Methods, devices, electronic equipment, and storage media for constructing 3D reconstructed digital twins of power systems
By using a Transformer model based on visual geometric constraints and grayscale masking technology, the problems of background noise and discontinuity of point cloud structure in the 3D reconstruction of transmission lines were solved, realizing the rapid and automated construction of 3D reconstructed power digital twins and the effective association of power asset data.
Patent Information
- Application Number
- CN202611345974.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-01
- Publication Date
- 2026-11-17
AI Technical Summary
Existing 3D reconstruction methods face problems such as high background noise, unclear target boundaries, and discontinuous point cloud structure in power transmission line inspection. Furthermore, they are difficult to associate with power asset data in a unified manner, which affects object positioning and auxiliary operation and maintenance in practical application scenarios.
A Transformer model based on visual geometric constraints is adopted, and the confidence information in the parameters of the 3D reconstruction model is weighted by combining grayscale mask to generate a 3D reconstruction power digital twin system. The model is constructed using power-themed vector semantic data.
It improves the accuracy of 3D reconstruction in complex backgrounds, reduces background noise, ensures the continuity of point cloud structure, and supports unified association and retrieval of power asset data, thereby enhancing the application value of the digital twin platform.
Smart Images

Figure CN123118910A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of digital twin construction technology, and in particular to a method, apparatus, electronic device, and storage medium for constructing a three-dimensional reconstructed power digital twin. Background Technology
[0002] Power transmission line inspections typically rely on drones to collect multi-view images, which are then used to generate spatial models of the lines and target facilities using 3D reconstruction methods. Existing 3D reconstruction methods generally face two problems in engineering scenarios: First, the background of the inspection images is complex, with the sky, vegetation, ground, and distant unrelated facilities creating a large number of noise points in the reconstruction results, weakening the discernibility of the target structure; second, some methods directly crop or occlude the image at the input end to suppress the background, which can easily destroy global geometric cues, leading to a decrease in camera pose estimation, depth prediction, and point cloud integrity.
[0003] VGGT, as a multi-view 3D reconstruction model that does not require scene-by-scene optimization, can directly regress camera pose, depth map, world coordinates per pixel, and confidence from input images, offering advantages such as fast inference speed and simple process. However, in scenarios with strong background interference, such as power transmission lines, when VGGT performs dense predictions on the entire image, the background region will also generate point cloud coordinates and confidence, easily leading to problems such as high background noise, unclear target boundaries, and discontinuous point cloud structure in the final reconstruction results. Simply using input-end occlusion may weaken the model's ability to perceive the overall geometric relationship of the scene, which is not conducive to improving the accuracy of 3D reconstruction. Furthermore, the digital twin platform for power transmission lines needs to unify and link power asset data, inspection images, reconstructed point clouds, line topology, alarm information, and maintenance records to form a searchable, locatable, and traceable spatial business closed loop. Simple 3D point cloud results without map, topology, and asset data support are difficult to directly serve practical application scenarios, object positioning, and auxiliary maintenance. Summary of the Invention
[0004] This invention provides a method, apparatus, electronic device, and storage medium for constructing a three-dimensional reconstruction digital twin of a power system, in order to solve problems such as high background noise, unclear target boundaries, and discontinuous point cloud structure in the reconstruction results.
[0005] According to one aspect of the present invention, a method for constructing a three-dimensional reconstructed power digital twin system is provided, the method comprising: Several first image groups are identified; a first image group is a collection of images generated by taking pictures of the same tower from different perspectives; the towers photographed in each first image group are different; Determine the grayscale mask of the target image region in each of the first images in the first image group; the target image region is the image region of the tower in the first image. The first image group is input into the first model to obtain the 3D reconstruction model parameters; the first model is a Transformer model based on visual geometric constraints; the 3D reconstruction model parameters include camera pose, depth map, 3D point cloud and confidence information; The confidence information in the parameters of the 3D reconstruction model is weighted based on grayscale mask, and the 3D reconstruction model is determined based on the weighting result and the parameters of the 3D reconstruction model. Based on pre-determined power-themed vector semantic data and a 3D reconstruction model, a 3D reconstructed power digital twin system is generated.
