System and method for monitoring, drawing and evaluating small and micro water body

Abstract

The invention discloses a small and micro water body monitoring drawing evaluation system and method, and belongs to the technical field of water body monitoring, and the method comprises the steps: obtaining an initial image group of a to-be-drawn region, sequentially splicing all initial images in the initial image group, and generating a region image of the to-be-drawn region; summarizing all the point location areas, generating water body point locations of the to-be-drawn area, and setting the water body point locations of which the actual areas are located in a preset area interval as small and micro water body point locations; summarizing directory information of all small and micro water body point locations in the to-be-drawn area based on a preset label, inputting all supplementary images contained in the supplementary image set into the image processing model for learning and training, outputting pollution level prediction of the small and micro water body point locations within prediction time, and writing the pollution level prediction into the directory information; and generating a troubleshooting route based on the water quality pollution degree and the position coordinates, and sending the troubleshooting route to a manager. According to the invention, the integrity and accuracy of monitoring and evaluating the point location of the small and micro water body can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of water body monitoring, and in particular relates to a system and method for monitoring, mapping and evaluating small and micro water bodies. Background Art

[0002] With the acceleration of urbanization, the protection and management of small and micro water bodies has become an important part of water environment management. Small and micro water bodies include but are not limited to ditches, ponds, fish ponds, etc. Their health is directly related to the regional ecological environment and the quality of life of residents. The existing monitoring methods are inefficient and costly, and it is difficult to meet the current monitoring needs. It is necessary to improve the monitoring efficiency and data accuracy to provide a scientific basis for the protection and management of small and micro water bodies.

[0003] Similar prior art includes a Chinese patent application with publication number CN116597355A, which discloses a real-time monitoring method, system and device for micro-water body status based on a visual system, which is applied to the field of water environment management technology; obtaining micro-water body monitoring data; comparing the current data with a preset threshold value and outputting the water body status; using the current data, historical data and water body status as data sets to train the water body status prediction model, obtain a calibrated water body status prediction model, obtain the water body status at a future moment according to the calibrated water body status prediction model, make a warning level judgment based on the water body status at a future moment, and feedback to the management personnel. The invention obtains the current water body status through images such as floating objects and water body color, smell, water quality, etc.; at the same time, combines historical data with the current water body status to construct a water body status prediction model, predicts the water body status at a future moment, obtains the water body status at a future moment, makes a warning judgment based on the water body status level at a future moment, and feeds back to the management personnel for warning. There is also a Chinese patent application with publication number CN117078490A, which discloses a risk assessment method for urban micro-water bodies based on synchronous analysis of multiple factors, including: collecting direct water body data and indirect impact data of the target water body, respectively recorded as Class A data and Class B data, wherein Class A data contains time series data sequences of several direct indicators, and Class B data contains time series data sequences of several indirect indicators, obtaining the information richness of each time series data sequence, recording the time series data sequence of any direct indicator as the target direct data sequence, and recording the time series data sequence of any indirect indicator as the target indirect data sequence, obtaining the correlation between the target direct data sequence and the target indirect data sequence, thereby obtaining the correlation between the target indirect data sequence and Class A data, and thus obtaining the risk assessment value of the target water body. The invention adaptively reduces the dimension parameters and improves the accuracy of the risk assessment value of urban micro-water bodies.

[0004] However, the above-mentioned prior art does not check the specific coordinates of small and micro water bodies, which may lead to monitoring omissions. The use of multiple types of data to conduct risk assessment on small and micro water bodies makes the data collection method too complicated and requires higher equipment costs. Summary of the invention

[0005] To solve the above problems, the present invention provides a small water body monitoring, mapping and evaluation system and method to solve the problems in the prior art.

[0006] In order to achieve the above-mentioned invention object, the present invention proposes a small water body monitoring, mapping and evaluation method, comprising: S1: using a drone to obtain an initial image group of the area to be drawn based on a preset shooting trajectory and shooting method, stitching all initial images in the initial image group in sequence, and generating a regional image of the area to be drawn; S2: constructing an extraction model, extracting the point area of ​​the initial image based on the extraction model, calculating the measured area of ​​the point area based on the shooting method, marking the position coordinates of all the point areas in the regional image, summarizing all the point areas, generating water body points of the area to be mapped, and obtaining the actual area of ​​the water body point based on the measured area, and setting the water body point whose actual area is within the preset area interval as a small water body point; S3: Based on the preset labels, the directory information of all the small and micro water body points in the area to be mapped is summarized. The drone obtains the supplementary image set corresponding to the small and micro water body points based on the preset time period, builds an image processing model, inputs all the supplementary images contained in the supplementary image set into the image processing model for learning and training, and outputs the pollution level prediction of the small and micro water body points within the prediction time, and writes the pollution level prediction into the directory information; S4: Mark the water pollution levels of all the small water bodies based on the list information, generate a screening route based on the water pollution level and the location coordinates, send the screening route to management personnel, and complete the monitoring and evaluation of the small water bodies in the area to be mapped.

