An unmanned aerial vehicle-based whole-cycle and whole-region land comprehensive improvement system
By using drones to collect data, generate 3D models, and compare construction deviations, the problem of lax supervision in land consolidation projects has been solved, intelligent construction monitoring and early warning have been achieved, and the quality and efficiency of the projects have been improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JIGANG INT ENG & TECH CO LTD
- Filing Date
- 2026-02-28
- Publication Date
- 2026-06-09
Smart Images
Figure CN122175532A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent governance and ecological restoration of national land space, and in particular to a comprehensive land consolidation system based on unmanned aerial vehicles (UAVs) covering the entire lifecycle and all areas. Background Technology
[0002] Dynamic monitoring during the construction of land consolidation projects can ensure that the project construction meets design specifications, correct construction deviations in a timely manner, eliminate safety hazards, reduce rework and resource waste, improve project quality and long-term operation capabilities, and thus provide a solid guarantee for project implementation and subsequent operation and maintenance.
[0003] Currently, the construction process of land consolidation projects is generally plagued by a lack of effective supervision. Project quality largely relies on manual inspections and sampling checks of facilities, with low levels of automation and intelligent monitoring. This leads to blind spots in supervision, making it difficult to detect construction problems such as deviations in building structure placement in a timely and accurate manner. Consequently, the monitoring capacity for facilities is low, easily resulting in rework, project delays, and hidden quality defects. Summary of the Invention
[0004] In view of this, this application provides a full-cycle, all-area land consolidation system based on unmanned aerial vehicles (UAVs), the main purpose of which is to solve the technical problem of low monitoring capability of engineering facilities.
[0005] According to a first aspect of the present invention, a full-cycle, all-area land consolidation system based on unmanned aerial vehicles (UAVs) is provided, the system comprising UAVs and a central control unit; The drone is used to collect environmental data of the engineering facility to be monitored through a data acquisition device and send the environmental data to the central control unit; The central control unit is configured to perform the following processes: The central control unit generates a three-dimensional model of the engineering facility based on the environmental data, and identifies the target functional structures in the three-dimensional model; The engineering stage of the engineering facility is determined, and a preset reference 3D model of the engineering facility corresponding to the engineering stage is obtained. A reference functional structure is identified in the preset reference 3D model, wherein the target functional structure and the reference functional structure correspond to the same physical structure on the engineering facility. Determine the first relative position of the target functional structure in the three-dimensional model, determine the second relative position of the reference functional structure in the preset reference three-dimensional model, and determine whether the first relative position and the second relative position are consistent; If the first relative position is inconsistent with the second relative position, an alarm message is sent to the remote host computer.
[0006] In an optional embodiment, the central control unit is further configured to respond to an engineering phase update command sent by a host computer, control the UAV to fly to the engineering facility to collect environmental data of the engineering facility, and send the environmental data to the host computer.
[0007] In an optional embodiment, the data acquisition device includes a camera and a lidar; the camera is used to acquire visible light images of the engineering facility, and the lidar is used to acquire laser point cloud data of the engineering facility; the drone is used to send the visible light images and the laser point cloud data as environmental data to the central control unit; the central control unit generates a three-dimensional model of the engineering facility based on the environmental data, including: the central control unit generates a digital orthophoto and a digital elevation model of the engineering facility based on the laser point cloud data and the visible light images; and constructs a three-dimensional model of the engineering facility based on the digital orthophoto and the digital elevation model.
[0008] In an optional embodiment, the engineering facility includes a ditch; the data acquisition device further includes a multispectral camera, the image acquisition range of which is the same as that of the multispectral camera; the drone is also used to acquire visible light images of the ditch through the camera and multispectral images of the ditch through the multispectral camera, and send the visible light images and the multispectral images to the central control unit; the central control unit is also used to determine the ditch slope protection image in the visible light image, and based on the position of the ditch slope protection image in the visible light image, determine the slope protection multispectral image in the multispectral image, and determine the vegetation coverage rate of the ditch slope protection based on the slope protection multispectral image; the central control unit is also used to send a vegetation coverage warning message to the host computer when the vegetation coverage rate is lower than a preset coverage threshold.
[0009] In an optional embodiment, the central control unit is further configured to control the drone to fly to the ditch at preset time intervals to collect visible light images and multispectral images of the ditch, and to determine the vegetation coverage of the ditch slope based on the visible light images and multispectral images.
[0010] In an optional embodiment, the central control unit is further configured to determine the position deviation value between the first relative position and the second relative position, and send the position deviation value to the host computer.
[0011] In an optional embodiment, the central control unit is further configured to control the UAV to fly to the engineering facility at preset time intervals to collect environmental data of the engineering facility, generate a three-dimensional model of the engineering facility based on the environmental data, and send the three-dimensional model to a remote host computer.
