Aerial survey data processing system and method based on unmanned aerial vehicle

Abstract

The invention discloses an aerial survey data processing system and method based on an unmanned aerial vehicle, and belongs to the technical field of aerial survey of unmanned aerial vehicles. The system comprises a data acquisition module which acquires multiple groups of image data of a reservoir to be measured through an unmanned aerial vehicle; the data transmission and preprocessing module is used for transmitting the collected initial image data to a ground monitoring center in real time; the image scoring and screening module is used for performing accuracy scoring through an image accuracy scoring mechanism; the water flow change information extraction module is used for extracting water flow change information of the water inlet period of the reservoir from the final image data; and the prediction modeling module is used for constructing a reservoir water storage prediction model based on the final water flow change information. The accuracy and reliability of water flow change information extraction are improved, the precision of the reservoir water storage prediction model is improved, then the water demand of the reservoir is accurately predicted, and the problems that in the prior art, the precision of the water storage prediction model is not high, and the water demand of the reservoir is not adjusted in time are solved.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial survey technology, and in particular to an unmanned aerial survey data processing system and method based on the unmanned aerial vehicle. Background Art

[0002] UAV low-altitude aerial survey technology is a new technology that uses UAVs as carriers to obtain ground impact data and use relevant data software for post-processing to produce geographic information products. With the rapid development of information technology, in recent years, UAV low-altitude aerial survey technology has been widely used in land, water conservancy, agriculture, disaster emergency and other fields with its advantages of flexibility and low cost. In the field of surveying and mapping, it has achieved unprecedented development and gradually become an important means to enhance the surveying and mapping emergency service guarantee capabilities.

[0003] Reservoirs have long been responsible for the task of avoiding salty water and storing fresh water, and avoiding pollution and storing clean water. As the water supply increases year by year, the problem of small storage capacity is particularly prominent. Therefore, it is particularly important to make good use of the limited storage capacity of the reservoir and give full play to the dilution effect of the reservoir. The reservoir has been in operation for nearly 30 years. The water supply characteristics and water flow patterns are changing. The water storage capacity of the reservoir is mainly regulated by human experience, and the accuracy is not high. As a new type of terrain detection method, drone aerial survey technology has flexible and changeable operation modes and is easy to operate; the equipment and operation costs are low, and it is less affected by climate and site; it can obtain high-resolution remote sensing images, which has greatly improved surveying and mapping technology.

[0004] For example, the invention patent with publication number: CN116576826B discloses an intelligent processing system and method for aerial survey data based on a drone, including: S10: determining the blocked area according to the aerial survey data of the drone at the corresponding flight position, adjusting the aerial photography angle and direction of the drone in combination with the flight altitude and aerial photography focal length of the drone at the corresponding flight position, and determining the blocked area of ​​the blocked area, and predicting the risk factor of the blocked area in combination with the aerial survey data collected after the drone camera adjusts the angle; S20: restoring the blocked area; S30: sending warning information to the control terminal according to the adjusted risk factor, and the control terminal adjusts the motion trajectory of the drone according to the warning information.

[0005] For example, the invention patent with publication number: CN104296730A discloses a hydrological flow data processing system and method, which includes: a registration module, a setting module, a functional module and a display module. The registration module is used to save and record user information, and to operate the device normally with the support of an "authorization dongle"; the setting module is used to set the algorithm and mode executed by the device; the functional module is used to classify, integrate and process data according to the mode; the display module is used to display the processing progress and results of the data in real time, and supports data export operations.

[0006] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0007] In the prior art, high-altitude multi-angle and multi-altitude tracking and shooting by drones is used to build a reservoir water storage capacity prediction model based on water flow change information, and then the reservoir water storage capacity is adjusted. However, there are problems such as low accuracy of the water storage capacity prediction model and untimely adjustment of the reservoir water storage capacity. Summary of the invention

[0008] The embodiments of the present application solve the problems of low accuracy of water storage prediction models and untimely adjustment of water demand in reservoirs in the prior art by providing a drone-based aerial survey data processing system and method. This improves the accuracy of the reservoir water storage prediction model, thereby accurately predicting the water demand of the reservoir and facilitating staff to adjust the water storage capacity of the reservoir in a timely manner.

