Urban compensation modeling method and system combined with image recognition

Abstract

The invention discloses a city compensation modeling method and system combined with image recognition, and relates to the technical field of city compensation modeling, and the method comprises the steps: obtaining a historical city three-dimensional model and a historical unmanned plane collection path set of a target city; obtaining K-1 path interaction confrontation areas; obtaining K-1 interactive confrontation sub-region sets; determining K-1 fuzzy region sets; determining centralized to-be-compensated ambiguity; determining an optimized unmanned aerial vehicle flight parameter set; and carrying out image acquisition on the K-1 fuzzy region sets according to the historical unmanned aerial vehicle acquisition path set, carrying out three-dimensional modeling according to an acquisition result, and carrying out compensation updating on a historical city three-dimensional model according to a modeling result to obtain a target compensation city three-dimensional model. The technical problems of low city compensation modeling accuracy and low compensation modeling efficiency in the prior art are solved, and the technical effects of improving the city compensation modeling reliability and improving the modeling quality are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of urban compensation modeling, and in particular to an urban compensation modeling method and system combined with image recognition. Background Art

[0002] Image recognition technology can accurately monitor the expansion of a city or changes in buildings, roads, greenery and other facilities within a region. However, when using drones to collect urban images, due to the high flight speed or large sampling interval, the overlapping area between adjacent paths is insufficient, resulting in missing details in some urban areas, resulting in low accuracy in urban modeling. When compensating for areas with low accuracy in urban modeling, a long analysis time is required, further increasing the cycle of model update. Summary of the invention

[0003] The present application provides a city compensation modeling method and system combined with image recognition, which is used to solve the technical problems of low accuracy and low efficiency of city compensation modeling in the prior art.

[0004] In view of the above problems, the present application provides a city compensation modeling method and system combined with image recognition.

[0005] In a first aspect of the present application, a method for urban compensation modeling combined with image recognition is provided, the method comprising:

[0006] Obtain a historical city 3D model and a set of historical drone collection paths for the target city, wherein K historical drone collection paths in the set are parallel to each other and the interval distance between any two adjacent historical drone collection paths is the same, and K is an integer greater than or equal to 1;

[0007] Based on K environmental parameters of K historical drone collection paths of the target city and the model parameters of the drone, the overlapping area of ​​the drone field of view of two adjacent historical drone collection paths during image collection is determined to obtain K-1 path interaction confrontation areas;

[0008] Obtaining a preset sampling interval and a preset driving speed of the UAV, and dividing the K-1 path interaction confrontation areas in combination with the K-1 area lengths of the K-1 path interaction confrontation areas to obtain a set of K-1 interaction confrontation sub-areas;

[0009] Using the K-1 interactive confrontation sub-region sets as indexes, searching the texture map file of the historical city three-dimensional model, and performing image resolution and overlap rate analysis on the search results to determine K-1 fuzzy region sets, wherein the K-1 fuzzy region sets include K-1 fuzziness degree sets;

[0010] Performing concentrated ambiguity compensation analysis on the K-1 ambiguity sets to determine concentrated ambiguities to be compensated;

[0011] Collecting a set of historical UAV flight parameters for constructing the historical city three-dimensional model, optimizing the set of historical UAV flight parameters according to the concentrated ambiguity to be compensated, and determining an optimized set of UAV flight parameters;

[0012] The UAV is controlled based on the UAV flight parameter set, images of the K-1 fuzzy area sets are collected according to the historical UAV collection path set, and three-dimensional modeling is performed according to the collection results. The historical city three-dimensional model is compensated and updated according to the modeling results to obtain a target compensated city three-dimensional model.

[0013] A second aspect of the present application provides an urban compensation modeling system combined with image recognition, the system comprising:

[0014] A collection path set acquisition module is used to acquire a historical city three-dimensional model and a historical drone collection path set of a target city, wherein K historical drone collection paths in the historical drone collection path set are parallel to each other and the interval distance between any two adjacent historical drone collection paths is the same, and K is an integer greater than or equal to 1;

[0015] A path interaction confrontation area acquisition module is used to determine the overlapping area of ​​the drone field of view of two adjacent historical drone collection paths when performing image collection based on K environmental parameters of the K historical drone collection paths of the target city and the model parameters of the drone, and obtain K-1 path interaction confrontation areas;

[0016] An interactive confrontation sub-area set acquisition module is used to obtain a preset sampling interval and a preset driving speed of the UAV, and divide the K-1 path interactive confrontation areas into K-1 path interactive confrontation areas in combination with the K-1 area lengths to obtain a K-1 interactive confrontation sub-area set;

[0017] A fuzzy region set determination module is used to retrieve the texture map file of the historical city three-dimensional model with the K-1 interactive confrontation sub-region sets as indexes, and perform image resolution and overlap rate analysis on the retrieval results to determine K-1 fuzzy region sets, wherein the K-1 fuzzy region sets include K-1 fuzziness degree sets;

[0018] A centralized ambiguity determination module to be compensated is used to perform centralized ambiguity compensation analysis on the K-1 ambiguity sets to determine centralized ambiguities to be compensated;

[0019] An optimized UAV flight parameter set determination module is used to collect a historical UAV flight parameter set for constructing the historical city three-dimensional model, optimize the historical UAV flight parameter set according to the concentrated ambiguity to be compensated, and determine an optimized UAV flight parameter set;

[0020] The target compensated urban three-dimensional model acquisition module is used to control the drone based on the drone flight parameter set, collect images of the K-1 fuzzy area sets according to the historical drone collection path set, and perform three-dimensional modeling according to the collection results, and compensate and update the historical urban three-dimensional model according to the modeling results to obtain the target compensated urban three-dimensional model.