[0012] In an optional embodiment, the central control unit is further configured to perform the following processing: the central control unit acquires a visible light image of the engineering facility from the UAV, and identifies a structure image of the target functional structure in the visible light image; sends the structure image to a pre-trained defect identification model to obtain a defect identification result of the target functional structure output by the defect identification model, and sends the defect identification result to the host computer, wherein the defect identification result includes defect results and non-defect results.
[0013] In an optional embodiment, after obtaining the defect identification result of the target functional structure output by the defect identification model, the central control unit is further configured to: when the defect identification result is a defect result, determine the target functional structure as a defective functional structure, and determine the target location of the defective functional structure in the three-dimensional model; set a defect label at the target location in the three-dimensional model, and send the three-dimensional model to the host computer.
[0014] In an optional embodiment, the central control unit is also used to associate and store the environmental data, three-dimensional model, preset reference three-dimensional model, vegetation coverage, and location deviation value of the engineering facility.
[0015] The UAV-based full-cycle, full-area land consolidation system provided by this invention can collect environmental data of engineering facilities through UAVs, and automatically generate three-dimensional models of the engineering facilities by the central control unit. It can also identify the functional structures of the engineering facilities and compare their relative positions with the preset reference three-dimensional models of the corresponding engineering stages to determine whether there are construction deviations during the construction process. It can achieve full-cycle, full-area intelligent monitoring, greatly reduce the reliance on manual inspections, eliminate blind spots in supervision, and promptly and accurately detect construction deviations and issue alarm prompts. This effectively reduces the risk of rework and delays in the project and effectively improves the monitoring capability of engineering facilities.
[0016] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description
[0017] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings: Figure 1 This diagram illustrates the structure of a full-cycle, all-area land consolidation system based on unmanned aerial vehicles (UAVs) according to an embodiment of the present invention. Figure 2 This diagram illustrates the architecture of a central control unit provided by an embodiment of the present invention. Detailed Implementation
[0018] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the present application can be combined with each other.
[0019] Dynamic monitoring during the construction of comprehensive land consolidation projects can ensure that the project meets design specifications, promptly correct construction deviations, eliminate safety hazards, reduce rework and resource waste, improve project quality and long-term operational capabilities, and thus provide a solid guarantee for project implementation and subsequent operation and maintenance. Currently, the supervision methods in the construction of comprehensive land consolidation projects are generally inadequate. Project quality relies heavily on manual inspections and sampling checks of engineering facilities, with low levels of automation and intelligent monitoring, easily leading to blind spots in supervision. It is difficult to detect construction problems such as deviations in building structure placement, earthwork engineering, and design discrepancies in a timely and accurate manner, resulting in low monitoring capabilities for engineering facilities and easily causing rework, project delays, and other hidden quality defects.
[0020] To address the above problems, in one embodiment, such as Figure 1 As shown, a comprehensive land consolidation system based on unmanned aerial vehicles (UAVs) is provided. The system includes UAVs 100 and a central control unit 200. Each UAV 100 is equipped with a data acquisition device 110. The number of UAVs 100 can be single or multiple. Each UAV 100 can be a rotary-wing UAV and wirelessly connected to the central control unit 200. Furthermore, the central control unit 200 can be a computer device such as a server located at a control center. The central control unit 200 can control the UAVs 100 to fly to the engineering facilities to be monitored for data acquisition based on existing navigation software. Here, the engineering facilities can be man-made structures such as buildings under construction or completed ditches.
[0021] Furthermore, the data acquisition device 110 may include a multispectral camera, a camera, and a lidar; wherein, the multispectral camera is a device capable of simultaneously acquiring images of multiple specific bands such as near-infrared and short-wave infrared, used for acquiring multispectral images; wherein, multispectral images can reflect the spectral characteristics of objects and are commonly used for monitoring and analysis of vegetation, water bodies, soil, etc. Furthermore, the image acquisition angle of the camera is the same as that of the multispectral camera, making their image acquisition ranges identical. This ensures that the imaging position of the same object remains consistent in the images acquired by both cameras, preventing image misalignment or shift due to differences in acquisition angles. This facilitates subsequent precise registration and fusion processing of the two types of images, ensuring unified data benchmarks and accurate positioning during 3D model construction, target functional structure identification, and position comparison, thereby improving the accuracy and reliability of monitoring and deviation identification.
[0022] Furthermore, the UAV 100 is used to collect environmental data of the engineering facility to be monitored via the data acquisition device 110, and send the environmental data to the central control unit 200. Specifically, the UAV 100 can navigate to the engineering facility and conduct multiple sorties to collect environmental data of the remediation area where the engineering facility is located. Here, the UAV 100 can collect visible light images of the engineering facility at different angles using a camera as multi-angle oblique photography images, collect laser point cloud data of the engineering facility using a lidar, and collect multispectral images of the engineering facility, such as near-infrared and short-wave infrared, using a multispectral camera. The multispectral images, visible light images, and laser point cloud data are then sent to the central control unit 200 as environmental data. The point cloud density of the lidar can be greater than 50 points per square meter.
