A geological disaster analysis method and system based on remote sensing data fusion

Abstract

The present invention belongs to the technical field of geological disaster analysis and discloses a geological disaster analysis method and system based on remote sensing data fusion. The method obtains influencing factor data information within the disaster remote sensing data, and based on the acquired influencing factor data information within the disaster remote sensing data, utilizes a multi-influencing factor fusion method to obtain a geological disaster comprehensive analysis model and a geological disaster comprehensive analysis image. Based on the geological disaster comprehensive analysis image, comprehensive characteristics of the area affected by the disaster are obtained and visualized; the regional comprehensive characteristics include surface landform change characteristics in the disaster-affected area. The present invention analyzes geological disasters based on the fusion of multiple influencing factors from remote sensing data, obtains accurate analysis information, and further provides high-precision data information for monitoring or predicting geological disasters.

Description

Technical Field

[0001] The invention belongs to the technical field of geological disaster analysis and discloses a geological disaster analysis method and system based on remote sensing data fusion. Background Art

[0002] Remote sensing refers to a comprehensive technology that uses various sensing instruments to collect, process and finally image the electromagnetic wave information data radiated and reflected by distant targets without direct contact with the target objects based on the theory of electromagnetic radiation. After obtaining remote sensing data, the objects displayed in the remote sensing data can be visually interpreted based on manual or machine recognition to obtain data such as the height and shape of the objects. Therefore, remote sensing technology can be used to detect the topography and landforms in the area. In this context, when a disaster occurs, remote sensing technology can be used to survey the specific conditions in the disaster area, so as to quickly determine information such as the disaster occurrence point and disaster range, so as to formulate a more reasonable disaster relief plan and reduce the risk of the disaster relief process.

[0003] To solve the above problems, the invention patent: Geological Hazard Analysis Method and System Based on Remote Sensing Data Fusion (Publication No. CN116030354A, Publication Date 20230428) is published:

[0004] Step S1: Acquire a historical remote sensing dataset and a disaster remote sensing dataset, wherein the historical remote sensing dataset includes remote sensing data measured at the same time point before a disaster occurs, and the disaster remote sensing dataset includes remote sensing data measured at the same time point after a disaster occurs, fuse all remote sensing data in the historical remote sensing dataset to generate a first image, and fuse all data in the disaster remote sensing dataset to generate a second image;

[0005] Step S2: Segmenting the first image and the second image into a plurality of identical grid regions, performing visual interpretation on each grid region in the first image and the second image, respectively, to obtain first feature information data and second feature information data in the first image and the second image, wherein the first feature information data and the second feature information data both include a feature name, a feature category, and the coordinates of the region where the feature is located;

[0006] Step S3: generating a shooting number based on the shooting time of the first image and the second image, establishing a first table and a second table, wherein the first table and the second table respectively include the shooting numbers of the first image and the second image, and filling the first ground feature information data and the second ground feature information data into the first table and the second table respectively;

[0007] Step S4: comparing the first ground feature information data and the second ground feature information data in the first table and the second table, to determine the coordinates of the area affected by the disaster in the first image based on the comparison result;

[0008] Furthermore, geological disasters refer to geological actions caused by natural or human factors that cause damage and loss to human life, property and the environment, such as collapse, landslides, mudslides, ground fissures, soil erosion, land desertification and swamping, soil salinization, as well as earthquakes, volcanoes, geothermal damage, etc. Geological disaster prevention and control refers to changing the process of these geological disasters through effective geological engineering means to achieve the purpose of reducing or preventing disasters and minimizing the loss of life and property. Geological disasters occur under certain dynamic inductions, some of which are natural and some are man-made. Based on this, geological disasters can also be roughly divided into two categories according to their dynamic causes: natural geological disasters and man-made geological disasters. The location, scale and frequency of natural geological disasters are controlled by natural geological conditions and are not affected by the development of human history. Man-made geological disasters are restricted by human engineering development activities.

[0009] At present, when analyzing geological disaster information, most of the time, the geological disaster information is transmitted to the computer of the corresponding disaster prediction station for disaster analysis. However, this analysis method is relatively limited. It cannot achieve a comprehensive analysis of geological disasters by using geological Internet big data, cannot achieve the purpose of accurately grading and analyzing the disaster level, and cannot achieve a rapid assessment of the risk of geological disasters, which brings great inconvenience to people's monitoring of geological disasters.

[0010] To address the above-mentioned issues, an invention patent for an internet-based geological disaster analysis method (publication number CN109034588A, publication date 20181218) discloses the following: S1. First, the geological disaster monitoring module monitors the geological disaster situation and transmits the monitored disaster situation to the disaster situation management module. The disaster situation management module can then transmit the monitored geological disaster situation data to a geographic internet big data database for storage. S2. The disaster situation management module can then control the disaster severity classification unit and the disaster change classification unit within the geological disaster classification system to extract geographic information data stored in the geographic internet big data database to perform disaster classification. This relates to the field of geological disaster analysis technology. This internet-based geological disaster analysis method addresses the limitations of existing analysis methods by utilizing geological internet big data to conduct a comprehensive analysis of geological disaster situations, achieving the purpose of accurately grading and analyzing the disaster level, and realizing a rapid assessment of the risks of geological disasters.

