A method for intelligent identification of river water body survival state on orbit satellite

Abstract

This invention relates to an on-orbit intelligent method for determining the survival status of river water bodies: Vector data of the detected river water body is stored in an on-board image processor; the on-board camera payload images the Earth, and based on real-time satellite position information acquired on-board, the location information of the image coverage area is obtained; based on the image containing location information and the vector data of the detected river water body, images containing river water bodies are selected, i.e., target images; the spectral index related to water extraction in the target image is calculated, a threshold is determined for the spectral index, and the detected river water body information in the target image coverage area is extracted, resulting in a water extraction result map of the target image coverage area; the water extraction result map of the target image coverage area is partitioned and statistically analyzed with the extended river water body vector data to obtain the water area of ​​the target image coverage area; the survival coefficient of the detected river water body is calculated to determine and record the survival status of the detected river water body.

Description

Technical Field

[0001] This invention relates to an on-orbit intelligent method for determining the survival status of river water bodies. It is a method for determining the survival status of river water bodies on-orbit based on images captured in real time by an on-orbit satellite, and belongs to the field of on-orbit intelligent application of remote sensing data. Background Technology

[0002] Rivers are an important type of land cover, playing a vital role in water resource security, material transportation, and agricultural irrigation. Due to factors such as season and climate, the status of river water bodies often changes over time. Especially in areas with dry seasons, some river bodies may even experience short periods of waterlessness. Real-time information on the status of river water bodies is helpful for relevant departments in water resource early warning, management, and irrigation water allocation.

[0003] Satellite imagery, due to its wide coverage and real-time acquisition, is suitable as a primary data source for determining the survival status of river water bodies. Currently, the determination of river water survival status using satellite imagery data is mainly achieved through modeling and extraction methods. Starting from data acquisition, the process mainly includes data download, product processing, sample creation, model building, prediction and extraction, and survival status determination. After acquiring the downlink satellite imagery data, the entire process is primarily completed on the ground. Regarding sample creation and method selection, a relatively simple and efficient index thresholding method can be chosen, such as using the Normalized Difference Vegetation Index (NDVI) or Normalized Difference Water Index (NDBI) for thresholding and extraction. Alternatively, a more complex but more generalizable deep learning method, such as a convolutional neural network, can be selected.

[0004] The Guangzhou Institute of Geography, Guangdong Academy of Sciences and the Southern Laboratory of Marine Science and Engineering (Guangzhou) disclosed a method, device, and equipment for water body extraction based on Sentinel remote sensing data in patent CN201911295740.1. This invention uses Sentinel 1 and Sentinel 2 data products that have been processed and released as data sources. The 20m resolution data of Sentinel 2 is resampled to 10m resolution and combined with the original 10m resolution data of Sentinel 2 to construct and obtain the Normalized Difference Water Index (MNDWI). The first water body element is obtained by setting a threshold. Then, the mountain shadows in the extracted water body elements are excluded by combining the digital terrain elevation model and the Sentinel 1 data, and water body elements with too low pixel size are removed to obtain the final extraction result.

[0005] However, this method has the following main problems:

[0006] (1) The unclear mission area leads to low image utilization and high processing costs. Among the large number of images acquired by satellites orbiting the Earth, the presence of rivers and water bodies cannot be easily confirmed based solely on the image header information; other information is required. Calculating and extracting rivers and water bodies based on models for each acquired image will result in a large number of useless images being involved in the calculation, greatly reducing the effective utilization of images and consuming transmission and computing resources.

[0007] (2) The lag in data processing using data products transmitted to the ground and then processed. For ground-based river water body modeling, the data needs to be first transmitted from satellites, then processed to obtain data products, and then integrated with multi-source products such as digital elevation models (DEMs) before entering the river water body modeling and extraction process. This data transmission, processing, and integration result in a lag in obtaining the river water body's current status, lacking real-time capability.

[0008] (3) The reflectivity of a river when it dries up can easily lead to missed detections. When a river dries up, due to its reflectivity, the reflectivity information in optical or microwave images alone cannot be used to detect the river water, and therefore a single image cannot reflect the existing water condition. Summary of the Invention

[0009] The technical problem solved by this invention is to overcome the shortcomings of the prior art and provide an on-orbit intelligent method for determining the survival status of river water bodies, which can quickly obtain the survival status of river water bodies from satellite images, thereby improving the timeliness and convenience of obtaining the survival status of river water bodies.

[0010] The technical problem further solved by this invention is to automatically determine the survival status of river water bodies on the satellite using a small amount of stored data and satellite positioning and processing equipment, while reducing the amount of downlink data from the satellite.