[0021] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0022] The present application obtains a historical three-dimensional city model and a set of historical drone collection paths of a target city, wherein K historical drone collection paths in the set of historical drone collection paths are parallel to each other and the interval distance between any two adjacent historical drone collection paths is the same, and K is an integer greater than or equal to 1, and then based on K environmental parameters of the K historical drone collection paths of the target city and the model parameters of the drone, determines the overlapping area of ​​the drone field of view of two adjacent historical drone collection paths during image acquisition, obtains K-1 path interaction confrontation areas, and then obtains a preset sampling interval and a preset driving speed of the drone, and divides the K-1 path interaction confrontation areas in combination with the K-1 area lengths of the K-1 path interaction confrontation areas to obtain a set of K-1 interaction confrontation sub-areas, and then uses the K-1 interaction confrontation sub-area set as an index. , retrieve the texture map file of the historical city three-dimensional model, analyze the image resolution and overlap rate of the retrieval results, determine K-1 fuzzy area sets, where the K-1 fuzzy area sets include K-1 fuzziness sets, perform fuzziness concentration compensation analysis on the K-1 fuzziness sets, determine the concentrated fuzziness to be compensated, then collect the historical UAV flight parameter set for building the historical city three-dimensional model, optimize the historical UAV flight parameter set according to the concentrated fuzziness to be compensated, determine the optimized UAV flight parameter set, then control the UAV based on the UAV flight parameter set, collect images of the K-1 fuzzy area set according to the historical UAV collection path set, and perform three-dimensional modeling according to the collection results, and compensate and update the historical city three-dimensional model according to the modeling results to obtain the target compensated city three-dimensional model. The technical effect of efficiently identifying the fuzzy area and fuzziness degree in the city three-dimensional model and thus improving the accuracy and efficiency of compensation modeling is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0024] Figure 1 A schematic diagram of a flow chart of a city compensation modeling method combined with image recognition provided in an embodiment of the present application;

[0025] Figure 2 A schematic diagram of a process for determining the concentrated ambiguity to be compensated in a city compensation modeling method combined with image recognition provided in an embodiment of the present application;

[0026] Figure 3 A schematic diagram of the structure of an urban compensation modeling system combined with image recognition provided in an embodiment of the present application.

[0027] Explanation of the reference numerals: acquisition module 11 for collecting path set, module 12 for acquiring path interactive confrontation area, module 13 for acquiring interactive confrontation sub-area set, module 14 for determining fuzzy area set, module 15 for determining concentrated fuzziness to be compensated, module 16 for determining optimized UAV flight parameter set, module 17 for acquiring target compensated urban three-dimensional model. DETAILED DESCRIPTION

[0028] The present application provides a city compensation modeling method and system combined with image recognition, which is used to solve the technical problems of low accuracy and low efficiency of city compensation modeling in the prior art.

[0029] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0030] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0031] Embodiment 1, as Figure 1 As shown, the present application provides a city compensation modeling method combined with image recognition, wherein the method includes:

[0032] S100: Acquire a historical city three-dimensional model and a historical drone collection path set of a target city, wherein K historical drone collection paths in the historical drone collection path set are parallel to each other and the interval distance between any two adjacent historical drone collection paths is the same, and K is an integer greater than or equal to 1;

[0033] In a possible embodiment, the target city is any city that needs to be 3D modeled. The historical city 3D model is a 3D model obtained by scanning and modeling the target city using a drone equipped with a laser radar or a camera according to a set of historical drone collection paths in historical time. Obtaining the historical city 3D model of the target city provides basic reference data for subsequent compensation and updating.

[0034] Preferably, the historical drone collection path set includes K historical drone collection paths, wherein the K historical drone collection paths are parallel to each other and the interval between any two adjacent historical drone collection paths is the same, thereby ensuring omission or duplication when the drone collects images, and improving the accuracy and efficiency of stitching the images obtained after scanning the target city. K is an integer greater than or equal to 1. Optional,

[0035] S200: Based on K environmental parameters of K historical drone collection paths of the target city and model parameters of the drone, determine the overlapping area of ​​the drone field of view of two adjacent historical drone collection paths during image collection, and obtain K-1 path interaction confrontation areas;

[0036] In an embodiment of the present application, it is first necessary to obtain K environmental parameters of K historical drone collection paths of the target city. The environmental parameters of the historical drone collection paths include but are not limited to weather conditions (such as wind speed and humidity), lighting conditions, terrain undulations (such as building height and density), and other factors that may affect the flight stability and field of view of the drone. The drone model parameters are used to describe the type and performance of the drone, including flight altitude, flight speed, field of view angle (lens field of view angle), image resolution, camera type, etc. These parameters are crucial for calculating the field of view and image acquisition accuracy of the drone.

[0037] For each two adjacent paths in the historical drone collection path, it is necessary to calculate the image collection range (i.e., field of view) of the drone on each path based on the drone's flight altitude, viewing angle and other parameters. Usually, the field of view of a drone is calculated based on its camera field of view (FOV) and flight altitude, and those skilled in the art can determine it using geometric methods.

[0038] According to the spatial relationship between two adjacent historical drone collection paths and the range of the drone's field of view, the field of view overlap area of ​​the two paths is calculated, and the field of view overlap area is used as the K-1 path interaction confrontation area. By obtaining the K-1 path interaction confrontation area, the foundation is laid for model splicing after 3D modeling of the images obtained by drones according to K historical drone collection paths of the target city.

[0039] Since there are K historical drone collection paths, there will be K-1 adjacent historical drone collection path pairs. The drone field of view overlap area formed when the drone collects images on the K-1 adjacent historical drone collection path pairs is used as the K-1 path interaction confrontation area. Each path interaction confrontation area represents the field of view overlap between two adjacent paths.

[0040] By obtaining K-1 path interaction confrontation areas, the foundation is laid for subsequent image analysis and 3D modeling compensation, achieving the technical effect of better fusion, reconstruction and compensation of image data and improving the accuracy and efficiency of compensation modeling.

[0041] For example, assuming that there are three historical drone collection paths in the target city (i.e., K=3), it is necessary to calculate the overlapping area between two adjacent paths: an interactive confrontation area is formed between the first path and the second path; a second interactive confrontation area is formed between the second path and the third path; these two interactive confrontation areas will serve as the basis for subsequent compensation modeling analysis, and finally two path interactive confrontation areas will be obtained.

[0042] S300: Obtaining a preset sampling interval and a preset driving speed of the drone, dividing the K-1 path interaction confrontation area in combination with the K-1 area lengths of the K-1 path interaction confrontation area to obtain a set of K-1 interaction confrontation sub-areas;

[0043] Further, the preset sampling interval and the preset driving speed of the drone are obtained, and the K-1 path interaction confrontation area is divided into K-1 path interaction confrontation areas in combination with the K-1 area lengths of the K-1 path interaction confrontation areas to obtain a set of K-1 interaction confrontation sub-areas. The step S300 of the embodiment of the present application further includes:

[0044] Multiplying the preset sampling interval by the preset driving speed to obtain a preset sampling length;

[0045] Obtaining the number of K sub-region divisions according to the ratio of the preset sampling length to the length of the K-1 regions and the preset sampling length;

[0046] Based on the K sub-area division numbers and the preset sampling length, the K-1 path interaction confrontation areas are divided in sequence according to the path collection direction to obtain the K-1 interaction confrontation sub-area sets.