[0007] Furthermore, the method of using a drone to obtain an initial image group of the area to be mapped based on a preset shooting trajectory and shooting method includes the following steps: The GIS coordinates of the area to be drawn are obtained, and an area frame including the area to be drawn is constructed based on the GIS coordinates. The shooting range, flight altitude, flight speed and shooting parameters of the drone are set, and the area frame is divided into a plurality of grid points based on the shooting range. The drone takes the grid point corresponding to the vertex of the area frame as the starting point and performs flight shooting at any grid point in sequence in an S-shaped route to generate a plurality of initial images, and all the initial images are combined to generate the initial image group, wherein the S-shaped route is set as the shooting trajectory, the flight altitude, the flight speed and the shooting parameter combination are set as the shooting mode, and the initial images corresponding to any two adjacent grid points have an overlapping area.

[0008] Furthermore, a region image of the region to be drawn is generated based on the following steps: defining the initial images corresponding to the two grid points adjacent to the starting point as comparison images, extracting feature points of the comparison images respectively based on a feature extraction algorithm, performing feature point matching on the feature points of the two comparison images to extract feature point pairs, setting any feature point pair as a key point, and setting the maximum pixel area included in all the key points as the superposition area; Calculating a transformation matrix of the key point between the two contrast images based on the least square method, calculating predicted positions of the remaining key points in the two contrast images based on the transformation matrix, calculating residuals of all the key points based on the predicted positions and the actual positions of the remaining key points, adjusting parameters of the transformation matrix until the sum of all the residuals is minimized, and setting the adjusted transformation matrix as the image transformation matrix; The two contrast images are stitched together based on the image transformation matrix to generate a stitched image, and the stitched image is stitched together with the initial image corresponding to the next grid point based on this step to generate a new stitched image, and this step is repeated until the initial images corresponding to all the grid points are stitched together, and the last new stitched image is set as the regional image.

[0009] Furthermore, extracting the point area in the initial image based on the extraction model includes the following steps: The extraction model performs a deblurring process on the initial image based on a deblurring algorithm to generate a clear image, obtains the reflectivity ratio of any pixel in the clear image between a first preset band and a second preset band, and combines the reflectivity ratios corresponding to all pixels to generate a reflectivity ratio distribution image; Constructing a coordinate system, setting the numerical value corresponding to the reflectivity ratio in the reflectivity ratio distribution image as the horizontal coordinate, setting the number of pixels included in any reflectivity ratio as the vertical coordinate, and fitting and drawing in the coordinate system based on the horizontal coordinate and the vertical coordinate to generate a first reflectivity distribution curve; In the first reflectivity distribution curve, two adjacent inflection points are extracted, and the difference between the ordinates corresponding to the two inflection points is calculated, the two inflection points with the largest difference are set as dividing points, the reflectivity ratio corresponding to the minimum number of pixels in the dividing points is set as a first threshold, and the pixels in the reflectivity ratio distribution image whose reflectivity ratio is less than or equal to the first threshold are set as sub-regions; Based on this step, a second threshold corresponding to the reflectance ratio distribution image in the first preset band and the second preset band is generated, and the second threshold is greater than the first threshold, and the pixels in the sub-area whose new reflectance ratio is less than or equal to the second threshold are set as the point area.

[0010] Furthermore, obtaining the actual area of ​​the water body point based on the measured area comprises the following steps: Calculating the number of pixels contained in the point area and the pixel area of ​​any pixel based on the shooting method, and setting the product of the number of pixels corresponding to the point area and the pixel area as the measurement area; In the regional image, if there are adjacent pixels between the point regions included in any two initial images, the point regions in the two initial images are merged and set as the water body points; otherwise, the point regions included in the two initial images are respectively set as the water body points; The measured areas of the point regions corresponding to the water body points are accumulated to generate the actual area.

[0011] Furthermore, the process of summarizing the list information of all the small and micro water body points in the area to be mapped based on preset labels includes the following steps: The preset label includes a location area, an area size and a point ID, and the grid position of the grid point corresponding to the initial image in the area frame is obtained. The point position of the small and micro water body point in the area to be drawn is generated based on the grid position and the location coordinate positioning, and the point position is set as the information of the location area, and the actual area corresponding to the small and micro water body point is set as the information of the area size. The small and micro water body points are marked with serial numbers and set as the point ID information, and the information of all the small and micro water body points is summarized based on the preset label to generate the directory information.

[0012] Further, the pollution level prediction of the small water body point within the prediction time is output based on the following steps: The preset time period includes a plurality of time points, and the supplementary images corresponding to all the time points are combined and set as the supplementary image set; The image processing model extracts spectral data of all the supplementary images, clusters all the spectral data to generate a preset number of clusters, extracts the best band in each cluster based on a feature selection algorithm, obtains the spectral data belonging to the best band in the cluster, and sets it as a training set, inputs all the training sets into the image processing model for training and learning, and then generates a prediction model, the prediction model outputs a concentration prediction result within the prediction time based on the training set, and the concentration prediction result is set as the pollution level prediction.

[0013] Furthermore, the investigation route is generated based on the following steps: Setting a pollution concentration table, calculating the mean value of water quality distribution based on the actual area according to the pollution level prediction, obtaining the pollution level corresponding to the mean value of water quality distribution in the pollution concentration table, and setting the pollution level as the water quality pollution degree; In the directory information, all the small water body points are arranged in order from high to low based on the water pollution degree to generate a screening number, and based on the screening number, the location coordinates of the small water body points are connected in order to generate the screening route.