[0011] Through the above analysis, the problems and defects of the existing technology are as follows: most of the existing technologies are based on monitoring one influencing factor of geological disasters, but multiple influencing factors (collapse, landslide, mudslide, ground fissures, soil erosion, land desertification and swamping, soil salinization, as well as earthquakes, volcanoes, geothermal hazards, etc.) affect each other and are constantly changing, so a comprehensive judgment is needed, but the existing technology lacks a method for analyzing geological disasters based on the fusion of multiple influencing factors based on remote sensing data. Summary of the Invention

[0012] In order to overcome the problems existing in the related art, the embodiments disclosed in the present invention provide a geological disaster analysis method and system based on remote sensing data fusion.

[0013] The technical solution is as follows: A geological disaster analysis method based on remote sensing data fusion, including:

[0014] S1: Acquire the influencing factor data information in the disaster remote sensing data, which includes: collapse, landslide, debris flow, ground fissure, soil erosion, land desertification, swamping, soil salinization, as well as earthquake, volcano, and geothermal damage image information;

[0015] S2: Based on the acquired influencing factor data information, a multi-influencing factor fusion method is used to obtain a comprehensive analysis model of geological hazards and a comprehensive analysis image of geological hazards;

[0016] S3, based on the comprehensive analysis image of geological disasters, obtain the comprehensive characteristics of the area affected by the disaster and visualize them; the comprehensive characteristics of the area include the surface landform change characteristics of the area affected by the disaster.

[0017] In step S2, a multi-influencing factor fusion method is used to obtain a geological disaster comprehensive analysis model and a geological disaster comprehensive analysis image, including:

[0018] Step 1: Input the image training sample A of the influencing factors in the disaster remote sensing data and the image testing sample B of the influencing factors in the disaster remote sensing data, which are used for the training of the geological disaster comprehensive analysis model and the evaluation of the geological disaster comprehensive analysis model respectively;

[0019] Step 2: Initial pixel labeling for the impact factor image training sample A in the disaster remote sensing data Conduct normality tests;

[0020] Step 3: According to the current geological disaster area Number of message trees for each influencing factor in disaster remote sensing data ,by As fusion pixel markers, establish a message tree of each influencing factor in each disaster remote sensing data , integrating the loss function deviation of the current geological hazard comprehensive analysis model;

[0021] Step 4: After the fusion of the current geological disaster comprehensive analysis model is completed, calculate the fusion deviation under the current geological disaster section , , is the pixel label of the data to be fused, is the predicted value of the current geological disaster section for training sample A, and the expression is:

[0022] ;

[0023] Where, is the learning rate, is the cumulative predicted value of the current geological disaster section for training sample A, The first geological disaster area Message tree of each influencing factor in disaster remote sensing data Predicted values on the training set and update the data values to be fused ;

[0024] Step 5: Fusion bias of the influencing factor image training sample A in the disaster remote sensing data Conduct normality tests;

[0025] Step 6: When the loss function of the evaluation function on the impact factor image test sample B in the disaster remote sensing data no longer decreases within the specified number of early evolution stop stages, the training of the geological disaster comprehensive analysis model stops, and the optimal number of iteration stages of the geological disaster comprehensive analysis model is obtained. , complete the training of the geological disaster comprehensive analysis model, jump to step seven, otherwise the loss function on the impact factor image test sample B in the disaster remote sensing data continues to decrease, jump to step four, and continue the stage iteration;

[0026] Step 7: Predict the test set of each influencing factor in the disaster remote sensing data, and calculate the optimal number of iteration stages of the geological disaster comprehensive analysis model. , and obtain the prediction results of the test set of each influencing factor in the disaster remote sensing data , the expression is:

[0027] ;

[0028] Where, For the stage Pixel marking pre-disaster remote sensing data of each influencing factor image evolution and anti-disaster remote sensing data of each influencing factor image evolution, is the impact factor, For the stage Message tree of various influencing factors in disaster remote sensing data prediction results.