[0047] In a possible embodiment, the preset sampling interval is a time interval for collecting image data at a certain interval during the flight of the drone, which is preset by a person skilled in the art. The preset sampling interval is set by a person skilled in the art according to the required image resolution and image update frequency. A smaller sampling interval can improve the quality of image data, but may require higher computing power and storage capacity. The preset driving speed is the speed of the drone along the collection path during the flight, which is preset by a person skilled in the art. The driving speed affects the viewing angle relationship between the drone and the ground target, thereby affecting the clarity of the image and the coverage of the collection. Generally speaking, when the flight speed is lower, a higher image resolution can be obtained, but it will also result in a longer flight time.

[0048] Optionally, based on the preset sampling interval and driving speed, the flight distance (sampling length) corresponding to each sampling can be calculated. The formula is: sampling length = preset sampling interval × preset driving speed sampling length = preset sampling interval × preset driving speed. For example, assuming that the preset sampling interval of the drone is 10 seconds and the preset driving speed is 5 meters / second, the sampling length of each sampling is 50 meters. This sampling length means that the drone will collect images every 50 meters it flies.

[0049] Preferably, each path interaction confrontation area is the overlapping area of ​​the drone's field of view when the drone collects images of the target city according to two adjacent historical drone collection paths, and the drone collects according to the sampling interval and the preset driving speed. Therefore, each path interaction confrontation area can be equally divided into a set of interaction confrontation sub-areas. Among them, the area of ​​each interaction confrontation sub-area in each interaction confrontation sub-area set is the same. By dividing the K-1 path interaction confrontation areas, K-1 interaction confrontation sub-area sets are obtained. The number of interaction confrontation sub-areas in each interaction confrontation sub-area set is determined by the area length of the path interaction confrontation area. The longer the area length, the greater the number of interaction confrontation sub-areas divided accordingly.

[0050] In one possible embodiment, the length of the K-1 path interaction confrontation areas is counted, and the ratio of the K-1 area lengths to the preset sampling length is calculated to obtain the number of K sub-area divisions. The ratio indicates how many times images can be collected in the area according to the sampling interval and the flight speed. Preferably, the result of the comparison value calculation is rounded. For example, if the length of an area is 100 meters and the sampling length is 50 meters, then the ratio is 2, indicating that 2 images can be collected in the area.

[0051] Based on the number of K sub-area divisions and the preset sampling length, the K-1 path interactive confrontation areas are divided in turn according to the path acquisition direction to obtain the K-1 interactive confrontation sub-area sets. This means that the length of each area is the same as the preset sampling length. Assuming that the length of a path interactive confrontation area is 100 meters and the preset sampling length is 50, the path interactive confrontation area is divided into two 50-meter-long interactive confrontation sub-areas. Each interactive confrontation sub-area will become the basic unit for subsequent image acquisition and texture mapping.

[0052] S400: using the K-1 interactive adversarial sub-region sets as indexes, searching for texture mapping files of the historical city three-dimensional model, and performing image resolution and overlap rate analysis on the search results to determine K-1 fuzzy region sets, wherein the K-1 fuzzy region sets include K-1 fuzziness degree sets;

[0053] In one embodiment of the present application, the three-dimensional model of the historical city is also obtained by three-dimensional simulation of texture map files collected by a drone. Therefore, the three-dimensional model of the historical city is retrieved using the K-1 interactive adversarial sub-region sets as indexes to obtain texture map files corresponding to these sub-regions. These texture map files are image data in the three-dimensional model of the historical city that are used to describe the urban surface details (such as buildings, roads, trees, etc.) of the K-1 interactive adversarial sub-region sets.

[0054] Then, the image data in the search results are analyzed from the two dimensions of image resolution and overlap rate, and the sub-regions with blur in the K-1 interactive confrontation sub-region sets are determined, and added to the K-1 fuzzy region sets. The K-1 fuzzy region sets include K-1 fuzziness degree sets. The fuzziness degree is used to describe the image blur degree of the interactive confrontation sub-region. The lower the image resolution and overlap rate, the greater the fuzziness degree. When the image resolution is constant, the greater the overlap rate, the more images the interactive confrontation sub-region can collect, and the smaller the fuzziness degree.

[0055] By retrieving and analyzing the image quality of the texture map files of the historical city 3D model according to K-1 interactive adversarial sub-region sets, and combining the image resolution and overlap rate, the fuzzy areas can be identified, and the areas that need to be compensated or optimized can be effectively determined, thereby achieving the technical effect of improving the quality and accuracy of the city 3D model.

[0056] Further, using the K-1 interactive adversarial sub-region sets as indexes, searching for texture mapping files of the historical city three-dimensional model, and performing image resolution and overlap rate analysis on the search results to determine K-1 fuzzy region sets, step S400 of the embodiment of the present application further includes:

[0057] Using the K-1 interactive confrontation sub-region sets as indexes, the texture map files of the historical city three-dimensional model are retrieved to obtain K-1 sub-region drone-collected image group sets, each sub-region drone-collected image group including two region drone-collected images obtained by drones according to two adjacent historical drone-collected paths for corresponding interactive confrontation sub-regions;

[0058] Traversing the K-1 sub-region drone-collected image group sets to extract pixel width and pixel height, and obtaining K-1 sub-region image resolution group sets according to the extraction results;

[0059] Traversing the K-1 sub-region drone-collected image group sets to identify the overlapping area of ​​the images within the group, and respectively comparing the K-1 overlapping area identification result sets with the area of ​​any interactive adversarial sub-region in the K-1 interactive adversarial sub-region sets to obtain the K-1 sub-region image overlapping rate set;

[0060] Calling a fuzziness identifier to perform fuzziness identification on the K-1 sub-region image resolution group sets and the K-1 sub-region image overlap rate sets respectively, to obtain K-1 initial fuzziness sets;

[0061] It is determined whether the K-1 initial fuzzy degree sets are greater than or equal to the preset fuzzy degree. If so, the corresponding interactive confrontation sub-regions are used as fuzzy regions to obtain the K-1 fuzzy region sets.

[0062] Furthermore, step S400 in the embodiment of the present application further includes:

[0063] Collecting multiple sample sub-region image resolution groups, multiple sample sub-region image overlap rates and multiple sample initial blurs as training sample data sets;

[0064] The training sample data set is used to iteratively train the network framework of the ambiguity identifier based on the feedforward neural network, and the two-to-one mapping relationship between the sub-region image resolution group and the sub-region image overlap rate and the initial ambiguity is learned until the convergence requirement is met, thereby obtaining the trained ambiguity identifier.