[0009] The embodiment of the present application provides an aerial survey data processing system based on a drone, including: a data acquisition module: using a drone to collect multiple groups of image data of a reservoir to be measured from multiple angles and heights to obtain initial image data of the reservoir to be measured; a data transmission and preprocessing module: transmitting the collected initial image data to a ground monitoring center in real time via a wireless network, and performing enhancement processing on the initial image data to obtain enhanced initial image data; an image scoring and screening module: performing accuracy scoring on the enhanced initial image data through an image accuracy scoring mechanism, and screening out a group of enhanced initial image data with the highest accuracy as the final image data based on the scoring results; a water flow change information extraction module: based on the final image data, extracting water flow change information during the water inflow period of the reservoir from the final image data through an edge detection method, and obtaining the accuracy of the water flow change information through an extraction accuracy index formula Accuracy, if the accuracy of the water flow change information 1 meets the standard, the output is the final water flow change information; if the accuracy of the water flow change information 1 does not meet the standard, the water flow change information 2 during the reservoir water inflow period is extracted from the final image data through the optical flow analysis algorithm, if the accuracy of the water flow change information 2 meets the standard, the output is the final water flow change information; if the accuracy of the water flow change information 2 does not meet the standard, the final image data is enhanced to obtain the enhanced final image data, based on the enhanced final image data, the water flow change information 1 and the water flow change information 2 during the reservoir water inflow period are extracted again through the edge detection method and the optical flow analysis algorithm, and the accuracy of the water flow change information extracted by the two methods is obtained again through the extraction accuracy index formula, and the group with the highest accuracy is output as the final water flow change information; prediction modeling module: construct a reservoir water storage prediction model based on the final water flow change information.

[0010] Furthermore, the collected initial image data is transmitted to the ground monitoring center in real time, and the initial image data is enhanced by the following steps: the initial image data of the reservoir to be measured is received and stored by wireless transmission technology; the stored initial image data is subjected to image denoising by a filtering algorithm to obtain preprocessed initial image data; the preprocessed initial image data is enhanced by an adaptive histogram equalization algorithm to complete the enhancement of the initial image data.

[0011] Furthermore, the enhanced initial image data is scored for accuracy through an image accuracy scoring mechanism, including the following steps: obtaining the image quality index, lighting quality index, and water surface fluctuation index of the enhanced initial image, and performing standardization processing; obtaining the weight factor of the standardized image quality index, lighting quality index, and water surface fluctuation index for image accuracy through an objective weighting method; and calculating the accuracy of the initial image data through the image accuracy index formula.

[0012] Furthermore, the image accuracy score index formula is:

[0013]

[0014] Where TX is the image accuracy index, TZ is the image quality index, α1 is the weight factor of TZ to TX, GZ is the illumination quality index, α2 is the weight factor of GZ to TX, BD is the water surface fluctuation index, α3 is the weight factor of BD to TX, and e is a natural constant.

[0015] Furthermore, water flow change information during the reservoir water inflow period is extracted from the final image data through an edge detection method, comprising the following steps: converting the final image data into grayscale image data, removing noise in the grayscale image data, and obtaining processed grayscale image data; extracting water flow boundary information during the reservoir water inflow period from the processed grayscale image data; performing edge optimization based on the extracted water flow boundary information during the reservoir water inflow period, and obtaining water flow change information during the reservoir water inflow period.

[0016] Furthermore, the second water flow change information during the reservoir water inflow period is re-extracted from the final image data through an optical flow analysis algorithm, including the following steps: based on the final image data, a continuous image sequence during the reservoir water inflow period is extracted; pixel motion information between the front and rear images in the continuous image sequence is analyzed to obtain the second water flow boundary information during the reservoir water inflow period; based on the second water flow boundary information during the reservoir water inflow period, optimization is performed to obtain the optimized second water flow boundary information during the reservoir water inflow period.

[0017] Furthermore, the calculation formulas for the accuracy of the water flow change information 1 and the accuracy of the water flow change information 2 are as follows:

[0018]

[0019] In the formula, TQ1 is the accuracy of water flow change information one, WD1 is the stability of water flow change one, β1 is the weight factor of WD1 to TQ1, FX1 is the accuracy of water flow direction one, β2 is the weight factor of FX1 to TQ1, SD1 is the accuracy of water flow velocity one, β3 is the weight factor of SD1 to TQ1, and e is a natural constant; TQ2 is the accuracy of water flow change information two, WD2 is the stability of water flow change two, γ1 is the weight factor of WD2 to TQ2, FX2 is the accuracy of water flow direction two, γ2 is the weight factor of FX2 to TQ2, SD2 is the accuracy of water flow velocity two, and γ3 is the weight factor of SD2 to TQ2.