[0029] In step 2, the initial pixel marking of the impact factor image training sample A in the disaster remote sensing data is performed. Perform normality tests, including:

[0030] (1) Calculate the data distribution discrete bias rate based on the pixel labels of the image training samples of each influencing factor in the disaster remote sensing data , the expression is:

[0031] ;

[0032] Where, is the pixel label of the data to be fused, that is ; is the sample pixel label mean, is the sample size, Label the standard deviation of the sample pixels;

[0033] (2) Calculate the proportion of data distribution fusion suitable state according to the pixel labels of image training samples of each influencing factor in disaster remote sensing data , the expression is:

[0034] ;

[0035] (3) Image fusion statistics of various influencing factors in tectonic disaster remote sensing data , the expression is:

[0036] ;

[0037] test Statistical significance, examine the degree of bias to determine whether to change the data; if If the statistic is greater than the threshold, go to step (4), otherwise go to step (5);

[0038] (4) The data is obviously biased. Determine whether the upper limit of the number of data changes is reached. Let the upper limit of the number of changes be , if so, jump to step (5); otherwise, update The value of , perform the data Changes are made by the following formula The value traversal selects the image evolution of each influencing factor in the disaster remote sensing data to make the data pixel mark Likelihood function The largest Generate a variation expression:

[0039] ;

[0040] Where, is the disaster remote sensing data transformation amount in the geological disaster area, Take the logarithm of the transformed quantity, To conform to the transformation value of the normal distribution, the value range is 0.00-1.00;

[0041] Get the new data pixel label and use it as the first stage fusion pixel label ; Set the current geological disaster section judgment variable ;

[0042] (5) The data is not obviously biased, or the number of image evolutions of each influencing factor in the data pixel mark disaster remote sensing data reaches the target upper limit, and the original data pixel mark is taken As the first stage of fusion pixel labeling , set the current geological disaster section judgment variable .

[0043] In step 3, the loss function deviation of the current geological hazard comprehensive analysis model is integrated, including:

[0044] Step 1: Collect information about various influencing factors in disaster remote sensing data to the fusion center through the communication network;

[0045] Step 2: Based on the physical process and current geological hazards, a loss function deviation matrix is comprehensively analyzed to develop an insertion method for artificial influence factors, so that the fusion of virtual geological hazard interference objects and the deviation of the current geological hazard comprehensive analysis loss function become worse;

[0046] Step 3: The fusion center decodes the received signal, and the probability of successful decoding is related to the impact value of the impact factor;

[0047] Step 4: Derive the criteria for selection probability and successful decoding probability to obtain the impact factor insertion strategy.

[0048] In step 2, the method for inserting the artificial impact factor includes:

[0049] For objects without geological hazards, the parameter information of the dynamic model and observation equation is calculated. Based on the knowledge of remote sensing data of geological hazards, a random matrix is constructed according to the identification process of remote sensing data of geological hazards. The variance of artificial influencing factors is obtained by calculation based on the local information of each influencing factor message in each remote sensing data of geological hazards and mathematical operations.

[0050] For virtual geological disaster interference objects, it is difficult to obtain accurate matrices. Due to the existence of random matrices generated in real time, virtual geological disaster interference objects cannot obtain accurate geological disaster information.

[0051] Furthermore, the variance of the artificial impact factor is obtained, including:

[0052] In time Selected components within From the information of various influencing factors in disaster remote sensing data Sent to the fusion center, artificial impact factors will be inserted before sending Add to , sending signals through a communication channel.

[0053] In step 2, a comprehensive analysis of the loss function deviation matrix is conducted to develop an insertion method for the artificial impact factor, including:

[0054] 2.1, for any initial condition , the fusion mechanism achieves perfect expected fusion when both of the following conditions are met:

[0055] ;

[0056] Where, is the deviation of the current geological disaster comprehensive analysis loss function calculated by the virtual geological disaster interference object, are diagonal elements, is the matrix null space;

[0057] 2.2, Select the impact value of artificial impact factor , using the matrix of the orthogonal basis, we get the matrix null space The orthonormal basis of

[0058] 2.3, Artificial Impact Factor It is given by: Where, The component is the impact factor.

[0059] In step 6, the calculation of the loss function on the impact factor image test sample B in the disaster remote sensing data specifically includes:

[0060] Calculation ends at the current geological disaster section The predicted value of the geological disaster comprehensive analysis model on the impact factor image test sample B in the disaster remote sensing data, stage Pixel labeling Pre-disaster remote sensing data of various influencing factors in the image evolution Anti-disaster remote sensing data of various influencing factors in the image evolution The expression is:

[0061] ;

[0062] Where, Current geological disaster area Judgment variables; for The image evolution of each influencing factor in the disaster remote sensing data is the image evolution of each influencing factor in the disaster remote sensing data, for The image evolution of each influencing factor in the disaster remote sensing data is the image evolution of each influencing factor in the disaster remote sensing data; the expression is:

[0063] .

[0064] Furthermore, the loss function value of the current geological disaster section is calculated, and the expression is:

[0065] ;

[0066] Where, For the initial pixel labels of the test set B of each influencing factor in the disaster remote sensing data, The loss function used by the system is usually a square loss function for regression problems. .

