Multi-source cooperative remote sensing debris flow full-period monitoring analysis method

Abstract

The invention discloses a multi-source cooperative remote sensing debris flow full-period monitoring analysis method, which comprises the following steps of: 1, acquiring multi-source remote sensing data and land utilization vector data, and establishing a full-period monitoring basic database; 2, preprocessing the collected multi-source remote sensing data and land utilization vector data to form standardized space basic data of a unified space coordinate system; 3, analyzing and processing the standardized spatial basic data to obtain a disaster condition; and step 4, performing emergency evacuation plan planning, disaster damage assessment report making and post-disaster restoration progress monitoring based on the analyzed disaster condition. According to the invention, through four core innovations of standardized process, dynamic cooperative monitoring, low-cost change detection and disaster-bearing body refined analysis, an efficient and extensible debris flow full-cycle technical system is constructed, and the method has relatively high popularization universality.

Description

Technical Field

[0001] The present invention relates to the field of remote sensing monitoring of geological disasters, and in particular to a multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method. Background Art

[0002] In recent years, satellite and drone remote sensing technologies have provided differentiated capabilities for geological disaster monitoring: optical satellites, with their wide coverage, can quickly obtain macroscopic information on disaster conditions; synthetic aperture radar satellites, leveraging the penetrating properties of microwaves, can monitor millimeter-level deformations before disasters occur in rainy and foggy weather; and drone remote sensing, relying on low-altitude aerial photography, generates centimeter-level three-dimensional models, supporting detailed damage assessments. However, existing technologies often rely on single-source data or perform independent analyses in phases, which presents systemic flaws.

[0003] In terms of timeliness, optical satellites are often obscured by clouds and have long revisit cycles, leading to significant lags in emergency response during disasters. Although InSAR technology can monitor minute deformations before a disaster, its algorithm requires adjustment of deformation unwrapping parameters for different regions. Large-scale verification takes a long time, and the technology is expensive. In addition, it is difficult to capture the dramatic deformations during debris flow disasters, and phase incoherence is prone to occur in densely vegetated areas (Phase incoherence refers to the phenomenon in which the phase information of the radar echo signal loses consistency or correlation when observing the same area twice or more), making it unable to meet the needs of real-time dynamic tracking.

[0004] In terms of coverage capability, open source satellite data is delayed in acquisition, and the single operation range of drones is mostly less than 10km2, making it difficult to respond to large-scale disasters in a timely manner.

[0005] At the technical collaboration level, the efficiency of existing methods is limited by the discrete nature of multi-platform data processing processes. When the amount of multi-source data surges, the lack of fixed and convenient algorithms makes it difficult for the processing speed to match the timeliness requirements of emergency response. At the same time, traditional assessments mostly use static overlay analysis, which fails to couple spatial correlation information such as building structure risk levels and transportation network resilience, resulting in insufficient disaster adaptability in evacuation route planning and disaster damage assessment.

[0006] The prior art has the following problems:

[0007] (1) The data processing and analysis links in existing technologies rely too much on manual experience and judgment, and each link lacks standardized processes and unified technical standards, resulting in weak comparability of results and difficulty in cross-regional application.

[0008] (2) Existing single technologies have inherent limitations. For example, optical satellite remote sensing is restricted by cloudy and rainy weather, resulting in unstable effective data acquisition rate; synthetic aperture radar interferometry (InSAR) technology cannot effectively interpret violent deformation areas due to phase decoherence; drone aerial photography is limited by the coverage range of a single operation and cannot meet large-scale collection requirements.

[0009] (3) In general, InSAR technology requires more than 20 periods of radar data to resolve small deformations. Severe deformations after disasters are difficult to monitor due to phase decoherence, and the detection cost is high.

[0010] (4) Existing technologies over-rely on optical remote sensing spectral feature analysis and ignore the spatial attribute correlation of disaster-prone objects such as buildings and roads, resulting in systematic deviations in evacuation route planning and economic loss assessment.

[0011] Therefore, a single technical system is difficult to balance monitoring timeliness, spatial scale and data dimension, which restricts the effectiveness of the full-cycle response to geological disasters. Summary of the Invention

[0012] In view of this, the purpose of the present invention is to propose a multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method. In order to solve the core problems in the field of severe surface deformation geological disaster monitoring, such as insufficient multi-source remote sensing data collaboration ability, intermittent full-cycle response, and insufficient disaster analysis efficiency, the present invention aims to build a set of standardized, adaptive, and scalable "air-space-ground" ("air" refers to UAV remote sensing equipment, "space" refers to optical and radar satellite remote sensing equipment, and "ground" refers to land use vector data). ) An integrated, full-cycle monitoring technology system. Through four core innovations: standardized processes, dynamic collaborative monitoring, low-cost change detection, and refined analysis of hazard-bearing bodies, we have built an efficient and scalable full-cycle debris flow technology system with strong universal applicability.

[0013] In order to achieve the above technical objectives, the technical solution adopted by the present invention is:

[0014] A multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method includes the following steps:

[0015] Step 1: Acquire multi-source remote sensing data and land use vector data to establish a full-cycle monitoring basic database;

[0016] Step 2: Preprocess the collected multi-source remote sensing data and land use vector data to form standardized spatial basic data in a unified spatial coordinate system;

[0017] Step 3: Analyze and process the standardized spatial basic data to obtain the disaster situation;

[0018] Step 4: Based on the analyzed disaster situation, emergency evacuation plan planning, disaster damage assessment report and post-disaster restoration progress monitoring are carried out.

[0019] Furthermore, the step 1 specifically includes:

[0020] Step 11: Obtain the location information and time information of the geological disaster event;

[0021] Step 12: Using the location information and time information to obtain optical satellite remote sensing image data before the geological disaster occurs;

[0022] Step 13: Using the location information and time information, obtain radar satellite remote sensing image data before and after the geological disaster occurs;

[0023] Step 14: Using the location information, obtain land use vector data within the event area and its impact area;

[0024] Step 15: Determine whether the event area conditions and weather conditions meet the requirements for drone aerial photography. If so, proceed to step 16; if not, proceed to step 17;

[0025] Step 16: Supplement the acquisition of drone remote sensing image data after the geological disaster occurs through drone aerial photography, and proceed to step 2;

[0026] Step 17: Arrange the acquired raw data, including optical satellite remote sensing image data, radar satellite remote sensing image data, and land use vector data, and proceed to step 2.