[0006] According to another aspect of the present invention, a three-dimensional reconstruction apparatus for a power digital twin system is provided, the apparatus comprising: The image acquisition module is used to determine several first image groups; a first image group is a collection of images generated by taking pictures of the same tower from different perspectives; the towers photographed in each first image group are different; The mask calculation module is used to determine the grayscale mask of the target image region in each of the first images in the first image group; the target image region is the image region of the tower in the first image. The parameter acquisition module is used to input the first image group into the first model to obtain the 3D reconstruction model parameters; the first model is a Transformer model based on visual geometric constraints; the 3D reconstruction model parameters include camera pose, depth map, 3D point cloud and confidence information; The model building module is used to weight the confidence information in the parameters of the 3D reconstruction model based on the grayscale mask, and to determine the 3D reconstruction model based on the weighting result and the 3D reconstruction model parameters. The system construction module is used to generate a 3D reconstructed digital twin system for power based on pre-determined power-themed vector semantic data and a 3D reconstruction model.
[0007] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory that is communicatively connected to at least one processor; wherein, The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to execute the method for constructing a three-dimensional reconstructed power digital twin system according to any embodiment of the present invention.
[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the method for constructing a three-dimensional reconstructed power digital twin system according to any embodiment of the present invention.
[0009] According to another aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method for constructing a three-dimensional reconstructed power digital twin system according to any embodiment of the present invention.
[0010] The technical solution of this invention involves determining several first image groups; each first image group is a collection of images generated from first images of the same tower taken from different viewpoints; the towers in each first image group are different; determining grayscale masks for target image regions in each first image of the first image group; the target image region is the image region of the tower in the first image; inputting the first image groups into a first model to obtain three-dimensional reconstruction model parameters; the first model is a Transformer model based on visual geometric constraints; the three-dimensional reconstruction model parameters include camera pose, depth map, three-dimensional point cloud, and confidence information; weighting the confidence information in the three-dimensional reconstruction model parameters based on the grayscale mask, and determining the three-dimensional reconstruction model based on the weighting result and the three-dimensional reconstruction model parameters; generating a three-dimensional reconstructed power digital twin system based on pre-determined power thematic vector semantic data and the three-dimensional reconstruction model; using VGGT for rapid construction of the three-dimensional reconstruction model, and using the grayscale mask to process only the towers as much as possible to avoid data noise from irrelevant data, thereby achieving rapid and automated construction of the three-dimensional reconstructed power digital twin.
[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description
[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0013] Figure 1 This is a flowchart of a method for constructing a three-dimensional reconstruction power digital twin system provided by an embodiment of the present invention; Figure 2 This is a schematic diagram comparing the confidence levels under different masking strategies applicable to embodiments of the present invention. Figure 3 This is a flowchart of another method for constructing a three-dimensional reconstruction power digital twin system provided by an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of a three-dimensional reconstruction power digital twin system construction device provided in an embodiment of the present invention; Figure 5 This is a structural block diagram of an electronic device for a method of constructing a three-dimensional reconstruction power digital twin system provided in an embodiment of the present invention. Detailed Implementation
[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0016] In one embodiment, Figure 1 This is a flowchart illustrating a method for constructing a three-dimensional reconstructed power digital twin system according to an embodiment of the present invention. This embodiment is applicable to the construction of a three-dimensional reconstructed power digital twin system. The method can be executed by a three-dimensional reconstructed power digital twin system construction device, which can be implemented in hardware and / or software and can be configured in an electronic device with data processing capabilities. Figure 1 As shown, the method includes: S110. Determine several first image groups; a first image group is a set of images consisting of the first images generated after taking pictures of the same tower from different perspectives; the towers photographed in each first image group are different.