[0011] The solution of this invention is: an on-orbit intelligent method for determining the survival status of river water bodies, which includes the following steps:

[0012] The vector data of the detected river body is stored in the on-board image processor.

[0013] The onboard camera payload images the Earth and obtains the location information of the image coverage area based on the satellite position information acquired in real time on the satellite.

[0014] Based on the image containing location information and the vector data of the detected river water body, the image containing the river water body is selected, i.e., the target image. The vector data of the river water body in the area covered by the target image is extracted and buffered and expanded to obtain the expanded river water body vector data.

[0015] Calculate the spectral index related to water body extraction in the target image, apply a threshold to the spectral index, extract the detected river water information in the target image coverage area, and draw the water body extraction result map of the target image coverage area.

[0016] The water body extraction result map of the target image coverage area is compared with the extended river water body vector data by partitioning and statistically analyzing the water body area of ​​the target image coverage area;

[0017] Based on the area of ​​the detected water body covered by the target image, the survival coefficient of the detected river water body is calculated, and the survival status of the detected river water body is determined and recorded by the survival coefficient of the detected river water body.

[0018] Preferably, the vector data of the detected river water body is in shp format, which includes a coordinate information file, an attribute table, a data projection information file, and a geometric feature index file of the river water body to be detected. The geometric feature index file includes the order of the points that make up the line or surface, and the attribute table includes the name, number, threshold width, and average width of the river water body.

[0019] Preferably, the specific steps for selecting images containing river water bodies are as follows:

[0020] Based on the location information of the image coverage area, determine the coordinate range of the image coverage area;

[0021] If the coordinates of the detected river body in the vector data are within the coordinate range of the target image coverage area, the image is considered not to contain the detected river; if they are, the image is considered to contain the detected river.

[0022] Preferably, the method for extracting the detected river water information in the target image coverage area and drawing the water extraction result map of the target image coverage area is as follows:

[0023] If the spectral index of a target image pixel that is related to water extraction is greater than a preset threshold, the pixel is considered to be located in a water body; otherwise, the pixel is considered to be located in a non-water body.

[0024] The pixel value of water body pixels is assigned to 1, and non-water body pixels are marked as 0, resulting in the water extraction result map of the target image coverage area.

[0025] Preferably, the extended river water vector data of the target image coverage area is obtained by the following method:

[0026] The river water vector data stored in the spaceborne storage device or uploaded to the spaceborne storage device is compared with the target image to determine the coordinates, and the river water vectors contained in the area covered by the target image are selected and recorded as the river water vectors to be processed.

[0027] The river water vector to be processed is buffered and expanded to obtain extended river water vector data covering the target image area.

[0028] Preferably, the river water vector data is point, line, or area vector data;

[0029] Preferably, when the stored vector data of the detected river water body is point-like or line-like vector data, the buffer expansion includes the following steps:

[0030] Using half of the threshold width of the river water body in the attribute table of the river water body vector data to be processed as the buffer radius, the buffer of the river water body vector to be processed is expanded to obtain the first river water body buffer vector;

[0031] The buffer radius is calculated by adding 2 pixels to half the average width of the river water body in the attribute table of the river water body vector data to be processed. The buffer is then expanded to obtain the second river water body buffer vector.

[0032] The second river water buffer vector is the extended river water vector data of the target image coverage area.

[0033] Preferably, when the stored vector data of the detected river water body is area vector data, the buffer expansion includes the following steps:

[0034] Using half of the threshold width of the river water body in the attribute table of the river water body vector data to be processed as the buffer radius, the buffer of the river water body vector to be processed is expanded to obtain the first river water body buffer vector;

[0035] The first river water buffer vector serves as the extended river water vector data for the target image coverage area.

[0036] Preferably, the survival coefficient of the detected river water body is: extracted water area / threshold water area.

[0037] Preferably, the condition of the river water being tested is determined by the following method:

[0038] (a) If the river water survival coefficient of the target river is greater than the preset survival coefficient threshold, it is marked as having a normal survival status;

[0039] (b) If the river water survival coefficient of the target river body is less than or equal to the preset survival coefficient threshold, it is marked as an abnormal survival status.

[0040] The advantages of this invention compared to the prior art are:

[0041] (1) The present invention can determine the water body existence status in real time based on the real-time acquired images, thereby improving the timeliness of water body existence status information acquisition.

[0042] (2) This invention transfers the calculation of river water body existence status to the satellite, which reduces the amount of downlink data when the satellite meets the corresponding mission requirements and effectively alleviates data redundancy.