[0067] Another object of the present invention is to provide a geological hazard analysis system based on remote sensing data fusion, which is implemented by the geological hazard analysis method based on remote sensing data fusion, and includes:

[0068] An influencing factor data information acquisition module is used to obtain influencing factor data information in disaster remote sensing data, wherein the influencing factor data information includes collapse, landslide, debris flow, ground fissure, soil erosion, land desertification and swamping, soil salinization, as well as earthquake, volcano, and geothermal damage image information;

[0069] The multi-influencing factor fusion module is used to obtain a geological disaster comprehensive analysis model and a geological disaster comprehensive analysis image based on the influencing factor data information in the acquired disaster remote sensing data using the multi-influencing factor fusion method;

[0070] The module for acquiring comprehensive characteristics of the disaster-affected area is used to obtain comprehensive characteristics of the disaster-affected area based on the comprehensive analysis image of geological disasters, and to visualize them; the comprehensive characteristics of the area include the surface landform change characteristics of the disaster-affected area.

[0071] Combining all the above technical solutions, the beneficial effects of the present invention are as follows: the present invention analyzes geological disasters based on the fusion of multiple influencing factors of remote sensing data, obtains accurate analysis information, further provides high-precision data information for the monitoring or prediction of geological disasters, and provides a theoretical basis for geological disaster monitoring and the corresponding guidance plans adopted in the later stage. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure;

[0073] Figure 1 This is a flow chart of a geological disaster analysis method based on remote sensing data fusion provided by an embodiment of the present invention;

[0074] Figure 2 This is a schematic diagram of geological disaster analysis based on remote sensing data fusion provided by an embodiment of the present invention;

[0075] In the figure: 1. Impact factor data information acquisition module; 2. Multiple impact factor fusion module; 3. Comprehensive feature acquisition module of disaster-affected areas. DETAILED DESCRIPTION

[0076] To make the above-mentioned objects, features, and advantages of the present invention more readily apparent, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings. The following description sets forth numerous specific details to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art may make similar modifications without departing from the scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0077] Example 1, as Figure 1 As shown in the figure, the geological disaster analysis methods based on remote sensing data fusion include:

[0078] S1: Acquire the influencing factor data information in the disaster remote sensing data, which includes: collapse, landslide, debris flow, ground fissure, soil erosion, land desertification, swamping, soil salinization, as well as earthquake, volcano, and geothermal damage image information;

[0079] S2: Based on the acquired influencing factor data information, a multi-influencing factor fusion method is used to obtain a comprehensive analysis model of geological hazards and a comprehensive analysis image of geological hazards;

[0080] S3, based on the comprehensive analysis image of geological disasters, obtain the comprehensive characteristics of the area affected by the disaster and visualize them; the comprehensive characteristics of the area include the surface landform change characteristics of the area affected by the disaster.

[0081] In step S2, a multi-influencing factor fusion method is used to obtain a geological disaster comprehensive analysis model and a geological disaster comprehensive analysis image, including:

[0082] Step 1: Input the image training sample A of the influencing factors in the disaster remote sensing data and the image testing sample B of the influencing factors in the disaster remote sensing data, which are used for the training of the geological disaster comprehensive analysis model and the evaluation of the geological disaster comprehensive analysis model respectively;

[0083] Step 2: Initial pixel labeling for the impact factor image training sample A in the disaster remote sensing data Conduct normality tests;

[0084] Step 3: According to the current geological disaster area Number of message trees for each influencing factor in disaster remote sensing data ,by As fusion pixel markers, establish a message tree of each influencing factor in each disaster remote sensing data , integrating the loss function deviation of the current geological disaster comprehensive analysis model; specifically including:

[0085] Step 1: Collect information about various influencing factors in disaster remote sensing data to the fusion center through the communication network;

[0086] Step 2: Based on the physical process and current geological hazards, the loss function deviation matrix is comprehensively analyzed to develop an insertion method for artificial influence factors, so that the fusion of virtual geological hazard interference objects and the current geological hazard comprehensive analysis loss function deviation become worse; for geological hazard non-interference objects, the parameter information of the dynamic model and observation equation is calculated. Based on the knowledge of remote sensing data of geological disasters, a random matrix is constructed according to the identification process of remote sensing data of geological disasters. The variance of artificial influencing factors is calculated according to the local information of each influencing factor message in each remote sensing data of geological disasters and mathematical operations. For virtual geological disaster interference objects, it is difficult to obtain an accurate matrix. With the existence of random matrices generated in real time, virtual geological disaster interference objects cannot obtain accurate geological disaster information. Obtaining the variance of artificial influencing factors includes: Selected components within From the information of various influencing factors in disaster remote sensing data Sent to the fusion center, artificial impact factors will be inserted before sending Add to , sending a signal through a communication channel.

[0087] Step 3: The fusion center decodes the received signal, and the probability of successful decoding is related to the impact value of the impact factor;

[0088] Step 4: Derive the criteria for selection probability and successful decoding probability to obtain the impact factor insertion strategy.

