Flood disaster identification method, device, equipment, medium and product

Abstract

The invention provides a flood disaster identification method and device, equipment, a medium and a product, and the method comprises the steps: obtaining multispectral remote sensing data of a first region before a disaster occurs and radar remote sensing data of the first region when the disaster occurs, the second area is a final danger range determined according to a disaster real-time danger range; determining a first water body range before a disaster occurs according to the multispectral remote sensing data of the first area; determining a second water body range when a disaster occurs according to the radar remote sensing data of the first area; and determining a flood disaster range according to the first water body range and the second water body range. The second water body range is extracted through the radar remote sensing data when the disaster occurs, weather interference in the disaster process can be reduced, the water body recognition accuracy is ensured, and therefore the flood disaster occurrence range can be recognized more accurately and rapidly by comparing the second water body range with the pre-disaster water body range recognized through the multispectral remote sensing data.

Description

Technical Field

[0001] The present application belongs to the field of image recognition technology, and specifically, relates to a flood disaster identification method, device, equipment, medium and product. Background Art

[0002] Climate change has led to an increase in the frequency and severity of flood disasters, which has posed a major threat to people's lives and property safety. From the characteristics of flood disasters, floods occur quickly and affect a wide area. Therefore, building a method for rapid response and accurate identification of the scope of flood disasters is of great significance for disaster assessment, emergency command and rescue, and post-disaster reconstruction analysis.

[0003] From the current research on flood identification, most methods use satellite remote sensing images to extract the water area before and after the flood disaster, and then identify the scope of the flood. According to the imaging principle, satellite remote sensing images are mainly divided into optical remote sensing images (multispectral remote sensing images) and radar remote sensing images. At present, most of them extract the water area before and after the flood from optical remote sensing images and then identify the scope of the flood disaster. Some combine deep learning with satellite remote sensing images to identify the water area.

[0004] However, the above method has the following problems in water body extraction: if optical remote sensing images are used to extract the scope of water bodies after floods occur, since extremely bad weather will occur in most cities when floods occur, optical remote sensing images are easily affected by cloud and fog obstruction and continuous heavy rainfall, resulting in large errors in water body identification and extraction using the first optical remote sensing image data, and poor extraction results. If a method combining deep learning with satellite remote sensing images is used to identify and extract water bodies, and then monitor the scope of flood disasters, although such methods can guarantee a high accuracy rate, the computational complexity is generally high and cannot meet the high timeliness requirements of flood area monitoring. Therefore, the recognition methods in the prior art cannot take into account both recognition accuracy and timeliness at the same time. Summary of the invention

[0005] In view of the above problems, the present application aims to provide a flood disaster identification method, device, equipment, medium and product, which are used to solve the problem that existing flood disaster identification methods cannot take into account both identification accuracy and timeliness.

[0006] In a first aspect, the present application provides a flood disaster identification method, comprising:

[0007] Acquire multispectral remote sensing data of a first area before a disaster occurs and radar remote sensing data of the first area when a disaster occurs, wherein the first area is larger than a second area, and the second area is a final endangered area determined according to the real-time endangered area of ​​the disaster;

[0008] Determine the scope of the first water body before the disaster occurs based on the multispectral remote sensing data of the first area;

[0009] Determine the scope of the second water body when the disaster occurs based on the radar remote sensing data of the first area;

[0010] Determine the scope of flood disaster based on the scope of the first water body and the second water body.

[0011] In a possible embodiment, determining the first water body range before the disaster occurs according to the multispectral remote sensing data of the first area includes:

[0012] Preprocessing the multispectral remote sensing data of the first area to obtain first image data including at least the real-time endangered range of the disaster;

[0013] A first water body range is extracted according to the first image data.

[0014] In a possible embodiment, the preprocessing of the multispectral remote sensing data of the first area to obtain first image data including at least the real-time endangered range of the disaster specifically includes:

[0015] Convert the multispectral remote sensing data of the first area into Geotiff data of different bands;

[0016] Merge Geotiff data of different bands into multi-band TIF files;

[0017] The multi-band TIF file is cropped to obtain the first image data including at least the real-time endangered scope of the disaster.

[0018] In a possible embodiment, extracting the first water body range according to the first image data includes:

[0019] Normalizing the first image data to obtain normalized first image data;

[0020] Determine a first threshold range for distinguishing between water bodies and non-water bodies according to the normalized first image data;

[0021] The first water body range is extracted from the normalized first image data according to the first threshold range.

[0022] In a possible embodiment, determining the range of the second water body when the disaster occurs according to the radar remote sensing data of the first area includes:

[0023] determining a bipolar index based on a backscatter coefficient of radar remote sensing data of the first area;

[0024] Determine a second threshold range according to histogram statistics of the bipolar exponential distribution;

[0025] The second water body range when the disaster occurs is extracted from the radar remote sensing data according to the second threshold range.

[0026] In a possible embodiment, determining the bipolar index according to the backscatter coefficient of the radar remote sensing data of the first area includes:

[0027] Clipping the radar remote sensing data of the first area to obtain second image data at least including the real-time endangered range of the disaster;

[0028] The bipolar index is calculated based on the VV and VH polarization images of the second image data.