[0023] Furthermore, the central control unit 200 is configured to perform the following processes: First, the central control unit 200 generates a three-dimensional model of the engineering facility based on the environmental data as a digital current status base map, and identifies the target functional structures in the three-dimensional model. Specifically, the central control unit 200 can generate a digital orthophoto map (DOM) and a digital elevation model (DEM) of the engineering facility based on laser point cloud data and visible light images from multiple angles. Here, the central control unit 200 can receive laser point cloud data and visible light images from multiple angles collected by UAVs, using the laser point cloud data as the elevation control benchmark, and through point cloud filtering, surface point classification, and raster interpolation, accurately construct a digital elevation model with centimeter-level resolution that reflects the real terrain undulations of the engineering facility, providing a reliable elevation basis for terrain correction. Furthermore, aerial triangulation, bundle adjustment, geometric distortion correction, and image color mosaicking are performed using visible light images, and then combined with the generated digital elevation model to perform projection difference correction and orthophoto correction on the image, eliminating the positional deviation caused by terrain undulations, and generating a digital orthophoto map with centimeter-level resolution and high-precision geographic coordinates.
[0024] Furthermore, the central control unit 200 constructs a three-dimensional model of the engineering facility based on the digital orthophoto and the digital elevation model. Specifically, the central control unit 200 can use the digital orthophoto as texture data and the digital elevation model as the terrain geometric skeleton, register and fuse the two in a unified spatial coordinate system, and through triangulation construction, surface fitting and texture mapping processing, attach the digital orthophoto to the three-dimensional terrain surface formed by the digital elevation model, automatically generating a three-dimensional model that can realistically reflect the spatial shape and surface texture of the engineering facility.
[0025] Furthermore, a target functional structure refers to a physical component within an engineering facility that possesses a defined function and undertakes a specific structural role. It is an independently identifiable structural unit used to achieve a specific engineering objective. For example, if the engineering facility is a building, the target functional structure could be walls, beams, columns, and floor slabs; if the engineering facility is an irrigation canal or similar ditch, the target functional structure could be the canal body, slope, bottom, outlet, inlet, retaining wall, and slope protection. Here, the number of identified target functional structures can be single or multiple. The types and quantities of target functional structures to be identified can be determined based on the actual situation.
[0026] Furthermore, the central control unit 200 can automatically extract and locate the target functional structures that perform corresponding functions in the engineering facility in the generated three-dimensional model based on the preset component features, geometric shapes and spatial location information, and clarify their spatial form and distribution location in the three-dimensional model. As an example, when the engineering facility is an irrigation canal, the central control unit 200 can, based on preset parameters such as the elongated geometric features of the canal body, the inclination angle range of the canal slope, and the height and location distribution of the water-retaining sill, distinguish and locate target functional structures such as the canal body, canal slope, and water-retaining sill from the model through texture feature extraction, edge contour recognition, and spatial coordinate analysis of the 3D model, accurately determining the 3D contour, size, and relative distribution position of each structure in the 3D model. Furthermore, if the engineering facility is a building, based on preset characteristics such as the planar extension features of the wall body, the columnar geometric shape of the beams and columns, and the horizontal distribution of the floor slabs, the central control unit 200 can extract and locate target functional structures such as walls, beams, columns, and floor slabs from the 3D model through image recognition and model analysis methods such as model mesh segmentation and feature point matching, clarifying their specific spatial form and distribution position in the model.
[0027] Then, the central control unit 200 can determine the engineering stage of the engineering facility and obtain a preset 3D model of the engineering facility corresponding to the engineering stage. An engineering stage refers to the entire implementation process of an engineering project, from preliminary planning, surveying and design, construction, to completion acceptance and operation and maintenance. These stages are sequentially connected and together constitute the complete construction and operation process of the project. Here, after completing a certain engineering stage, the construction personnel or relevant staff of the engineering facility can send an engineering stage update command to the central control unit 200 via the host computer 300 to indicate the current engineering stage of the engineering facility. The engineering stage update command is an instruction to indicate that the engineering facility has entered a new engineering stage, and it includes the name of the engineering stage. For example, if the engineering facility is a multi-story building, the construction personnel can send an engineering stage update command to the central control unit 200 after the building is topped out, sending the information that the building is in the topping-out stage, so that the central control unit 200 can determine that the engineering facility is in the topping-out stage.