[0020] Furthermore, data enhancement is performed on the final image data to obtain enhanced final image data, which includes the following steps: linearly expanding the pixel values ​​of the final image data to obtain linearly expanded final image data; based on the linearly expanded final image data, enhancing the edge portion of the final image data to obtain enhanced final image data.

[0021] Furthermore, constructing a reservoir water storage capacity prediction model based on the final water flow change information includes the following steps: obtaining historical reservoir water flow change information to be measured, the historical reservoir water flow change information to be measured includes historical water flow direction, historical water flow velocity, and historical reservoir water storage, and corresponding the historical water flow direction, historical water flow velocity, and historical reservoir water storage one by one according to the timestamp; preprocessing the historical reservoir water flow change information to be measured; dividing the training set and the test set based on the preprocessed historical reservoir water flow change information to be measured; inputting the training set into a polynomial regression model for training to obtain a preliminary reservoir water storage capacity prediction model; inputting the test set into the preliminary reservoir water storage capacity prediction model for testing to obtain a reservoir water storage capacity prediction model.

[0022] The embodiment of the present application also includes a method for processing aerial survey data based on a drone, comprising the following steps: using a drone to collect multiple groups of image data of a reservoir to be measured from multiple angles and heights to obtain initial image data of the reservoir to be measured; transmitting the collected initial image data to a ground monitoring center in real time via a wireless network, and performing enhancement processing on the initial image data to obtain enhanced initial image data; performing accuracy scoring on the enhanced initial image data through an image accuracy scoring mechanism, and selecting a group of enhanced initial image data with the highest accuracy as final image data based on the scoring results; based on the final image data, extracting water flow change information 1 during the water inflow period of the reservoir from the final image data through an edge detection method, and obtaining the accuracy of the water flow change information 1 by extracting an accuracy index formula; if the accuracy of the water flow change information 1 is If the accuracy of the water flow change information 1 does not meet the standard, the water flow change information 2 during the reservoir water inflow period is extracted from the final image data through the optical flow analysis algorithm, and if the accuracy of the water flow change information 2 meets the standard, the final water flow change information is output; if the accuracy of the water flow change information 2 does not meet the standard, the final image data is enhanced to obtain the enhanced final image data, and based on the enhanced final image data, the water flow change information 1 and the water flow change information 2 during the reservoir water inflow period are extracted again through the edge detection method and the optical flow analysis algorithm, and the accuracy of the water flow change information extracted by the two methods is obtained again through the extraction accuracy index formula, and the group with the highest accuracy is output as the final water flow change information; a reservoir water storage prediction model is constructed based on the final water flow change information.

[0023] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0024] 1. The enhanced initial image data is scored through the image accuracy scoring mechanism, and the image data with the highest accuracy is screened out based on the scoring results, thereby ensuring the quality of the final image data, and then realizing the automation and high-precision output of image data screening, which can significantly improve the accuracy of water flow change information extraction and effectively solve the problems of difficult to ensure image quality and large errors in the existing technology.

[0025] 2. The water flow change information during the reservoir inflow period is extracted through edge detection method and optical flow analysis algorithm, and the results are further screened and optimized through the accuracy index formula, thereby improving the accuracy and reliability of water flow change information extraction, and then realizing the construction of a more accurate reservoir water storage prediction model, solving the problem of untimely regulation of reservoir water storage.

[0026] 3. By using drones to collect multiple sets of image data of the reservoir to be measured from multiple angles and heights, all-round and dynamically changing reservoir image information is obtained, thereby achieving more comprehensive data collection, which can capture the different states and environmental changes of the reservoir, and effectively solve the problems of information loss and error accumulation caused by single angles and limited viewing angles in existing technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 A structural diagram of an aerial survey data processing system based on a drone provided in an embodiment of the present application.

[0028] Figure 2 A flowchart of a method for processing aerial survey data based on a drone provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] The embodiments of the present application solve the problems of low accuracy of water storage prediction models and untimely adjustment of reservoir water storage in the prior art by providing a drone-based aerial survey data processing system and method. By improving the accuracy of the reservoir water storage prediction model, accurate prediction of reservoir water demand is achieved, making it convenient for staff to adjust the reservoir water storage in a timely manner.

[0030] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0031] like Figure 1 As shown, it is a structural diagram of the drone-based aerial survey data processing system provided in an embodiment of the present application, including: a data acquisition module: a plurality of groups of image data of the reservoir to be measured are collected from multiple angles and heights by a drone to obtain initial image data of the reservoir to be measured; a data transmission and preprocessing module: the collected initial image data are transmitted to a ground monitoring center in real time via a wireless network, and the initial image data are enhanced to obtain enhanced initial image data; an image scoring and screening module: the enhanced initial image data are scored for accuracy through an image accuracy scoring mechanism, and a group of enhanced initial image data with the highest accuracy is screened out as final image data based on the scoring results.