[0027] Furthermore, the processing of the obtained raw data in step 17 specifically includes:

[0028] The optical satellite remote sensing image data includes panchromatic images and multispectral images before the disaster, the radar satellite remote sensing image data includes non-interference single-view images before and after the disaster, and the land use vector data includes building and road vector data;

[0029] Organize and check the pre-disaster panchromatic images, multispectral images and related parameter data in the event area, the pre-disaster and post-disaster non-interference single-view images and related parameter data in the radar satellite remote sensing image data, and the building and road vector data.

[0030] Furthermore, the step 2 specifically includes:

[0031] Step 21: Based on the acquired UAV remote sensing image data, aerial triangulation, DSM generation and DEM extraction, as well as MESH modeling and DOM production processing are performed on the UAV remote sensing image data in sequence, and then step 3 is entered;

[0032] Step 22: determine whether the sorted original data is optical satellite remote sensing image data. If so, perform RPC parameter solution, orthorectification, and image fusion processing on the optical satellite remote sensing image data in sequence, and then proceed to step 24; if not, proceed to step 23;

[0033] Step 23: determine whether the sorted original data is radar satellite remote sensing image data. If so, perform multi-view processing, spatial filtering, geocoding and radiometric correction processing on the radar satellite remote sensing image data in sequence, and then proceed to step 24; if not, proceed to step 25;

[0034] Step 24: Based on the results of processing the optical satellite remote sensing image data, perform RPC parameter generation and spatial registration on the processed radar satellite remote sensing image data, and then proceed to step 3.

[0035] Step 25: Extract geometry and attribute-related information, perform format conversion, and coordinate conversion on the land use vector data, and then proceed to step 3.

[0036] Furthermore, in step 21, aerial triangulation, DSM generation and DEM extraction, as well as MESH modeling and DOM production processing are sequentially performed on the UAV remote sensing image data, specifically including:

[0037] Step 211: Perform aerial triangulation on the UAV remote sensing image data to generate three-dimensional coordinate results;

[0038] Step 212: Evaluate the three-dimensional coordinate results to determine whether the accuracy of the three-dimensional coordinate results meets the requirements for debris flow or landslide geological disaster monitoring. If so, proceed to step 213; if not, add control points and return to step 211.

[0039] Step 213: Create DSM and DEM using the three-dimensional coordinate results.

[0040] Step 214: Use the generated DSM and DEM to create a MESH model and generate a DOM, and then proceed to step 3;

[0041] In step 22, RPC parameter calculation, orthorectification and image fusion processing are sequentially performed on the optical satellite remote sensing image data, specifically including:

[0042] Step 221: performing RPC parameter calculation on the multispectral image and panchromatic image before the disaster in the optical satellite remote sensing image data;

[0043] Step 222, calculate the accuracy of the multispectral image and the panchromatic image after the block adjustment before the disaster, and judge whether the accuracy meets the monitoring requirements. If so, proceed to step 223; if not, return to step 221;

[0044] Step 223: orthorectify the panchromatic image and the multispectral image to eliminate terrain distortion.

[0045] Step 224: perform image fusion on the panchromatic image and the multispectral image to obtain a multi-band orthoimage. The result is used as the base map of the radar satellite remote sensing image data, and then proceed to step 24.

[0046] In step 23, the radar satellite remote sensing image data is sequentially subjected to multi-view processing, spatial filtering, geocoding, and radiometric correction processing, specifically including:

[0047] Step 231: pre-processing the non-interference single-view images before and after the disaster in the radar satellite remote sensing image data, including orbit correction and / or multi-view processing;

[0048] Step 232: spatially filter the non-interferometric monoscopic images before and after the disaster using a spatial filter to remove noise while maintaining the available spatial resolution.

[0049] Step 233: geocoding the pre-disaster and post-disaster non-interference single-view images to convert the radar satellite remote sensing image data into real geographic spatial locations that are easily discernible by visual inspection; and performing radiometric correction on the pre-disaster and post-disaster non-interference single-view images to convert the digital signals into quantitative backscatter coefficients.

[0050] In step 24, based on the result of processing the optical satellite remote sensing image data, RPC parameter generation and spatial registration are sequentially performed on the processed radar satellite remote sensing image data, specifically including:

[0051] Step 241: Using the processed results of the optical satellite remote sensing image data as a basis, overlay the processed results of the radar satellite remote sensing image data to determine whether there is a deviation in the geometric positions of the non-interference single-view images before and after the disaster. If so, proceed to step 242; if not, proceed to step 3;

[0052] Step 242: using professional software for processing satellite remote sensing image data, generating RPC parameter files of non-interferometric single-view images before and after the disaster based on the reference DEM;

[0053] Step 243: Based on the results of processing the optical satellite remote sensing image data, spatial registration is performed on the non-interferometric single-view images before and after the disaster, so that the spatial position of the radar satellite remote sensing image data meets the monitoring requirements, and then the process proceeds to step 3.

[0054] In step 25, the land use vector data is sequentially subjected to geometric and attribute related information extraction, format conversion, and coordinate conversion processing, specifically including:

[0055] Step 251: Select and extract relevant land feature attributes from the disaster information of the land use vector data;

[0056] Step 252: determine whether the format of the land use vector data is incompatible with the professional format of geographic information used. If so, format conversion is required and the process proceeds to step 253; if not, the process proceeds to step 254;

[0057] Step 253: Convert the land use vector data into a usable data format.

[0058] Step 254: determine whether the coordinate system of the land use vector data is inconsistent with the coordinate system of the image data. If so, proceed to step 255; if not, proceed to step 3;

[0059] Step 255: Perform coordinate conversion on the land use vector data to unify the coordinate systems of the land use vector data and the image data, and then proceed to step 3.

[0060] Furthermore, the aerial triangulation includes the layout and measurement of image control points, acquisition of ground reference coordinates, multi-view image acquisition, image internal orientation and distortion correction, relative orientation, regional network adjustment solution of exterior orientation elements, absolute orientation and calculation of three-dimensional coordinates of encrypted points.

[0061] Furthermore, the RPC parameter calculation specifically includes:

[0062] Using professional software for processing optical satellite remote sensing image data, select benchmark images and reference DEMs of appropriate spatial resolution as references for multispectral and panchromatic image correction;

[0063] Control point encryption, single-chip solution, tie point layout, block adjustment and image matching are performed on multispectral and panchromatic images to obtain the corresponding RPC parameter files.