[0017] A pole or tower can be an overhead line support structure that suspends conductors via insulators. A first image group can be a collection of several first images, each corresponding to the same pole or tower, but the shooting angles of the first images in the first image group are different.
[0018] S120. Determine the grayscale mask of the target image region in each of the first images in the first image group; the target image region is the image region of the tower in the first image.
[0019] A grayscale mask can be a single-channel grayscale mask image, with pixel values covering the entire grayscale range (integer range 0 to 255, normalized floating-point range 0.0 to 1.0). The pixel grayscale value represents the weight of the corresponding original image pixel in the calculation; a pixel value of 0 represents a completely masked area, extreme values of 255 / 1.0 represent completely preserved areas, and grayscale values in the middle of the range represent semi-weighted transition areas. Binary hard masks are a special subset of grayscale masks, containing only two discrete grayscale values: 0 and 255. Soft masks and Gaussian feathered masks are continuous grayscale masks.
[0020] Since each first image in the first image group is a photograph of a tower, a grayscale mask is created at the location of the tower in the image. This grayscale mask is then used to remove content outside the tower in subsequent images.
[0021] Optionally, determining the grayscale mask for the target image region in each of the first images in the first image group includes: Determine the bounding box coordinates of each first image in the first image group; the bounding box is used to mark the image range where the tower is located in the first image; Based on the bounding box coordinates, determine the grayscale mask corresponding to the target image region in the first image.
[0022] Since each first image in the same first image group contains towers, but the positions of the towers in the images may differ due to different shooting angles, it is necessary to label the towers in each first image. The labeling is done using bounding boxes, and the position information of the bounding boxes is the bounding box coordinates.
[0023] After labeling the tower with the annotation box, the target area in the first image is determined using the bounding box coordinates, and the grayscale mask at the corresponding position of the bounding box coordinates is calculated using the bounding box coordinates.
[0024] For grayscale masks, hard masks, soft masks, and Gaussian feathering masks can be further calculated.
[0025] Optionally, based on the bounding box coordinates, determine the grayscale mask corresponding to the target image region in the first image, including: Based on the bounding box coordinates, the target image region in the first image is determined; Based on SAM2 segmentation, the target image region in the first image is segmented and a fixed threshold binary classification is performed to generate an initial mask. Perform connected component processing on the initial mask to determine the grayscale mask.
[0026] SAM2 segmentation, short for SegmentAnythingModel2, is a cue-driven visual segmentation (PVS) model for unified image processing. Connected component processing can be an image processing flow that extracts independent target regions from a binary mask image based on pixel connectivity rules, and performs a series of morphological and statistical operations such as filtering, correction, filling, segmentation, and deletion on each connected region.
[0027] When determining the grayscale mask corresponding to the target image region in the first image, the target image region in the first image can be determined using bounding box coordinates. Then, SAM2 segmentation is used to segment the target image region in the first image to generate the initial mask corresponding to the location of the tower. Finally, connected component processing is performed to generate the grayscale mask corresponding to the target image region.
[0028] Optionally, determining the bounding box coordinates of each first image in the first image group includes: In the first image group, a first number of second images and a second number of third images are determined; the sum of the second and third numbers is the total number of images in the first image group. Label the bounding boxes of the towers in the second image and determine their coordinates; Boundary box coordinates for each third image are generated by projecting bounding box coordinates from the bounding box coordinates of the second image.
[0029] To improve the efficiency of bounding box annotation, the first images in the first image group can be divided into a first number of second images and a second number of third images. The second images are manually annotated, and the bounding box is projected using the homography matrix to determine the bounding box coordinates of each third image.
[0030] Let the bounding box of the second image be... Its four corner points are Then, from the homography matrix After projection, it satisfies The bounding box of the third image is formed by the four corner points after projection: .