[0028] Here, the central control unit 200 is also used to respond to the engineering phase update command sent by the host computer 300, and control the UAV to fly to the engineering facility to collect environmental data of the engineering facility. Specifically, when the central control unit 200 receives the engineering phase update command, it can automatically generate the corresponding flight path and data collection task, control the UAV to fly to the area where the engineering facility is located according to the preset path, and collect data from the target area and facilities in all directions through the data acquisition device, thereby obtaining environmental data for 3D modeling and structure recognition.
[0029] Furthermore, for the engineering facility, pre-constructed reference 3D models corresponding to each engineering stage can be built in advance. These pre-constructed reference 3D models are standard engineering models of the engineering facility at a specific engineering stage, used to represent the design style of the engineering facility at that specific engineering stage. If the engineering facility is constructed entirely according to the design, the 3D model at the same engineering stage will be consistent with the pre-constructed reference 3D model. Further, the pre-constructed reference 3D model can be a Building Information Model (BIM) or a Site Implementation Model (SIM). Specifically, after determining the current engineering stage of the engineering facility, a pre-constructed reference 3D model for that current stage can be obtained.
[0030] Furthermore, a reference functional structure is identified in the preset reference 3D model, wherein the target functional structure and the reference functional structure correspond to the same physical structure on the engineering facility. Here, the reference functional structure corresponding to the target functional structure can be identified in the preset reference 3D model in the same way as the target functional structure. Specifically, the reference functional structure and the target functional structure are the same actual physical structure in the engineering facility. As an example, when the engineering facility is an irrigation canal, if the target functional structure identified in the 3D model is a canal slope protection, then the reference functional structure in the preset reference 3D model is the same canal slope protection; if the target functional structure is an irrigation canal outlet, then the reference functional structure in the preset reference 3D model is the same outlet. The two correspond one-to-one and are the same physical component.
[0031] Next, the central control unit 200 determines the first relative position of the target functional structure in the three-dimensional model, determines the second relative position of the comparison functional structure in the preset comparison three-dimensional model, and determines whether the first relative position and the second relative position are consistent. Here, the central control unit 200 can extract information such as the spatial coordinates, contour boundary, and key point position of the target functional structure in the three-dimensional model to obtain its first relative position relative to the overall engineering facility. Similarly, it extracts the corresponding parameters of the comparison functional structure in the preset comparison three-dimensional model to obtain the second relative position. By quantitatively comparing the spatial coordinates, dimensions, posture, and relative relationship of the two, it determines whether the positions of the two are consistent.
[0032] As an example, a specific point can be identified in the 3D model, and a corresponding reference point can be identified in a preset reference 3D model. The point in the 3D model and the reference point in the preset reference 3D model correspond to the same structure within the engineering facility, ensuring accurate spatial correspondence. Further, the point in the 3D model is set to the origin of the 3D coordinate system to determine the coordinates of the target functional structure in the 3D coordinate system, and these coordinates are used as the first relative position. Similarly, the reference point in the preset reference 3D model is set to the origin of the same 3D coordinate system to determine the coordinates of the reference functional structure in the 3D coordinate system, and these coordinates are used as the second relative position. Here, the spatial orientation of the 3D model and the preset reference 3D model remains consistent in the 3D coordinate system to ensure comparability of their position coordinates. Further, the coordinates of the first and second relative positions are compared to determine if their positions are consistent.
[0033] Finally, if the first relative position is inconsistent with the second relative position, an alarm message is sent to the remote host computer 300. Specifically, if the first relative position is inconsistent with the second relative position, it indicates that a construction error has occurred during the construction of the engineering facilities. In this case, an alarm message can be promptly sent to the computer terminal used by the relevant supervisory personnel to prompt the relevant staff to make rectifications.
[0034] Furthermore, the central control unit 200 is also used to determine the positional deviation value between the first relative position and the second relative position, and send the positional deviation value to the host computer 300. Specifically, the central control unit 200 can quantify the positional deviation value between the target functional structure and the reference functional structure by calculating the coordinate difference in the spatial coordinate system, and send the positional deviation value to the host computer in real time, so as to intuitively present the degree of deviation in construction to the management personnel.
[0035] As an example, if the engineering facility is an irrigation canal and the target functional structure is the canal bottom, if the canal bottom is found to be deeper than the bottom in the preset reference 3D model, it indicates that the irrigation canal excavation is too deep or the filling is insufficient. In this case, an alarm message can be sent to the host computer for real-time construction deviation warning. Simultaneously, the positional deviation between the actual and designed positions of the canal bottom can be calculated and sent to the host computer 300 so that relevant personnel are aware of the degree of engineering deviation. Based on the deviation between the actual and designed positions of the canal bottom, the required earthwork volume can be calculated, facilitating project management by relevant personnel.