[0064] Furthermore, the step 3 specifically includes:

[0065] Step 31: interpret and vectorize the disaster-affected area based on the MESH model generated by the drone aerial photography and the DOM;

[0066] Step 32: performing normalization processing, regularized difference enhancement, low-pass filtering denoising processing, disaster-affected area threshold extraction, post-classification clustering processing, and disaster-affected area vectorization processing on the spatially registered pre-disaster and post-disaster non-interference single-view images.

[0067] Step 33: Based on the interpretation vectorization results of the disaster-stricken area, the vectorization processing results of the disaster-stricken area, and the coordinate conversion processing results, multi-source data intersection analysis, disaster attribute feature extraction, and three-dimensional disaster situation visualization are performed in sequence.

[0068] Furthermore, the step 31 specifically includes:

[0069] Based on the MESH model and DOM generated by drone aerial photography, the disaster-affected area is extracted through remote sensing interpretation technology, and the disaster-affected area is converted into a vector result; the remote sensing interpretation technology is visual interpretation or automatic interpretation of remote sensing images, and the process proceeds to step 331;

[0070] The step 32 specifically includes:

[0071] Step 321: Based on the spatially registered non-interference single-view images before and after the disaster, the maximum statistical value DN of each pixel in the non-interference single-view images before and after the disaster is calculated. max and minimum statistical value DN min , use the raster calculator to perform normalization, the formula is:

[0072] SAR Nor =(DN-DN min ) / (DN max -DN min )

[0073] Among them, SAR Nor is the normalized image, DN is the pixel value of the original image position;

[0074] Step 322: enhance the radar remote sensing index image NDERI using normalized difference to enhance the difference in change between the non-interference single-view image before and after the disaster. The formula is as follows:

[0075]

[0076] Among them, NDERI is the normalized difference enhanced radar remote sensing index image, SAR post It represents the pixel value of the image position of the radar image obtained after the geological disaster occurs and normalized in the range of [0-1]. pre It represents the pixel value of the image position in the [0-1] interval normalized by the same method of the baseline radar image when no geological disaster occurs. The small constant of 0.00001 added to the numerator is used to avoid the result of the calculation being zero in the result image.

[0077] Step 323: determine whether the noise of the enhanced image affects the threshold extraction. If so, perform low-pass filtering and denoising on the enhanced image and proceed to step 324; if not, directly proceed to step 324;

[0078] Step 324: perform threshold extraction on the disaster-stricken area using image segmentation to form a binary image;

[0079] Step 325: determine whether the noise of the binary image affects the extraction of the disaster area. If so, perform clustering on the binary image to remove the noise and proceed to step 326; if not, directly proceed to step 326;

[0080] Step 326: Convert the raster data of the affected area into vector data, retain the area with a value of 1, and remove the area with a value of 0, and proceed to step 327;

[0081] Step 327: Extract the vector range of the disaster-stricken area. Use geographic information professional software to select and extract the disaster-stricken area, and then proceed to step 331.

[0082] The step 33 specifically includes:

[0083] Step 331: Overlay analysis is performed on the obtained interpretation vectorization processing results of the disaster-stricken area, the extracted vector range of the disaster-stricken area, and the vector data results processed in step 254 or step 255 in professional geographic information software, and the required vector range of the event area is extracted through spatial intersection analysis;

[0084] Step 332: Determine whether disaster-affected area mapping is required based on the urgency of the emergency response. If so, proceed to step 333; if not, proceed to step 334.

[0085] Step 333: Using the optical satellite remote sensing image data before the geological disaster as a base map, overlay the extracted results of the radar satellite remote sensing image data or the extracted results of the drone shaking image data to carry out a schematic mapping of the disaster area, and then proceed to step 4;

[0086] Step 334: Extract the disaster-affected attributes of key concern based on the emergency event characteristics, and determine whether there are MESH results and hardware and software equipment that can display three-dimensional images. If so, proceed to step 335; if not, proceed to step 336;

[0087] Step 335: Utilize multiple 2D and 3D results, combined with software and hardware equipment, to perform visualization, and proceed to step 4;

[0088] Step 336: Determine whether the report needs to be edited. If so, proceed to step 4; if not, proceed to step 337;

[0089] Step 337: Determine whether post-disaster restoration monitoring is needed. If so, proceed to step 4; if not, end the process.

[0090] Furthermore, the step 4 specifically includes:

[0091] Step 41: Combine the schematic mapping of the disaster area and the visualization of the multiple results in step 335 to carry out emergency evacuation plan planning, and return to step 336;

[0092] Step 42: When the result of step 336 is that the report needs to be edited, a report template is prepared and the process proceeds to step 43;

[0093] Step 43: Prepare a disaster damage assessment report, and return to step 337;

[0094] Step 44: When the judgment result of step 337 is to carry out post-disaster restoration monitoring, the restoration status of the disaster-stricken area is analyzed by inverting the NDVI vegetation index using optical satellite remote sensing image data and combining it with the change monitoring technology of image segmentation.

[0095] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art:

[0096] (1) Standardized full-process design to solve the problem of technological fragmentation: This invention realizes the unified process processing of multi-source remote sensing data and ensures the spatial matching accuracy of multi-source data through the design and construction of a standardized data processing chain throughout the entire process, providing a unified technical framework for geological disaster monitoring. Through the spatial reference matching algorithm (constructing RPC parameters), automatic spatial registration of radar, optical and drone data is achieved, the data analysis prerequisite of a unified spatial reference is established, and the vector attributes of land use such as buildings and roads are synchronously integrated to improve the collaborative accuracy and efficiency of multi-source data.

[0097] (2) Dynamic collaborative monitoring of "air-space-ground" to balance efficiency and accuracy: This invention innovatively proposes a dynamic resource scheduling mechanism, using radar satellites to penetrate clouds for wide-area monitoring, and dispatching drones to target key areas with conditions to fill in blind spots and make up for the lack of satellite resolution. This solution breaks through the performance boundaries of a single sensor, significantly improving the effective data acquisition rate under cloudy and rainy conditions, and significantly optimizing the all-weather adaptability and spatial resolution balance of geological disaster emergency monitoring. By establishing a multi-source priority response mechanism, based on pre-disaster optical satellite remote sensing image data, high-resolution (must be 3 meters or better than 3 meters) and high-reentry radar satellites are used to extract the boundaries of the debris flow accumulation area during the disaster through pre-processing and difference enhancement technology. According to weather conditions, drones are used to collect information on the disaster area and model and generate DSM, DOM, and Mesh, forming a collaborative monitoring system of "optical satellite, radar satellite, and drone aerial photography". A closed-loop framework is designed for pre-disaster optical satellite background survey, radar satellite-drone fusion monitoring during the disaster, and post-disaster multi-source quantitative assessment to track the entire process from disaster to repair.