[0043] (3) Based on the information provided by the river water vector data and the GNSS positioning information, the present invention can determine whether there is a target river water body in the image before the image data is processed through a simple mapping relationship, thereby greatly reducing the number of invalid images involved in the calculation.

[0044] (4) The present invention uses vector data as a file, and the range of the target river water body can be determined by vector data. The introduction and utilization of vector data reduces the impact of false detection of urban building shadows and missed detection of dried river water bodies on the determination of the survival status.

[0045] (5) The satellite of this invention can continuously acquire images and continuously determine the survival status of river water bodies based on the images, thereby providing early warning of areas where the survival status of river water bodies changes. Attached Figure Description

[0046] Figure 1 This is a schematic flowchart of an on-orbit intelligent determination method for the survival status of river water bodies according to an embodiment of the present invention. Detailed Implementation

[0047] The present invention will be further described below with reference to the embodiments.

[0048] The on-orbit intelligent determination method for the survival status of river water bodies of the present invention mainly includes data preparation, on-orbit data screening and processing, and determination and transmission of the survival status of river water bodies.

[0049] During the data preparation process, the vector data and the on-orbit intelligent discrimination program for the existence status of river water bodies are placed into the on-board storage device through pre-storage or data uploading. The camera payload, positioning device and FPGA complete the earth imaging, recording positioning and data processing, and finally obtain image data with location information.

[0050] During the on-board data screening and processing, the coordinate information of the vector data storage is compared with the position information of the image to screen out target images containing river water bodies, and the spectral index is calculated. A threshold is set to complete the water body extraction and obtain the water body extraction result.

[0051] In the process of determining the survival status of river water bodies, the water body extraction results are overlaid with vector data for analysis to obtain records of the survival status of river water bodies. These records are then encoded and transmitted to the ground to obtain the final results of the survival status of river water bodies.

[0052] The focus of this invention is on the introduction and utilization of vector data, on-board image screening, and river water body survival determination, specifically including the following steps:

[0053] The vector data of the detected river body is stored in the on-board image processor.

[0054] The onboard camera payload images the Earth and obtains the location information of the image coverage area based on the satellite position information acquired in real time on the satellite.

[0055] Based on the image containing location information and the vector data of the detected river water body, the image containing the river water body is selected, i.e., the target image. The vector data of the river water body in the area covered by the target image is extracted and buffered and expanded to obtain the expanded river water body vector data.

[0056] Calculate the spectral index related to water body extraction in the target image, apply a threshold to the spectral index, extract the detected river water information in the target image coverage area, and draw the water body extraction result map of the target image coverage area.

[0057] The water body extraction result map of the target image coverage area is compared with the extended river water body vector data by partitioning and statistically analyzing the water body area of ​​the target image coverage area;

[0058] Based on the area of ​​the detected water body covered by the target image, the survival coefficient of the detected river water body is calculated, and the survival status of the detected river water body is determined and recorded by the survival coefficient of the detected river water body.

[0059] The spectral indices related to water extraction are the near-infrared normalized water index (NDWI), the mid-infrared normalized water index (MNDWI), or the normalized vegetation index (NDVI).

[0060] The measured river water vector data is stored in the onboard storage device in advance or via data uploading.

[0061] The detected river water body vector data is Shp format vector data, which includes the coordinate information file, attribute table, data projection information file, and geometric feature index file of the river water body to be detected. The geometric feature index file includes the order of the points that make up the line or surface, and the attribute table includes the name, number, threshold width, and average width of the river water body.

[0062] The specific steps for selecting images containing river water bodies are as follows:

[0063] S3.1 Determine the coordinate range of the image coverage area based on the location information of the image coverage area;

[0064] S3.2. Search the vector data for coordinates of the detected river body to see if there are any coordinates within the coordinate range of the target image coverage area. If not, the image is considered not to contain the detected river; if so, the image is considered to contain the detected river.

[0065] The method for extracting the detected river water information from the target image coverage area and plotting the water extraction result map of the target image coverage area is as follows:

[0066] If the spectral index of a target image pixel that is related to water extraction is greater than a preset threshold, the pixel is considered to be located in a water body; otherwise, the pixel is considered to be located in a non-water body.

[0067] The pixel value of water body pixels is assigned to 1, and non-water body pixels are marked as 0, resulting in the water extraction result map of the target image coverage area.

[0068] The extended river water vector data of the target image coverage area is obtained through the following method:

[0069] The river water vector data stored in the spaceborne storage device or uploaded to the spaceborne storage device is compared with the target image to determine the coordinates, and the river water vectors contained in the area covered by the target image are selected and recorded as the river water vectors to be processed.

[0070] The river water vector to be processed is buffered and expanded to obtain extended river water vector data covering the target image area.