[0029] In a possible embodiment, extracting the second water body range when the disaster occurs from the radar remote sensing data according to the second threshold range includes:

[0030] According to the second threshold range, water body information that meets the preset confidence level is proposed as a positive sample, and non-water body information is proposed as a negative sample, and the historical radar remote sensing data is trained by the positive samples and the negative samples to obtain a water body extraction model;

[0031] The second water body range when the disaster occurs is extracted from the second image data through a water body extraction model.

[0032] In a possible embodiment, after determining the flood disaster range according to the first water body range and the second water body range, the method further includes:

[0033] Determine the type of affected areas and the level of damage according to the scope of the flood disaster.

[0034] In a second aspect, the present application also provides a flood disaster identification device, comprising:

[0035] A remote sensing data acquisition module, which acquires multispectral remote sensing data of a first area before a disaster occurs and radar remote sensing data of the first area when a disaster occurs, wherein the first area is larger than a second area, and the second area is a final endangered area determined according to the real-time endangered area of ​​the disaster;

[0036] A first water body range determination module, which determines the first water body range before the disaster occurs based on the multispectral remote sensing data of the first area;

[0037] A second water body range determination module, which determines the range of the second water body when the disaster occurs based on the radar remote sensing data of the first area;

[0038] The flood disaster scope determination module determines the flood disaster scope according to the first water body scope and the second water body scope.

[0039] In a third aspect, the present application further provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor;

[0040] The memory stores computer-executable instructions;

[0041] The processor executes the computer-executable instructions stored in the memory to implement a flood disaster identification method as described in any one of the first aspects above.

[0042] In a fourth aspect, the present application further provides a computer storage medium, wherein the computer-readable storage medium stores computer execution instructions, and when the computer execution instructions are executed by a processor, they are used to implement a flood disaster identification method as described in any one of the first aspects above.

[0043] In a fifth aspect, the present application further provides a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, the method described in any one of claims 1 to 7 is implemented.

[0044] The present application provides a flood disaster identification method, device, equipment, medium and product, the method comprising: obtaining multispectral remote sensing data of a first area before the disaster occurs and radar remote sensing data of the first area when the disaster occurs, the first area is larger than the second area, and the second area is the final endangered range determined according to the real-time endangered range of the disaster; determining the first water body range before the disaster occurs according to the multispectral remote sensing data of the first area; determining the second water body range when the disaster occurs according to the radar remote sensing data of the first area; determining the flood disaster range according to the first water body range and the second water body range. Since the water body range when the disaster occurs is extracted through radar remote sensing data, the accuracy of water body identification is not affected even when there is weather interference when the disaster occurs. Combining it with the water body range before the disaster identified by multispectral remote sensing data can more accurately and quickly identify the flood disaster range, and provide scientific, precise and fast guidance information for disaster rescue, emergency disposal and post-disaster emergency material needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0046] Figure 1 An application scenario diagram of a flood disaster identification method provided in this application;

[0047] Figure 2 A flow chart of a flood disaster identification method provided in an embodiment of the present application;

[0048] Figure 3 A flow chart of another flood disaster identification method provided in an embodiment of the present application;

[0049] Figure 4 A flow chart of a third flood disaster identification method provided in an embodiment of the present application;

[0050] Figure 5 is the statistical histogram of the bipolar index;

[0051] Figure 6 A schematic diagram of the structure of a flood disaster identification device provided in an embodiment of the present application;

[0052] Figure 7 A hardware structure diagram of an electronic device provided in an embodiment of the present application.

[0053] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments.

[0054] Description of reference numerals:

[0055] 101-remote sensing data acquisition device; 102-flood detection device; 103-processing end; 601-remote sensing data acquisition module; 602-first water body range determination module; 603-second water body range determination module; 604-flood disaster range determination module; 701-processor; 702-memory; 703-communication interface; 704-communication bus. DETAILED DESCRIPTION

[0056] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0057] In the embodiments of the present application, words such as "first" and "second" are used to distinguish between identical or similar items with substantially the same functions and effects. Those skilled in the art will understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit the difference. It should be noted that in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" or "for example" in this application should not be interpreted as being more preferred or more advantageous than other embodiments or design. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way. In the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more.

[0058] It should be noted that the "at..." in the embodiment of the present application can be the instant when a certain situation occurs, or can be a period of time after the situation occurs, and the embodiment of the present application does not specifically limit this. In addition, the flood disaster identification method provided in the embodiment of the present application is only used as an example, and the flood disaster identification method can also include more or less content.

[0059] In order to clearly describe the technical solutions of the embodiments of the present application, some terms and technologies involved in the embodiments of the present application are briefly introduced below:

[0060] Multispectral remote sensing data: refers to the acquisition of multiple single bands of ground radiation. The first image data obtained will contain spectral information of multiple bands. If the information of the three RGB bands is displayed, it is an RGB color image.

[0061] Radar remote sensing data: It is a remote sensing image formed by emitting radar pulses to obtain backscattered signals from ground objects. The imaging principle of radar remote sensing data makes the imaging quality of radar images highly resistant to interference.

[0062] Otsu method: also known as the maximum inter-class variance method, its basic principle is to assume that the detection image consists of two parts, the foreground and the background, and to find a suitable threshold through statistical methods so that the foreground and the background can be separated to the greatest extent. Otsu method is one of the common water body extraction methods. Its core is that the statistical values ​​of the target and the background form a bimodal shape, which requires that the statistical target and background areas are as close as possible.

