade I3≈0.470, and ≥30° grade I4≈0.693. These original information values ​​can then be assigned to all pixels of the corresponding grade for subsequent spatial correction calculations.

[0036] In an optional embodiment, the step of calculating a spatial correction amount based on the neighborhood range and using the spatial correction amount to correct the information content of the evaluation factor includes:

[0037] The arithmetic mean of the information content of all pixels within a defined neighborhood, excluding the center pixel, is used as the spatial correction value. The corrected information content is calculated using the following formula:

[0038]

[0039] in For the corrected information content, The original amount of information. k1 is a spatial correction value, and k1 is a coefficient greater than 0 and less than 1. Preferably, k1 = 0.6.

[0040] Specifically, taking pixel P as an example, assuming the neighborhood range is defined as a 3×3 window, its original information content... The information content of the eight neighboring pixels (excluding the center pixel P) within the 3×3 window is 2.0, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, and 2.7, respectively. Therefore, the spatial correction value is 2.5. That is, the arithmetic mean of the information content of the 8 neighboring pixels, is calculated as follows: .

[0041] The weighting coefficients 0.6 and 0.4 are preset parameters, representing 60% of the original information and 40% of the neighborhood influence. Substituting the example data, the information content of pixel P after correction is: This method integrates the information of the central pixel with the average features of the spatial neighborhood, achieving smoothing and correction of the original information.

[0042] S3. Based on the absolute value of the difference in information content before and after correction of each evaluation factor, the evaluation factors are divided into a set of trigger factors and a set of response factors. When superimposing information content for each calculation unit, if the sum of information content of each trigger factor in the unit exceeds a preset trigger threshold, the weight of information content of each factor in the response factor set is increased in the superposition operation to obtain the comprehensive information content of the unit.

[0043] For each evaluation factor, the absolute values ​​of the differences between the corrected and uncorrected information values ​​of that factor across all pixels in the entire study area are calculated. Factors with larger sums, such as heavy rainfall and rapid changes in vegetation cover, indicate strong spatial heterogeneity and are greatly influenced by their neighbors, and are thus classified into the trigger factor set. Factors with smaller sums, such as lithology and topographic relief, indicate relatively uniform and stable spatial distribution, and are thus classified into the response factor set. When calculating the comprehensive information value of each raster unit, the corrected information values ​​of all trigger factors within that unit, such as rainfall and NDVI changes, are summed to obtain the total trigger information value. This sum is compared with a trigger threshold, such as 0.8, set through statistical analysis or expert experience. If the sum is greater than 0.8, the corrected information values ​​of all response factors within that unit, such as slope and lithology, are multiplied by a weighting coefficient greater than 1, such as 1.2, and then added to the total trigger information value; if the sum is not greater than 0.8, the weighting coefficient for all response factors is 1. The sum is then used to obtain the comprehensive information value of that unit.

[0044] To classify the factors based on their sensitivity to spatial correction, in an optional embodiment, the step of dividing the evaluation factors into a trigger factor set and a response factor set based on the absolute value of the difference in information content before and after correction for each evaluation factor includes:

[0045] For each evaluation factor, calculate the average of the absolute values ​​of the information content difference of the evaluation factor before and after correction across all pixels. ,when When the value exceeds the factor partitioning threshold, the evaluation factor is assigned to the trigger factor set.

[0046] when When the evaluation factor is less than or equal to the factor division threshold, it is classified into the response factor set.

[0047] For an evaluation factor, such as rainfall, there are indicative maps before and after correction. The absolute value of the difference between corresponding pixel values ​​in the two layers is calculated pixel by pixel. For example, for pixel P1, if the indicative value before correction is 3.2 and after correction is 2.8, then the absolute value of the difference is 0.4. This calculation is performed on all pixels throughout the entire study area.

[0048] After calculating the differences for all pixels, the absolute values ​​of all the differences are arithmetically averaged to obtain the global average information content change of the evaluation factor. Assuming the entire area has 10,000 pixels, and the sum of the absolute values ​​of all differences is 2,000, then the rainfall factor... The value is compared with the preset factor division threshold of 0.15. Since 0.2 > 0.15, it indicates that the information content of the rainfall factor changes drastically during the spatial correction process, exhibiting strong spatial heterogeneity; therefore, it is included in the trigger factor set. If another factor, such as slope... If the calculated result is 0.1, and 0.1 ≤ 0.15, then it is included in the response factor set, such as... Figure 3 .