[0014] The present invention also provides a small water body monitoring, mapping and evaluation system, which is used to implement the small water body monitoring, mapping and evaluation method described above, and the system mainly includes: A drawing module, using a drone to obtain an initial image group of the area to be drawn based on a preset shooting trajectory and shooting method, sequentially splicing all initial images in the initial image group, and generating a regional image of the area to be drawn; An extraction module, constructing an extraction model, extracting the point area of ​​the initial image based on the extraction model, calculating the measured area of ​​the point area based on the shooting method, marking the position coordinates of all the point areas in the area image, summarizing all the point areas, generating water body points of the area to be mapped, and obtaining the actual area of ​​the water body point based on the measured area, and setting the water body point whose actual area is within the preset area interval as a small water body point; The analysis module summarizes the directory information of all the small and micro water body points in the area to be mapped based on preset labels, and the drone obtains the supplementary image set corresponding to the small and micro water body points based on a preset time period, builds an image processing model, inputs all the supplementary images contained in the supplementary image set into the image processing model for learning and training, and outputs the pollution level prediction of the small and micro water body points within the prediction time, and writes the pollution level prediction into the directory information; The monitoring module marks the water pollution levels of all the small water bodies based on the list information, generates an investigation route based on the water pollution level and the location coordinates, and sends the investigation route to the management personnel to complete the monitoring and evaluation of the small water bodies in the area to be mapped.

[0015] Compared with the prior art, the beneficial effects of the present invention are at least as follows: The present invention first generates a regional image of the area to be drawn through an initial image group, which can improve the integrity of the area to be drawn, and then accurately obtains the actual area and position coordinates of the point area through an extraction model, which can accurately obtain the directory information of small and micro water body points and improve the integrity of the directory information. Finally, the water quality of small and micro water body points is predicted and the pollution level prediction is output through an image processing model, thereby improving the accuracy of monitoring small and micro water body points.

[0016] The present invention also constructs an inspection route for all small and micro water body points through the directory information, which can avoid missing the monitoring and evaluation of small and micro water body points. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flowchart of the steps of a method for monitoring and drawing an evaluation method of a small water body according to the present invention; Figure 2 A schematic diagram of the shooting trajectory of the drone in the present invention; Figure 3 This is a structural diagram of a small water body monitoring, drawing and evaluation system of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0019] It is understood that the terms "first", "second", etc. used in this application may be used herein to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, without departing from the scope of this application, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script.

[0020] like Figure 1 As shown, a small water body monitoring, mapping and evaluation method includes: S1: using a drone to obtain an initial image group of the area to be drawn based on a preset shooting trajectory and shooting method, stitching all initial images in the initial image group in sequence, and generating a regional image of the area to be drawn.

[0021] Specifically, the present invention uses a spectral imaging instrument carried by an unmanned aerial vehicle to collect images of the area to be mapped, and realizes the monitoring, mapping and evaluation of small water bodies through an image analysis process. In this embodiment, the shooting trajectory refers to the flight trajectory of the unmanned aerial vehicle, and the shooting method refers to the shooting parameters of the spectral imaging instrument carried by the unmanned aerial vehicle, wherein the shooting method is related to the flight altitude and flight speed of the unmanned aerial vehicle. The area to be mapped refers to the entire area where water body monitoring is required, and the initial image group is composed of a plurality of initial images, and the initial image refers to the spectral image generated by the spectral imaging instrument, wherein the initial image mainly includes the image generated in the process of shooting the area to be mapped. Since the area to be mapped is a large area, the initial images can be stitched in sequence to generate an image containing the complete area to be mapped, that is, the regional image.

[0022] S2: Construct an extraction model, extract the point area of ​​the initial image based on the extraction model, calculate the measured area of ​​the point area based on the shooting method, mark the position coordinates of all point areas in the regional image, summarize all point areas, generate water body points in the area to be drawn, and obtain the actual area of ​​the water body points based on the measured area, and set the water body points whose actual areas are within the preset area range as small water body points.

[0023] Specifically, in this embodiment, the extraction model refers to a model for image feature extraction, and the point area can be identified by feature extraction. The point area refers to an area that may be water. The pixel area of ​​any pixel in the initial image can be calculated by the flight altitude and shooting parameters contained in the shooting method, and the measured area of ​​the point area can be calculated by the number of pixels corresponding to the point area. The position coordinates of the point area can be generated by marking the pixel coordinate points of the point area in the regional image. The water body point refers to all the point areas summarized in the area to be drawn. Since the same point area may be shot and generated in different initial images, it is necessary to check some suspected repeated point areas and set them as water body points. The actual area of ​​the water body point can be calculated by the measured area. For example, the same point area B is shot and generated in the initial images A1 and A2, wherein the measured area of ​​the point area B in the initial image A1 is C1, and the measured area of ​​the point area B in the initial image A2 is C2. It is necessary to calculate the actual area C of the point area B by comparing the number of pixels of the point area B in the regional image. The preset area interval refers to the area of ​​the limited micro-water body. For example, the preset area interval is , then the water body points with actual area less than or equal to D are set as small water body points.