[0065] In a possible embodiment, the K-1 sets of interactive confrontation sub-regions are used as indexes to retrieve the corresponding texture map files in the historical city three-dimensional model, and obtain the K-1 sub-region drone acquisition image group sets. Each sub-region drone acquisition image group includes two regional drone acquisition images in which the drone acquires images of the corresponding interactive confrontation sub-region according to two adjacent historical drone acquisition paths. Exemplarily, each sub-region drone acquisition image group includes two images: one is a captured image of the interactive confrontation sub-region from path A, and the other is a captured image of the interactive confrontation sub-region from path B. Among them, path A is adjacent to path B.

[0066] Perform resolution analysis on the images of each sub-region drone image group set, extract the pixel width (number of horizontal pixels) and pixel height (number of vertical pixels) of each sub-region drone image, and then calculate the image resolution based on the pixel width and pixel height to obtain the image resolution group set of K-1 sub-regions.

[0067] For each K-1 sub-region UAV image group, the overlapping area is identified, the overlapping area of ​​the image is calculated, and the ratio of the overlapping area to the area of ​​any interactive adversarial sub-region in the corresponding interactive adversarial sub-region set is calculated to obtain the image overlap rate and obtain the image overlap rate set of K-1 sub-regions.

[0068] In a possible embodiment, the blur identifier is a pre-trained functional model for intelligently analyzing the image resolution group and the image overlap rate to determine the blur. The blur identifier will combine the image resolution group and the image overlap rate to calculate the blur value of the image of each interactive adversarial sub-region and obtain the initial blur set of K-1 sub-regions.

[0069] Optionally, multiple sample sub-region image resolution groups, multiple sample sub-region image overlap rates, and multiple sample initial blurs are collected to form a training sample data set. A blur identifier is constructed using a feedforward neural network-based architecture. The network will learn a two-to-one mapping relationship between image resolution and image overlap rate and initial blur. The neural network is iteratively trained using the training sample data set, and the network weights are continuously adjusted until the network converges and meets the preset accuracy requirements. A trained blur identifier that can accurately identify blurry areas is obtained.

[0070] The trained ambiguity identifier is used to perform ambiguity identification on the K-1 sub-region image resolution group sets and the K-1 sub-region image overlap rate sets, respectively, to determine and output all ambiguities, and to obtain the K-1 initial ambiguity sets.

[0071] Then, according to the size of the initial fuzziness in the K-1 initial fuzziness sets, determine which interactive confrontation sub-regions need to be marked as fuzzy regions. Determine whether the K-1 initial fuzziness sets are greater than or equal to the preset fuzziness, and if so, use the corresponding interactive confrontation sub-regions as fuzzy regions to obtain the K-1 fuzzy region sets.

[0072] By combining the image resolution, overlap rate and fuzziness recognition technology, the fuzzy areas in the historical city 3D model can be effectively identified, and the fuzziness identifier can be used to further improve the recognition accuracy. In the process of optimizing image acquisition and modeling, these fuzzy areas will be used as key areas for supplementary acquisition or optimization processing, thereby improving the quality and accuracy of the final city compensation 3D model.

[0073] S500: performing ambiguity centralized compensation analysis on the K-1 ambiguity sets to determine centralized ambiguities to be compensated;

[0074] Further, such as Figure 2 As shown, the K-1 ambiguity sets are subjected to ambiguity centralized compensation analysis to determine the centralized ambiguity to be compensated. The step S500 of the embodiment of the present application further includes:

[0075] Taking each ambiguity in the K-1 ambiguity sets as an analysis particle point, constructing a compensation analysis space, wherein the compensation analysis space includes a plurality of analysis particle points;

[0076] Calculating the fuzziness mean of the multiple analysis particle points to obtain a central analysis particle point;

[0077] Taking the central analysis particle point as the starting point, constructing a neighborhood in the compensation analysis space according to a preset iteration step length, and determining a central analysis neighborhood, wherein a plurality of analysis particle points in the central analysis neighborhood are analysis particle points whose distances to the starting point are within the preset iteration step length;

[0078] Iteratively expand the central analysis neighborhood according to the preset iteration step to obtain a target analysis neighborhood;

[0079] The fuzziness mean of multiple analysis particle points in the target analysis neighborhood is calculated to obtain the concentrated fuzziness to be compensated.

[0080] Further, the central analysis neighborhood is iteratively expanded according to the preset iteration step to obtain a target analysis neighborhood. Step S500 of the embodiment of the present application further includes:

[0081] Iterating leftwards on the leftmost analysis particle point of the central analysis neighborhood and iterating rightwards on the rightmost analysis particle point of the central analysis neighborhood respectively according to the preset iteration step length to obtain a first left iterative analysis neighborhood and a first right iterative analysis neighborhood;

[0082] Determine whether the concentration density of the first left iterative analysis neighborhood and the concentration density of the first right iterative analysis neighborhood are both greater than or equal to the concentration density of the center analysis neighborhood; if so, obtain a union of the first left iterative analysis neighborhood and the first right iterative analysis neighborhood to obtain a first-stage iterative analysis neighborhood;

[0083] Again, iterating leftwards the leftmost analysis particle point of the iterative analysis neighborhood of the first stage and iterating rightwards the rightmost analysis particle point of the iterative analysis neighborhood of the first stage respectively according to the preset iteration step size;

[0084] After multiple bilateral iterations, when the concentration density of the left stage iterative analysis neighborhood and the concentration density of the right stage iterative analysis neighborhood obtained in the current iteration are both less than the concentration density of the stage iterative analysis neighborhood obtained in the previous iteration, the iteration is stopped and the stage iterative analysis neighborhood obtained in the previous iteration is used as the target analysis neighborhood.

[0085] Furthermore, step S500 of the embodiment of the present application further includes:

[0086] When only the concentration density of the first left iterative analysis neighborhood or the concentration density of the first right iterative analysis neighborhood is greater than or equal to the concentration density of the central analysis neighborhood, the first left iterative analysis neighborhood or the first right iterative analysis neighborhood is used as the first-stage iterative analysis neighborhood;

[0087] Iterate the leftmost analysis particle point of the first-stage iterative analysis neighborhood to the left or iterate the rightmost analysis particle point of the first-stage iterative analysis neighborhood to the right according to the preset iteration step;

[0088] After multiple one-sided iterations, when the concentration density of the left stage iterative analysis neighborhood or the concentration density of the right stage iterative analysis neighborhood obtained in the current iteration is less than the concentration density of the stage iterative analysis neighborhood obtained in the previous iteration, the iteration is stopped and the stage iterative analysis neighborhood obtained in the previous iteration is used as the target analysis neighborhood.