[0049] In an optional embodiment, when the sum of the information content of each triggering factor within the unit exceeds a preset triggering threshold, increasing the weight of the information content of each factor in the response factor set during the superposition operation includes:

[0050] When the sum of the information content of all triggering factors in the calculation unit is greater than the triggering threshold, the weighting coefficient of the information content of each factor in the response factor set in the unit is set to 1.5.

[0051] When the sum of the information content of the triggering factors is less than or equal to the triggering threshold, the weighting coefficient is set to 1.0.

[0052] Specifically, the judgment is made on a single computational unit, such as a single pixel. Based on the above division results, it is assumed that the triggering factor set includes two factors: rainfall and fault zone density, and the response factor set includes two factors: vegetation cover and slope.

[0053] For any computational unit P within the study area, the information content of all triggering factors in that unit is extracted and summed. Assuming that the information content of rainfall in unit P is 0.9 and the information content of fault zone density is 0.5, the sum of the information content of the triggering factors in that unit is 1.4. This sum is compared with a preset triggering threshold of 1.2. Since 1.4 > 1.2, the weighting coefficient of all response factors within that unit is set to 1.5. When calculating the comprehensive information content, the information content of vegetation cover and slope in unit P will be multiplied by 1.5. If the sum of the information content of the triggering factors in another computational unit Q is 1.1, and 1.1 ≤ 1.2, then the weighting coefficient of the response factors remains at 1.0, i.e., no adjustment is made. Figure 4 .

[0054] S4. For each calculation unit, aggregate the temporal stability coefficients of each evaluation factor within the unit to obtain the comprehensive temporal stability index of the unit; weight and fuse the comprehensive information quantity with the comprehensive temporal stability index to generate a spatial discrimination result map of geological hazards in the area to be judged.

[0055] For each raster calculation unit, the arithmetic mean of the temporal stability coefficient values ​​of all evaluation factors contained in the unit is calculated to obtain the comprehensive temporal stability index of the unit. The weights here can be determined based on the importance of each factor to the impact of geological hazards, for example, through the analytic hierarchy process (AHP). The comprehensive information quantity calculated above is then weighted and fused with the comprehensive temporal stability index. For example, if the weight of the comprehensive information quantity is set to 0.7 and the weight of the comprehensive temporal stability index is set to 0.3, then the geological hazard discrimination index is: 0.7 multiplied by the comprehensive information quantity plus 0.3 multiplied by the comprehensive temporal stability index. This fusion calculation is performed on all calculation units within the study area to obtain the discrimination index for each unit. In the geographic information system software, the natural breakpoint classification method is used to classify the discrimination indices of all units into multiple geological hazard susceptibility levels (high, medium, and low), assigning different colors to each level to generate a visualized spatial discrimination result map of geological hazards.

[0056] To integrate the temporal stability information of multiple evaluation factors into a comprehensive index for a computational unit, in an optional embodiment, for each computational unit, the temporal stability coefficients of each evaluation factor within the unit are aggregated to obtain the comprehensive temporal stability index of the unit, including:

[0057] The arithmetic mean of the temporal stability coefficients of all evaluation factors in all pixels within the calculation unit is the comprehensive temporal stability index of the unit.

[0058] Define the scope of the computational unit, for example, a 10×10 pixel grid. Assume that four evaluation factors are used in the study: slope, rainfall, vegetation cover, and geological structure.

[0059] For a specific 10×10 computational unit, which contains 100 pixels, a temporal stability coefficient has been calculated for each pixel on each evaluation factor. Therefore, there are a total of 400 temporal stability coefficient values ​​within this computational unit. These 400 values ​​are extracted; for example, the stability coefficients of the first pixel on the four factors are 0.1, 0.3, 0.15, and 0.05, respectively, and those of the second pixel are 0.12, 0.33, 0.16, and 0.04, and so on. The sum of these 400 values, divided by 400, gives the overall temporal stability index of the computational unit. If the sum is 60, the overall temporal stability index of the unit is 0.15.