[0089] Step 4: After the fusion of the current geological disaster comprehensive analysis model is completed, calculate the fusion deviation under the current geological disaster section , , is the pixel label of the data to be fused, is the predicted value of the current geological disaster section for training sample A, and the expression is:

[0090] ;

[0091] Where, is the learning rate, is the cumulative predicted value of the current geological disaster section for training sample A, The first geological disaster area Message tree of each influencing factor in disaster remote sensing data Predicted values on the training set and update the data values to be fused ;

[0092] Step 5: Fusion bias of the influencing factor image training sample A in the disaster remote sensing data Conduct normality tests;

[0093] Step 6: When the loss function of the evaluation function on the impact factor image test sample B in the disaster remote sensing data no longer decreases within the specified number of early evolution stop stages, the training of the geological disaster comprehensive analysis model stops, and the optimal number of iteration stages of the geological disaster comprehensive analysis model is obtained. , complete the training of the geological disaster comprehensive analysis model, jump to step seven, otherwise the loss function on the impact factor image test sample B in the disaster remote sensing data continues to decrease, jump to step four, and continue the stage iteration;

[0094] The calculation of the loss function on the impact factor image test sample B in the disaster remote sensing data specifically includes:

[0095] Calculation ends at the current geological disaster section The predicted value of the geological disaster comprehensive analysis model on the impact factor image test sample B in the disaster remote sensing data, stage Pixel labeling Pre-disaster remote sensing data of various influencing factors in the image evolution Anti-disaster remote sensing data of various influencing factors in the image evolution The expression is:

[0096] ;

[0097] Where, Current geological disaster area Judgment variables; for The image evolution of each influencing factor in the disaster remote sensing data is the image evolution of each influencing factor in the disaster remote sensing data, for The image evolution of each influencing factor in the disaster remote sensing data is the image evolution of each influencing factor in the disaster remote sensing data; the expression is:

[0098] ;

[0099] Furthermore, the loss function value of the current geological disaster section is calculated, and the expression is:

[0100] ;

[0101] Where, For the initial pixel labels of the test set B of each influencing factor in the disaster remote sensing data, The loss function used by the system is usually a square loss function for regression problems. .

[0102] Step 7: Predict the test set of each influencing factor in the disaster remote sensing data, and calculate the optimal number of iteration stages according to the comprehensive analysis model of geological hazards. , and obtain the prediction results of the test set of each influencing factor in the disaster remote sensing data , the expression is:

[0103] ;

[0104] Where, For the stage Pixel marking pre-disaster remote sensing data of each influencing factor image evolution and anti-disaster remote sensing data of each influencing factor image evolution, is the impact factor, For the stage Message tree of various influencing factors in disaster remote sensing data prediction results.

[0105] In step 2, the initial pixel marking of the impact factor image training sample A in the disaster remote sensing data is performed. Perform normality tests, including:

[0106] (1) Calculate the data distribution discrete bias rate based on the pixel labels of the image training samples of each influencing factor in the disaster remote sensing data , the expression is:

[0107] ;

[0108] Where, is the pixel label of the data to be fused, that is ; is the sample pixel label mean, is the sample size, Label the standard deviation of the sample pixels;

[0109] (2) Calculate the proportion of data distribution fusion suitable state according to the pixel labels of image training samples of each influencing factor in disaster remote sensing data , the expression is:

[0110] ;

[0111] (3) Image fusion statistics of various influencing factors in tectonic disaster remote sensing data , the expression is:

[0112] ;

[0113] test Statistical significance, examine the degree of bias to determine whether to change the data; if If the statistic is greater than the threshold, go to step (4), otherwise go to step (5);

[0114] (4) The data is obviously biased. Determine whether the upper limit of the number of data changes is reached. Let the upper limit of the number of changes be , if so, jump to step (5); otherwise, update The value of , perform the data Changes are made by the following formula The value traversal selects the image evolution of each influencing factor in the disaster remote sensing data to make the data pixel mark Likelihood function The largest Generate a variation expression:

[0115] ;

[0116] Where, is the disaster remote sensing data transformation amount in the geological disaster area, Take the logarithm of the transformed quantity, To conform to the transformation value of the normal distribution, the value range is 0.00-1.00;

[0117] Get the new data pixel label and use it as the first stage fusion pixel label ; Set the current geological disaster section judgment variable ;

[0118] (5) The data is not obviously biased, or the number of image evolutions of each influencing factor in the data pixel mark disaster remote sensing data reaches the target upper limit, and the original data pixel mark is taken As the first stage of fusion pixel labeling , set the current geological disaster section judgment variable .

[0119] In step 2, a comprehensive analysis of the loss function deviation matrix is conducted to develop an insertion method for the artificial impact factor, including:

[0120] 2.1, for any initial condition , the fusion mechanism achieves perfect expected fusion when both of the following conditions are met:

[0121] ;

[0122] Where, is the deviation of the current geological disaster comprehensive analysis loss function calculated by the virtual geological disaster interference object, are diagonal elements, is the matrix null space;

[0123] 2.2, Select the impact value of artificial impact factor , using the matrix of the orthogonal basis, we get the matrix null space The orthonormal basis of

[0124] 2.3, Artificial Impact Factor It is given by:

[0125] ;

[0126] Where, The component is the impact factor.