[0036] Furthermore, in the early stages of engineering facility construction, aerial surveys using drones can quickly acquire centimeter-resolution digital orthophotos and digital elevation models of the land consolidation area, constructing a realistic 3D model. This allows for precise mapping of the topography, land cover, and the geometric and attribute information of existing engineering facilities such as ditches and roads, generating a high-precision digital baseline map. Further, based on preliminary survey data, a digital twin of the consolidation area can be constructed, synchronously mapped to the physical world. This twin can then be used for virtual design and multi-scenario simulation of consolidation plans, including water flow simulation, earthwork balance simulation, and landscape visual simulation. Moreover, during the construction and maintenance phases, this twin can receive real-time monitoring data from drones and update dynamically, achieving virtual-real linkage and synchronous interaction.
[0037] The UAV-based full-cycle, full-area land consolidation system provided in this embodiment can collect environmental data of engineering facilities through UAVs, and the central control unit can automatically generate a three-dimensional model of the engineering facilities, identify the functional structures of the engineering facilities, and compare their relative positions with the preset reference three-dimensional models of the corresponding engineering stages to determine whether there are construction deviations in the construction process. It can achieve full-cycle, full-area intelligent monitoring, greatly reduce the reliance on manual inspections, eliminate blind spots in supervision, and promptly and accurately detect construction deviations and issue alarm prompts, effectively reducing the risk of rework and delays in the project, and effectively improving the monitoring capability of engineering facilities.
[0038] In an optional embodiment, the central control unit is further configured to control the UAV to fly to the engineering facility for inspection at preset time intervals, collect environmental data of the engineering facility, generate a 3D model of the engineering facility based on the environmental data, and send the 3D model to a remote host computer. Here, the central control unit may include a construction guidance unit, which can generate the 3D model based on the environmental data collected by the UAV and send it to the computer terminal used by construction workers to provide 3D visualization operation guidance for construction machinery such as excavators. Furthermore, the construction guidance unit can also plan inspection paths for the UAV's inspection operations, pre-setting the UAV's flight path as a dedicated path adapted to construction inspection needs, which can improve the efficiency and coverage of UAV inspections, achieve accurate monitoring of the construction area, and ensure data integrity and reliability.
[0039] In an optional embodiment, the engineering facility includes a ditch. Furthermore, the drone is also used to acquire multispectral images of the ditch via the multispectral camera and send the visible light image and the multispectral image to the central control unit; here, the drone can acquire both the multispectral and visible light images of the ditch simultaneously, so that the same object appears in the same position in both the multispectral and visible light images.
[0040] Furthermore, the central control unit is also used to extract the slope protection area of the ditch from the visible light image of the engineering facility and determine the corresponding ditch slope protection image; here, the central control unit can determine the location of the ditch slope protection in the visible light image based on a pre-trained first convolutional neural network model. The first convolutional neural network model can be trained using ditch slope protection images labeled with ditch slope protection tags, and is able to identify the ditch slope protection in the image.
[0041] Furthermore, based on the position of the ditch slope in the visible light image, the central control unit determines the multispectral image of the ditch slope in the multispectral image, and determines the vegetation coverage of the ditch slope based on the multispectral image. Here, since the camera and the multispectral camera have the same image acquisition angle, the imaging position of the ditch slope in the same area remains consistent in both images. This allows the central control unit to accurately locate and extract the corresponding area's multispectral image of the slope in the multispectral image based on the contour, coordinates, and other positional information of the ditch slope in the visible light image, ensuring the consistency and accuracy of the monitoring area. Furthermore, the central control unit can perform spectral feature analysis on the extracted slope multispectral image. By identifying the spectral information corresponding to vegetation, it calculates the pixel ratio of vegetation and the overall ratio of the slope area, thereby quantifying the vegetation coverage of the ditch slope.
[0042] Furthermore, the central control unit is also used to send a vegetation coverage warning to the host computer when the vegetation coverage rate is lower than a preset coverage threshold. Here, the coverage threshold can be 60%. Specifically, the central control unit can compare the calculated vegetation coverage rate of the ditch slope with the system's preset coverage threshold. When the vegetation coverage rate is lower than the threshold, it determines that the slope vegetation growth is substandard, there is a risk of soil erosion or slope damage, and automatically sends a vegetation coverage warning to the host computer so that management personnel can carry out maintenance, repair, and other disposal work in a timely manner.
[0043] Furthermore, the central control unit is also used to control the drone to fly to the ditch at preset time intervals to collect visible light and multispectral images of the ditch, and to determine the vegetation coverage of the ditch slope based on the visible light and multispectral images. Specifically, after the completion of the ditch and other projects, the central control unit will formulate a periodic operation monitoring plan for the drone according to preset time intervals, automatically control the drone to fly to the ditch regularly, collect visible light and multispectral images of the ditch, and determine the vegetation coverage of the ditch slope based on the two types of images; after each monitoring is completed, the system compares the real-time vegetation coverage with a preset threshold through the early warning module, automatically analyzes the ecological restoration effect and the slope degradation risk, and identifies potential problem areas; at the same time, it generates targeted maintenance task work orders based on the analysis results to drive precise maintenance, and can verify the maintenance effect through the drone monitoring data of the next cycle, forming an intelligent closed loop of "monitoring-evaluation-maintenance-re-monitoring".