[0060] In an optional embodiment, the step of weightedly fusing the comprehensive information quantity with the comprehensive temporal stability index to generate a spatial discrimination result map of geological hazards in the area to be discriminated includes:

[0061] The formula for calculating the Geological Hazard Discriminant Index (GDI) is as follows:

[0062]

[0063] in The total amount of information, The comprehensive time-domain stability index is defined as k2, which is a coefficient greater than 0 and less than 1, preferably k2 = 0.7.

[0064] Specifically, for a given computational unit, two input values ​​are required: the total amount of information. and comprehensive time-domain stability index .

[0065] Among them, the comprehensive information volume It is obtained by superimposing the information content of all corrected and weighted evaluation factors within the unit, with an assumed value of 5.0. Comprehensive time-domain stability index. The value obtained from the above calculation reflects the overall temporal volatility of all factor information within the unit, with an assumed value of 0.2. These two values ​​are then substituted into a preset weighted fusion formula. Weight coefficients of 0.7 and 0.3 indicate that, in the discrimination, the contribution of comprehensive information accounts for 70%, while the contribution of temporal stability accounts for 30%. Substituting the example data, the calculated GDI is 3.56. This GDI value is the geological hazard susceptibility index of the unit. After repeating the above GDI calculation for all calculation units within the region, the natural breakpoint method is used in QGIS software to divide the GDI values ​​of all units into four levels: GDI < 2.0 (low susceptibility area, assigned light green), 2.0-3.0 (medium susceptibility area, assigned yellow), 3.0-4.0 (high susceptibility area, assigned orange), and > 4.0 (extremely high susceptibility area, assigned red). Geographic base maps of regional administrative boundaries, rivers, and main roads are overlaid to generate a visualized spatial discrimination result map of geological hazards.

[0066] In an optional embodiment, the present invention also proposes a geological hazard identification system that integrates remote sensing and big data analysis, comprising the following modules:

[0067] The calculation module is used to acquire multi-source remote sensing data and historical geological disaster point data of the area to be judged, and extract multiple geological disaster evaluation factors from them; based on the time series of the remote sensing data, the module calculates the temporal stability coefficient of each evaluation factor at the pixel level, which represents the degree of fluctuation of each evaluation factor in the time dimension;

[0068] The correction module is used to determine the neighborhood range for spatial autocorrelation calculation based on the temporal stability coefficient of any evaluation factor, calculate the spatial correction amount based on the neighborhood range, and use the spatial correction amount to correct the information content of the evaluation factor to obtain the corrected information content of the evaluation factor.

[0069] The overlay module is used to divide the evaluation factors into a set of trigger factors and a set of response factors based on the absolute value of the difference in information content before and after the correction of each evaluation factor. When overlaying information content for each calculation unit, if the sum of information content of each trigger factor in the unit exceeds a preset trigger threshold, the weight of information content of each factor in the response factor set is increased in the overlay operation to obtain the comprehensive information content of the unit.

[0070] The generation module is used to aggregate the temporal stability coefficients of each evaluation factor within each calculation unit to obtain the comprehensive temporal stability index of the unit; and to perform weighted fusion of the comprehensive information and the comprehensive temporal stability index to generate a spatial discrimination result map of geological hazards in the area to be judged.

[0071] In an optional embodiment, calculating the temporal stability coefficient of each evaluation factor at the pixel level based on the time series of the remote sensing data includes:

[0072] For remote sensing data with a time series of length n, for any pixel, calculate the standard deviation of the time series of each evaluation factor for that pixel. and mean Through formula The calculation result is used as the temporal stability coefficient of the pixel.

[0073] In an optional embodiment, determining the neighborhood range for spatial autocorrelation calculation based on the temporal stability coefficient of the evaluation factor includes:

[0074] For any evaluation factor, the temporal stability coefficient at each pixel is determined to be less than the stability threshold. When the temporal stability coefficient at the pixel is less than the stability threshold, the neighborhood range for spatial autocorrelation calculation centered at the pixel is determined to be a 3×3 pixel window.