[0071] The river water vector data can be point, line, or area vector data;

[0072] When the stored vector data of the detected river water body is point-like or line-like vector data, the buffer expansion includes the following steps:

[0073] Using half of the threshold width of the river water body in the attribute table of the river water body vector data to be processed as the buffer radius, the buffer of the river water body vector to be processed is expanded to obtain the first river water body buffer vector;

[0074] The buffer radius is calculated by adding 2 pixels to half the average width of the river water body in the attribute table of the river water body vector data to be processed. The buffer is then expanded to obtain the second river water body buffer vector.

[0075] The second river water buffer vector is the extended river water vector data of the target image coverage area.

[0076] When the stored vector data of the detected river water body is area vector data, the buffer expansion includes the following steps:

[0077] Using half of the threshold width of the river water body in the attribute table of the river water body vector data to be processed as the buffer radius, the buffer of the river water body vector to be processed is expanded to obtain the first river water body buffer vector;

[0078] The first river water buffer vector serves as the extended river water vector data for the target image coverage area.

[0079] The survival coefficient of the detected river water body is: extracted water area / threshold water area.

[0080] The condition of the tested river water body is determined by the following method:

[0081] (b) If the river water survival coefficient of the target river is greater than the preset survival coefficient threshold, it is marked as having a normal survival status.

[0082] (c) If the river water survival coefficient of the target river is less than or equal to the preset survival coefficient threshold, it is marked as an abnormal survival status.

[0083] The preset survival coefficient threshold can be set to different values ​​depending on the actual situation. For example, in a certain region, a threshold of 50% is considered to indicate that the water body survival is abnormal and drought is occurring if the threshold is less than 50%. The thresholds of 40% and 30% can be set according to the actual situation in different regions. In fact, they should be set according to different regions, because the abnormal water body survival situation varies from place to place. In the southern region, there is no situation where rivers are dry during the dry season, so the survival coefficient threshold can be increased.

[0084] Example 1:

[0085] This embodiment illustrates the application of a method for detecting the normalized state of linear river water bodies. The specific implementation steps are as follows:

[0086] 1. Data preparation process:

[0087] (1) Before the satellite is launched into orbit, the algorithm program corresponding to the on-orbit intelligent identification method of the river water body survival status is configured on the on-board image processor by external import; external import means storing the data and program on the computer, which can be done via USB or network cable.

[0088] (2) The linear river water vector data is imported into the storage device of the onboard image processor from an external source.

[0089] The detected river water body vector data is in shp format, a commonly used data format in the GIS field. It contains coordinate information files, attribute tables, data projection information files, and geometric feature index files of the river water body to be detected. The geometric feature index files include the order of the points that make up the line or surface, and the attribute tables contain the name, number, threshold width, and average width of the river water body.

[0090] SHP format vector data mainly consists of files such as *.shp, *.dbf, *.shx, and *.prj.

[0091] The main coordinate information is stored in *.shp files, which are used to store the geometric entities of the elements;

[0092] The attribute table information is contained in the *.dbf file. The *.dbf file is the attribute data file, which stores the attribute file data of each geometry in the format of dBaseIII+ data table.

[0093] The geometric feature index file is a *.shx file, which is an image index format file. The geometry position index records the position of each geometry within the *.shp file, enabling faster forward or backward searching of a geometry.

[0094] The data projection information file is a *.prj file. The *.prj file is a geographic projection format file used to save geographic coordinate system and projection information. It is a text file that stores a well-known text projection descriptor.

[0095] (3) During the satellite's operation in orbit, the camera payload is powered on and performs Earth imaging in pushbroom mode;

[0096] (4) When the camera payload is imaging, the on-board GNSS equipment synchronously receives the satellite position information and sends it to the FPGA on the GNSS equipment on the satellite in the form of broadcast;

[0097] (5) The FPGA completes the standardized processing of the acquired image, such as calibration, according to the preset processing flow, obtains the location information of the image coverage area, and sends it to the image processor, which then writes the location information of the image coverage area into the header file of the image file.

[0098] 2. On-board data filtering and processing:

[0099] (1) Based on the location information of the image coverage area, determine the coordinate range of the image coverage area, and combine it with the river water vector data to determine whether there is a river water body in the image. The process is as follows:

[0100] (a) Based on the location information of the image coverage area contained in the image header, two sets of coordinates of the image boundary are calculated (x upper left, y upper left) and (x lower right, y lower right), thus determining the coordinate range of the image coverage area;

[0101] (b) Search the linear river water body vector data for coordinates of the detected river water body. If there are no coordinates within the coordinate range of the target image coverage area, the image is considered not to contain the detected river. If there are, the image is considered to contain the detected river and is recorded as the target image.