[0044] The embodiments provided in this application can accurately extract the slope protection area and calculate its vegetation coverage by simultaneously acquiring visible light and multispectral images of engineering facilities and combining them with artificial intelligence recognition. This enables automated and high-precision monitoring of the ecological status of ditch slope protection, timely issuance of early warnings, effective reduction of manual inspection costs, improvement of engineering management efficiency and response speed, and protection of the structural stability and ecological restoration effect of irrigation canals.
[0045] In an optional embodiment, the central control unit is further configured to associate and store environmental data, 3D models, preset reference 3D models, vegetation coverage, and location deviation values of the engineering facilities. Specifically, the central control unit can associate and store environmental data of the engineering facilities collected by the UAV, real-scene 3D models, preset reference 3D models, location deviation data, and vegetation coverage, enabling traceability of the construction process and maintenance records. Furthermore, it can provide complete data support for subsequent work order generation, review and inspection, and ecological restoration assessment, forming a closed-loop management system with full-cycle traceability. The embodiments provided in this application can associate and store multi-source data, enabling traceability throughout the construction and maintenance process, providing complete data support for work order generation, review and inspection, and ecological assessment, improving the standardization and intelligence of management, forming a closed-loop management system with full-cycle traceability, and ensuring project quality and ecological restoration effectiveness.
[0046] In an optional embodiment, the central control unit is further configured to perform the following processes: First, the central control unit acquires visible light images of the engineering facilities from the UAV and identifies the target functional structure's image within the visible light image. Here, the central control unit may be equipped with a second convolutional neural network model, which can identify the type of the target functional structure within the visible light image and identify its image. This second convolutional neural network model can be trained using multi-category images labeled with structure types, including images of ditch slopes, roads, field embankments, buildings, and vegetation.
[0047] Then, the central control unit sends the structure image to the pre-trained defect recognition model to obtain the defect recognition result of the target functional structure output by the defect recognition model, and sends the defect recognition result to the host computer. The defect recognition result includes defect results and non-defect results.
[0048] Here, the central control unit can also be pre-set with multiple pre-trained defect recognition models. Each defect recognition model is used to receive images of a type of target functional structure and identify whether the target functional structure has defects. As an example, if the target functional structure includes ditch slope protection, the defect recognition model corresponding to the ditch slope protection can be trained using images of defective ditch slope protection and defect labels, and can identify the defect information of the ditch slope protection in the images. Here, after determining the type of target functional structure in the structure image, the central control unit can send it to the corresponding defect recognition model to obtain the defect recognition result.
[0049] Furthermore, if the defect identification model identifies a defect in the target functional structure, the defect identification result of the target functional structure is a defect result; conversely, if the defect identification model does not identify a defect in the target functional structure, the defect identification result of the target functional structure is a non-defect result.
[0050] Furthermore, after identifying defects in target functional structures, spatial analysis tools such as Geographic Information System (GIS) can be used to spatially locate the identified structures, calculate the defect area, and analyze slope and elevation. This quantifies the degree of damage, degradation level, and distribution density, generating quantitative results containing location, extent, and severity information, providing a basis for subsequent early warning and maintenance. The embodiments provided in this application can automatically identify various functional structures and accurately detect their defects by combining convolutional neural networks with a dedicated defect recognition model. By combining GIS for spatial location and quantitative analysis, the efficiency and accuracy of defect detection can be significantly improved, reducing reliance on manual inspections and enabling intelligent monitoring of engineering facilities. This provides reliable data support for early warning and maintenance, significantly improving the automation and refinement of engineering operation and maintenance management.
[0051] In an optional embodiment, after obtaining the defect identification results of the target functional structure output by the defect identification model, the central control unit is further configured to: First, when the defect identification result is a defect, the target functional structure is identified as a defective functional structure, and the target location of the defective functional structure is determined in the 3D model. Specifically, when the defect identification result of the target functional structure is determined to have a defect, the central control unit marks the corresponding target functional structure as a defective functional structure, and, in combination with the spatial mapping relationship between the visible light image and the 3D model, locates the 3D coordinates, outline range, and area of the defective functional structure in the real-world 3D model, thus determining its target location in the engineering facility.
[0052] Then, defect labels are set at the target locations in the 3D model, and the 3D model is sent to the host computer. Specifically, the central control unit can add defect labels to the target locations corresponding to the defective functional structures in the 3D model. These labels can contain relevant information such as defect type, location, and severity. The 3D model with the defect labels is then sent to the host computer, allowing managers to intuitively and quickly view the defect locations and related information in the 3D model.