[0089] In a possible embodiment, the K-1 fuzzy degree sets reflect the fuzziness of the K-1 fuzzy area sets in the K-1 interactive confrontation sub-area sets in the historical city three-dimensional model. By performing fuzzy concentration compensation analysis on the K-1 fuzzy degree sets, the general fuzziness of the K-1 fuzzy area sets is determined, and the concentrated fuzziness to be compensated is obtained. The technical effect of providing data support for subsequent urban compensation modeling is achieved.

[0090] By taking each ambiguity in the K-1 ambiguity sets as an analysis particle point, a compensation analysis space is constructed, wherein the compensation analysis space includes a plurality of analysis particle points. Then, the ambiguity mean of the plurality of analysis particle points is calculated, and the average ambiguity of the compensation analysis space when considering the analysis particle points at the edge positions is determined, thereby obtaining a central analysis particle point.

[0091] Furthermore, taking the central analysis particle point as the starting point, a neighborhood is constructed in the compensation analysis space according to a preset iteration step (the difference in ambiguity in a single iteration pre-set by a technician in this field), and the central analysis neighborhood is determined, wherein the multiple analysis particle points in the central analysis neighborhood are analysis particle points whose distance to the starting point is within the preset iteration step.

[0092] Then, the central analysis neighborhood is iteratively expanded according to the preset iteration step until the edge of the area where the analysis particle points are densely distributed is reached to obtain the target analysis neighborhood. By calculating the ambiguity mean of multiple analysis particle points in the target analysis neighborhood, the general ambiguity of the K-1 ambiguity sets is determined when the ambiguities with a relatively biased distribution and less frequent occurrence are not considered, and the calculation result is used as the concentrated ambiguity to be compensated.

[0093] Preferably, the leftmost analysis particle point of the central analysis neighborhood is iterated to the left and the rightmost analysis particle point of the central analysis neighborhood is iterated to the right according to the preset iteration step, respectively, to obtain a first left iterative analysis neighborhood and a first right iterative analysis neighborhood. Wherein, the distance from the analysis particle point in the first left iterative analysis neighborhood to the leftmost analysis particle of the central analysis neighborhood is less than or equal to the preset iteration step. The distance from the analysis particle point in the first right iterative analysis neighborhood to the rightmost analysis particle of the central analysis neighborhood is less than or equal to the preset iteration step.

[0094] In one embodiment, the central neighborhood ambiguity difference between the leftmost analysis particle and the rightmost analysis particle of the central analysis neighborhood is calculated, and the number of analysis particle points in the central analysis neighborhood is counted, and the ratio of the counted number of particle points to the central neighborhood ambiguity difference is used as the concentration density of the central analysis neighborhood. The concentration density of the central analysis neighborhood reflects the concentration density of the analysis particle points in the central analysis neighborhood. Based on the same acquisition principle, the concentration density of the first left iterative analysis neighborhood and the concentration density of the first right iterative analysis neighborhood are calculated.

[0095] Determine whether the concentration density of the first left iterative analysis neighborhood and the concentration density of the first right iterative analysis neighborhood are both greater than or equal to the concentration density of the central analysis neighborhood. If so, it indicates that the distribution of analysis particle points in the first left iterative analysis neighborhood and the first right iterative analysis neighborhood are both relatively dense. At this time, the first left iterative analysis neighborhood and the first right iterative analysis neighborhood are unioned to obtain the first stage iterative analysis neighborhood.

[0096] Then, the leftmost analysis particle point of the first stage iterative analysis neighborhood is iterated to the left and the rightmost analysis particle point of the first stage iterative analysis neighborhood is iterated to the right according to the preset iteration step size again, after multiple bilateral iterations, until the concentration density of the left stage iterative analysis neighborhood and the concentration density of the right stage iterative analysis neighborhood obtained in the current iteration are both less than the concentration density of the stage iterative analysis neighborhood obtained in the previous iteration, at this time, both sides have reached the edge of the area where the analysis particle points are densely distributed, the iteration is stopped, and the stage iterative analysis neighborhood obtained in the previous iteration is used as the target analysis neighborhood.

[0097] In one embodiment, when only the concentration density of the first left iterative analysis neighborhood or the concentration density of the first right iterative analysis neighborhood is greater than or equal to the concentration density of the central analysis neighborhood, it indicates that the analysis density of the analysis particle points of the first left iterative analysis neighborhood or the analysis density of the analysis particle points of the first right iterative analysis neighborhood is greater than that of the central analysis neighborhood. At this time, the first left iterative analysis neighborhood or the first right iterative analysis neighborhood is used as the first stage iterative analysis neighborhood. The leftmost analysis particle point of the first stage iterative analysis neighborhood is iterated to the left or the rightmost analysis particle point of the first stage iterative analysis neighborhood is iterated to the right according to the preset iteration step.

[0098] After multiple one-sided iterations, when the concentration density of the left stage iterative analysis neighborhood or the concentration density of the right stage iterative analysis neighborhood obtained in the current iteration is less than the concentration density of the stage iterative analysis neighborhood obtained in the previous iteration, the iteration is stopped and the stage iterative analysis neighborhood obtained in the previous iteration is used as the target analysis neighborhood.

[0099] By determining the concentrated fuzziness to be compensated, data support is provided for subsequent urban compensation modeling, thereby achieving the technical effect of improving the accuracy of compensation modeling.

[0100] S600: collecting a set of historical UAV flight parameters for constructing the historical city three-dimensional model, optimizing the set of historical UAV flight parameters according to the concentrated ambiguity to be compensated, and determining an optimized UAV flight parameter set;

[0101] S700: Control the drone based on the drone flight parameter set, collect images of the K-1 fuzzy area sets according to the historical drone collection path set, perform three-dimensional modeling based on the collection results, compensate and update the historical city three-dimensional model based on the modeling results, and obtain a target compensated city three-dimensional model.

[0102] In one embodiment, a set of historical UAV flight parameters for model construction is extracted from the historical city three-dimensional model. The set of historical UAV flight parameters includes flight path, flight altitude, speed, attitude, etc. Based on the results of the centralized analysis of ambiguity, the flight parameters (such as flight altitude, flight attitude, etc.) are adjusted to reduce the error caused by the fuzzy area.

[0103] Optionally, multiple sample UAV flight parameter sets, multiple sample sets of ambiguities to be compensated, and multiple sample optimized UAV flight parameter sets are obtained as sample data, and the network framework constructed based on the feedforward neural network is supervised and trained using the sample data until convergence, thereby obtaining a trained flight parameter optimizer. The ambiguities to be compensated in the set and the historical UAV flight parameter set are input into the flight parameter optimizer for intelligent analysis to obtain the optimized UAV flight parameter set. This ensures that when the UAV using the optimized UAV flight parameter set is used for image acquisition, the influence of the fuzzy area in the historical model can be effectively compensated.