[0063] Random forest method: Random forest algorithm is an algorithm that integrates multiple trees through the idea of ​​ensemble learning. Its basic unit is the decision tree, and its essence belongs to a major branch of machine learning - ensemble learning method. From an intuitive point of view, each decision tree is a classifier (assuming it is a classification problem), then for an input sample, N trees will have N classification results. Random forest integrates all classification voting results and designates the category with the most votes as the final output.

[0064] Climate change has led to an increase in the frequency and severity of flood disasters, which has posed a major threat to people's lives and property safety. From the characteristics of flood disasters, floods occur quickly and affect a wide area. Therefore, building a method for rapid response and accurate identification of the scope of flood disasters is of great significance for disaster assessment, emergency command and rescue, and post-disaster reconstruction analysis.

[0065] From the current research on flood identification, most methods use satellite remote sensing images to extract the water area before and after the flood disaster, and then identify the scope of the flood. According to the imaging principle, satellite remote sensing images are mainly divided into optical remote sensing images (multispectral remote sensing images) and radar remote sensing images. At present, most of them extract the water area before and after the flood from optical remote sensing images and then identify the scope of the flood disaster. Some combine deep learning with satellite remote sensing images to identify the water area.

[0066] However, the above methods have the following problems in water body extraction: if optical remote sensing images are used to extract the scope of water bodies after floods, since extreme weather is likely to occur when floods occur, optical remote sensing images are easily affected by cloud and fog and continuous heavy rainfall, resulting in large errors in water body identification and extraction using the first optical remote sensing image data, and poor extraction results. If deep learning combined with satellite remote sensing images is used to identify and extract water bodies, and then the scope of floods is monitored, although such methods can guarantee a high accuracy rate, the computational complexity is generally high and cannot meet the high timeliness requirements of flood area monitoring.

[0067] In order to solve the above problems, the present application provides a flood disaster identification method, device, equipment, medium and product. The flood disaster identification method uses two types of satellite remote sensing data (multispectral remote sensing data and radar remote sensing data) at the same time and combines corresponding identification methods to more accurately and efficiently identify the scope of flood disasters.

[0068] Figure 1The figure shows a specific application scenario of a flood disaster identification method provided by the present application. In this scenario, the remote sensing data acquisition device 101 is used to collect remote sensing data of the area before and after the disaster occurs, and the flood detection device 102 is used to determine whether a flood occurs in the area. When the flood detection device 102 detects the occurrence of a flood, it sends an early warning signal and the location information of the disaster site to the processing end 103. After the processing end 103 receives the early warning signal and the location information of the disaster site, it obtains the remote sensing data of the corresponding area before and during the disaster from the remote sensing data acquisition device 101, thereby determining the scope of the flood disaster. During the process of the disaster, the processing end 103 obtains the latest remote sensing data in real time, thereby updating the scope of the flood disaster in real time, and providing scientific guidance information for disaster rescue, emergency disposal, post-disaster emergency supplies and other demand assessments.

[0069] In the present application, the remote sensing data acquisition device 101 can collect remote sensing data through remote sensing satellites, aircraft and drones. The remote sensing satellites can be open source and free, or commercial.

[0070] The flood detection device 102 can comprehensively determine whether to issue a warning message based on multiple factors such as meteorological conditions, geographical location, warning information and water conditions. The processing end 103 outputs the flood disaster scope in real time after receiving the warning information.

[0071] The processing end 103 may be a mobile terminal such as a smart phone or a tablet computer, or may be a device such as a desktop computer or a server.

[0072] Figure 2 A flow chart of a flood disaster identification method provided in an embodiment of the present application, which is applied to Figure 1 The processing end 103 in FIG. Figure 2 As shown, the flood disaster identification method includes:

[0073] S201: Acquire multispectral remote sensing data of a first area before a disaster occurs and radar remote sensing data of the first area when a disaster occurs, wherein the first area is larger than the second area, and the second area is a final endangered area determined according to the real-time endangered area of ​​the disaster;

[0074] When a disaster occurs, the first thing to do is to determine the real-time scope of the disaster. The so-called real-time scope of the disaster is the geographical scope (such as longitude and latitude or radius) affected by the flood disaster after receiving the disaster warning signal, not the water body coverage. It should be noted that the real-time scope of the disaster can be the initial geographical scope affected by the flood disaster when the disaster occurs, or the real-time geographical scope at any time after the disaster occurs (while still in the disaster).

[0075] Then, based on the real-time endangered scope of the disaster, combined with local geospatial information, geological information and future weather information, the final endangered scope that may be affected by future flood disasters, that is, the second area, is inferred. Similarly, the final endangered scope here is also a geographical scope, not a water body coverage scope.

[0076] After the second area is determined, it is necessary to select a suitable first area, which is a geographical range that is significantly larger than the second area. For example, the first area can be selected as a 20 km × 20 km longitude and latitude range, or a 20 km circle with the second area as the center, or a longitude and latitude range of the town or county-level administrative area that includes the second area.

[0077] During the disaster, when the real-time scope of the disaster changes, the scope of the second area changes at any time due to other factors. Therefore, in order to avoid the first area from changing at any time, the scope is generally much larger than the second area.