[0053] The embodiments provided in this application can automatically identify defects through artificial intelligence and accurately mark the defect location and information on a three-dimensional model, visualizing abstract data and enabling managers to intuitively grasp the defect status, thus greatly improving the efficiency and response speed of defect investigation.
[0054] Furthermore, such as Figure 2 As shown, the central control unit 200 may include an intelligent UAV monitoring and data fusion module 210, a full-cycle intelligent diagnosis and decision-making module 220, a digital twin and simulation optimization engine module 230, and an intelligent construction guidance and maintenance execution module 240.
[0055] The intelligent drone monitoring and data fusion module 210 is used for drone flight navigation to control the drone's flight status and for fusion processing of environmental data collected by the drone to obtain a three-dimensional model of the engineering facility. Furthermore, the full-cycle intelligent diagnosis and decision-making module 220 is used to acquire high-precision geospatial data from drones during the project's early stages, construction phase, and operation and maintenance phases. This module also includes an early diagnosis module and a dynamic monitoring and early warning module, which can automatically identify problems based on drone data and conduct dynamic comparative analysis and early warning during the construction and operation and maintenance phases.
[0056] Furthermore, the digital twin and simulation optimization engine module 230 is used to construct a digital twin of the remediation area based on the preliminary survey data collected by UAVs and synchronously mapped to the physical world. This twin allows for virtual design and multi-scenario simulation of remediation plans, such as water flow simulation, earthwork balance simulation, and landscape visual simulation. During the construction and maintenance phases, the twin receives real-time monitoring data from UAVs and updates dynamically, achieving virtual-real linkage. Furthermore, the intelligent construction guidance and maintenance execution module 240 is used to provide 3D visualization operation guidance for construction machinery based on the 3D model of the engineering facilities, or to plan UAV flight paths as construction inspection routes. Based on the UAV's later monitoring and early warning information, it automatically generates and dispatches maintenance work orders to the mobile terminals of relevant responsible persons, achieving closed-loop management.
[0057] Furthermore, the actual workflow of the drone-based full-cycle, full-area land consolidation system is explained: First, during the precise survey and planning phase, the central control unit controls drones equipped with cameras and LiDAR to perform gridded aerial surveys of the project area, acquiring 3D data with a point cloud density greater than 50 points / square meter. Based on the LiDAR point cloud and oblique image data, a DOM model with a resolution of 2 cm, a DEM model with an accuracy of 5 cm, and a realistic 3D model are generated. These models are then compared with a pre-set ideal 3D model to determine if any anomalies exist in the project area. Furthermore, when anomalies are found, an ecological restoration plan is developed based on a digital twin engine, and earthwork balance and water flow simulation are completed.
[0058] Then, during the construction phase of the ecological slope protection project, orthophotos of the engineering facilities are collected using drones at preset time intervals to generate corresponding 3D models. Furthermore, each generated 3D model is overlaid and compared with functional structures such as the slope protection design boundaries in the Building Information Model (BIM) to determine if any offset has occurred. If an offset is detected, the project team immediately conducts on-site verification and correction to avoid subsequent rework. Simultaneously, drones can monitor the transportation of excavated soil, providing objective and visual evidence for engineering measurement.
[0059] Finally, in the intelligent management and assessment phase, after project completion, the system is configured to use drones to conduct quarterly inspections of the completed facilities to determine if there are issues such as insufficient vegetation cover or water stress. The system automatically generates management work orders with specific coordinates and on-site photos, which are then pushed to the mobile terminals of the management personnel. Furthermore, after on-site handling, the management personnel can report the results via their terminals, allowing the drones to conduct a focused review of the area during the next inspection cycle, forming a closed-loop management system. Simultaneously, the system continuously tracks the NDVI change curve across the entire area, quantitatively assessing the overall trend of ecological restoration.
[0060] The comprehensive land consolidation system based on unmanned aerial vehicles (UAVs) provided in this application adopts a technical solution that combines UAVs with artificial intelligence and digital twins. This significantly improves the accuracy and efficiency of surveying, rapidly acquires centimeter-level three-dimensional geographic information, overcomes the limitations of complex terrain, and provides a precise and reliable data foundation for engineering design. Furthermore, through high-frequency UAV inspections and artificial intelligence change detection, it achieves panoramic dynamic monitoring of the entire construction process, reduces management blind spots and quality hazards, and normalizes and automates inspections, enabling proactive early warning and long-term maintenance. This changes the current situation of neglecting management during reconstruction, forms a digital management closed loop, greatly improves the scientific nature of decision-making, and effectively reduces labor costs and safety risks.
[0061] The serial numbers in this application are for descriptive purposes only and do not represent the superiority or inferiority of any particular implementation scenario. The above disclosures are merely a few specific implementation scenarios of this application; however, this application is not limited thereto, and any variations conceived by those skilled in the art should fall within the protection scope of this application.