[0032] Water flow change information extraction module: Based on the final image data, the water flow change information 1 during the water inflow period of the reservoir is extracted from the final image data by the edge detection method, and the accuracy of the water flow change information 1 is obtained by the extraction accuracy index formula. If the accuracy of the water flow change information 1 meets the standard, it is output as the final water flow change information; if the accuracy of the water flow change information 1 does not meet the standard, the water flow change information 2 during the water inflow period of the reservoir is extracted from the final image data by the optical flow analysis algorithm. If the accuracy of the water flow change information 2 meets the standard, it is output as the final water flow change information; if the accuracy of the water flow change information 2 does not meet the standard, the final image data is enhanced to obtain the enhanced final image data. Based on the enhanced final image data, the water flow change information 1 and the water flow change information 2 during the water inflow period of the reservoir are extracted again by the edge detection method and the optical flow analysis algorithm, and the accuracy of the water flow change information extracted by the two methods is obtained again by the extraction accuracy index formula, and the group with the highest accuracy is output as the final water flow change information; Prediction modeling module: Construct a reservoir water storage prediction model based on the final water flow change information.

[0033] In this embodiment, the drone collects multiple sets of image data of the reservoir to be measured from multiple angles and heights to ensure that the data of the reservoir at different viewing angles can reflect the actual situation of the reservoir. It can capture all kinds of changes in the water surface of the reservoir in a comprehensive and three-dimensional manner, enhance the accuracy of reservoir monitoring, and avoid affecting the analysis results due to a single viewing angle or insufficient data.

[0034] The enhanced initial image data is scored for accuracy through the image accuracy scoring mechanism, and a group of enhanced initial image data with the highest accuracy is selected as the final image data based on the scoring results. This ensures that only high-quality image data is used for subsequent analysis, effectively avoiding erroneous extraction and analysis caused by poor image quality, and improving the efficiency and reliability of the overall system; the accuracy scoring of the initial image data through the image accuracy scoring mechanism includes the following steps: all image information in the initial image data is scored for accuracy through the image accuracy scoring mechanism, and the scoring parameters include the image quality index of the image, the illumination quality index of the image, and the water surface fluctuation index of the image, and then the image accuracy index formula is used to score the accuracy of all image information based on the above three parameters, and the group of images with the highest scores at each position are selected as the final image data based on the scoring results.

[0035] The water flow change information during the reservoir inflow period is extracted from the final image data by the edge detection method, which includes the following steps: using the Sobel operator to calculate the gradient values ​​of the final image data in the x-axis and y-axis directions, and obtaining the change rate of each pixel point; according to the calculated gradient value, enhancing the part of the image with large changes (i.e., the edge part), which represents the boundary of the object in the image; finding all the edge points in the edge map; retaining the local maximum value by non-maximum suppression (NMS), reducing pseudo edges; using the thresholding method to connect the edge points, and obtaining a complete edge map.

[0036] Extracting water flow change information during the reservoir inflow period from the final image data through optical flow analysis includes the following steps: for each frame of final image data, calculating the spatial gradient (such as partial derivatives in the x and y directions) and the temporal gradient (change of the image on the time axis) of the image, and establishing the optical flow equation based on the constant brightness assumption using the spatial and temporal gradients of the final image data. By solving the optical flow equation, the movement speed and direction of each pixel point are obtained, and then the least squares method is usually used to optimize and estimate the optical flow. Based on the estimated optical flow, the movement direction and speed of the object in the image are calculated to identify the dynamic changing features such as the water flow.

[0037] Water flow change information is extracted through two different algorithms to ensure the accuracy of the extraction results and reduce the errors caused by a single algorithm.

[0038] Furthermore, the collected initial image data is transmitted to the ground monitoring center in real time, and the initial image data is enhanced by the following steps: the initial image data of the reservoir to be measured is received and stored by wireless transmission technology; the stored initial image data is subjected to image denoising by a filtering algorithm to obtain preprocessed initial image data; the preprocessed initial image data is enhanced by an adaptive histogram equalization algorithm to complete the enhancement of the initial image data.