[0078] For example, the real-time endangered area when a disaster occurs, and the longitude and latitude ranges of the first area and the second area can be obtained from the local management department of the place where the disaster occurs.

[0079] After obtaining the geographic scope of the first area (such as longitude and latitude), it is necessary to obtain remote sensing data of the first area. In order to more accurately determine the water body scope of the flood disaster, it is necessary to obtain multispectral remote sensing data of the first area before the disaster occurs and radar remote sensing data of the first area when the disaster occurs. Since the first area is much larger than the second area, in general, the size of the first area will not be affected by the real-time endangered area of ​​the disaster or the second area (the final endangered area) that changes due to the real-time endangered area of ​​the disaster.

[0080] The multispectral remote sensing data and radar remote sensing data of the first area can be obtained from some satellites that specialize in observing earth remote sensing data, or they can be collected by airplanes or drones as needed.

[0081] S202: Determine the range of a first water body before a disaster occurs according to multispectral remote sensing data of the first area;

[0082] After obtaining the multispectral remote sensing data of the first area, it is necessary to extract the water coverage range therein, so as to obtain the first water range before the disaster occurs. The first water range represents the normal water coverage range in the first area before the disaster.

[0083] First, the calculation of the water body range can be done by extracting the water body area through index method, clustering, machine learning, deep learning and other methods.

[0084] It should be noted that although the multispectral remote sensing data of the first area comes from before the disaster, in order to ensure the recognition accuracy of the first water body range, historical multispectral remote sensing data collected during the period when the weather conditions were good and the cloud coverage was low before the disaster was preferred.

[0085] S203: Determine the range of the second water body when the disaster occurs according to the radar remote sensing data of the first area;

[0086] After obtaining the radar remote sensing data of the first area, it is necessary to extract the real-time coverage of the water body to obtain the second water body range when the disaster occurs. The second water body range represents the real-time coverage of the water body in the first area.

[0087] When flood disasters occur, due to the relatively bad weather conditions, the water body identification effect of multispectral remote sensing data may be poor. In this embodiment, radar remote sensing data is selected to extract the water body coverage range in view of its strong anti-interference ability, so that the extracted first water body range can truly reflect the water body coverage range.

[0088] S204: Determine the flood disaster scope according to the first water body scope and the second water body scope.

[0089] After obtaining the range of the first water body and the range of the second water body, the two ranges can be compared to determine the current flood disaster range.

[0090] It should be noted that, in the above embodiment, the execution order of steps S202 and S203 can be swapped or performed simultaneously, which has no effect on determining the scope of the flood disaster.

[0091] In the case where there is a long interval between two adjacent remote sensing data collections, if it is not possible to use airplanes or drones, and there is no authority to call on satellites for real-time collection, then an algorithm simulation can be performed based on the local real-time flood disaster scope, water flow conditions, disaster area terrain, weather, etc., to estimate the flood disaster scope within the interval time.

[0092] This embodiment extracts the first water body range through multi-spectral remote sensing data before the disaster occurs, and extracts the second water body range through radar remote sensing data when the disaster occurs. Since radar remote sensing data can truly reflect the water body coverage under weather interference, the accuracy of water body identification will not be affected even if there is weather interference during the disaster. Combining it with the water body range before the disaster identified by multi-spectral remote sensing data can more accurately and quickly identify the scope of flood disasters.

[0093] Figure 3 Another flood disaster identification method flow chart provided in the present application embodiment. Figure 2Based on the embodiment shown, the flood disaster identification method is further described. Figure 3 As shown, the flood disaster identification method includes:

[0094] S301: Acquire multispectral remote sensing data of a first area before a disaster occurs and radar remote sensing data of the first area when a disaster occurs, wherein the first area is larger than the second area, and the second area is a final endangered area determined according to the real-time endangered area of ​​the disaster;

[0095] The main process in this step is the same as S201.

[0096] Exemplarily, multispectral remote sensing data is obtained from the Sentinel-2 satellite of the Copernicus Data Open Access Center. The Sentinel-2 satellite carries a Multi-Spectral Imager (MSI) at an altitude of 786km, covering 13 spectral bands with a width of 290km. The ground resolutions are 10m, 20m and 60m respectively, the revisit period of one satellite is 10 days, and the two satellites complement each other with a revisit period of 5 days. From visible light and near-infrared to short-wave infrared, it has different spatial resolutions. Among optical data, Sentinel-2 data is the only one that contains three bands in the red edge range, which is very effective for monitoring vegetation health information. Figure 1 As shown, among the 13 bands of the Sentinel-2 satellite, the blue (B2), green (B3), red (B4) and near-infrared (B8) bands have a resolution of 10 meters; the ground sampling distance of the red end (B5), near-infrared NIR (B6, B7 and B8a) and shortwave infrared SWIR (B11 and B12) is 20 meters; the pixel size of the coastal atmospheric aerosol (B1) and cirrus band (B10) is 60 meters.