[0024] S3: Based on the preset labels, the directory information of all small and micro water body points in the area to be mapped is summarized. The UAV obtains the supplementary image set corresponding to the small and micro water body points based on the preset time period, builds an image processing model, inputs all the supplementary images contained in the supplementary image set into the image processing model for learning and training, and outputs the pollution level prediction of the small and micro water body points within the prediction time, and writes the pollution level prediction into the directory information.

[0025] Specifically, in this embodiment, the preset label refers to various label types, and the database of all small and micro water body points is summarized by the preset label, which is the directory information. The supplementary image set is composed of multiple supplementary images, wherein the preset time period refers to a set time period, and the drone realizes fixed-point shooting within the preset time period. The supplementary image is a spectral image generated by shooting the small and micro water body points by the spectral imaging instrument carried by the drone. Since the position coordinates of the small and micro water body points have been obtained, the supplementary image is an accurate shot of the small and micro water body points, so that the imaging of the supplementary image completely includes the small and micro water body points. The image processing model refers to a machine learning model, which can train and learn the supplementary image set, and predict the water quality changes of the small and micro water body points within the prediction time through the change process, generate a pollution level prediction, and then summarize the pollution level prediction and write it into the directory information for convenient statistical analysis.

[0026] According to this scheme, the accuracy of water quality monitoring of small water bodies can be improved.

[0027] S4: Mark the water pollution levels of all small and micro water bodies based on the list information, generate inspection routes based on the water pollution levels and location coordinates, send the inspection routes to management personnel, and complete the monitoring and evaluation of small and micro water bodies in the area to be mapped.

[0028] Specifically, in this embodiment, the water pollution degree refers to the pollution index, and the water pollution degree can be marked by the pollution level prediction. The inspection route refers to the inspection planning route for manual monitoring of small and micro water body points. By setting the inspection route, the monitoring efficiency of the management personnel can be improved, and the monitoring and evaluation of small and micro water body points can be completed quickly.

[0029] As a preferred technical solution of the present invention, using a drone to obtain an initial image group of the area to be mapped based on a preset shooting trajectory and shooting method includes the following steps: The GIS coordinates of the area to be drawn are obtained, and an area frame including the area to be drawn is constructed based on the GIS coordinates. The shooting range, flight altitude, flight speed and shooting parameters of the drone are set. The area frame is divided into multiple grid points based on the shooting range. The drone takes the grid point corresponding to the vertex of the area frame as the starting point and performs flight shooting at any grid point in sequence in an S-shaped route to generate multiple initial images. All the initial images are combined to generate an initial image group, wherein the S-shaped route is set as the shooting trajectory, the flight altitude, flight speed and shooting parameter combination are set as the shooting mode, and the initial images corresponding to any two adjacent grid points have an overlapping area.

[0030] Specifically, in this embodiment, GIS coordinates, i.e. Geographic Information System (Geographic Information System) coordinates, refer to coordinates used to locate geographic elements such as points, lines, and surfaces in geographic space, including but not limited to latitude and longitude coordinates, plane coordinates, elevation coordinates, etc. The area frame can be set as a rectangular frame containing the area to be drawn according to the GIS coordinates. The shooting range refers to the size of the shooting area of ​​the spectral imaging instrument carried by the drone. The flight altitude refers to the vertical height between the drone and the shooting ground, i.e. the shooting altitude. The flight speed refers to the speed at which the drone flies between grid points. The flight speed can be used to set the time for the drone to shoot above the grid points. The shooting parameters refer to the imaging parameters of the spectral imaging instrument, including but not limited to resolution, etc. The shooting trajectory of the drone is as follows: Figure 2 As shown, the area frame M2 corresponding to the area to be drawn M1 is divided into multiple grid points, wherein the drone flies and shoots from the grid point p1 as the starting point to the adjacent grid points in the direction of the arrow →, thereby forming a flight trajectory corresponding to the S-shaped route, which is set as the shooting trajectory. This shooting trajectory enables the initial images corresponding to any two adjacent grid points to have overlapping areas, which is conducive to the accuracy of drawing the regional image of the area to be drawn.

[0031] As a preferred technical solution of the present invention, a regional image of the area to be drawn is generated based on the following steps: The initial images corresponding to the two grid points adjacent to the starting point are defined as contrast images. The feature points of the contrast images are extracted respectively based on the feature extraction algorithm. The feature points of the two contrast images are matched to extract feature point pairs. Any feature point pair is set as a key point, and the maximum pixel area contained in all key points is set as the superposition area.

[0032] The transformation matrix of the key points between the two contrast images is calculated based on the least squares method, the predicted positions of the remaining key points in the two contrast images are calculated based on the transformation matrix, the residuals of all key points are calculated based on the predicted positions and the actual positions of the remaining key points, the parameters of the transformation matrix are adjusted until the sum of all residuals is minimized, and the adjusted transformation matrix is ​​set as the image transformation matrix.

[0033] The two contrast images are stitched together based on the image transformation matrix to generate a stitched image, and the stitched image is stitched together with the initial image corresponding to the next grid point based on this step to generate a new stitched image. This step is repeated until the initial images corresponding to all grid points are stitched together, and the last new stitched image is set as the regional image.