[0039] In this embodiment, the initial image data is denoised by a Gaussian filtering algorithm to obtain relatively clean, noise-free initial image data, thereby improving the signal-to-noise ratio of the initial image data. The higher the signal-to-noise ratio, the better the image quality, thereby ensuring that subsequent image processing and analysis will not be disturbed by noise, thereby enhancing the accuracy and reliability of image processing.

[0040] The initial image data is enhanced by an adaptive histogram equalization algorithm. This algorithm can effectively improve image details by redistributing the grayscale of the initial image data and enhancing the contrast of the initial image data, making subtle changes more obvious, enhancing the contrast and clarity of the initial image data, and making important features in the image more obvious, including water flow edges, change areas, etc., to facilitate the subsequent extraction of water flow change information.

[0041] Furthermore, the accuracy scoring of the enhanced initial image data through the image accuracy scoring mechanism includes the following steps: obtaining the image quality index, lighting quality index, and water surface fluctuation index of the enhanced initial image, and performing standardization processing; obtaining the weight factor of the standardized image quality index, lighting quality index, and water surface fluctuation index for image accuracy through an objective weighting method; and calculating the accuracy of the initial image data through the image accuracy index formula.

[0042] In this embodiment, obtaining the image quality index of the enhanced initial image includes the following steps: first, denoising the enhanced initial image, then calculating the difference between the original image and the denoised image to obtain the signal-to-noise ratio; then comparing the enhanced initial image with the initial image to calculate the SSIM (structural similarity index) value. Combining the signal-to-noise ratio and the SSIM value, the image quality index is obtained by weighted averaging.

[0043] Obtaining the light quality index of the enhanced initial image includes the following steps: converting the enhanced initial image into a grayscale image and calculating its brightness distribution, using histogram equalization to analyze the distribution of the brightness of the enhanced initial image, and the enhanced initial image with uniform illumination usually has a smoother brightness histogram. Dividing the enhanced initial image into multiple small blocks and calculating the contrast of each small block. The brightness distribution and local contrast are weighted and synthesized to obtain the final light quality index.

[0044] The water surface fluctuation index of the enhanced initial image is obtained by the following steps: extracting the water surface texture of the enhanced initial image by using the Canny operator, calculating the contrast, energy and other characteristics of the water surface texture by using the gray level co-occurrence matrix, and then performing differential analysis on the continuous images by using the optical flow analysis method to detect the motion on the water surface, and obtaining the amplitude and direction of the water surface fluctuation by calculating the motion difference between adjacent frame images. The water surface texture analysis is combined with the results of the water surface motion detection to obtain the water surface fluctuation index.

[0045] Furthermore, the image accuracy score index formula is:

[0046]

[0047] Where TX is the image accuracy index, TZ is the image quality index, α1 is the weight factor of TZ to TX, GZ is the illumination quality index, α2 is the weight factor of GZ to TX, BD is the water surface fluctuation index, α3 is the weight factor of BD to TX, and e is a natural constant.

[0048] In this embodiment, when TZ=0.8, α1=0.5, GZ=0.75, α2=0.2, BD=0.45, α3=0.3, it is obtained that TX=0.153. The closer TX is to 1, the higher the accuracy of the enhanced initial image data.

[0049] Furthermore, water flow change information during the reservoir water inflow period is extracted from the final image data through an edge detection method, comprising the following steps: converting the final image data into grayscale image data, removing noise in the grayscale image data, and obtaining processed grayscale image data; extracting water flow boundary information during the reservoir water inflow period from the processed grayscale image data; performing edge optimization based on the extracted water flow boundary information during the reservoir water inflow period, and obtaining water flow change information during the reservoir water inflow period.

[0050] In this embodiment, the water flow change information is extracted by edge detection method, which can effectively identify the change of water flow during the reservoir inflow period. This method ensures the accuracy and clarity of water flow change information through steps such as conversion to grayscale image, denoising, edge extraction and optimization, and ultimately provides reliable data support for reservoir monitoring and analysis.

[0051] Furthermore, the second water flow change information during the reservoir water inflow period is re-extracted from the final image data through an optical flow analysis algorithm, including the following steps: based on the final image data, a continuous image sequence during the reservoir water inflow period is extracted; pixel motion information between the front and rear images in the continuous image sequence is analyzed to obtain the second water flow boundary information during the reservoir water inflow period; based on the second water flow boundary information during the reservoir water inflow period, optimization is performed to obtain the optimized second water flow boundary information during the reservoir water inflow period.