[0127] Through the above embodiments, it can be seen that, first, the distribution of geological disaster data presents diversity, and the single formula disaster remote sensing data's image evolution of each influencing factor is difficult to handle a variety of biased situations. In addition, in the fusion process of the existing technology, there is a lack of immediate data correction strategy when there is bias or long tail, and different correction strategies often have relevant requirements for data. If the image evolution of each influencing factor in the corresponding disaster remote sensing data is not performed, the correction strategy cannot proceed. Performing the image evolution of each influencing factor in the corresponding disaster remote sensing data often results in forced mapping of the data, which destroys the inherent statistical laws of the data, affects the subsequent algorithm fusion ability, and cannot improve the prediction accuracy. How to reasonably perform data distribution correction and ensure the fusion ability of the algorithm is crucial. The present invention establishes the next tree of each influencing factor message tree in the disaster remote sensing data based on the deviation direction of the fusion loss function. The process of establishing the message tree group of each influencing factor in the entire disaster remote sensing data is a tightly coupled process. Changing the fusion deviation at a certain stage will affect the establishment of the message tree of each influencing factor in all subsequent disaster remote sensing data. How to achieve the purpose of flexibly adjusting the data pixel label distribution through patterned decoupling, and alleviate the accumulation of bias to ensure the algorithm fusion effect and improve the prediction accuracy.

[0128] This invention uses multi-stage pixel labeling to analyze the evolution of various influencing factors within disaster remote sensing data, effectively mitigating the impact of data bias on the algorithm system. This allows the algorithm to dynamically adjust for bias during iteration, improving its flexibility and fusion capabilities. This invention minimizes the impact of data bias on the algorithm, enhancing its flexibility and fusion capabilities.

[0129] In the process of alleviating the bias of the deviation, the present invention aims at the requirement of data positivity for the image evolution of each influencing factor in the disaster remote sensing data. In the multi-stage fusion process, the image evolution of each influencing factor in the disaster remote sensing data is first performed to compress the deviation to an interval for the stage fusion deviation, thereby alleviating the data bias while ensuring the positivity of the data, and further enhancing the nonlinear expression ability of the algorithm. Subsequently, the image evolution of each influencing factor in the disaster remote sensing data is performed to better alleviate the bias of the data. The present invention overcomes the risk of over-fusion by introducing a timely stopping strategy. At the same time, the timely stopping strategy with stage as the granularity better avoids the influence of special values than the traditional timely stopping strategy with the message tree of each influencing factor in the disaster remote sensing data as the granularity, thereby making the algorithm more robust.

[0130] The method of the present invention greatly improves the system performance on the test sets of various influencing factors in disaster remote sensing data in multiple time windows. The algorithm's fusion ability under biased data is greatly enhanced, and at the same time, it shows good robustness, effectively solving the adaptability problem of the algorithm under biased data conditions.

[0131] Example 2, as Figure 2 As shown, an embodiment of the present invention provides a geological disaster analysis system based on remote sensing data fusion, the system comprising:

[0132] The influencing factor data information acquisition module 1 is used to obtain the influencing factor data information in the disaster remote sensing data, and the influencing factor data information includes collapse, landslide, debris flow, ground fissure, soil erosion, land desertification and swamping, soil salinization, as well as earthquake, volcano, and geothermal damage image information;

[0133] The multi-influencing factor fusion module 2 is used to obtain a geological disaster comprehensive analysis model and a geological disaster comprehensive analysis image based on the influencing factor data information in the acquired disaster remote sensing data using a multi-influencing factor fusion method;

[0134] The module 3 for acquiring comprehensive characteristics of the area affected by the disaster is used to acquire comprehensive characteristics of the area affected by the disaster based on the comprehensive analysis image of geological disasters, and to visualize them; the comprehensive characteristics of the area include the surface landform change characteristics of the area affected by the disaster.

[0135] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0136] The information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present invention. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0137] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment.

[0138] An embodiment of the present invention also provides a computer device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor implements the steps of any of the above-mentioned method embodiments when executing the computer program.

[0139] An embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0140] An embodiment of the present invention also provides an information data processing terminal, which is used to provide a user input interface to implement the steps in the above-mentioned method embodiments when executed on an electronic device. The information data processing terminal is not limited to mobile phones, computers, and switches.

[0141] An embodiment of the present invention further provides a server, which is used to provide a user input interface to implement the steps in the above method embodiments when executed on an electronic device.

[0142] An embodiment of the present invention provides a computer program product. When the computer program product is run on an electronic device, the electronic device can implement the steps of the above-mentioned method embodiments when executing the computer program product.