[0098] (3) Innovative application of normalized difference technology to achieve low-cost detection of severe deformation: This paper proposes a Normalized Difference Enhanced Radar Index (NDERI) that can extract the scope of severe surface deformation-type geological disasters using only two phases of radar data. This index is based on the characteristic differences in surface roughness, water content, and geometric structure of the multi-phase synthetic aperture radar (SAR) backscatter coefficient (σ) before and after the occurrence of geological disasters. The normalized difference method is used to further enhance and amplify the difference in the backscatter coefficient (σ) before and after the occurrence of geological disasters, thereby achieving increased sensitivity to changes in surface characteristics caused by geological disasters (such as landslides and debris flows). It can then be used to effectively extract the affected area of ​​debris flow disasters, providing a highly efficient and low-cost change detection solution for geological disaster emergency response. The physical principle is that when disasters such as landslides and debris flows occur on the surface, the broken rock or deposits will cause a significant increase in roughness, enhancing the diffuse scattering of radar waves (σ increases), while smooth surfaces (such as vegetation-covered areas) are dominated by specular reflection (σ is lower); increased soil moisture content will increase the dielectric constant (especially in the C / L band), further increasing the σ value.

[0099] (4) Deep integration of disaster-prone body attributes to improve the scientific nature of emergency decision-making: The present invention proposes a disaster assessment method based on the fusion of multi-source vector data. Based on land use data and census-related databases (such as the national census), house vector data containing building structure type, construction age and ownership properties are extracted, and traffic network data containing road grade and maintenance status are integrated. Through geocoding, spatial superposition with disaster-affected area raster data is achieved to generate a disaster-prone body attribute association matrix; UAV oblique photogrammetry is used to construct a real-life three-dimensional model with a resolution better than 5 cm, and parameters such as building safety level and road interruption status are mapped to three-dimensional grid vertices to form an interactively queryable three-dimensional disaster model to assist in evacuation path decision-making; through a preset evaluation template, the analysis results of multi-source remote sensing inversion data are associated to generate a standardized evaluation report for monitoring the repair progress. BRIEF DESCRIPTION OF THE DRAWINGS

[0100] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0101] Figure 1 It is a schematic diagram of the overall framework of a multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method provided by an embodiment of the present invention.

[0102] Figure 2This is a specific execution flow chart of a multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0103] The present invention will be described in further detail below with reference to the accompanying drawings and examples. It is particularly noted that the following examples are intended only to illustrate the present invention and are not intended to limit the scope of the present invention. Similarly, the following examples are only some embodiments of the present invention and are not intended to be exhaustive. All other embodiments obtained by those of ordinary skill in the art without creative effort are intended to fall within the scope of protection of the present invention.

[0104] See Figure 1 and Figure 2 The present invention provides a multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method, comprising the following steps:

[0105] Step S10, data acquisition: multi-source remote sensing data and land use vector data are acquired through various technical means to establish a full-cycle monitoring basic database; data of different spatial resolutions are selected as needed, the optical satellite remote sensing image data includes pre-disaster panchromatic and multispectral images, the radar satellite remote sensing image data includes pre-disaster and post-disaster non-interference single-view images, and the land use vector data includes building and road vector data;

[0106] Step S20, data processing: Preprocess the collected multi-source remote sensing data and land use vector data, including orthorectification, spatial filtering, aerial triangulation, information extraction, etc., to form standardized spatial basic data with a unified spatial coordinate system; such as DOM (satellite images), SAR (Synthetic Aperture Radar, an advanced radar technology that can obtain high-resolution ground images under various weather conditions) images (before and after the disaster), land use vector data, etc.; if operating conditions permit (weather, personnel, traffic, etc.), drone aerial photography can be used to supplement the collection of disaster conditions in key areas to generate DSM, DOM (aerial photography), MESH, etc.

[0107] Step S30, Data Analysis: Analyze and process standardized spatial basic data to determine the disaster situation. By performing normalization, regularized difference enhancement, spatial filtering, threshold extraction, and other change detection steps on pre- and post-disaster radar satellite remote sensing imagery, the vector extent of the affected area is determined. Land use vector data is then overlaid to further extract and analyze the affected area's attributes. Supplementary drone spatial basic data can be used in geographic information system software for more accurate spatial overlay analysis and three-dimensional disaster visualization.

[0108] Step S40: Decision-making application: Based on the analyzed disaster situation, emergency evacuation plans are planned, damage assessment reports are produced, and post-disaster repair progress is monitored. The analyzed disaster situation is used to support decision-making in emergency evacuation plan planning; report templates are created to quickly generate disaster assessment reports, calculating the affected area, number of houses, road conditions, agricultural facilities, etc.; and post-disaster repair accuracy testing is regularly conducted.

[0109] This method integrates data from radar satellite remote sensing (which monitors the surface through clouds), multispectral satellite remote sensing (high-precision surface coverage), and unmanned aerial vehicle remote sensing (3D modeling and orthophoto generation), combined with land use vector data such as houses and roads in the affected area. This system constructs an integrated "air-space-ground" dynamic monitoring system for geological disasters, enabling closed-loop management of debris flow disasters throughout their entire lifecycle: pre-disaster regional background surveys, tracking of hazard evolution during a disaster, and post-disaster quantitative loss assessment and repair tracking. A computer program stored in a medium performs multi-source data fusion, change detection, and disaster analysis, providing emergency response departments with decision-making support such as damage assessment reports and evacuation route planning, providing timely and high-precision decision-making support for geological disaster emergency response.

[0110] Preferably, the step S10 specifically includes:

[0111] Step S101, obtaining the approximate location of the geological disaster event: obtaining the location information and time information of the debris flow (or landslide) geological disaster event through emergency news or notifications from relevant departments, and then proceeding to step S102;

[0112] Step S102, obtaining optical satellite remote sensing image data before the geological disaster occurs: using the acquired location information and time information to collect optical satellite remote sensing image data before the geological disaster occurs (before the disaster), and proceeding to step S103; paying attention to the size and scale of the event to select satellite remote sensing images of different spatial resolutions that match the scale, such as panchromatic images with spatial resolutions of 0.5, 1 meter, 2 meters, and 10 meters, and multispectral images with corresponding spatial resolutions.