[0104] Using the optimized UAV flight parameter set obtained in step S600, the UAV is controlled to fly according to the historical flight path. According to the historical UAV acquisition path set, the UAV is organized to collect images of the K-1 fuzzy area set in step S500 to ensure that the image data covers the fuzzy area and improve the image quality as much as possible. Using the image data collected by the UAV, the image processing algorithm is used for 3D reconstruction to obtain a high-precision 3D model of the target area. Combined with the collected 3D data, the historical city 3D model is compensated and updated. By comparing the difference between the updated 3D model and the original model, error correction and detail supplementation are performed to finally obtain the target compensated 3D model of the city.

[0105] By optimizing flight parameters and precise image acquisition, the influence of blurred areas was compensated, the accuracy of 3D modeling was improved, and effective compensation and updating of historical city 3D models was achieved.

[0106] In summary, the embodiments of the present application have at least the following technical effects:

[0107] 1. This application can determine the overlapping areas between adjacent paths by combining historical drone collection paths with environmental parameters, thereby effectively filling in the vacant areas during the image collection process. The analysis and compensation of blurred areas ensure that the details of the historical city 3D model are reconstructed more accurately, avoiding model distortion caused by insufficient viewing angle or resolution, thereby greatly improving the accuracy of the compensation model.

[0108] 2. By analyzing historical UAV flight parameters and sampling intervals, flight speed and other factors, the flight path and parameters of the UAV can be accurately optimized, ensuring the overlap rate and image resolution of the images obtained when collecting images in the fuzzy area, and correcting the fuzzy areas in the historical city three-dimensional model, achieving the technical effect of improving modeling efficiency and compensating modeling quality.

[0109] Embodiment 2 is based on the same inventive concept as the city compensation modeling method combined with image recognition in the above embodiment. Figure 3 As shown, the present application provides a city compensation modeling system combined with image recognition, and the system and method embodiments in the present application embodiments are based on the same inventive concept. The system includes:

[0110] The collection path set acquisition module 11 is used to obtain the historical city three-dimensional model and the historical drone collection path set of the target city, wherein K historical drone collection paths in the historical drone collection path set are parallel to each other and the interval distance between any two adjacent historical drone collection paths is the same, and K is an integer greater than or equal to 1;

[0111] A path interaction confrontation area acquisition module 12 is used to determine the overlapping area of ​​the drone field of view of two adjacent historical drone collection paths when performing image collection based on K environmental parameters of the K historical drone collection paths of the target city and the model parameters of the drone, and obtain K-1 path interaction confrontation areas;

[0112] The interactive confrontation sub-area set obtaining module 13 is used to obtain a preset sampling interval and a preset driving speed of the UAV, and divide the K-1 path interactive confrontation areas into K-1 path interactive confrontation areas in combination with the K-1 area lengths to obtain a K-1 interactive confrontation sub-area set;

[0113] A fuzzy region set determination module 14 is used to retrieve the texture map file of the historical city three-dimensional model with the K-1 interactive confrontation sub-region sets as indexes, and perform image resolution and overlap rate analysis on the retrieval results to determine K-1 fuzzy region sets, wherein the K-1 fuzzy region sets include K-1 fuzziness degree sets;

[0114] A centralized ambiguity determination module 15 for performing centralized ambiguity compensation analysis on the K-1 ambiguity sets to determine centralized ambiguities for compensation;

[0115] An optimized UAV flight parameter set determination module 16 is used to collect a historical UAV flight parameter set for constructing the historical city three-dimensional model, optimize the historical UAV flight parameter set according to the concentrated ambiguity to be compensated, and determine an optimized UAV flight parameter set;

[0116] The target compensated urban three-dimensional model acquisition module 17 is used to control the drone based on the drone flight parameter set, perform image acquisition of the K-1 fuzzy area set according to the historical drone acquisition path set, and perform three-dimensional modeling according to the acquisition results, and compensate and update the historical urban three-dimensional model according to the modeling results to obtain the target compensated urban three-dimensional model.

[0117] Furthermore, the fuzzy region set determination module 14 is used to perform the following steps:

[0118] Using the K-1 interactive confrontation sub-region sets as indexes, the texture map files of the historical city three-dimensional model are retrieved to obtain K-1 sub-region drone-collected image group sets, each sub-region drone-collected image group including two region drone-collected images obtained by drones according to two adjacent historical drone-collected paths for corresponding interactive confrontation sub-regions;

[0119] Traversing the K-1 sub-region drone-collected image group sets to extract pixel width and pixel height, and obtaining K-1 sub-region image resolution group sets according to the extraction results;

[0120] Traversing the K-1 sub-region drone-collected image group sets to identify the overlapping area of ​​the images within the group, and respectively comparing the K-1 overlapping area identification result sets with the area of ​​any interactive adversarial sub-region in the K-1 interactive adversarial sub-region sets to obtain the K-1 sub-region image overlapping rate set;

[0121] Calling a fuzziness identifier to perform fuzziness identification on the K-1 sub-region image resolution group sets and the K-1 sub-region image overlap rate sets respectively, to obtain K-1 initial fuzziness sets;

[0122] It is determined whether the K-1 initial fuzzy degree sets are greater than or equal to the preset fuzzy degree. If so, the corresponding interactive confrontation sub-regions are used as fuzzy regions to obtain the K-1 fuzzy region sets.

[0123] Furthermore, the fuzzy region set determination module 14 is used to perform the following steps:

[0124] Collecting multiple sample sub-region image resolution groups, multiple sample sub-region image overlap rates and multiple sample initial blurs as training sample data sets;

[0125] The training sample data set is used to iteratively train the network framework of the ambiguity identifier based on the feedforward neural network, and the two-to-one mapping relationship between the sub-region image resolution group and the sub-region image overlap rate and the initial ambiguity is learned until the convergence requirement is met, thereby obtaining the trained ambiguity identifier.