[0052] In this embodiment, the dynamic change information of water flow can be extracted from the continuous image sequence through the optical flow algorithm. This method can analyze the dynamic characteristics of water flow such as speed and direction, so as to obtain the water flow change information during the reservoir inflow period. Compared with the edge detection of static images, the optical flow algorithm can capture the subtle changes of water flow more accurately, and is particularly suitable for processing complex fluid dynamics. At the same time, optical flow analysis is applicable to a variety of water surface scenes, whether it is a calm water surface or a water surface with large fluctuations, it can provide effective water flow dynamic information, and can significantly improve the robustness and stability of water flow boundary extraction.

[0053] Furthermore, the calculation formulas for the accuracy of the water flow change information 1 and the accuracy of the water flow change information 2 are as follows:

[0054]

[0055] In the formula, TQ1 is the accuracy of water flow change information one, WD1 is the stability of water flow change one, β1 is the weight factor of WD1 to TQ1, FX1 is the accuracy of water flow direction one, β2 is the weight factor of FX1 to TQ1, SD1 is the accuracy of water flow velocity one, β3 is the weight factor of SD1 to TQ1, and e is a natural constant; TQ2 is the accuracy of water flow change information two, WD2 is the stability of water flow change two, γ1 is the weight factor of WD2 to TQ2, FX2 is the accuracy of water flow direction two, γ2 is the weight factor of FX2 to TQ2, SD2 is the accuracy of water flow velocity two, and γ3 is the weight factor of SD2 to TQ2.

[0056] In this embodiment, when WD1=0.75, β1=0.4, FX1=0.8, β2=0.3, SD1=0.4, β3=0.3, it is concluded that TQ1=0.541; when WD2=0.9, γ1=0.4, FX2=0.75, γ2=0.3, SD2=0.6, γ3=0.3, it is concluded that TQ2=0.648; by comparing TQ1 and TQ2 with the set water flow change information accuracy threshold, if it is lower than the set threshold, it means that the water flow change information accuracy does not meet the standard.

[0057] Furthermore, data enhancement is performed on the final image data to obtain enhanced final image data, which includes the following steps: linearly expanding the pixel values ​​of the final image data to obtain linearly expanded final image data; based on the linearly expanded final image data, enhancing the edge portion of the final image data to obtain enhanced final image data.

[0058] In this embodiment, the final image data is enhanced, and the image contrast, clarity and edge information visibility can be significantly improved through linear expansion and edge enhancement. These improvements make the final image data more reliable and accurate in the subsequent process of extracting water flow change information 1 and water flow change information 2 during the reservoir inflow period through edge detection and optical flow analysis algorithms, improve the robustness of the system in complex environments, and optimize the accuracy of water flow monitoring and change detection.

[0059] Furthermore, constructing a reservoir water storage capacity prediction model based on the final water flow change information includes the following steps: obtaining historical water flow change information of the reservoir to be measured, the historical water flow change information of the reservoir to be measured includes historical water flow direction, historical water flow velocity, and historical reservoir water storage, and corresponding the historical water flow direction, historical water flow velocity, and historical reservoir water storage one by one according to the timestamp; preprocessing the historical water flow change information of the reservoir to be measured; dividing the training set and the test set based on the preprocessed historical water flow change information of the reservoir to be measured; inputting the training set into the polynomial regression model for training to obtain a preliminary prediction model for reservoir water storage; inputting the test set into the preliminary prediction model for reservoir water storage for testing to obtain a reservoir water storage prediction model.

[0060] In this embodiment, by constructing a reservoir water storage prediction model based on the final water flow change information, the reservoir water storage can be accurately predicted, the intelligent level of reservoir management can be improved, and strong decision-making support can be provided for reservoir managers to optimize reservoir operations and water resource scheduling. With the help of historical data and polynomial regression models, the system can more effectively understand the impact of water flow changes on reservoir water storage, and improve management efficiency, reduce risks, and enhance the ability to respond to emergencies in practical applications.

[0061] like Figure 2 As shown, it is a flow chart of the aerial survey data processing method based on drone provided in an embodiment of the present application. This embodiment also provides an aerial survey data processing method based on drone, comprising the following steps: obtaining image data of the reservoir to be measured: collecting multiple groups of image data of the reservoir to be measured from multiple angles and heights by drone to obtain initial image data of the reservoir to be measured.

[0062] Preprocessing of the initial image data: The collected initial image data is transmitted to the ground monitoring center in real time via a wireless network, and the initial image data is enhanced to obtain enhanced initial image data.

[0063] Obtaining final image data: The enhanced initial image data is scored for accuracy through an image accuracy scoring mechanism, and a group of enhanced initial image data with the highest accuracy is selected as the final image data based on the scoring results.