[0113] Step S103, obtain radar satellite remote sensing image data before and after the geological disaster occurs: use the acquired location information and time information to collect radar satellite remote sensing image data before (before the disaster) and after (after the disaster) the geological disaster occurs, and proceed to step S104; note: the spatial resolution should match the scale of the event, and it is recommended that the spatial resolution of the radar satellite remote sensing image data used be 3 meters or better than 3 meters.

[0114] Step S104: Obtaining land use vector data for the event area: Using the location information, obtain land use vector data for the event area and its impact area, and proceed to step S105. The data is vector data that can express the area of ​​buildings, roads, green spaces, and other key attributes, such as the risk level of building structures, the resilience of transportation networks, and other related information.

[0115] Step S105: Determine whether the event area conditions and weather conditions meet the requirements for drone aerial photography: If the optical satellite remote sensing image data, radar satellite remote sensing image data, and land use vector data have been collected, determine whether the event area conditions and weather conditions meet the requirements for drone aerial photography. If so, proceed to step S106; if not, proceed to step S107;

[0116] Step S106, UAV aerial photography supplementary acquisition to obtain high-resolution basic data: Through route design, flight altitude design, aerial film overlap design, aerial photography mode selection, etc., UAV aerial photography of the event area is carried out, and UAV remote sensing image data after the geological disaster incident is obtained through UAV aerial photography supplementary acquisition, that is, raw data that can be used to produce DSM, MESH, and DOM is obtained, and then the process proceeds to step S201;

[0117] Step S107 , sorting the acquired original data: sorting the acquired original data, including optical satellite remote sensing image data, radar satellite remote sensing image data and land use vector data, and proceeding to step S206 .

[0118] The collated raw data specifically includes:

[0119] The optical satellite remote sensing image data includes panchromatic images and multispectral images before the disaster, the radar satellite remote sensing image data includes non-interference single-view images before and after the disaster, and the land use vector data includes building and road vector data;

[0120] Organize and check the pre-disaster panchromatic images, multispectral images and related parameter data in the event area, the pre-disaster and post-disaster non-interference single-view images and related parameter data in the radar satellite remote sensing image data, and the building and road vector data.

[0121] Preferably, the step S20 specifically includes:

[0122] Step S201, aerial triangulation: aerial triangulation is performed on the drone remote sensing image data taken by the drone to generate three-dimensional coordinate results, and then proceed to step S202; wherein, the aerial triangulation includes the layout and measurement of image control points, acquisition of ground reference coordinates, multi-view image acquisition, image internal orientation and distortion correction, relative orientation, regional block adjustment solution of exterior orientation elements, absolute orientation and calculation of three-dimensional coordinates of encryption points.

[0123] Step S202, determining whether the accuracy meets the monitoring requirements: Evaluate the three-dimensional coordinate results to determine whether the accuracy of the three-dimensional coordinate results meets the requirements for debris flow or landslide geological disaster monitoring. If so, proceed to step S203; if not, proceed to step S205;

[0124] Step S203, DSM production and DEM extraction: Use the three-dimensional coordinate results to produce DSM and DEM, and then proceed to step S204;

[0125] Step S204, MESH modeling and DOM production: Use the produced DSM and DEM to make a MESH model and generate a DOM, and then go to step S301;

[0126] Among them, DSM is Digital Surface Model, which represents the elevation information of the surface, including natural and man-made features such as buildings and trees.

[0127] DEM (Digital Elevation Model) is a digital model used to represent terrain elevation information. It has extremely important application value in the fields of geographic information system (GIS), surveying and mapping, environmental science, urban planning, etc.

[0128] MESH is a three-dimensional grid model that constructs the three-dimensional form of objects or scenes through geometric elements such as points, lines, and surfaces, and can provide rich spatial geometric information.

[0129] DOM (Digital Orthophoto Map) is an important geospatial data product. It processes aerial or satellite images, eliminates factors such as terrain undulations and projection distortion in the images, and makes them have the same geometric accuracy and scale as maps.

[0130] Step S205, supplementing control points: when the aerial triangulation accuracy does not meet the monitoring requirements, supplement control points and enter S201;

[0131] Step S206, determining whether it is optical satellite remote sensing image data: determining whether the sorted original data is optical satellite remote sensing image data, if so, proceeding to step S207; if not, proceeding to step S211;

[0132] Step S207, RPC parameter calculation: Perform RPC parameter calculation on the pre-disaster multispectral image and panchromatic image in the optical satellite remote sensing image data, and proceed to step S208; wherein the RPC parameter calculation specifically includes:

[0133] Using professional software for processing optical satellite remote sensing image data, select benchmark images and reference DEMs of appropriate spatial resolution as references for multispectral and panchromatic image correction (note that benchmark images and reference DEMs need to be collected and obtained separately);

[0134] Control point encryption, single-chip solution, tie point layout, block adjustment and image matching are performed on multispectral and panchromatic images to obtain the corresponding RPC parameter files.

[0135] Step S208, judging whether the accuracy meets the monitoring requirements: Calculate the accuracy of the multispectral image and panchromatic image after block adjustment before the disaster, and judge whether the accuracy meets the monitoring requirements. If so, proceed to step S209; if not, proceed to step S207;

[0136] Step S209, orthorectification of multispectral and panchromatic images: orthorectify the panchromatic and multispectral images to eliminate terrain distortion, and then proceed to step S210;

[0137] Step S210, multispectral and panchromatic image fusion: perform image fusion on the panchromatic image and the multispectral image to obtain a multi-band orthoimage, and use the result as the base map of the radar satellite remote sensing image data, and proceed to step S215;

[0138] Step S211: determine whether the sorted original data is radar satellite remote sensing image data. If so, proceed to step S212; if not, proceed to step S218;

[0139] Step S212: pre-processing the non-interference single-view images before and after the disaster in the radar satellite remote sensing image data, including orbit correction and / or multi-view processing, and then proceeding to step S213;

[0140] Step S213, spatial filtering: Select an appropriate spatial filter and perform spatial filtering on the non-interference monoscopic images before and after the disaster. When removing noise, the available spatial resolution must be maintained, and then proceed to step S214;

[0141] Step S214, geocoding and radiation correction: geocode the non-interference single-view images before and after the disaster, and convert the radar satellite remote sensing image data into real geographic spatial positions that are easy to identify visually; and perform radiation correction on the non-interference single-view images before and after the disaster, convert the digital signals into quantitative backscatter coefficients, and enter step S215; Note: The above processing of radar satellite remote sensing image data includes the processing of both pre- and post-disaster images.