[0075] When the temporal stability coefficient of the pixel is greater than or equal to the stability threshold, the neighborhood range is determined to be a 5×5 pixel window.

[0076] In an optional embodiment, the step of correcting the information content of the evaluation factor using the spatial correction amount includes:

[0077] The arithmetic mean of the information content of all pixels within a defined neighborhood, excluding the center pixel, is used as the spatial correction value. The corrected information content is calculated using the following formula:

[0078]

[0079] in For the corrected information content, The original amount of information. k1 is a spatial correction value, where k1 is a coefficient greater than 0 and less than 1.

[0080] In an optional embodiment, dividing the evaluation factors into a trigger factor set and a response factor set based on the absolute value of the difference in information content before and after correction of each evaluation factor includes:

[0081] For each evaluation factor, calculate the average of the absolute values ​​of the information content difference of the evaluation factor before and after correction across all pixels. ,when When the value exceeds the factor partitioning threshold, the evaluation factor is assigned to the trigger factor set.

[0082] when When the evaluation factor is less than or equal to the factor division threshold, it is classified into the response factor set.

[0083] In an optional embodiment, when the sum of the information content of each triggering factor within the unit exceeds a preset triggering threshold, increasing the weight of the information content of each factor in the response factor set during the superposition operation includes:

[0084] When the sum of the information content of all triggering factors in the calculation unit is greater than the triggering threshold, the weighting coefficient of the information content of each factor in the response factor set in the unit is set to 1.5.

[0085] When the sum of the information content of the triggering factors is less than or equal to the triggering threshold, the weighting coefficient is set to 1.0.

[0086] In an optional embodiment, the step of aggregating the time-domain stability coefficients of each evaluation factor within each computational unit to obtain a comprehensive time-domain stability index for the unit includes:

[0087] The arithmetic mean of the temporal stability coefficients of all evaluation factors in all pixels within the calculation unit is the comprehensive temporal stability index of the unit.

[0088] In an optional embodiment, the weighted fusion of the comprehensive information content and the comprehensive time-domain stability index includes:

[0089] The formula for calculating the Geological Hazard Discriminant Index (GDI) is as follows:

[0090]

[0091] in The total amount of information, Here, k2 is the comprehensive time-domain stability index, where k2 is a coefficient greater than 0 and less than 1.

[0092] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0093] The functional modules shown in the above-described block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0094] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0095] The aspects of this application have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0096] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A geological disaster discrimination method combining remote sensing and big data analysis, characterized in that, Includes the following steps: Acquire multi-source remote sensing data and historical geological hazard point data of the area to be judged, and extract multiple geological hazard evaluation factors from them; Based on the time series of the remote sensing data, the temporal stability coefficient of each evaluation factor at the pixel level is calculated. For any evaluation factor, the neighborhood range for spatial autocorrelation calculation is determined based on the time-domain stability coefficient of the evaluation factor, and a spatial correction amount is calculated based on the neighborhood range. The information content of the evaluation factor is corrected using the spatial correction amount to obtain the corrected information content of the evaluation factor. Based on the absolute value of the difference in information content before and after correction of each evaluation factor, the evaluation factors are divided into a set of triggering factors and a set of response factors. When superimposing information on each computing unit, if the sum of the information of each triggering factor within the unit exceeds a preset triggering threshold, the weight of the information of each factor in the response factor set is increased during the superposition operation to obtain the comprehensive information of the unit. For each calculation unit, the time-domain stability coefficients of each evaluation factor within the unit are aggregated to obtain the comprehensive time-domain stability index of the unit. The comprehensive information quantity and the comprehensive temporal stability index are weighted and fused to generate a spatial discrimination result map of geological hazards in the area to be discriminated.

2. The method of claim 1, wherein, The calculation of the temporal stability coefficient of each evaluation factor at the pixel level based on the time series of the remote sensing data includes: For remote sensing data with time series length n, for any pixel, calculate the standard deviation of each evaluation factor time series on the pixel and mean value , through the formula Take the calculation result as the time domain stability coefficient of the pixel.