[0143] If the integrated unit is implemented as a software functional unit and sold or used as a standalone product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application can implement all or part of the process steps in the above-mentioned method embodiments by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium can include at least: any entity or device capable of carrying the computer program code to the camera / terminal device, a recording medium, computer memory, read-only memory (ROM), random access memory (RAM), an electrical carrier signal, a telecommunications signal, and a software distribution medium. Examples include a USB flash drive, a removable hard drive, a magnetic disk, or an optical disk.

[0144] The above description is only a preferred specific implementation method of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A geological disaster analysis method based on remote sensing data fusion, characterized in that: The method includes: S1: Acquire the influencing factor data information in the disaster remote sensing data, which includes: collapse, landslide, debris flow, ground fissure, soil erosion, land desertification, swamping, soil salinization, as well as earthquake, volcano, and geothermal damage image information; S2: Based on the acquired influencing factor data information, a multi-influencing factor fusion method is used to obtain a comprehensive analysis model of geological hazards and a comprehensive analysis image of geological hazards; S3, based on the comprehensive analysis image of geological disasters, obtain the comprehensive characteristics of the area affected by the disaster and visualize them; the comprehensive characteristics of the area include the surface landform change characteristics of the area affected by the disaster; Step S2 specifically includes: Step 1: Input the image training sample A of the influencing factors in the disaster remote sensing data and the image testing sample B of the influencing factors in the disaster remote sensing data, which are used for the training of the geological disaster comprehensive analysis model and the evaluation of the geological disaster comprehensive analysis model respectively; Step 2: Initial pixel labeling for the impact factor image training sample A in the disaster remote sensing data Conduct normality tests; Step 3: According to the current geological disaster area Number of message trees for each influencing factor in disaster remote sensing data ,by As fusion pixel markers, establish a message tree of each influencing factor in each disaster remote sensing data , integrating the loss function deviation of the current geological hazard comprehensive analysis model; Step 4: After the fusion of the current geological disaster comprehensive analysis model is completed, calculate the fusion deviation under the current geological disaster section , , is the pixel label of the data to be fused, is the predicted value of the current geological disaster section for training sample A, and the expression is: ； Where, is the learning rate, is the cumulative predicted value of the current geological disaster section for training sample A, The first geological disaster area Message tree of each influencing factor in disaster remote sensing data Predicted values on the training set and update the data values to be fused ; Step 5: Fusion bias of the influencing factor image training sample A in the disaster remote sensing data Conduct normality tests; Step 6: When the loss function of the evaluation function on the impact factor image test sample B in the disaster remote sensing data no longer decreases within the specified number of early evolution stop stages, the training of the geological disaster comprehensive analysis model stops, and the optimal number of iteration stages of the geological disaster comprehensive analysis model is obtained. , complete the training of the geological disaster comprehensive analysis model, jump to step seven, otherwise the loss function on the impact factor image test sample B in the disaster remote sensing data continues to decrease, jump to step four, and continue the stage iteration; Step 7: Predict the test set of each influencing factor in the disaster remote sensing data, and calculate the optimal number of iteration stages of the geological disaster comprehensive analysis model. , and obtain the prediction results of the test set of each influencing factor in the disaster remote sensing data , the expression is: ； Where, For the stage Pixel marking pre-disaster remote sensing data of each influencing factor image evolution and anti-disaster remote sensing data of each influencing factor image evolution, is the impact factor, For the stage Message tree of various influencing factors in disaster remote sensing data prediction results.

2. The geological disaster analysis method based on remote sensing data fusion according to claim 1 is characterized in that: In step 2, the initial pixel marking of the impact factor image training sample A in the disaster remote sensing data is performed. Perform normality tests, including: (1) Calculate the data distribution discrete bias rate based on the pixel labels of the image training samples of each influencing factor in the disaster remote sensing data , the expression is: ； Where, is the pixel label of the data to be fused, that is ; is the sample pixel label mean, is the sample size, Label the standard deviation of the sample pixels; (2) Calculate the proportion of data distribution fusion suitable state according to the pixel labels of image training samples of each influencing factor in disaster remote sensing data , the expression is: ； (3) Image fusion statistics of various influencing factors in tectonic disaster remote sensing data , the expression is: ； test Statistical significance, examine the degree of bias to determine whether to change the data; if If the statistic is greater than the threshold, go to step (4), otherwise go to step (5); (4) The data is obviously biased. Determine whether the upper limit of the number of data changes is reached. Set the upper limit of the number of changes to be , if so, jump to step (5); otherwise, update The value of , perform the data Changes are made by the following formula The value traversal selects the image evolution of each influencing factor in the disaster remote sensing data to make the data pixel mark Likelihood function The largest Generate a variation expression: ； Where, is the disaster remote sensing data transformation amount in the geological disaster area, Take the logarithm of the transformed quantity, To conform to the transformation value of the normal distribution, the value range is 0.00-1.00; Get the new data pixel label and use it as the first stage fusion pixel label ; Set the current geological disaster section judgment variable ; (5) The data is not obviously biased, or the number of image evolutions of each influencing factor in the data pixel mark disaster remote sensing data reaches the target upper limit, and the original data pixel mark is taken As the first stage of fusion pixel labeling , set the current geological disaster section judgment variable .