[0142] Step S215: Determine the deviation of the geometric position of the radar satellite remote sensing image data: Using the processed result of the optical satellite remote sensing image data as a basis, overlay the processed result of the radar satellite remote sensing image data to determine whether there is a deviation in the geometric position of the non-interference monoscopic image before and after the disaster. If so, proceed to step S216; if not, proceed to step S302;

[0143] Step S216, generating RPC parameters for the two phases of radar satellite remote sensing image data: using professional software for processing satellite remote sensing image data, generating RPC parameter files for non-interference single-view images before and after the disaster based on the reference DEM, and then proceeding to step S217;

[0144] Step S217, spatially registering the two phases of radar satellite remote sensing image data: Based on the results of processing the optical satellite remote sensing image data, spatially register the pre-disaster and post-disaster non-interference single-view images so that the spatial position of the radar satellite remote sensing image data meets the monitoring requirements, and then proceed to step S302;

[0145] Step S218, extracting information from the land use vector data: selecting and extracting relevant land feature attributes from the disaster information of the land use vector data, and proceeding to step S219;

[0146] Step S219: determine whether the format of the land use vector data is incompatible with the professional format of geographic information used. If so, format conversion is required and the process proceeds to step S220; if not, the process proceeds to step S221;

[0147] Step S220, format conversion processing: convert the land use vector data into a usable data format, and then proceed to step S221;

[0148] Step S221: determine whether the coordinate system of the land use vector data is inconsistent with the coordinate system of the image data. If so, proceed to step S222; if not, proceed to step S311;

[0149] Step S222: Perform coordinate conversion processing on the land use vector data to unify the coordinate systems of the land use vector data and the image data, and then proceed to step S311.

[0150] Preferably, the step S30 specifically includes:

[0151] Step S301: Interpret and vectorize the disaster-affected area based on the MESH model and DOM generated by drone aerial photography: Using the MESH model and DOM generated by drone aerial photography as a basis, remote sensing interpretation technology is used to extract high-precision disaster damage and convert the disaster-affected area into vector results, and then proceed to step S311; wherein the remote sensing interpretation technology is visual interpretation or automatic interpretation of remote sensing images;

[0152] Step S302: normalize the two phases of radar satellite remote sensing image data: based on the spatially registered pre-disaster and post-disaster non-interference single-view images, calculate the maximum statistical value DN of each pixel of the pre-disaster and post-disaster non-interference single-view images. max and minimum statistical value DN min , perform normalization processing using the grid calculator and proceed to step S303; the normalization formula is:

[0153] SAR Nor =(DN-DN min ) / (DN max -DN min )

[0154] Among them, SAR Nor is the normalized image, DN is the pixel value at the (x, y) position in the original image;

[0155] Step S303: Perform normalized difference enhancement on the two phases of radar satellite remote sensing image data: use the Normalized Difference Enhanced Radar Index (NDERI) to enhance the difference in the non-interference single-view image before and after the disaster. The formula is as follows:

[0156]

[0157] Among them, NDERI is the normalized difference enhanced radar remote sensing index image, SAR post It represents the pixel value of the image (x, y) position of the radar image obtained after the geological disaster occurs and normalized in the range of [0-1]. pre represents the pixel value of the image (x, y) position normalized in the interval [0-1] of the baseline radar image processed by the same method when no geological disaster occurs. The small constant of 0.00001 added to the numerator is used to avoid the result image from having a zero calculation result and to ensure the order of magnitude stability of the calculation result; proceed to S304;

[0158] Step S304: determine whether the enhanced image noise affects the threshold extraction. If so, proceed to step S305; if not, proceed to step S306;

[0159] Step S305: Low-pass filtering is performed on the enhanced image to remove noise. A filter such as a mean filter or a median filter can be selected to smooth and reduce noise. (Low-pass filtering is a technique that achieves image smoothing or noise removal by suppressing high-frequency signals and retaining low-frequency signals. Its core goal is to weaken rapidly changing details in the image (such as edges, noise, and texture) while retaining slowly changing macro features (such as overall regional brightness and large-scale object outlines). Note: The filtering process should reduce the impact on the differentiation of the affected areas in the enhanced image, and then proceed to step S306.

[0160] Step S306: Use image segmentation to perform threshold extraction on the affected area to form a binary image. The principle of threshold extraction is "if i1>threshold then 1else 0", where i1 is the image to be extracted and threshold is the threshold that can be used for extraction. The threshold needs to be obtained through experience or continuous testing, and then proceed to step S307.

[0161] Step S307: determine whether the noise of the binary image affects the extraction of the disaster area. If so, proceed to step S308; if not, proceed to step S309;

[0162] Step S308: cluster the binary image to remove noise, and then proceed to step S309;

[0163] Step S309, raster spot vectorization: convert the raster data of the disaster area into vector data, retain the area with a value of 1, and eliminate the area with a value of 0, and proceed to step S310;

[0164] Step S310: Extract the vector range of the disaster-stricken area, use geographic information professional software to select and extract the disaster-stricken area, and then proceed to step S311;

[0165] Step S311, multi-source data intersection analysis: The vectorized interpretation results of the disaster area obtained in step S301 (if any), the vector range of the disaster area extracted in step S310, and the vector data results processed in step S221 or step S222 are overlapped and analyzed in professional geographic information software. The required vector range of the event area is extracted through spatial intersection analysis, and the process proceeds to step S312;

[0166] Step S312: Based on the urgency of the emergency response, determine whether it is necessary to map the affected area and obtain first-hand disaster information. If so, proceed to step S313; if not, proceed to step S314;

[0167] Step S313: Using the optical satellite remote sensing image data before the geological disaster as a base map, the extracted results of the radar satellite remote sensing image data or the extracted results of the drone shaking image data are superimposed to carry out schematic mapping of the disaster area, and then proceed to step S401;

[0168] Step S314, extraction of disaster attribute features: Extracting disaster attributes of key concern based on emergency event features, and proceeding to step S315;

[0169] Step S315: Determine whether there is a MESH result and whether there are hardware and software devices capable of displaying 3D images. If so, proceed to step S316; if not, proceed to step S317.

[0170] Step S316, 3D disaster visualization display: Utilize multiple 2D and 3D results, combined with software and hardware equipment, to perform visualization display, and proceed to step S401;

[0171] Step S317: Determine whether the report needs to be edited. If so, proceed to step S402; if not, proceed to step S318;

[0172] Step S318: Determine whether post-disaster restoration monitoring is required. If so, proceed to step S404; if not, end the process.