[0097] Table 1 Information diagram of different bands of Sentinel-2

[0098] Band Resolution (m) Central wavelength describe B1 60 443 nm Ultra blue (coastal and aerosol) B2 10 490 nm Blue B3 10 560 nm Green B4 10 665 nm Red B5 20 705 nm Visible and Near Infrared (VNIR) B6 20 740 nm Visible and Near Infrared (VNIR) B7 20 783 nm Visible and Near Infrared (VNIR) B8 10 842 nm Visible and Near Infrared (VNIR) B8 20 865 nm Visible and Near Infrared (VNIR) B9 60 940 nm Shortwave infrared (SWIR) B10 60 1375 nm Shortwave infrared (SWIR) B11 20 1610 nm Shortwave infrared (SWIR) B12 20 2190 nm Shortwave infrared (SWIR)

[0099] Radar remote sensing data are obtained from Sentinel-1, an Earth observation satellite in the European Space Agency's Copernicus program (GMES). It consists of two satellites, A and B, and carries a C-band synthetic aperture radar that can provide continuous images (day, night, and all kinds of weather).

[0100] It should be noted that any remote sensing satellite has a revisit period. For example, Sentinel-2 consists of two satellites working together, so the revisit period is five days. In the event of a major flood disaster, in order to obtain the latest flood disaster scope as soon as possible, it is necessary to continuously collect remote sensing data of the disaster area, and the general satellite revisit period cannot meet the requirement. At this time, aircraft or drones can be used to collect remote sensing data of the disaster area regularly (for example, every half a day), or under the condition of permission, the corresponding satellite can be directly called to collect remote sensing data of the disaster area in real time.

[0101] The present embodiment preferably uses Sentinel-1 and Sentinel-2 satellites because they have been in use for a relatively long time and have more historical data, which facilitates the analysis of the previous and next time series.

[0102] S302: Preprocessing the multispectral remote sensing data of the first area to obtain first image data including at least the real-time endangered range of the disaster;

[0103] Since the scope of the first area is much larger than the real-time disaster hazard range, in order to reduce the interference of complex landforms on the extraction of the first water body range, it is necessary to convert the multispectral remote sensing data of the first area into first image data that at least includes the real-time disaster hazard range before extracting the water body.

[0104] For example, the multispectral remote sensing data of the first area is obtained from the Sentinel-2 satellite, and each band obtained from the satellite is JPEG2000 data. The data obtained by some remote sensing satellites may be in GRID, TIF, etc. No matter what kind of data is collected by multispectral remote sensing data, preprocessing is required.

[0105] The following takes the JPEG2000 data obtained by the Sentinel-2 satellite as an example to explain the preprocessing process in detail:

[0106] (1) converting the multispectral remote sensing data of the first area into Geotiff data of different bands;

[0107] Geotiff data contains geospatial information, such as geographic coordinates, projection information, geographic boundaries, resolution, etc. These metadata enable GEOTIFF images to be seamlessly integrated with geographic information systems (GIS) and geospatial analysis tools for applications such as map making, geolocation, and spatial analysis. During conversion, JPEG2000 data from the Sentinel-2 satellite is converted into Geotiff data.

[0108] After conversion, considering the extraction accuracy of water body range in the later stage, all the different bands with resolution greater than 10m in Table 1 are resampled to 10m resolution. For multispectral remote sensing data from other different satellites, different resolutions can be resampled to the same resolution that meets the requirements.

[0109] (2) merging the Geotiff data of different bands in the first area into a TIF file;

[0110] After obtaining Geotiff data of different bands, the bands are superimposed to obtain a multi-band TIF file. TIF file is a universal image file format that can be used to store various types of image data. After superposition and conversion, TIF file discards some useless information in the water body extraction process, such as geographic spatial information, which is conducive to improving the efficiency of the first water body range extraction in the later stage.

[0111] (3) Crop the TIF file to obtain first image data that at least includes the real-time hazard range of the disaster.

[0112] After cropping the TIF file, we can get the geographical scope where most of the area is covered by water bodies. In this way, the TIF file with a higher proportion of water body range can further improve the efficiency of water body range extraction in the later extraction.

[0113] After the above preprocessing, the first water body range directly extracted from the first image data has higher recognition accuracy and extraction efficiency, which is conducive to providing disaster guidance information in a timely manner.

[0114] S303: Extracting a first water body range according to the first image data.

[0115] In this step, the first water body range can be extracted by the method mentioned in step S202. This embodiment uses the index method to extract the first water body range from the first image data for detailed description.

[0116] (1) normalizing the first image data to obtain normalized first image data;

[0117] For example, the normalized difference water index NDWI (Normalized Difference Water Index) is calculated based on the first image data. The construction of the normalized difference water index NDWI utilizes the characteristic that the reflectivity of the water spectrum in the near-infrared range is almost zero, which is suitable for natural scenes with uniformly distributed elements and relatively single types. The calculation method is as follows:

[0118] NDWI=(GREEN-NIR) / (GREEN+NIR)

[0119] Among them, GREEN represents the reflectivity of the green light band, and NIR represents the reflectivity of the near-infrared band.

[0120] However, the above normalized water index has poor water extraction effect in the range with more building background. Therefore, the improved normalized water index MNDWI (Modified NDWI) can be used, and the calculation method is as follows:

[0121] MNDWI=(GREEN-SWIR) / (GREEN+SWIR)

[0122] Among them, SWIR is the reflectivity of the shortwave infrared band.

[0123] Therefore, the improved normalized difference water index (MNDWI) is more suitable for extracting the range of water bodies containing buildings.