[0034] Specifically, in this embodiment, the contrast image refers to the initial image corresponding to the grid points adjacent to the starting point, and the feature extraction algorithm refers to the SIFT (Scale-Invariant Feature Transform) algorithm, which can detect and describe local features in the image. These features are invariant to image transformations such as rotation, scaling, and brightness changes, thereby extracting feature points and matching them. A feature point pair refers to the matching feature points in two contrast images, that is, there is a corresponding transformation relationship between the feature point pairs. The maximum pixel area refers to the pixel range contained by all corresponding key points in the two contrast images, and the pixel range belongs to the same photographed area in the two contrast images.

[0035] The least squares method is a mathematical optimization technique that can find the homography matrix between two contrasting images, that is, the transformation matrix. The difference between the actual position of any key point and the predicted position rotated by the transformation matrix is ​​any residual. For example, the actual position of the key point e1 in the contrasting image E1 is (x1, y1), and the predicted position of the key point e1 in the contrasting image E1 generated by the transformation matrix is ​​(x2, y2). The residual H1 between (x1, y1) and (x2, y2) can be calculated by the first formula, where the first formula is: The residuals corresponding to all key points are accumulated to calculate the sum of the residuals. The parameters of the transformation matrix are adjusted to minimize the sum of the residuals, thereby generating the image transformation matrix of the two contrast images.

[0036] The image transformation matrix can be used to stitch two contrasting images through rotation combination to generate a stitched image, and then stitch the above steps with the initial image corresponding to the next adjacent grid point to generate a new stitched image. Repeat this step to set the last new stitched image as the regional image of the area to be drawn.

[0037] As a preferred technical solution of the present invention, extracting the point area in the initial image based on the extraction model includes the following steps: The extraction model deblurs the initial image based on the deblurring algorithm to generate a clear image, obtains the reflectivity ratio of any pixel in the clear image between the first preset band and the second preset band, and combines the reflectivity ratios corresponding to all pixels to generate a reflectivity ratio distribution image.

[0038] A coordinate system is constructed, the numerical value corresponding to the reflectivity ratio in the reflectivity ratio distribution image is set as the horizontal coordinate, the number of pixels contained in any reflectivity ratio is set as the vertical coordinate, and the first reflectivity distribution curve is generated by fitting and drawing in the coordinate system based on the horizontal coordinate and the vertical coordinate.

[0039] In the first reflectivity distribution curve, two adjacent inflection points are extracted, and the difference between the vertical coordinates corresponding to the two inflection points is calculated, the two inflection points with the largest difference are set as dividing points, the reflectivity ratio corresponding to the minimum number of pixels in the dividing points is set as the first threshold, and the pixels in the reflectivity ratio distribution image whose reflectivity ratio is less than or equal to the first threshold are set as sub-areas.

[0040] Based on this step, a reflectance ratio distribution image is generated in the first preset band and the second preset band corresponding to the second threshold, and the second threshold is greater than the first threshold, and the pixels in the sub-area whose new reflectance ratio is less than or equal to the second threshold are set as the point area.

[0041] Specifically, in this embodiment, the deblurring algorithm is a method for correcting the clarity of an image, including but not limited to blur detection, machine learning methods, etc., and the initial image can be subjected to radiation correction, atmospheric correction, and geometric correction to eliminate distortion and distortion, and the initial image after deblurring is defined as a clear image. Since the initial image is a spectral image, the spectral image is an image that records the radiation reflected or emitted by ground objects in different bands. These images can be multispectral images, containing several wider bands, or hyperspectral images, containing hundreds of continuous and narrow bands. A band refers to radiation within a specific wavelength range in a spectral image, and each band generally corresponds to a specific region of the electromagnetic spectrum, such as visible light, near infrared, short-wave infrared, medium-wave infrared, or long-wave infrared bands. Reflectivity refers to the ratio of radiation reflected from the surface of a ground object to incident radiation. Reflectivity is band-specific, and different substances have different reflectivities in different bands, which forms a unique spectral feature for each substance.

[0042] The first preset band refers to the red band, and the second preset band refers to the green band. In spectral imaging, the green band and the red band are very important bands because the green band contains information related to vegetation health, biomass and other surface features, and the red band contains a variety of information related to surface features. The reflectivity ratio refers to the ratio of the reflectivity corresponding to the first preset band and the second preset band of the same pixel in the clear image. All reflectivity ratios can be based on the corresponding pixel coordinates to generate a reflectivity ratio distribution image.

[0043] In order to count the change curve between the number of pixels and the reflectivity in the reflectivity ratio distribution image, a first reflectivity distribution curve can be drawn through a coordinate system. Since there is a large difference between the reflectivity of the ground and the reflectivity of the water in physical properties, the inflection point of the first reflectivity distribution curve can be used for screening to obtain the dividing point, wherein the dividing point includes two inflection points, and the inflection point reflectivity ratio with the minimum number of pixels is set as the first threshold. In the reflectivity ratio distribution image, pixels less than or equal to the first threshold are likely to be water and shadow areas, and areas greater than the first threshold are surface areas. Therefore, it is necessary to further distinguish the sub-areas.