[0031] Optionally, determining the bounding box coordinates of each first image in the first image group includes: In the first image group, a first number of second images and a second number of third images are determined; the sum of the second and third numbers is the total number of images in the first image group. Label the bounding boxes of the towers in the second image and determine their coordinates; The bounding box coordinates in the second image are normalized according to the width and height of the first image and mapped to the third image to determine the bounding box coordinates of each third image.
[0032] In actual processing, insufficient feature matching may occur. Therefore, when the number of feature matching is insufficient, normalized coordinate migration is used as a fallback strategy, that is, the coordinates of the bounding box of the second image are normalized according to the width and height of the image before being mapped to the third image. ; ; ; ; in, These are the width and height of the second image, respectively. These represent the width and height of the third image, respectively.
[0033] S130. Input the first image group into the first model to obtain the three-dimensional reconstruction model parameters; the first model is a Transformer model based on visual geometric constraints; the three-dimensional reconstruction model parameters include camera pose, depth map, three-dimensional point cloud and confidence information.
[0034] The Visual Geometry Grounded Transformer (VGGT) model takes any number of scene images as input and predicts camera parameters, pixel-wise depth, dense 3D point cloud, and cross-view point trajectory simultaneously in a single inference. The model achieves multi-view feature fusion through a global attention mechanism, breaking away from the step-by-step iterative optimization process of traditional motion recovery structures and supporting rapid 3D scene reconstruction in single and multiple views.
[0035] By directly inputting the first image group into the first model, camera pose, depth map, 3D point cloud and confidence information can be output. VGGT maintains the original network structure and parameter form without modifying its backbone structure.
[0036] S140. The confidence information in the parameters of the 3D reconstruction model is weighted based on the grayscale mask, and the 3D reconstruction model is determined based on the weighting result and the parameters of the 3D reconstruction model.
[0037] After obtaining the grayscale mask, the grayscale mask is applied to the confidence tensor output by VGGT for weighted processing to suppress the background of the 3D point cloud and depth results, thereby constructing a 3D reconstruction model that contains only towers as much as possible.
[0038] As mentioned earlier, grayscale masks can be further used to calculate hard masks, soft masks, and Gaussian feathered masks.
[0039] Figure 2 This is a schematic diagram comparing the confidence levels under different masking strategies applicable to embodiments of the present invention. See also... Figure 2 Therefore, when performing weighted processing, the background region has zero weight in hard mask mode; the background region retains the minimum weight in soft mask mode; in Gaussian feathering mode, the binary mask is first Gaussian blurred and then mapped to a continuous weight field so that the weight at the target boundary transitions smoothly, reducing point cloud boundary breaks and step artifacts.
[0040] S150. Based on pre-determined power-themed vector semantic data and a three-dimensional reconstruction model, a three-dimensional reconstructed power digital twin system is generated.
[0041] After the 3D reconstruction model is built, line numbers, tower numbers, scene numbers, equipment model addresses, and spatial index information are bound together, supporting spatial retrieval, viewpoint positioning, and result viewing based on scene numbers and object numbers. The filtered 3D point cloud is loaded, rendered, and displayed to form a 3D scene representation of the target area of the transmission line. Inspection and maintenance information is recorded, queried, and associated, and maintenance content is stored according to object type and object number. Historical record queries are supported by tower, site, or line. The system provides front-end interfaces for asset management, alarm management, map management, recognition processing, and 3D reconstruction processing, returning query results and displaying resource addresses, thus completing the construction of the 3D reconstruction power digital twin system.