[0124] (2) determining a first threshold range for distinguishing between a water body and a non-water body according to the normalized first image data;

[0125] The optimal threshold for distinguishing water bodies and non-water bodies can be selected from the first threshold range, and the first threshold range is determined according to the recognition effects of different thresholds. In a specific embodiment, after testing, when the first threshold range is 0.2 to 10, the normalized water index NDWI can clearly distinguish water bodies and non-water bodies, and the recognition and extraction effect is good.

[0126] (3) Extracting a first water body range from the normalized first image data according to a first threshold range.

[0127] In this embodiment, after comparing the water body extraction in a certain area, it is found that the NDWI and MNDWI extraction methods have similar effects. When making a specific selection, NDWI is used for water body identification and extraction, and finally, the irregular image spots are reduced and the noise is reduced by filtering and other methods.

[0128] S304: Determine the range of the second water body when the disaster occurs according to the radar remote sensing data of the first area;

[0129] For radar remote sensing data, the threshold method, Otsu method and random forest method can be used to extract the second water body range.

[0130] S305: Determine the flood disaster scope according to the first water body scope and the second water body scope.

[0131] It should be noted that in the above embodiment, the execution order of steps S302 to S303 and step S304 can be exchanged, for example, S304 is executed first, and then S302 and S303, or they can be performed simultaneously, which has no effect on determining the scope of the flood disaster.

[0132] The flood disaster identification method in the above embodiment can improve the recognition accuracy of the water body range and the water body extraction efficiency by preprocessing the multispectral remote sensing data and then extracting the first water body range through the index method.

[0133] Figure 4 Another flood disaster identification method flow chart provided in the present application embodiment. Figure 2Based on the embodiment shown, the flood disaster identification method is further described. Figure 4 As shown, the flood disaster identification method includes:

[0134] S401: Acquire multispectral remote sensing data of a first area before a disaster occurs and radar remote sensing data of the first area when a disaster occurs, wherein the first area is larger than the second area, and the second area is a final endangered area determined according to the real-time endangered area of ​​the disaster;

[0135] The specific implementation process of this step can refer to the description of step S301, which will not be repeated here.

[0136] S402: Determine the range of a first water body before a disaster occurs based on multispectral remote sensing data of a first area;

[0137] The specific implementation process of this step can refer to the description of steps S302 to S303 to achieve the extraction of the first water body range, which will not be repeated here.

[0138] In addition to using Sentinel-2 to obtain multispectral remote sensing data, it can also be obtained from the following remote sensing satellites: for example, the LANDSAT satellite jointly operated by the National Aeronautics and Space Administration (NASA) and the United States Geological Survey (USGS), and the Ziyuan-3 (ZY-3) developed and operated by the China National Space Administration (CNSA).

[0139] S403: Determine a bipolar index according to radar remote sensing data of the first area;

[0140] For radar remote sensing data obtained from the Sentinel-1 satellite, the dual-polarized water index (SDWI) is a water index obtained by band calculation using the VV and VH backscatter coefficients of the Sentinel-1 satellite.

[0141] The calculation formula of SDWI index is as follows:

[0142] SDWI=ln(k×VV×VH)+s

[0143] Among them, VV and VH are polarization images in radar remote sensing data, k is the gain coefficient, which is generally 10, and s is the initial threshold.

[0144] In addition to using Sentinel-1 to obtain radar remote sensing data, it can also be obtained from the following remote sensing satellites: such as Haisi-1, Chaohu-1, Fucheng-1 and other commercial satellites.

[0145] The VV (vertical-vertical polarization) and VH (vertical-horizontal polarization) polarization images are from radar remote sensing data of the first area. To improve calculation efficiency, the bipolar index is determined according to the backscattering coefficient of the radar remote sensing data of the first area, which may also include:

[0146] (1) cropping the radar remote sensing data of the first area to obtain second image data that at least includes the real-time endangered area of ​​the disaster;

[0147] The second image data obtained by clipping has a smaller range than the radar remote sensing data of the first area. On the basis of extracting only the second water body range, less data needs to be processed, which is more conducive to accelerating the water body extraction efficiency.

[0148] (2) Calculating the bipolar index based on the VV and VH polarization images of the second image data.

[0149] For the specific calculation formula, please refer to the above description. At this time, VV and VH are polarization images in the second image data.

[0150] S404: Determine a second threshold range according to histogram statistics of the bipolar exponential distribution;

[0151] After determining the bipolar directional index SDWI of radar remote sensing data, histogram statistics can be performed on it. Figure 5 is the statistical histogram of the bipolar index, from which the separation range of the two peaks can be obtained, one peak is the water body and the other peak is the non-water body. Figure 5 For example, the second threshold range for distinguishing water bodies from non-water bodies is approximately between 8 and 8.5.

[0152] S405: Extracting the second water body range when the disaster occurs from the radar remote sensing data according to the second threshold range.

[0153] According to the second threshold range obtained in step S404, this embodiment uses 8.2 as the separation index threshold of water bodies and non-water bodies, thereby extracting the second water body range.

[0154] In another embodiment, extracting the second water body range when the disaster occurs from the radar remote sensing data according to the second threshold range includes:

[0155] According to the second threshold range, water body information that meets the preset confidence level is proposed as a positive sample, and non-water body information is proposed as a negative sample, and the historical radar remote sensing data is trained by the positive samples and the negative samples to obtain a water body extraction model;

[0156] The second water body range when the disaster occurs is extracted from the second image data through a water body extraction model.