[0126] Furthermore, the centralized ambiguity to be compensated determining module 15 is used to perform the following steps:

[0127] Taking each ambiguity in the K-1 ambiguity sets as an analysis particle point, constructing a compensation analysis space, wherein the compensation analysis space includes a plurality of analysis particle points;

[0128] Calculating the fuzziness mean of the multiple analysis particle points to obtain a central analysis particle point;

[0129] Taking the central analysis particle point as the starting point, constructing a neighborhood in the compensation analysis space according to a preset iteration step length, and determining a central analysis neighborhood, wherein a plurality of analysis particle points in the central analysis neighborhood are analysis particle points whose distances to the starting point are within the preset iteration step length;

[0130] Iteratively expand the central analysis neighborhood according to the preset iteration step to obtain a target analysis neighborhood;

[0131] The fuzziness mean of multiple analysis particle points in the target analysis neighborhood is calculated to obtain the concentrated fuzziness to be compensated.

[0132] Furthermore, the centralized ambiguity to be compensated determining module 15 is used to perform the following steps:

[0133] Iterating leftwards on the leftmost analysis particle point of the central analysis neighborhood and iterating rightwards on the rightmost analysis particle point of the central analysis neighborhood respectively according to the preset iteration step length to obtain a first left iterative analysis neighborhood and a first right iterative analysis neighborhood;

[0134] Determine whether the concentration density of the first left iterative analysis neighborhood and the concentration density of the first right iterative analysis neighborhood are both greater than or equal to the concentration density of the center analysis neighborhood; if so, obtain a union of the first left iterative analysis neighborhood and the first right iterative analysis neighborhood to obtain a first-stage iterative analysis neighborhood;

[0135] Iterate the leftmost analysis particle point of the first stage iterative analysis neighborhood to the left and iterate the rightmost analysis particle point of the first stage iterative analysis neighborhood to the right again according to the preset iteration step size. After multiple iterations, when the concentration density of the left stage iterative analysis neighborhood and the concentration density of the right stage iterative analysis neighborhood obtained in the current iteration are both less than the concentration density of the stage iterative analysis neighborhood obtained in the previous iteration, stop the iteration and use the stage iterative analysis neighborhood obtained in the previous iteration as the target analysis neighborhood.

[0136] Furthermore, the centralized ambiguity to be compensated determining module 15 is used to perform the following steps:

[0137] When the concentration density of the first left iterative analysis neighborhood or the concentration density of the first right iterative analysis neighborhood is greater than or equal to the concentration density of the central analysis neighborhood, the first left iterative analysis neighborhood or the first right iterative analysis neighborhood is used as the first stage iterative analysis neighborhood;

[0138] Iterate the leftmost analysis particle point of the first stage iterative analysis neighborhood to the left or iterate the rightmost analysis particle point of the first stage iterative analysis neighborhood to the right according to the preset iteration step size. After multiple one-sided iterations, when the concentration density of the left stage iterative analysis neighborhood or the concentration density of the right stage iterative analysis neighborhood obtained in the current iteration is less than the concentration density of the stage iterative analysis neighborhood obtained in the previous iteration, stop the iteration and use the stage iterative analysis neighborhood obtained in the previous iteration as the target analysis neighborhood.

[0139] Furthermore, the interactive confrontation sub-region set obtaining module 13 is used to perform the following steps:

[0140] Multiplying the preset sampling interval by the preset driving speed to obtain a preset sampling length;

[0141] Calculate the ratio of the length of the K-1 regions to the preset sampling length to obtain the number of K sub-region divisions;

[0142] Based on the K sub-area division numbers and the preset sampling length, the K-1 path interaction confrontation areas are divided in sequence according to the path collection direction to obtain the K-1 interaction confrontation sub-area sets.

[0143] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0144] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0145] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A city compensation modeling method combined with image recognition, characterized in that: The method comprises: Obtain a historical city 3D model and a set of historical drone collection paths for the target city, wherein K historical drone collection paths in the set are parallel to each other and the interval distance between any two adjacent historical drone collection paths is the same, and K is an integer greater than or equal to 1; Based on K environmental parameters of K historical drone collection paths of the target city and the model parameters of the drone, the overlapping area of ​​the drone field of view of two adjacent historical drone collection paths during image collection is determined to obtain K-1 path interaction confrontation areas; Obtaining a preset sampling interval and a preset driving speed of the UAV, and dividing the K-1 path interaction confrontation areas in combination with the K-1 area lengths of the K-1 path interaction confrontation areas to obtain a set of K-1 interaction confrontation sub-areas; Using the K-1 interactive confrontation sub-region sets as indexes, searching the texture map file of the historical city three-dimensional model, and performing image resolution and overlap rate analysis on the search results to determine K-1 fuzzy region sets, wherein the K-1 fuzzy region sets include K-1 fuzziness degree sets; Performing concentrated ambiguity compensation analysis on the K-1 ambiguity sets to determine concentrated ambiguities to be compensated; Collecting a set of historical UAV flight parameters for constructing the historical city three-dimensional model, optimizing the set of historical UAV flight parameters according to the concentrated ambiguity to be compensated, and determining an optimized set of UAV flight parameters; The UAV is controlled based on the UAV flight parameter set, images of the K-1 fuzzy area sets are collected according to the historical UAV collection path set, and three-dimensional modeling is performed according to the collection results. The historical city three-dimensional model is compensated and updated according to the modeling results to obtain a target compensated city three-dimensional model.

2. The urban compensation modeling method combined with image recognition as claimed in claim 1, characterized in that: Using the K-1 interactive confrontation sub-region sets as indexes, the texture map files of the historical city three-dimensional model are retrieved, and the image resolution and overlap rate of the retrieval results are analyzed to determine the K-1 fuzzy region sets, including: Using the K-1 interactive confrontation sub-region sets as indexes, the texture map files of the historical city three-dimensional model are retrieved to obtain K-1 sub-region drone-collected image group sets, each sub-region drone-collected image group including two region drone-collected images obtained by drones according to two adjacent historical drone-collected paths for corresponding interactive confrontation sub-regions; Traversing the K-1 sub-region drone-collected image group sets to extract pixel width and pixel height, and obtaining K-1 sub-region image resolution group sets according to the extraction results; Traversing the K-1 sub-region drone-collected image group sets to identify the overlapping area of ​​the images within the group, and respectively comparing the K-1 overlapping area identification result sets with the area of ​​any interactive adversarial sub-region in the K-1 interactive adversarial sub-region sets to obtain the K-1 sub-region image overlapping rate set; Calling a fuzziness identifier to perform fuzziness identification on the K-1 sub-region image resolution group sets and the K-1 sub-region image overlap rate sets respectively, to obtain K-1 initial fuzziness sets; It is determined whether the K-1 initial fuzzy degree sets are greater than or equal to the preset fuzzy degree. If so, the corresponding interactive confrontation sub-regions are used as fuzzy regions to obtain the K-1 fuzzy region sets.