[0102] To distinguish between different images, the header information of each image includes the coordinate range it covers and its index number. The index number is used to facilitate subsequent calculations to index the target image.

[0103] (c) The river water vector data stored in the on-board storage device or uploaded to the on-board storage device is compared with the target image to determine the coordinates, and the river water vectors contained in the area covered by the target image are selected and recorded as the river water vector data to be processed.

[0104] The previous step completed image filtering. This step refers to filtering out the corresponding vector data during the image filtering process, and then copying the filtered data to create new vector data.

[0105] (d) Pair the target image with the river water body vector to be processed; specifically: the image data selected in the first two steps are matched with the new vector data sequentially using index numbers.

[0106] (2) For image data containing river water bodies, calculate the Normalized Difference Water Index (NDWI) to obtain the NDWI data for the image area. The NDWI calculation formula is as follows:

[0107] NDWI=(ρGreen–ρNIR) / (ρGreen+ρNIR)

[0108] In the formula, ρGreen is the reflectance value of the green band of the image, and ρNIR is the reflectance value of the near-infrared band of the image.

[0109] (3) Set the threshold to 0.2, and perform water extraction pixel by pixel according to the following logic formula. Water extraction refers to the calculation process of obtaining the water area in the image by calculating the image pixel values.

[0110] (a) Extraction is performed according to the following logical expression:

[0111] NDWI > 0.2 indicates water bodies, while NDWI <= 0.2 indicates non-water bodies.

[0112] (b) The pixel value of the water pixel is assigned a value of 1;

[0113] (c) Non-water body pixels are marked as 0;

[0114] (4) Obtain the water extraction result map of the image area;

[0115] Obviously, the method for extracting the detected river water information from the target image coverage area and drawing the water extraction result map of the target image coverage area is as follows:

[0116] If the spectral index of a target image pixel that is related to water extraction is greater than a preset threshold, the pixel is considered to be located in a water body; otherwise, the pixel is considered to be located in a non-water body.

[0117] The pixel values ​​of water pixels are assigned a value of 1, and non-water pixels are marked as 0, resulting in a water extraction result image of the target image coverage area. This water extraction result image of the target image coverage area is a binary image.

[0118] 3. Determination and transmission of river water body existence status

[0119] (1) Perform two buffer expansions on the vector of the river water body to be processed: Buffer expansion is a GIS professional term, which means expanding a point or line outward.

[0120] (a) The first buffer expansion is: using half of the threshold width of the river water body in the attribute table of the river water body vector data to be processed as the buffer radius, the river water body vector to be processed is expanded to obtain the first river water body buffer vector;

[0121] (b) The second buffer expansion is as follows: the average width of the river water body in the attribute table of the river water body vector data to be processed is 1 / 2 plus 2 pixels as the buffer radius, and the buffer expansion is performed on the river water body vector to be processed to obtain the second river water body buffer vector; the pixel width refers to the image resolution size;

[0122] (c) Calculate the vector area of ​​the first river water buffer zone, and update the corresponding record in the attribute table of the second river water buffer zone vector as the "threshold water area" field;

[0123] (2) Overlay the second river water buffer vector with the water extraction result map and perform zone statistics. Update the attribute table of the second river water buffer vector as the "extracted water area" field.

[0124] (3) Analyze the attribute table of the second river water buffer zone vector:

[0125] (a) Calculate the extracted water area / threshold water area and update the attribute table of the second river water buffer vector as the "river water survival coefficient" field;

[0126] (b) If the river water survival coefficient of the target river body is >80%, it is marked as having a normal survival status;

[0127] (c) If the river water survival coefficient of the target river body is <= 80%, it is marked as an abnormal survival status;

[0128] (d) Update the attribute table of the second river water body buffer vector with the labeling result of the survival status as the "survival status" field;

[0129] (4) Compress and encode the attribute table of the second river water buffer vector to obtain the data to be transmitted;

[0130] (5) The data to be transmitted is transmitted via the satellite-to-ground antenna.

[0131] Example 2:

[0132] For important river water body data, linear and width-based data cannot adequately represent river water bodies, such as reservoirs. Therefore, for this type of river water body vector data, planar river water body vector data is used. Since planar river water body vector data already describes the river water body area well, only one buffer expansion was performed in Example 2, which is different from the buffer expansion performed in Example 1 for obtaining the river water body area.

[0133] This embodiment illustrates the application of a method for detecting abnormal water conditions in important river bodies. The specific implementation steps are as follows:

[0134] 1. Data preparation process:

[0135] (1) Before the satellite is launched into orbit, the algorithm program corresponding to the on-orbit intelligent discrimination method of river water status is configured on the on-board image processor by external import.