[0044] The first reflectivity ratio image is analyzed by the same steps, and the corresponding second threshold is set, and the second threshold should be greater than the first threshold, because the shadow part may be an algae area, which also belongs to the point area. The pixels in the sub-area whose new reflectivity ratio is less than or equal to the second threshold are combined to generate a point area, wherein the new reflectivity ratio refers to the reflectivity ratio generated by reprocessing the first reflectivity ratio image between the first preset band and the second preset band.

[0045] As a preferred technical solution of the present invention, obtaining the actual area of ​​a water body point based on the measured area includes the following steps: The number of pixels contained in the point area and the pixel area of ​​any pixel are calculated based on the shooting mode, and the product of the number of pixels corresponding to the point area and the pixel area is set as the measurement area.

[0046] In the regional image, if there are adjacent pixels between the point areas contained in any two initial images, the point areas in the two initial images are merged and set as water body points; otherwise, the point areas contained in the two initial images are set as water body points respectively.

[0047] The measured areas of the point areas corresponding to the water body points are accumulated to generate the actual area.

[0048] Specifically, in this embodiment, the pixel area refers to the actual area corresponding to each pixel, which can be converted and generated by the flight altitude of the drone and the resolution in the shooting parameters. The number of pixels contained in the point area is multiplied by the pixel area to calculate the measurement area.

[0049] Adjacent pixels refer to pixels that are adjacent to each other. If there are adjacent pixels, it means that two point areas in the regional image that do not belong to the same initial image belong to the same water body. Therefore, the two point areas are merged into one point area and defined as a water body point. Other point areas are also defined as water body points.

[0050] The actual area of ​​a water body point is the measured area of ​​the corresponding point area. If the water body point is generated by merging point areas, the actual area is the result of cumulative calculation of the measured areas of the point areas.

[0051] As a preferred technical solution of the present invention, summarizing the list information of all small and micro water body points in the area to be mapped based on preset labels includes the following steps: The preset labels include location area, area size and point ID. The grid position of the grid point corresponding to the initial image in the area box is obtained, and the point position of the small and micro water body point in the area to be drawn is generated based on the grid position and position coordinate positioning. The point position is set as the location area information, the actual area corresponding to the small and micro water body point is set as the area size information, the small and micro water body point is marked with a serial number and set as the point ID information, and the information of all small and micro water body points is summarized based on the preset labels to generate a list information.

[0052] Specifically, in this embodiment, the location area refers to the coordinate position of the small and micro water body point on the actual ground, the area size refers to the actual area of ​​the small and micro water body point, and the point ID refers to marking and distinguishing each small and micro water body point by setting the ID. The point position of each small and micro water body point in the area to be drawn can be determined by the grid position and the location coordinates, and the small and micro water body points in the area to be drawn can be quickly located.

[0053] As a preferred technical solution of the present invention, the pollution level prediction of small water body points within the prediction time is output based on the following steps: The preset time period includes multiple time points, and the supplementary images corresponding to all the time points are combined and set as a supplementary image set.

[0054] The image processing model extracts the spectral data of all supplementary images, clusters all the spectral data to generate a preset number of clusters, extracts the best band in each cluster based on the feature selection algorithm, obtains the spectral data belonging to the best band in the cluster, and sets it as a training set. All the training sets are input into the image processing model for training and learning to generate a prediction model. The prediction model outputs the concentration prediction result within the prediction time based on the training set, and the concentration prediction result is set as the pollution level prediction.

[0055] Specifically, in this embodiment, the supplementary images corresponding to each time point constitute a supplementary image set, wherein the supplementary image set is a set of images generated by photographing the same small micro-water body point.

[0056] The image processing model extracts the spectral data of all supplementary images through the algorithm corresponding to the machine learning model, where the spectral data refers to the information on the variation of the spectral reflectance or radiance of each pixel in the supplementary image with the wavelength. All spectral data can be clustered through the Gaussian Mixture Model (GMM), and a preset number of clustering results are set as cluster clusters. The best band refers to the band that is most effective in distinguishing and identifying specific features (such as the concentration of suspended green algae in the water body) from a large number of spectral bands contained in the cluster cluster. The training set is constructed from the spectral data corresponding to the best band to learn and train the image processing model, and the trained image processing model is set as the prediction model. The concentration prediction result refers to the band distribution corresponding to the pollutant concentration in the small and micro water body point. It can improve the accuracy of the prediction model, which is conducive to improving the accuracy of water quality assessment.

[0057] As a preferred technical solution of the present invention, a search route is generated based on the following steps: Set up a pollution concentration table, calculate the mean water quality distribution based on the actual area based on the pollution level prediction, obtain the pollution grade corresponding to the mean water quality distribution in the pollution concentration table, and set the pollution grade as the degree of water pollution.

[0058] In the list information, all small and micro water body points are arranged in order from high to low based on the degree of water pollution, and a screening number is generated. Based on the screening number, the location coordinates of the small and micro water body points are connected in sequence to generate a screening route.