3. The urban compensation modeling method combined with image recognition as claimed in claim 2, characterized in that: include: Collecting multiple sample sub-region image resolution groups, multiple sample sub-region image overlap rates and multiple sample initial blurs as training sample data sets; The training sample data set is used to iteratively train the network framework of the ambiguity identifier based on the feedforward neural network, and the two-to-one mapping relationship between the sub-region image resolution group and the sub-region image overlap rate and the initial ambiguity is learned until the convergence requirement is met, thereby obtaining the trained ambiguity identifier.

4. The urban compensation modeling method combined with image recognition as claimed in claim 1, characterized in that: Performing ambiguity centralized compensation analysis on the K-1 ambiguity sets to determine centralized ambiguities to be compensated includes: Taking each ambiguity in the K-1 ambiguity sets as an analysis particle point, constructing a compensation analysis space, wherein the compensation analysis space includes a plurality of analysis particle points; Calculating the fuzziness mean of the multiple analysis particle points to obtain a central analysis particle point; Taking the central analysis particle point as the starting point, constructing a neighborhood in the compensation analysis space according to a preset iteration step length, and determining a central analysis neighborhood, wherein a plurality of analysis particle points in the central analysis neighborhood are analysis particle points whose distances to the starting point are within the preset iteration step length; Iteratively expand the central analysis neighborhood according to the preset iteration step to obtain a target analysis neighborhood; The fuzziness mean of multiple analysis particle points in the target analysis neighborhood is calculated to obtain the concentrated fuzziness to be compensated.

5. The urban compensation modeling method combined with image recognition as claimed in claim 4, characterized in that: Iteratively expanding the central analysis neighborhood according to the preset iteration step to obtain a target analysis neighborhood includes: Iterating leftwards on the leftmost analysis particle point of the central analysis neighborhood and iterating rightwards on the rightmost analysis particle point of the central analysis neighborhood respectively according to the preset iteration step length to obtain a first left iterative analysis neighborhood and a first right iterative analysis neighborhood; Determine whether the concentration density of the first left iterative analysis neighborhood and the concentration density of the first right iterative analysis neighborhood are both greater than or equal to the concentration density of the center analysis neighborhood; if so, obtain a union of the first left iterative analysis neighborhood and the first right iterative analysis neighborhood to obtain a first-stage iterative analysis neighborhood; Again, iterating leftwards the leftmost analysis particle point of the iterative analysis neighborhood of the first stage and iterating rightwards the rightmost analysis particle point of the iterative analysis neighborhood of the first stage respectively according to the preset iteration step size; After multiple bilateral iterations, when the concentration density of the left stage iterative analysis neighborhood and the concentration density of the right stage iterative analysis neighborhood obtained in the current iteration are both less than the concentration density of the stage iterative analysis neighborhood obtained in the previous iteration, the iteration is stopped and the stage iterative analysis neighborhood obtained in the previous iteration is used as the target analysis neighborhood.

6. The urban compensation modeling method combined with image recognition as claimed in claim 5, characterized in that: include: When only the concentration density of the first left iterative analysis neighborhood or the concentration density of the first right iterative analysis neighborhood is greater than or equal to the concentration density of the central analysis neighborhood, the first left iterative analysis neighborhood or the first right iterative analysis neighborhood is used as the first-stage iterative analysis neighborhood; Iterate the leftmost analysis particle point of the first-stage iterative analysis neighborhood to the left or iterate the rightmost analysis particle point of the first-stage iterative analysis neighborhood to the right according to the preset iteration step; After multiple one-sided iterations, when the concentration density of the left stage iterative analysis neighborhood or the concentration density of the right stage iterative analysis neighborhood obtained in the current iteration is less than the concentration density of the stage iterative analysis neighborhood obtained in the previous iteration, the iteration is stopped and the stage iterative analysis neighborhood obtained in the previous iteration is used as the target analysis neighborhood.

7. The urban compensation modeling method combined with image recognition as claimed in claim 1, characterized in that: The preset sampling interval and the preset driving speed of the drone are obtained, and the K-1 path interaction confrontation areas are divided according to the K-1 area lengths of the K-1 path interaction confrontation areas to obtain a set of K-1 interaction confrontation sub-areas, including: Multiplying the preset sampling interval by the preset driving speed to obtain a preset sampling length; Calculate the ratio of the length of the K-1 regions to the preset sampling length to obtain the number of K sub-region divisions; Based on the K sub-area division numbers and the preset sampling length, the K-1 path interaction confrontation areas are divided in sequence according to the path collection direction to obtain the K-1 interaction confrontation sub-area sets.

8. An urban compensation modeling system combined with image recognition, characterized in that: The system is used to implement the urban compensation modeling method combined with image recognition as described in any one of claims 1 to 7, and the system includes: A collection path set acquisition module is used to acquire a historical city three-dimensional model and a historical drone collection path set of a target city, wherein K historical drone collection paths in the historical drone collection path set are parallel to each other and the interval distance between any two adjacent historical drone collection paths is the same, and K is an integer greater than or equal to 1; A path interaction confrontation area acquisition module is used to determine the overlapping area of ​​the drone field of view of two adjacent historical drone collection paths when performing image collection based on K environmental parameters of the K historical drone collection paths of the target city and the model parameters of the drone, and obtain K-1 path interaction confrontation areas; An interactive confrontation sub-area set acquisition module is used to obtain a preset sampling interval and a preset driving speed of the UAV, and divide the K-1 path interactive confrontation areas into K-1 path interactive confrontation areas in combination with the K-1 area lengths to obtain a K-1 interactive confrontation sub-area set; A fuzzy region set determination module is used to retrieve the texture map file of the historical city three-dimensional model with the K-1 interactive confrontation sub-region sets as indexes, and perform image resolution and overlap rate analysis on the retrieval results to determine K-1 fuzzy region sets, wherein the K-1 fuzzy region sets include K-1 fuzziness degree sets; A centralized ambiguity determination module to be compensated is used to perform centralized ambiguity compensation analysis on the K-1 ambiguity sets to determine centralized ambiguities to be compensated; An optimized UAV flight parameter set determination module is used to collect a historical UAV flight parameter set for constructing the historical city three-dimensional model, optimize the historical UAV flight parameter set according to the concentrated ambiguity to be compensated, and determine an optimized UAV flight parameter set; The target compensated urban three-dimensional model acquisition module is used to control the drone based on the drone flight parameter set, collect images of the K-1 fuzzy area sets according to the historical drone collection path set, and perform three-dimensional modeling according to the collection results, and compensate and update the historical urban three-dimensional model according to the modeling results to obtain the target compensated urban three-dimensional model.

Images