[0136] (2) The vector data of the river body in the area drawn by combining historical images is stored in the storage device of the on-board image processor by external import;

[0137] (3) During the satellite's operation in orbit, the camera payload is powered on and performs Earth imaging in push-broom mode;

[0138] (4) When the camera payload is imaging, the on-board GNSS equipment synchronously receives satellite position information and sends it to the FPGA via broadcast;

[0139] (5) The FPGA completes the standardized processing of the acquired image, such as calibration, according to the preset processing flow, obtains the location information of the image coverage area, and sends it to the image processor, which then writes the location information of the image coverage area into the header file of the image file.

[0140] 2. On-board data filtering and processing:

[0141] (1) Based on the image location information and combined with the stored river water vector data, determine whether there is a river water body in the image. The process is as follows:

[0142] (a) Based on the location information of the image coverage area contained in the image header, two sets of coordinates of the image boundary are calculated (x upper left, y upper left) and (x lower right, y lower right), thus determining the coordinate range of the image coverage area. Alternatively, the location information of the four boundaries of the image (north, south, east, west) can be used to determine the coordinate range of the image coverage area.

[0143] (b) Search the vector data of the river water body in the area to be detected for coordinates within the coordinate range of the target image coverage area. If there are no coordinates, the image is considered to contain the detected river. If there are coordinates, the image is considered to contain the detected river and the image is recorded as the target image.

[0144] The header information of each image includes the coordinate range it covers and its index number. The index number is used to facilitate subsequent calculations to index the target image.

[0145] (c) Use all river water surface vectors that exist in the image to be processed to create the river water vector data to be processed;

[0146] (d) Pair the target image with the vector data of the river water body to be processed;

[0147] (2) For image data containing rivers, calculate the Normalized Difference Vegetation Index (NDVI) to obtain the NDVI data for the image area. The NDVI calculation formula is as follows:

[0148] NDVI=(ρNIR–ρRed) / (ρNIR+ρRed)

[0149] In the formula, ρNIR is the near-infrared reflectance value of the image, and ρRed is the red reflectance value of the image.

[0150] (3) Set the threshold to 0, and extract water body pixel by pixel according to the following logic:

[0151] (a) Extraction is performed according to the following logical expression:

[0152] NDVI < 0 indicates a water body, NDVI > = 0 indicates a non-water body.

[0153] (b) The pixel value of the water pixel is assigned a value of 1;

[0154] (c) Non-water body pixels are marked as 0;

[0155] (4) Obtain the water extraction result map of the image area;

[0156] 3. Determination and transmission of river water body existence status

[0157] (1) Perform buffer expansion on the vector data of the river water body to be processed:

[0158] (a) Perform buffer analysis with a buffer radius of 10m to obtain the buffer vector of the first river water body;

[0159] (b) Calculate the area of ​​the areal vector in the vector data of the river water body to be processed, and update the corresponding record in the vector attribute table of the water body buffer as the "threshold water area" field;

[0160] (2) Overlay the first river water buffer vector with the water extraction result map and perform zone statistics. The result is updated as the "extracted water area" field in the attribute table of the first river water buffer vector. The zone statistics count the number of water pixels in the area covered by each vector graphic, and the area can be calculated from the number of pixels and the image resolution.

[0161] (3) Analyze the attribute table of the first river water buffer zone vector:

[0162] (a) Calculate the extracted water area / threshold water area and update the attribute table of the first river water buffer vector as the "river water survival coefficient" field;

[0163] (b) If the river water survival coefficient of the target river body is >70%, it is marked as having a normal survival status;

[0164] (c) If the river water survival coefficient of the target river body is <= 70%, it is marked as an abnormal survival status;

[0165] (d) Update the attribute table of the first river water body buffer vector with the labeling result of the survival status as the "survival status" field;

[0166] (4) Compress and encode the attribute table of the first river water buffer vector to obtain the data to be transmitted;

[0167] (5) The data to be transmitted is transmitted via the satellite-to-ground antenna.

[0168] Example 3

[0169] Compared with Examples 1 and 2, this example is aimed at updating river water vector data after the satellite is in orbit, for additional river water data not in the pre-stored data. The data can be sent from the ground station to the satellite through a high-speed uploading channel. The use of sampled point data can further reduce the file size and make it easier to update the vector data through uploading.

[0170] This embodiment illustrates the application of a method for detecting the survival status of sampled water bodies in a specific river. The specific implementation steps are as follows:

[0171] 1. Data preparation process:

[0172] (1) Before the satellite is launched into orbit, the algorithm program corresponding to the on-orbit intelligent discrimination method of river water status is configured on the on-board image processor by external import.