[0059] Specifically, in this embodiment, the pollution concentration table is a table used to measure the degree of water pollution. Different pollution degree intervals correspond to different pollution levels. The water quality distribution mean refers to the ratio of the area shown to be polluted in the pollution level prediction to the actual area. The pollution degree interval corresponding to the water quality distribution mean is obtained in the pollution concentration table, and the pollution level corresponding to the pollution degree interval is set as the pollution level of the small water body point.

[0060] The pollution levels corresponding to all small and micro water bodies are arranged from high to low by the inspection sequence number, so as to connect the location coordinates in sequence according to the inspection sequence number and generate the inspection route. Based on the inspection route, the management personnel can quickly realize the monitoring and evaluation of small and micro water bodies.

[0061] like Figure 3 As shown, the present invention also provides a small water body monitoring, mapping and evaluation system, which is used to implement the above-mentioned small water body monitoring, mapping and evaluation method, and the system mainly includes: The drawing module uses a drone to obtain an initial image group of the area to be drawn based on a preset shooting trajectory and shooting method, stitches all initial images in the initial image group in sequence, and generates a regional image of the area to be drawn.

[0062] The extraction module constructs an extraction model, extracts the point area of ​​the initial image based on the extraction model, calculates the measured area of ​​the point area based on the shooting method, marks the position coordinates of all point areas in the regional image, summarizes all point areas, generates water body points in the area to be drawn, and obtains the actual area of ​​the water body points based on the measured area, and sets the water body points whose actual areas are within the preset area range as small water body points.

[0063] The analysis module summarizes the directory information of all small and micro water body points in the area to be mapped based on preset labels. The drone obtains the supplementary image set corresponding to the small and micro water body points based on the preset time period, builds an image processing model, inputs all the supplementary images contained in the supplementary image set into the image processing model for learning and training, and outputs the pollution level prediction of the small and micro water body points within the prediction time, and writes the pollution level prediction into the directory information.

[0064] The monitoring module marks the water pollution levels of all small and micro water bodies based on the list information, generates investigation routes based on the water pollution levels and location coordinates, and sends the investigation routes to management personnel to complete the monitoring and evaluation of small and micro water bodies in the area to be mapped.

[0065] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0066] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the above-mentioned program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0067] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0068] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for those of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0069] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring, mapping and evaluating small and micro water bodies, characterized in that: The method comprises the following steps: S1: using a drone to obtain an initial image group of the area to be drawn based on a preset shooting trajectory and shooting method, stitching all initial images in the initial image group in sequence, and generating a regional image of the area to be drawn; S2: constructing an extraction model, extracting the point area of ​​the initial image based on the extraction model, calculating the measured area of ​​the point area based on the shooting method, marking the position coordinates of all the point areas in the regional image, summarizing all the point areas, generating water body points of the area to be mapped, and obtaining the actual area of ​​the water body point based on the measured area, and setting the water body point whose actual area is within the preset area interval as a small water body point; S3: Based on the preset labels, the directory information of all the small and micro water body points in the area to be mapped is summarized. The drone obtains the supplementary image set corresponding to the small and micro water body points based on the preset time period, builds an image processing model, inputs all the supplementary images contained in the supplementary image set into the image processing model for learning and training, and outputs the pollution level prediction of the small and micro water body points within the prediction time, and writes the pollution level prediction into the directory information; S4: Mark the water pollution levels of all the small water bodies based on the list information, generate a screening route based on the water pollution level and the location coordinates, send the screening route to management personnel, and complete the monitoring and evaluation of the small water bodies in the area to be mapped.

2. The method according to claim 1, characterized in that: The method of using a drone to obtain an initial image group of the area to be mapped based on a preset shooting trajectory and shooting method includes the following steps: The GIS coordinates of the area to be drawn are obtained, and an area frame including the area to be drawn is constructed based on the GIS coordinates. The shooting range, flight altitude, flight speed and shooting parameters of the drone are set, and the area frame is divided into a plurality of grid points based on the shooting range. The drone takes the grid point corresponding to the vertex of the area frame as the starting point and performs flight shooting at any grid point in sequence in an S-shaped route to generate a plurality of initial images, and all the initial images are combined to generate the initial image group, wherein the S-shaped route is set as the shooting trajectory, the flight altitude, the flight speed and the shooting parameter combination are set as the shooting mode, and the initial images corresponding to any two adjacent grid points have an overlapping area.

3. The method according to claim 2, characterized in that The regional image of the area to be drawn is generated based on the following steps: defining the initial images corresponding to the two grid points adjacent to the starting point as comparison images, extracting feature points of the comparison images respectively based on a feature extraction algorithm, performing feature point matching on the feature points of the two comparison images to extract feature point pairs, setting any feature point pair as a key point, and setting the maximum pixel area included in all the key points as the superposition area; Calculating a transformation matrix of the key point between the two contrast images based on the least square method, calculating predicted positions of the remaining key points in the two contrast images based on the transformation matrix, calculating residuals of all the key points based on the predicted positions and the actual positions of the remaining key points, adjusting parameters of the transformation matrix until the sum of all the residuals is minimized, and setting the adjusted transformation matrix as the image transformation matrix; The two contrast images are stitched together based on the image transformation matrix to generate a stitched image, and the stitched image is stitched together with the initial image corresponding to the next grid point based on this step to generate a new stitched image, and this step is repeated until the initial images corresponding to all the grid points are stitched together, and the last new stitched image is set as the regional image.