3. The geological disaster analysis method based on remote sensing data fusion according to claim 1 is characterized in that: In step 3, the loss function deviation of the current geological hazard comprehensive analysis model is integrated, including: Step 1: Collect information about various influencing factors in disaster remote sensing data to the fusion center through the communication network; Step 2: Based on the physical process and current geological hazards, a loss function deviation matrix is comprehensively analyzed to develop an insertion method for artificial influence factors, so that the fusion of virtual geological hazard interference objects and the deviation of the current geological hazard comprehensive analysis loss function become worse; Step 3: The fusion center decodes the received signal, and the probability of successful decoding is related to the impact value of the impact factor; Step 4: Derive the criteria for selection probability and successful decoding probability to obtain the impact factor insertion strategy.

4. The geological disaster analysis method based on remote sensing data fusion according to claim 3 is characterized in that: In step 2, the method for inserting the artificial impact factor includes: For objects without geological hazards, the parameter information of the dynamic model and observation equation is calculated. Based on the knowledge of remote sensing data of geological hazards, a random matrix is constructed according to the identification process of remote sensing data of geological hazards. The variance of artificial influencing factors is obtained by calculation based on the local information of each influencing factor message in each remote sensing data of geological hazards and mathematical operations. For virtual geological disaster interference objects, it is difficult to obtain accurate matrices. Due to the existence of random matrices generated in real time, virtual geological disaster interference objects cannot obtain accurate geological disaster information.

5. The geological disaster analysis method based on remote sensing data fusion according to claim 4 is characterized in that: Get the variance of artificial influence factors, including: In time Selected components within From the information of various influencing factors in disaster remote sensing data Sent to the fusion center, artificial impact factors will be inserted before sending Add to , sending a signal through a communication channel.

6. The geological disaster analysis method based on remote sensing data fusion according to claim 3 is characterized in that: In step 2, a comprehensive analysis of the loss function deviation matrix is conducted to develop an insertion method for the artificial impact factor, including: 2.1, for any initial condition , the fusion mechanism achieves perfect expected fusion when both of the following conditions are met: ； Where, is the deviation of the current geological disaster comprehensive analysis loss function calculated by the virtual geological disaster interference object, are diagonal elements, is the matrix null space; 2.2, Select the impact value of artificial impact factor , using the matrix of the orthogonal basis, we get the matrix null space The orthonormal basis of 2.3, Artificial Impact Factor It is given by: ； Where, The component is the impact factor.

7. The geological disaster analysis method based on remote sensing data fusion according to claim 1 is characterized in that: In step 6, the calculation of the loss function on the impact factor image test sample B in the disaster remote sensing data specifically includes: Calculation ends at the current geological disaster section The predicted value of the geological disaster comprehensive analysis model on the impact factor image test sample B in the disaster remote sensing data, stage Pixel labeling Pre-disaster remote sensing data of various influencing factors in the image evolution Anti-disaster remote sensing data of various influencing factors in the image evolution The expression is: ； Where, Current geological disaster area Judgment variables; for The image evolution of each influencing factor in the disaster remote sensing data is the image evolution of each influencing factor in the disaster remote sensing data, for The image evolution of each influencing factor in the disaster remote sensing data is the image evolution of each influencing factor in the disaster remote sensing data; the expression is: 。 8. The geological disaster analysis method based on remote sensing data fusion according to claim 7 is characterized in that: Calculate the loss function value of the current geological disaster section. The expression is: ； Where, For the initial pixel labels of the test set B of each influencing factor in the disaster remote sensing data, The loss function used by the system is usually a square loss function for regression problems. .

9. A geological disaster analysis system based on remote sensing data fusion, characterized in that: The system is implemented by the geological disaster analysis method based on remote sensing data fusion according to any one of claims 1 to 8, and the system includes: An influencing factor data information acquisition module (1) is used to acquire influencing factor data information in disaster remote sensing data, wherein the influencing factor data information includes collapse, landslide, debris flow, ground fissure, soil erosion, land desertification and swamping, soil salinization, and earthquake, volcano, and geothermal damage image information; A multi-influencing factor fusion module (2) is used to obtain a geological disaster comprehensive analysis model and a geological disaster comprehensive analysis image based on the influencing factor data information in the acquired disaster remote sensing data using a multi-influencing factor fusion method; The module (3) for obtaining comprehensive characteristics of the area affected by the disaster is used to obtain comprehensive characteristics of the area affected by the disaster based on the geological disaster comprehensive analysis image, and to visualize the characteristics; the comprehensive characteristics of the area include surface landform change characteristics of the area affected by the disaster.

Images