[0173] (2) The point-like river water vector data formed by sampling is stored in the storage device of the spaceborne image processor through data uploading. In addition to the river water number information, the vector data attribute table must contain at least the threshold width information.

[0174] (3) During the satellite's operation in orbit, the camera payload is powered on and performs Earth imaging in push-broom mode;

[0175] (4) When the camera payload is imaging, the on-board GNSS equipment synchronously receives satellite position information and sends it to the FPGA via broadcast;

[0176] (5) The FPGA completes the standardized processing of the acquired image, such as calibration, according to the preset processing flow, obtains the location information of the image coverage area, and sends it to the image processor, which then writes the location information of the image coverage area into the header file of the image file.

[0177] 2. On-board data filtering and processing:

[0178] (1) Based on the location information of the image coverage area, determine the coordinate range of the image coverage area, and combine it with the river water vector data to determine whether there is a river water body in the image. The process is as follows:

[0179] (a) Based on the location information contained in the image header, two sets of coordinates of the image boundary are calculated (x upper left, y upper left) and (x lower right, y lower right), thus determining the coordinate range of the image coverage area;

[0180] (b) Search the point-like river water body vector data for coordinates of the detected river water body to see if there are coordinates within the coordinate range of the target image coverage area. If not, the image is considered not to contain the detected river. If there are, the image is considered to contain the detected river and is recorded as the target image.

[0181] (c) The river water vector data stored in the on-board storage device or uploaded to the on-board storage device is compared with the target image to determine the coordinates, and the river water vectors contained in the area covered by the target image are selected and recorded as the river water vector data to be processed.

[0182] (c) Pair the target image with the river water vector to be processed;

[0183] (2) For image data containing river water bodies, calculate the Normalized Difference Water Index (MNDWI) to obtain the MNDWI data for the image area. The MNDWI calculation formula is as follows:

[0184] MNDWI=(ρGreen–ρMIR) / (ρGreen+ρMIR)

[0185] In the formula, ρGreen is the reflectance value of the green band of the image, and ρMIR is the reflectance value of the infrared band of the image.

[0186] (3) Set the threshold to 0, and extract water body pixel by pixel according to the following logic:

[0187] (a) Extraction is performed according to the following logical expression:

[0188] MNDWI > 0 indicates water bodies, and MNDWI <= 0 indicates non-water bodies.

[0189] (b) The pixel value of the water pixel is assigned a value of 1;

[0190] (c) Non-water body pixels are marked as 0;

[0191] (4) Obtain the water extraction result map of the image area;

[0192] 3. Determination and transmission of river water body existence status

[0193] (1) Double buffer expansion of the river water vector data to be processed:

[0194] (a) First buffer expansion: The buffer radius is calculated by using half of the threshold width of the river water body in the attribute table as the buffer radius to obtain the first river water body buffer vector;

[0195] (b) Second buffer expansion: The second river water body buffer vector is obtained by using half the average width of the river water body plus 2 pixels as the buffer radius;

[0196] (c) Calculate the vector area of ​​the first river water buffer zone, and update the corresponding record in the attribute table of the second river water buffer zone vector as the "threshold water area" field;

[0197] (2) Overlay the second river water buffer vector with the water extraction result map and perform zone statistics. Update the attribute table of the second river water buffer vector as the "extracted water area" field.

[0198] (3) Analyze the attribute table of the second river water buffer zone vector:

[0199] (a) Calculate the extracted water area / threshold water area and update the attribute table of the second river water buffer vector as the "river water survival coefficient" field;

[0200] (b) If the river water survival coefficient of the target river body is >50%, it is marked as having a normal survival status;

[0201] (c) If the river water survival coefficient of the target river body is <= 50%, it is marked as an abnormal survival status;

[0202] (d) Update the attribute table of the second river water body buffer vector with the labeling result of the survival status as the "survival status" field;

[0203] (4) Compress and encode the attribute table of the second river water buffer vector to obtain the data to be transmitted;

[0204] (5) The data to be transmitted is transmitted via the satellite-to-ground antenna.