[0173] Preferably, the step S40 specifically includes:

[0174] Step S401, emergency evacuation plan planning: combining the schematic mapping of the disaster area and the visualization of multiple results in step S316 to carry out emergency evacuation plan planning, and then proceed to step S317;

[0175] Step S402: Prepare a report template and proceed to step S403;

[0176] Step S403: Prepare a disaster damage assessment report and proceed to step S318;

[0177] Step S404, post-disaster restoration monitoring: Utilize optical satellite remote sensing image data to invert the NDVI vegetation index, combine it with image segmentation change monitoring technology, analyze the restoration status of the disaster-stricken area, and end the process.

[0178] The above descriptions are only some embodiments of the present invention and do not limit the scope of protection of the present invention. Any equivalent device or equivalent process transformation made by using the contents of the description and drawings of the present invention, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method, characterized in that: The steps include: Step 1: Acquire multi-source remote sensing data and land use vector data to establish a full-cycle monitoring basic database; Step 2: Preprocess the collected multi-source remote sensing data and land use vector data to form standardized spatial basic data in a unified spatial coordinate system; Step 3: Analyze and process the standardized spatial basic data to obtain the disaster situation; Step 4: Based on the analyzed disaster situation, emergency evacuation plan planning, disaster damage assessment report and post-disaster restoration progress monitoring are carried out.

2. The multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method according to claim 1, characterized in that: The step 1 specifically includes: Step 11: Obtain the location information and time information of the geological disaster event; Step 12: Using the location information and time information to obtain optical satellite remote sensing image data before the geological disaster occurs; Step 13: Using the location information and time information, obtain radar satellite remote sensing image data before and after the geological disaster occurs; Step 14: Using the location information, obtain land use vector data within the event area and its impact area; Step 15: Determine whether the event area conditions and weather conditions meet the requirements for drone aerial photography. If so, proceed to step 16; if not, proceed to step 17; Step 16: Supplement the acquisition of drone remote sensing image data after the geological disaster occurs through drone aerial photography, and proceed to step 2; Step 17: Arrange the acquired raw data, including optical satellite remote sensing image data, radar satellite remote sensing image data, and land use vector data, and proceed to step 2.

3. The multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method according to claim 2, characterized in that: The processing of the obtained raw data in step 17 specifically includes: The optical satellite remote sensing image data includes panchromatic images and multispectral images before the disaster, the radar satellite remote sensing image data includes non-interference single-view images before and after the disaster, and the land use vector data includes building and road vector data; Organize and check the pre-disaster panchromatic images, multispectral images and related parameter data in the event area, the pre-disaster and post-disaster non-interference single-view images and related parameter data in the radar satellite remote sensing image data, and the building and road vector data.

4. The multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method according to claim 3, characterized in that: The step 2 specifically includes: Step 21: Based on the acquired UAV remote sensing image data, aerial triangulation, DSM generation and DEM extraction, as well as MESH modeling and DOM production processing are performed on the UAV remote sensing image data in sequence, and then step 3 is entered; Step 22: determine whether the sorted original data is optical satellite remote sensing image data. If so, perform RPC parameter solution, orthorectification, and image fusion processing on the optical satellite remote sensing image data in sequence, and then proceed to step 24; if not, proceed to step 23; Step 23: determine whether the sorted original data is radar satellite remote sensing image data. If so, perform multi-view processing, spatial filtering, geocoding and radiometric correction processing on the radar satellite remote sensing image data in sequence, and then proceed to step 24; if not, proceed to step 25; Step 24: Based on the results of processing the optical satellite remote sensing image data, perform RPC parameter generation and spatial registration on the processed radar satellite remote sensing image data, and then proceed to step 3. Step 25: Extract geometry and attribute-related information, perform format conversion, and coordinate conversion on the land use vector data, and then proceed to step 3.

5. The multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method according to claim 4, characterized in that: In step 21, aerial triangulation, DSM generation and DEM extraction, MESH modeling and DOM production processing are sequentially performed on the UAV remote sensing image data, specifically including: Step 211: Perform aerial triangulation on the UAV remote sensing image data to generate three-dimensional coordinate results; Step 212: Evaluate the three-dimensional coordinate results to determine whether the accuracy of the three-dimensional coordinate results meets the requirements for debris flow or landslide geological disaster monitoring. If so, proceed to step 213; if not, add control points and return to step 211. Step 213: Create DSM and DEM using the three-dimensional coordinate results. Step 214: Use the generated DSM and DEM to create a MESH model and generate a DOM, and then proceed to step 3; In step 22, RPC parameter calculation, orthorectification and image fusion processing are sequentially performed on the optical satellite remote sensing image data, specifically including: Step 221: performing RPC parameter calculation on the multispectral image and panchromatic image before the disaster in the optical satellite remote sensing image data; Step 222, calculate the accuracy of the multispectral image and the panchromatic image after the block adjustment before the disaster, and judge whether the accuracy meets the monitoring requirements. If so, proceed to step 223; if not, return to step 221; Step 223: orthorectify the panchromatic image and the multispectral image to eliminate terrain distortion. Step 224: perform image fusion on the panchromatic image and the multispectral image to obtain a multi-band orthoimage. The result is used as the base map of the radar satellite remote sensing image data, and then proceed to step 24. In step 23, the radar satellite remote sensing image data is sequentially subjected to multi-view processing, spatial filtering, geocoding, and radiometric correction processing, specifically including: Step 231: pre-processing the non-interference single-view images before and after the disaster in the radar satellite remote sensing image data, including orbit correction and / or multi-view processing; Step 232: spatially filter the non-interferometric monoscopic images before and after the disaster using a spatial filter to remove noise while maintaining the available spatial resolution. Step 233: geocoding the pre-disaster and post-disaster non-interference single-view images to convert the radar satellite remote sensing image data into real geographic spatial locations that are easily discernible by visual inspection; and performing radiometric correction on the pre-disaster and post-disaster non-interference single-view images to convert the digital signals into quantitative backscatter coefficients. In step 24, based on the result of processing the optical satellite remote sensing image data, RPC parameter generation and spatial registration are sequentially performed on the processed radar satellite remote sensing image data, specifically including: Step 241: Using the processed results of the optical satellite remote sensing image data as a basis, overlay the processed results of the radar satellite remote sensing image data to determine whether there is a deviation in the geometric positions of the non-interference single-view images before and after the disaster. If so, proceed to step 242; if not, proceed to step 3; Step 242: using professional software for processing satellite remote sensing image data, generating RPC parameter files of non-interferometric single-view images before and after the disaster based on the reference DEM; Step 243: Based on the results of processing the optical satellite remote sensing image data, spatial registration is performed on the non-interferometric single-view images before and after the disaster, so that the spatial position of the radar satellite remote sensing image data meets the monitoring requirements, and then the process proceeds to step 3. In step 25, the land use vector data is sequentially subjected to geometric and attribute related information extraction, format conversion, and coordinate conversion processing, specifically including: Step 251: Select and extract relevant land feature attributes from the disaster information of the land use vector data; Step 252: determine whether the format of the land use vector data is incompatible with the professional format of geographic information used. If so, format conversion is required and the process proceeds to step 253; if not, the process proceeds to step 254; Step 253: Convert the land use vector data into a usable data format. Step 254: determine whether the coordinate system of the land use vector data is inconsistent with the coordinate system of the image data. If so, proceed to step 255; if not, proceed to step 3; Step 255: Perform coordinate conversion on the land use vector data to unify the coordinate systems of the land use vector data and the image data, and then proceed to step 3.