[0064] Obtaining the final water flow change information: Based on the final image data, the water flow change information 1 during the reservoir water inflow period is extracted from the final image data by the edge detection method, and the accuracy of the water flow change information 1 is obtained by the extraction accuracy index formula. If the accuracy of the water flow change information 1 meets the standard, it is output as the final water flow change information; if the accuracy of the water flow change information 1 does not meet the standard, the water flow change information 2 during the reservoir water inflow period is extracted from the final image data by the optical flow analysis algorithm, and if the accuracy of the water flow change information 2 meets the standard, it is output as the final water flow change information; if the accuracy of the water flow change information 2 does not meet the standard, the final image data is enhanced to obtain the enhanced final image data. Based on the enhanced final image data, the water flow change information 1 and the water flow change information 2 during the reservoir water inflow period are extracted again by the edge detection method and the optical flow analysis algorithm, respectively, and the accuracy of the water flow change information extracted by the two methods is obtained again by the extraction accuracy index formula, and the group with the highest accuracy is output as the final water flow change information.

[0065] Construct a reservoir water storage capacity prediction model: Construct a reservoir water storage capacity prediction model based on the final water flow change information.

[0066] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0067] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0068] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0069] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0070] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0071] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. The aerial survey data processing system based on UAV is characterized by: The following steps are involved: Data acquisition module: collects multiple sets of image data of the reservoir to be measured from multiple angles and heights through drones to obtain initial image data of the reservoir to be measured; Data transmission and preprocessing module: The collected initial image data is transmitted to the ground monitoring center in real time through the wireless network, and the initial image data is enhanced to obtain enhanced initial image data; Image scoring and screening module: The enhanced initial image data is scored for accuracy through the image accuracy scoring mechanism, and a group of enhanced initial image data with the highest accuracy is screened out as the final image data based on the scoring results; Water flow change information extraction module: Based on the final image data, the water flow change information 1 of the reservoir water inflow period is extracted from the final image data by edge detection method, and the accuracy of the water flow change information 1 is obtained by extracting the accuracy index formula. If the accuracy of the water flow change information 1 meets the standard, the final water flow change information is output; If the accuracy of the water flow change information 1 does not meet the standard, the water flow change information 2 during the reservoir inflow period is extracted from the final image data through the optical flow analysis algorithm. If the accuracy of the water flow change information 2 meets the standard, the final water flow change information is output; If the accuracy of the water flow change information 2 does not meet the standard, the final image data is enhanced to obtain the enhanced final image data. Based on the enhanced final image data, the water flow change information 1 and the water flow change information 2 during the reservoir inflow period are extracted again by the edge detection method and the optical flow analysis algorithm, and the accuracy of the water flow change information extracted by the two methods is obtained again by the extraction accuracy index formula, and the group with the highest accuracy is output as the final water flow change information; Prediction modeling module: Build a reservoir water storage prediction model based on the final water flow change information.

2. The drone-based aerial survey data processing system according to claim 1, characterized in that: The collected initial image data is transmitted to the ground monitoring center in real time, and the initial image data is enhanced and processed, including the following steps: Receive and store the initial image data of the reservoir to be measured through wireless transmission technology; Performing image denoising processing on the stored initial image data by using a filtering algorithm to obtain pre-processed initial image data; The pre-processed initial image data is enhanced by using an adaptive histogram equalization algorithm to complete the enhancement processing of the initial image data.

3. The drone-based aerial survey data processing system according to claim 1, characterized in that: The enhanced initial image data is scored for accuracy through an image accuracy scoring mechanism, including the following steps: Obtaining the image quality index, illumination quality index, and water surface fluctuation index of the enhanced initial image and performing standardization processing; The weight factors of the standardized image quality index, illumination quality index and water surface fluctuation index on image accuracy are obtained through objective weighting method. The accuracy of the initial image data is calculated using the image accuracy index formula.

4. The drone-based aerial survey data processing system according to claim 3, characterized in that: The image accuracy scoring index formula is: Where TX is the image accuracy index, TZ is the image quality index, α1 is the weight factor of TZ to TX, GZ is the illumination quality index, α2 is the weight factor of GZ to TX, BD is the water surface fluctuation index, α3 is the weight factor of BD to TX, and e is a natural constant.

5. The drone-based aerial survey data processing system according to claim 1, characterized in that: Extracting water flow change information during the reservoir inflow period from the final image data by edge detection method includes the following steps: Convert the final image data into grayscale image data, remove noise in the grayscale image data, and obtain processed grayscale image data; Extracting water flow boundary information during the reservoir inflow period from the processed grayscale image data; Based on the water flow boundary information 1 extracted during the reservoir inflow period, edge optimization is performed to obtain the water flow change information 1 during the reservoir inflow period.