[0205] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. An on-orbit intelligent method for determining the survival status of river water bodies, characterized in that... Includes the following steps: The detected river water vector data is stored in the on-board image processor device; the detected river water vector data is in shp format. The onboard camera payload images the Earth and obtains the location information of the image coverage area based on the satellite position information acquired in real time on the satellite. Based on the image containing location information and the vector data of the detected river water body, the image containing the river water body is selected, i.e., the target image. The vector data of the river water body in the area covered by the target image is extracted and buffered and expanded to obtain the expanded river water body vector data. The specific steps for selecting images containing river water bodies are as follows: Based on the location information of the image coverage area, determine the coordinate range of the image coverage area; specifically: During satellite operation, the camera payload is powered on and performs Earth imaging in pushbroom mode; When the camera payload is imaging, the onboard GNSS equipment synchronously receives satellite position information and transmits it to the FPGA on the GNSS equipment on the satellite via broadcast. The FPGA completes the calibration process of the acquired image according to the preset processing flow, obtains the location information of the image coverage area, and sends it to the image processor, which then writes the location information of the image coverage area into the header file of the image file. Based on the location information of the image coverage area contained in the image header, two sets of coordinates of the image boundary are calculated (x upper left, y upper left) and (x lower right, y lower right), thus determining the coordinate range of the image coverage area; If the coordinate information of the detected river body in the vector data is found to contain coordinates within the coordinate range of the target image coverage area, the image is considered not to contain the detected river if no coordinates are found; otherwise, the image is considered to contain the detected river. Calculate the spectral index related to water body extraction in the target image, apply a threshold to the spectral index, extract the detected river water information in the target image coverage area, and draw the water body extraction result map of the target image coverage area. The water body extraction result map of the target image coverage area is compared with the extended river water body vector data by partitioning and statistically analyzing the water body area of ​​the target image coverage area; Based on the detected water body area covered by the target image, calculate the water body survival coefficient of the detected river, determine the survival status of the detected river water body using the water body survival coefficient, and record it. The survival coefficient of the detected river water body is: extracted water area / threshold water area.

2. The on-orbit intelligent determination method for the survival status of river water bodies according to claim 1, characterized in that... The vector data of the detected river water body includes a coordinate information file, an attribute table, a data projection information file, and a geometric feature index file. The geometric feature index file includes the order of the points that make up the line or surface, and the attribute table includes the name, number, threshold width, and average width of the river water body.

3. The on-orbit intelligent determination method for the survival status of river water bodies according to claim 1, characterized in that... The method for extracting the detected river water information from the target image coverage area and plotting the water extraction result map of the target image coverage area is as follows: If the spectral index of a target image pixel that is related to water extraction is greater than a preset threshold, the pixel is considered to be located in a water body; otherwise, the pixel is considered to be located in a non-water body. The pixel value of water body pixels is assigned to 1, and non-water body pixels are marked as 0, resulting in the water extraction result map of the target image coverage area.

4. The on-orbit intelligent determination method for the survival status of river water bodies according to claim 1, characterized in that... The extended river water vector data of the target image coverage area is obtained through the following method: The river water vector data stored in the spaceborne storage device or uploaded to the spaceborne storage device is compared with the target image to determine the coordinates, and the river water vectors contained in the area covered by the target image are selected and recorded as the river water vectors to be processed. The river water vector to be processed is buffered and expanded to obtain extended river water vector data covering the target image area.

5. The on-orbit intelligent determination method for the survival status of river water bodies according to claim 1, characterized in that... The river water vector data can be point, line, or area vector data.

6. The on-orbit intelligent determination method for the survival status of river water bodies according to claim 4, characterized in that... When the stored vector data of the detected river water body is point-like or line-like vector data, the buffer expansion includes the following steps: Using half of the threshold width of the river water body in the attribute table of the river water body vector data to be processed as the buffer radius, the buffer of the river water body vector to be processed is expanded to obtain the first river water body buffer vector; The buffer radius is calculated by adding 2 pixels to half the average width of the river water body in the attribute table of the river water body vector data to be processed. The buffer is then expanded to obtain the second river water body buffer vector. The second river water buffer vector is the extended river water vector data of the target image coverage area.

7. The on-orbit intelligent determination method for the survival status of river water bodies according to claim 4, characterized in that... When the stored vector data of the detected river water body is area vector data, the buffer expansion includes the following steps: Using half of the threshold width of the river water body in the attribute table of the river water body vector data to be processed as the buffer radius, the buffer of the river water body vector to be processed is expanded to obtain the first river water body buffer vector; The first river water buffer vector serves as the extended river water vector data for the target image coverage area.

8. The on-orbit intelligent determination method for the survival status of river water bodies according to claim 1, characterized in that... The sustainability of the tested river water body was determined using the following method: (a) If the river water survival coefficient of the target river is greater than the preset survival coefficient threshold, it is marked as having a normal survival status; (b) If the river water survival coefficient of the target river is less than or equal to the preset survival coefficient threshold, it is marked as an abnormal survival status.

Images