4. The method according to claim 1, characterized in that: Extracting the point area in the initial image based on the extraction model comprises the following steps: The extraction model performs a deblurring process on the initial image based on a deblurring algorithm to generate a clear image, obtains the reflectivity ratio of any pixel in the clear image between a first preset band and a second preset band, and combines the reflectivity ratios corresponding to all pixels to generate a reflectivity ratio distribution image; Constructing a coordinate system, setting the numerical value corresponding to the reflectivity ratio in the reflectivity ratio distribution image as the horizontal coordinate, setting the number of pixels included in any reflectivity ratio as the vertical coordinate, and fitting and drawing in the coordinate system based on the horizontal coordinate and the vertical coordinate to generate a first reflectivity distribution curve; In the first reflectivity distribution curve, two adjacent inflection points are extracted, and the difference between the ordinates corresponding to the two inflection points is calculated, the two inflection points with the largest difference are set as dividing points, the reflectivity ratio corresponding to the minimum number of pixels in the dividing points is set as a first threshold, and the pixels in the reflectivity ratio distribution image whose reflectivity ratio is less than or equal to the first threshold are set as sub-regions; Based on this step, a second threshold corresponding to the reflectance ratio distribution image in the first preset band and the second preset band is generated, and the second threshold is greater than the first threshold, and the pixels in the sub-area whose new reflectance ratio is less than or equal to the second threshold are set as the point area.

5. The method according to claim 4, characterized in that The obtaining of the actual area of ​​the water body point based on the measured area comprises the following steps: Calculating the number of pixels contained in the point area and the pixel area of ​​any pixel based on the shooting method, and setting the product of the number of pixels corresponding to the point area and the pixel area as the measurement area; In the regional image, if there are adjacent pixels between the point regions included in any two initial images, the point regions in the two initial images are merged and set as the water body points; otherwise, the point regions included in the two initial images are respectively set as the water body points; The measured areas of the point regions corresponding to the water body points are accumulated to generate the actual area.

6. The method according to claim 2, characterized in that The step of summarizing the list information of all the small and micro water body points in the area to be mapped based on preset tags includes the following steps: The preset label includes a location area, an area size and a point ID, and the grid position of the grid point corresponding to the initial image in the area frame is obtained. The point position of the small and micro water body point in the area to be drawn is generated based on the grid position and the location coordinate positioning, and the point position is set as the information of the location area, and the actual area corresponding to the small and micro water body point is set as the information of the area size. The small and micro water body points are marked with serial numbers and set as the point ID information, and the information of all the small and micro water body points is summarized based on the preset label to generate the directory information.

7. The method according to claim 1, characterized in that Output the pollution level prediction of the small water body point within the prediction time based on the following steps: The preset time period includes a plurality of time points, and the supplementary images corresponding to all the time points are combined and set as the supplementary image set; The image processing model extracts spectral data of all the supplementary images, clusters all the spectral data to generate a preset number of clusters, extracts the best band in each cluster based on a feature selection algorithm, obtains the spectral data belonging to the best band in the cluster, and sets it as a training set, inputs all the training sets into the image processing model for training and learning, and then generates a prediction model, the prediction model outputs a concentration prediction result within the prediction time based on the training set, and the concentration prediction result is set as the pollution level prediction.

8. The method according to claim 7, characterized in that The investigation route is generated based on the following steps: Setting a pollution concentration table, calculating the mean value of water quality distribution based on the actual area according to the pollution level prediction, obtaining the pollution level corresponding to the mean value of water quality distribution in the pollution concentration table, and setting the pollution level as the water quality pollution degree; In the directory information, all the small water body points are arranged in order from high to low based on the water pollution degree to generate a screening number, and based on the screening number, the location coordinates of the small water body points are connected in order to generate the screening route.

9. A small water body monitoring, mapping and evaluation system, used to implement the method according to any one of claims 1 to 8, characterized in that: The system includes the following modules: A drawing module, using a drone to obtain an initial image group of the area to be drawn based on a preset shooting trajectory and shooting method, sequentially splicing all initial images in the initial image group, and generating a regional image of the area to be drawn; An extraction module, constructing an extraction model, extracting the point area of ​​the initial image based on the extraction model, calculating the measured area of ​​the point area based on the shooting method, marking the position coordinates of all the point areas in the area image, summarizing all the point areas, generating water body points of the area to be mapped, and obtaining the actual area of ​​the water body point based on the measured area, and setting the water body point whose actual area is within the preset area interval as a small water body point; The analysis module summarizes the directory information of all the small and micro water body points in the area to be mapped based on preset labels, and the drone obtains the supplementary image set corresponding to the small and micro water body points based on a preset time period, builds an image processing model, inputs all the supplementary images contained in the supplementary image set into the image processing model for learning and training, and outputs the pollution level prediction of the small and micro water body points within the prediction time, and writes the pollution level prediction into the directory information; The monitoring module marks the water pollution levels of all the small water bodies based on the list information, generates an investigation route based on the water pollution level and the location coordinates, and sends the investigation route to the management personnel to complete the monitoring and evaluation of the small water bodies in the area to be mapped.

Images