[0157] In this embodiment, the second threshold range is used as a screening condition for positive and negative samples. The cropped second image data area is sampled in 32×32 dimensions. The water body range extracted from the historical radar remote sensing data is trained with positive and negative samples to obtain a water body extraction model. The water body extraction model is applied to the second image data to obtain the second water body range when the disaster occurs.

[0158] S406: Determine the flood disaster scope according to the first water body scope and the second water body scope;

[0159] S407: Determine the type of affected areas and the level of damage within the flood disaster range based on the flood disaster range.

[0160] After determining the scope of the flood disaster according to the scope of the first water body and the scope of the second water body, the type of endangered area and the degree of hazard can also be determined according to the scope of the flood disaster. This can provide scientific, accurate and rapid guidance information for disaster rescue, emergency disposal and post-disaster emergency material needs.

[0161] It should be noted that, in the above embodiment, the execution order of step S402 and steps S403 to S405 can be exchanged, for example, S403 to S405 are executed first, and then S402 is executed, or they can be performed simultaneously, which has no effect on determining the scope of the flood disaster.

[0162] In this embodiment, the radar remote sensing data is cropped and then combined with the threshold method and machine learning algorithm to extract the second water body range, which has higher accuracy and timeliness, and can provide scientific, precise and rapid guidance information for disaster rescue, emergency response, and post-disaster emergency material demand assessment.

[0163] Figure 6 A structural schematic diagram of a flood disaster identification device provided in an embodiment of the present application, which device can be in the form of software and / or hardware, and the flood disaster identification device includes: a remote sensing data acquisition module 601, a first water body range determination module 602, a second water body range determination module 603 and a flood disaster range determination module 604.

[0164] The remote sensing data acquisition module 601 is used to acquire multispectral remote sensing data of a first area before a disaster occurs and radar remote sensing data of the first area when a disaster occurs, the first area is larger than the second area, and the second area is a final endangered area determined according to the real-time endangered area of ​​the disaster;

[0165] The first water body range determination module 602 is used to determine the first water body range before the disaster occurs according to the multispectral remote sensing data of the first area;

[0166] The second water body range determination module 603 is used to determine the second water body range when the disaster occurs according to the radar remote sensing data of the first area;

[0167] The flood disaster range determination module 604 is used to determine the flood disaster range according to the first water body range and the second water body range.

[0168] In another embodiment, the first water body range determination module 602 may also be configured as follows:

[0169] Preprocessing the multispectral remote sensing data of the first area to obtain first image data including at least the real-time endangered range of the disaster;

[0170] A first water body range is extracted according to the first image data.

[0171] The multispectral remote sensing data of the first area is preprocessed to obtain first image data including at least the real-time endangered range of the disaster, specifically including:

[0172] Convert the multispectral remote sensing data of the first area into Geotiff data of different bands;

[0173] Merge Geotiff data of different bands into multi-band TIF files;

[0174] The multi-band TIF file is cropped to obtain the first image data including at least the real-time endangered scope of the disaster.

[0175] In the first water body range determination module 602, extracting the first water body range according to the first image data includes:

[0176] Normalizing the first image data to obtain normalized first image data;

[0177] Determine a first threshold range for distinguishing between water bodies and non-water bodies according to the normalized first image data;

[0178] The first water body range is extracted from the normalized first image data according to the first threshold range.

[0179] In another embodiment, the second water body range determining module 603 may also be configured as follows:

[0180] determining a bipolar index based on radar remote sensing data of the first area;

[0181] Determine a second threshold range according to histogram statistics of the bipolar exponential distribution;

[0182] The second water body range when the disaster occurs is extracted from the radar remote sensing data according to the second threshold range.

[0183] The bipolar index is determined according to the backscatter coefficient of the radar remote sensing data of the first area, including:

[0184] Clipping the radar remote sensing data of the first area to obtain second image data at least including the real-time endangered range of the disaster;

[0185] The bipolar index is calculated based on the VV and VH polarization images of the second image data.

[0186] In the second water body range determination module 603, the second water body range when the disaster occurs is extracted from the radar remote sensing data according to the second threshold range, including:

[0187] According to the second threshold range, water body information that meets the preset confidence level is proposed as a positive sample, and non-water body information is proposed as a negative sample, and the historical radar remote sensing data is trained by the positive samples and the negative samples to obtain a water body extraction model;

[0188] The second water body range when the disaster occurs is extracted from the second image data through a water body extraction model.

[0189] In another embodiment, after outputting the flood disaster range, the flood disaster range determination module 604 further determines the type of endangered area and the hazard level of the flood disaster range according to the flood disaster range.

[0190] The flood disaster identification device provided in the embodiment of the present application has the same implementation principle and technical effects as those in the aforementioned flood disaster identification method embodiment. For the sake of brief description, for matters not mentioned in the device embodiment, reference may be made to the corresponding contents in the aforementioned flood disaster identification method embodiment.

[0191] Figure 7 This is a hardware structure diagram of an electronic device provided in an embodiment of the present application. The embodiment provides an electronic device, including: a processor 701, and a memory 702 communicatively connected to the processor 701; the memory 702 stores computer execution instructions; the processor 701 executes the computer execution instructions stored in the memory 702, so as to implement the description of the flood disaster identification method in the corresponding embodiment above.