6. The multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method according to claim 4, characterized in that: The aerial triangulation includes the layout and measurement of image control points, acquisition of ground reference coordinates, multi-view image acquisition, image internal orientation and distortion correction, relative orientation, regional block adjustment solution of exterior orientation elements, absolute orientation and calculation of three-dimensional coordinates of encrypted points.

7. The multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method according to claim 4, characterized in that: The RPC parameter calculation specifically includes: Using professional software for processing optical satellite remote sensing image data, select benchmark images and reference DEMs of appropriate spatial resolution as references for multispectral and panchromatic image correction; Control point encryption, single-chip solution, tie point layout, block adjustment and image matching are performed on multispectral and panchromatic images to obtain the corresponding RPC parameter files.

8. The multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method according to claim 5, characterized in that: The step 3 specifically includes: Step 31: interpret and vectorize the disaster-affected area based on the MESH model generated by the drone aerial photography and the DOM; Step 32: performing normalization processing, regularized difference enhancement, low-pass filtering denoising processing, disaster-affected area threshold extraction, post-classification clustering processing, and disaster-affected area vectorization processing on the spatially registered pre-disaster and post-disaster non-interference single-view images. Step 33: Based on the interpretation vectorization results of the disaster-stricken area, the vectorization processing results of the disaster-stricken area, and the coordinate conversion processing results, multi-source data intersection analysis, disaster attribute feature extraction, and three-dimensional disaster situation visualization are performed in sequence.

9. The multi-source collaborative remote sensing debris flow full-cycle monitoring and analysis method according to claim 8, characterized in that: The step 31 specifically includes: Based on the MESH model and DOM generated by drone aerial photography, the disaster-affected area is extracted through remote sensing interpretation technology, and the disaster-affected area is converted into a vector result; the remote sensing interpretation technology is visual interpretation or automatic interpretation of remote sensing images, and the process proceeds to step 331; The step 32 specifically includes: Step 321: Based on the spatially registered non-interference single-view images before and after the disaster, the maximum statistical value DN of each pixel in the non-interference single-view images before and after the disaster is calculated. max and minimum statistical value DN min , use the raster calculator to perform normalization, the formula is: SAR Nor =(DN-DN min ) / (DN max -DN min ) Among them, SAR Nor is the normalized image, DN is the pixel value of the original image position; Step 322: enhance the radar remote sensing index image NDERI using normalized difference to enhance the difference in change between the non-interference single-view image before and after the disaster. The formula is as follows: Among them, NDERI is the normalized difference enhanced radar remote sensing index image, SAR post It represents the pixel value of the image position of the radar image obtained after the geological disaster occurs and normalized in the range of [0-1]. pre It represents the pixel value of the image position in the [0-1] interval normalized by the same method of the baseline radar image when no geological disaster occurs. The small constant of 0.00001 added to the numerator is used to avoid the result of the calculation being zero in the result image. Step 323: determine whether the noise of the enhanced image affects the threshold extraction. If so, perform low-pass filtering and denoising on the enhanced image and proceed to step 324; if not, directly proceed to step 324; Step 324: perform threshold extraction on the disaster-stricken area using image segmentation to form a binary image; Step 325: determine whether the noise of the binary image affects the extraction of the disaster area. If so, perform clustering on the binary image to remove the noise and proceed to step 326; if not, directly proceed to step 326; Step 326: Convert the raster data of the affected area into vector data, retain the area with a value of 1, and remove the area with a value of 0, and proceed to step 327; Step 327: Extract the vector range of the disaster-stricken area. Use geographic information professional software to select and extract the disaster-stricken area, and then proceed to step 331. The step 33 specifically includes: Step 331: Overlay analysis is performed on the obtained interpretation vectorization processing results of the disaster-stricken area, the extracted vector range of the disaster-stricken area, and the vector data results processed in step 254 or step 255 in professional geographic information software, and the required vector range of the event area is extracted through spatial intersection analysis; Step 332: Determine whether disaster-affected area mapping is required based on the urgency of the emergency response. If so, proceed to step 333; if not, proceed to step 334. Step 333: Using the optical satellite remote sensing image data before the geological disaster as a base map, overlay the extracted results of the radar satellite remote sensing image data or the extracted results of the drone shaking image data to carry out a schematic mapping of the disaster area, and then proceed to step 4; Step 334: Extract the disaster-affected attributes of key concern based on the emergency event characteristics, and determine whether there are MESH results and hardware and software equipment that can display three-dimensional images. If so, proceed to step 335; if not, proceed to step 336; Step 335: Utilize multiple 2D and 3D results, combined with software and hardware equipment, to perform visualization, and proceed to step 4; Step 336: Determine whether the report needs to be edited. If so, proceed to step 4; if not, proceed to step 337; Step 337: Determine whether post-disaster restoration monitoring is needed. If so, proceed to step 4; if not, end the process.

10. The multi-source coordinated remote sensing debris flow full-cycle monitoring and analysis method according to claim 9, characterized in that: The step 4 specifically includes: Step 41: Combine the schematic mapping of the disaster area and the visualization of the multiple results in step 335 to carry out emergency evacuation plan planning, and return to step 336; Step 42: When the result of step 336 is that the report needs to be edited, a report template is prepared and the process proceeds to step 43; Step 43: Prepare a disaster damage assessment report, and return to step 337; Step 44: When the judgment result of step 337 is to carry out post-disaster restoration monitoring, the restoration status of the disaster-stricken area is analyzed by inverting the NDVI vegetation index using optical satellite remote sensing image data and combining it with the change monitoring technology of image segmentation.

Images