6. The drone-based aerial survey data processing system according to claim 1, characterized in that: The water flow change information of the reservoir inflow period is extracted from the final image data again by using the optical flow analysis algorithm, including the following steps: Based on the final image data, a continuous image sequence of the reservoir inflow period is extracted; Analyze the pixel motion information between the previous and next images in the continuous image sequence to obtain the water flow boundary information 2 during the reservoir inflow period; Based on the water flow boundary information 2 during the reservoir water inflow period, optimization is performed to obtain the optimized water flow boundary information 2 during the reservoir water inflow period.

7. The drone-based aerial survey data processing system according to claim 1, characterized in that: The calculation formulas for the accuracy of water flow change information 1 and the accuracy of water flow change information 2 are: Wherein, TQ1 is the accuracy of water flow change information, WD1 is the stability of water flow change, β1 is the weight factor of WD1 to TQ1, FX1 is the accuracy of water flow direction, β2 is the weight factor of FX1 to TQ1, SD1 is the accuracy of water flow velocity, β3 is the weight factor of SD1 to TQ1, and e is a natural constant; TQ2 is the accuracy of water flow change information 2, WD2 is the water flow change stability 2, γ1 is the weight factor of WD2 to TQ2, FX2 is the water flow direction accuracy 2, γ2 is the weight factor of FX2 to TQ2, SD2 is the water flow velocity accuracy 2, and γ3 is the weight factor of SD2 to TQ2.

8. The drone-based aerial survey data processing system according to claim 1, characterized in that: Performing data enhancement on the final image data to obtain the enhanced final image data includes the following steps: Performing linear expansion on the pixel values ​​of the final image data to obtain the final image data after linear expansion; Based on the final image data after linear expansion, the edge portion of the final image data is enhanced to obtain enhanced final image data.

9. The drone-based aerial survey data processing system according to claim 1, characterized in that: The reservoir storage capacity prediction model is constructed based on the final water flow change information, including the following steps: Obtaining historical water flow change information of the reservoir to be measured, wherein the historical water flow change information of the reservoir to be measured includes historical water flow direction, historical water flow velocity, and historical reservoir water storage, and corresponding the historical water flow direction, historical water flow velocity, and historical reservoir water storage one by one according to the timestamp; Preprocess the historical information of water flow changes in the reservoir to be measured; Based on the preprocessed historical water flow change information of the reservoir to be measured, the training set and the test set are divided; The training set is input into the polynomial regression model for training, and a preliminary prediction model of reservoir water storage capacity is obtained; The test set is input into the preliminary prediction model of reservoir water storage capacity for testing to obtain the reservoir water storage capacity prediction model.

10. A method for processing aerial survey data based on a drone, applied to the aerial survey data processing system based on a drone according to any one of claims 1 to 9, characterized in that: The following steps are involved: Using drones to collect multiple sets of image data of the reservoir to be measured from multiple angles and heights, and obtaining initial image data of the reservoir to be measured; The collected initial image data is transmitted to the ground monitoring center in real time via a wireless network, and the initial image data is enhanced to obtain enhanced initial image data; The enhanced initial image data is scored for accuracy through an image accuracy scoring mechanism, and a group of enhanced initial image data with the highest accuracy is selected as the final image data based on the scoring results; Based on the final image data, the water flow change information 1 during the reservoir water inflow period is extracted from the final image data by edge detection method, and the accuracy of the water flow change information 1 is obtained by extracting the accuracy index formula. If the accuracy of the water flow change information 1 meets the standard, the final water flow change information is output; If the accuracy of the water flow change information 1 does not meet the standard, the water flow change information 2 during the reservoir inflow period is extracted from the final image data through the optical flow analysis algorithm. If the accuracy of the water flow change information 2 meets the standard, the final water flow change information is output; If the accuracy of the water flow change information 2 does not meet the standard, the final image data is enhanced to obtain the enhanced final image data. Based on the enhanced final image data, the water flow change information 1 and the water flow change information 2 during the reservoir inflow period are extracted again by the edge detection method and the optical flow analysis algorithm, and the accuracy of the water flow change information extracted by the two methods is obtained again by the extraction accuracy index formula, and the group with the highest accuracy is output as the final water flow change information; A reservoir storage capacity prediction model is constructed based on the final water flow change information.

Images