[0192] Figure 7 The electronic device shown further includes a communication interface 703 and a communication bus 704 , wherein the processor 701 , the memory 702 and the communication interface 703 communicate with each other via the communication bus 704 .

[0193] The memory 702 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk storage. The communication bus 704 may be an ISA bus, a PCI bus, or an EISA bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one bidirectional arrow is used in the figure, but it does not mean that there is only one bus or one type of communication bus.

[0194] The communication interface 703 is used to connect to at least one user terminal and other network units through a network interface, and send the encapsulated IPv4 message or IPv4 message to the user terminal through the network interface.

[0195] The processor 701 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 701. The above processor 701 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The disclosed methods, steps and logic block diagrams in the embodiments of the present disclosure can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present disclosure can be directly embodied as a hardware decoding processor for execution, or a combination of hardware and software modules in the decoding processor for execution. The software module may be located in a storage medium mature in the art, such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 702, and the processor 701 reads the information in the memory 702 and completes the steps of the method of the above embodiment in combination with its hardware.

[0196] The embodiment of the present application also provides a computer storage medium, in which a computer-readable storage medium stores computer executable instructions, which are used to implement the description of the flood disaster identification method in the above corresponding embodiment when executed by the processor, so it will not be repeated here. In addition, the description of the beneficial effects of using the same method will not be repeated. For technical details not disclosed in the computer storage medium embodiment involved in the present invention, please refer to the description of the method embodiment of the present invention.

[0197] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the above-mentioned program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The above-mentioned storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0198] The embodiment of the present application also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it is used to implement the description of the flood disaster identification method in the corresponding embodiment above. Therefore, it will not be repeated here.

[0199] It should be understood that the present application is not limited to the precise structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A flood disaster identification method, characterized in that: include: Acquire multispectral remote sensing data of a first area before a disaster occurs and radar remote sensing data of the first area when a disaster occurs, wherein the first area is larger than a second area, and the second area is a final endangered area determined according to the real-time endangered area of ​​the disaster; Determine the scope of the first water body before the disaster occurs based on the multispectral remote sensing data of the first area; Determine the scope of the second water body when the disaster occurs based on the radar remote sensing data of the first area; Determine the scope of flood disaster based on the scope of the first water body and the second water body.

2. The method according to claim 1, characterized in that The determining of the first water body range before the disaster occurs according to the multispectral remote sensing data of the first area includes: Preprocessing the multispectral remote sensing data of the first area to obtain first image data including at least the real-time endangered range of the disaster; A first water body range is extracted according to the first image data.

3. The method according to claim 2, characterized in that The preprocessing of the multispectral remote sensing data of the first area to obtain first image data including at least the real-time endangered range of the disaster specifically includes: Convert the multispectral remote sensing data of the first area into Geotiff data of different bands; Merge Geotiff data of different bands into multi-band TIF files; The multi-band TIF file is cropped to obtain the first image data including at least the real-time endangered scope of the disaster.

4. The method according to claim 3, characterized in that The step of extracting the first water body range according to the first image data includes: Normalizing the first image data to obtain normalized first image data; Determine a first threshold range for distinguishing between water bodies and non-water bodies according to the normalized first image data; The first water body range is extracted from the normalized first image data according to the first threshold range.

5. The method according to claim 1, characterized in that The determining the scope of the second water body when the disaster occurs according to the radar remote sensing data of the first area includes: determining a bipolar index based on radar remote sensing data of the first area; Determine a second threshold range according to histogram statistics of the bipolar exponential distribution; The second water body range when the disaster occurs is extracted from the radar remote sensing data according to the second threshold range.

6. The method according to claim 5, characterized in that The step of determining the bipolar index according to the backscatter coefficient of the radar remote sensing data of the first area includes: Clipping the radar remote sensing data of the first area to obtain second image data at least including the real-time endangered range of the disaster; The bipolar index is calculated based on the VV and VH polarization images of the second image data.

7. The method according to claim 6, characterized in that The step of extracting the second water body range when the disaster occurs from the radar remote sensing data according to the second threshold range includes: According to the second threshold range, water body information that meets the preset confidence level is proposed as a positive sample, and non-water body information is proposed as a negative sample, and the historical radar remote sensing data is trained by the positive samples and the negative samples to obtain a water body extraction model; The second water body range when the disaster occurs is extracted from the second image data through a water body extraction model.

8. The method according to claim 1, characterized in that After determining the flood disaster range according to the first water body range and the second water body range, the method further includes: Determine the type of affected areas and the level of damage according to the scope of the flood disaster.

9. A flood disaster identification device, characterized in that: include: A remote sensing data acquisition module, which acquires multispectral remote sensing data of a first area before a disaster occurs and radar remote sensing data of the first area when a disaster occurs, wherein the first area is larger than a second area, and the second area is a final endangered area determined according to the real-time endangered area of ​​the disaster; A first water body range determination module, which determines the first water body range before the disaster occurs based on the multispectral remote sensing data of the first area; A second water body range determination module, which determines the range of the second water body when the disaster occurs based on the radar remote sensing data of the first area; The flood disaster scope determination module determines the flood disaster scope according to the first water body scope and the second water body scope.

10. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 8 when executed by a processor.

12. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images