3. The method of claim 1, wherein, The step of determining the neighborhood range for spatial autocorrelation calculation based on the temporal stability coefficient of the evaluation factor includes: For any evaluation factor, the temporal stability coefficient at each pixel is determined to be less than the stability threshold. When the temporal stability coefficient at the pixel is less than the stability threshold, the neighborhood range for spatial autocorrelation calculation centered at the pixel is determined to be a 3×3 pixel window. When the temporal stability coefficient of the pixel is greater than or equal to the stability threshold, the neighborhood range is determined to be a 5×5 pixel window.

4. The method according to claim 1, characterized in that, The step of correcting the information content of the evaluation factor using the spatial correction amount includes: The arithmetic mean of the information content of all pixels within a defined neighborhood, excluding the center pixel, is used as the spatial correction value. The corrected information content is calculated using the following formula: in For the corrected information content, The original amount of information. k1 is a spatial correction value, where k1 is a coefficient greater than 0 and less than 1.

5. The method according to claim 1, characterized in that, The step of dividing the evaluation factors into a trigger factor set and a response factor set based on the absolute value of the difference in information content before and after correction for each evaluation factor includes: For each evaluation factor, calculate the average of the absolute values ​​of the information content difference of the evaluation factor before and after correction across all pixels. ,when When the value exceeds the factor partitioning threshold, the evaluation factor is assigned to the trigger factor set. when When the evaluation factor is less than or equal to the factor division threshold, it is classified into the response factor set.

6. The method according to claim 2, characterized in that, When the sum of the information content of each triggering factor within the unit exceeds a preset triggering threshold, the weight of the information content of each factor in the response factor set is increased during the superposition operation, including: When the sum of the information content of all triggering factors in the calculation unit is greater than the triggering threshold, the weighting coefficient of the information content of each factor in the response factor set in the unit is set to 1.

5. When the sum of the information content of the triggering factors is less than or equal to the triggering threshold, the weighting coefficient is set to 1.

0.

7. The method according to claim 1, characterized in that, For each computational unit, the temporal stability coefficients of each evaluation factor within the unit are aggregated to obtain the comprehensive temporal stability index of the unit, including: The arithmetic mean of the temporal stability coefficients of all evaluation factors in all pixels within the calculation unit is the comprehensive temporal stability index of the unit.

8. The method according to claim 1, characterized in that, The weighted fusion of the comprehensive information content and the comprehensive time-domain stability index includes: The formula for calculating the Geological Hazard Discriminant Index (GDI) is as follows: in The total amount of information, Here, k2 is the comprehensive time-domain stability index, where k2 is a coefficient greater than 0 and less than 1.

9. A geological hazard identification system integrating remote sensing and big data analysis, characterized in that, Includes the following modules: The calculation module is used to acquire multi-source remote sensing data and historical geological disaster point data of the area to be judged, and extract multiple geological disaster evaluation factors from them; Based on the time series of the remote sensing data, the temporal stability coefficient of each evaluation factor at the pixel level is calculated. The correction module is used to determine the neighborhood range for spatial autocorrelation calculation based on the temporal stability coefficient of any evaluation factor, calculate the spatial correction amount based on the neighborhood range, and use the spatial correction amount to correct the information content of the evaluation factor to obtain the corrected information content of the evaluation factor. The overlay module is used to divide the evaluation factors into a set of triggering factors and a set of response factors based on the absolute value of the difference in information content before and after each evaluation factor is corrected. When superimposing information on each computing unit, if the sum of the information of each triggering factor within the unit exceeds a preset triggering threshold, the weight of the information of each factor in the response factor set is increased during the superposition operation to obtain the comprehensive information of the unit. The generation module is used to aggregate the time-domain stability coefficients of each evaluation factor within each calculation unit to obtain the comprehensive time-domain stability index of the unit. The comprehensive information quantity and the comprehensive temporal stability index are weighted and fused to generate a spatial discrimination result map of geological hazards in the area to be discriminated.

10. The system according to claim 9, characterized in that, The calculation of the temporal stability coefficient of each evaluation factor at the pixel level based on the time series of the remote sensing data includes: For remote sensing data with a time series of length n, for any pixel, calculate the standard deviation of the time series of each evaluation factor for that pixel. and mean Through formula The calculation result is used as the temporal stability coefficient of the pixel.

Images