Claims
1. A comprehensive land consolidation system based on unmanned aerial vehicles (UAVs) covering the entire lifecycle and all areas, characterized in that: The system includes a drone and a central control unit; The drone is used to collect environmental data of the engineering facility to be monitored through a data acquisition device and send the environmental data to the central control unit; The central control unit is configured to perform the following processes: The central control unit generates a three-dimensional model of the engineering facility based on the environmental data, and identifies the target functional structures in the three-dimensional model; The engineering stage of the engineering facility is determined, and a preset reference 3D model of the engineering facility corresponding to the engineering stage is obtained. A reference functional structure is identified in the preset reference 3D model, wherein the target functional structure and the reference functional structure correspond to the same physical structure on the engineering facility. Determine the first relative position of the target functional structure in the three-dimensional model, determine the second relative position of the reference functional structure in the preset reference three-dimensional model, and determine whether the first relative position and the second relative position are consistent; If the first relative position is inconsistent with the second relative position, an alarm message is sent to the remote host computer.
2. The UAV-based full-cycle, full-area land consolidation system according to claim 1, characterized in that, The central control unit is also used to respond to the engineering phase update command sent by the host computer, control the UAV to fly to the engineering facility to collect environmental data of the engineering facility, and send the environmental data to the host computer.
3. The UAV-based full-cycle, full-area land consolidation system according to claim 1, characterized in that, The data acquisition device includes a camera and a lidar; the camera is used to acquire visible light images of the engineering facility, and the lidar is used to acquire laser point cloud data of the engineering facility. The drone is used to send the visible light image and the laser point cloud data as environmental data to the central control unit. The central control unit generates a three-dimensional model of the engineering facility based on the environmental data, including: The central control unit generates a digital orthophoto and a digital elevation model of the engineering facility based on the laser point cloud data and the visible light image. A three-dimensional model of the engineering facility is constructed based on the digital orthophoto and the digital elevation model.
4. The UAV-based full-cycle, full-area land consolidation system according to claim 3, characterized in that, The engineering facilities include ditches; The data acquisition device also includes a multispectral camera, and the image acquisition range of the camera is the same as that of the multispectral camera. The drone is also used to acquire visible light images of the ditch through the camera and multispectral images of the ditch through the multispectral camera, and send the visible light images and the multispectral images to the central control unit; The central control unit is also used to determine the ditch slope protection image of the ditch in the visible light image, and based on the position of the ditch slope protection image in the visible light image, determine the slope protection multispectral image of the ditch slope in the multispectral image, and determine the vegetation coverage of the ditch slope based on the slope protection multispectral image. The central control unit is also used to send a vegetation coverage warning message to the host computer when the vegetation coverage rate is lower than a preset coverage threshold.
5. The UAV-based full-cycle, full-area land consolidation system according to claim 4, characterized in that, The central control unit is also used to control the drone to fly to the ditch at preset time intervals to collect visible light images and multispectral images of the ditch, and to determine the vegetation coverage of the ditch slope based on the visible light images and multispectral images.
6. The UAV-based full-cycle, full-area land consolidation system according to claim 5, characterized in that, The central control unit is also used to determine the position deviation value between the first relative position and the second relative position, and send the position deviation value to the host computer.
7. The UAV-based full-cycle, full-area land consolidation system according to claim 1, characterized in that, The central control unit is also used to control the UAV to fly to the engineering facility at preset time intervals to collect environmental data of the engineering facility, generate a three-dimensional model of the engineering facility based on the environmental data, and send the three-dimensional model to a remote host computer.
8. The UAV-based full-cycle, full-area land consolidation system according to claim 3, characterized in that, The central control unit is also configured to perform the following processes: The central control unit acquires visible light images of the engineering facility from the UAV and identifies the structure image of the target functional structure in the visible light image; The image of the structure is sent to a pre-trained defect recognition model to obtain the defect recognition result of the target functional structure output by the defect recognition model, and the defect recognition result is sent to the host computer. The defect recognition result includes defect results and non-defect results.
9. The UAV-based full-cycle, full-area land consolidation system according to claim 8, characterized in that, After obtaining the defect identification result of the target functional structure output by the defect identification model, the central control unit is further configured to: When the defect identification result is a defect result, the target functional structure is determined as a defective functional structure, and the target location of the defective functional structure is determined in the three-dimensional model; A defect label is set at the target location of the 3D model, and the 3D model is sent to the host computer.
10. The UAV-based full-cycle, full-area land consolidation system according to claim 6, characterized in that, The central control unit is also used to associate and store the environmental data, three-dimensional model, preset reference three-dimensional model, vegetation coverage rate and location deviation value of the engineering facility.