[0042] The technical solution provided by this invention involves determining several first image groups; each first image group is a collection of images generated from first images of the same tower taken from different viewpoints; the towers in each first image group are different; determining the grayscale mask of the target image region in each first image of the first image group; the target image region is the image region of the tower in the first image; inputting the first image groups into a first model to obtain three-dimensional reconstruction model parameters; the first model is a Transformer model based on visual geometric constraints; the three-dimensional reconstruction model parameters include camera pose, depth map, three-dimensional point cloud, and confidence information; weighting the confidence information in the three-dimensional reconstruction model parameters based on the grayscale mask, and determining the three-dimensional reconstruction model based on the weighting result and the three-dimensional reconstruction model parameters; generating a three-dimensional reconstructed power digital twin system based on pre-determined power topic vector semantic data and the three-dimensional reconstruction model; using VGGT for rapid construction of the three-dimensional reconstruction model, and using the grayscale mask to process only the tower as much as possible to avoid data noise from irrelevant data, thereby achieving rapid and automated construction of the three-dimensional reconstructed power digital twin.
[0043] In one embodiment, Figure 3This is a flowchart of another method for constructing a three-dimensional reconstruction power digital twin system provided by an embodiment of the present invention. The relationship between this embodiment and the above embodiments is further explained in that, when the confidence information includes three-dimensional point cloud confidence and depth confidence, the process of weighting the confidence information in the three-dimensional reconstruction model parameters based on a grayscale mask is further elaborated. For example... Figure 3 As shown, the method includes: S210. Using drones, take real-world photos of each tower from several angles to obtain several first image sets.
[0044] When determining the first image set of a power pole, manually photographing the pole from different angles would severely impact the efficiency of generating the first image set. To address this, a drone can be used to photograph the pole, and during the photographing process, video can be recorded from different angles. Based on the video footage, the first images from different angles can be selected from different frames. Alternatively, the drone can be controlled to move to various angles to capture images.
[0045] S220. Determine the grayscale mask of the target image region in each of the first images in the first image group; the target image region is the image region of the tower in the first image.
[0046] S230. Input the first image group into the first model to obtain the three-dimensional reconstruction model parameters; the first model is a Transformer model based on visual geometric constraints; the three-dimensional reconstruction model parameters include camera pose, depth map, three-dimensional point cloud and confidence information.
[0047] S240. Element-wise multiplication is used to weight the confidence scores of the grayscale mask and the 3D point cloud.
[0048] S250. Element-wise multiplication is used to weight the grayscale mask and the depth confidence score.
[0049] The confidence score of the 3D point cloud output by VGGT is denoted as... The depth confidence score is denoted as Semantic mask is denoted as Then the weighted result at the output end can be expressed as: ; ; In the formula, This indicates element-wise multiplication.
[0050] Hard mask satisfies A soft mask can be represented as A Gaussian feathering mask can be represented as .in, To preserve the coefficients for the background, The value is the size of the Gaussian blur kernel. Using the above method, the weights of the target edge regions decrease smoothly from the center outwards, thereby suppressing background noise and reducing 3D point cloud fragmentation.
[0051] S260. Based on the weighted results and the parameters of the three-dimensional reconstruction model, determine the three-dimensional reconstruction model.
[0052] S270. Based on predetermined power-themed vector semantic data and a three-dimensional reconstruction model, a three-dimensional reconstructed power digital twin system is generated.
[0053] The technical solution provided by this invention uses a drone to take real-world photos of each tower from several angles to obtain several first image groups, thereby avoiding the problem of low efficiency in manual acquisition and improving the acquisition efficiency of the first image groups. Furthermore, by using element-wise multiplication to weight the confidence scores of the grayscale mask and the 3D point cloud, and by using element-wise multiplication to weight the confidence scores of the grayscale mask and the depth confidence score, the confidence scores are weighted.
[0054] In one embodiment, Figure 4 This is a schematic diagram of a three-dimensional reconstruction device for a digital twin power system provided in an embodiment of the present invention. Figure 4 As shown, the device includes: an image acquisition module 310, a mask calculation module 320, a parameter acquisition module 330, a model construction module 340, and a system construction module 350, wherein: Image acquisition module 310 is used to determine several first image groups; a first image group is a set of images consisting of first images generated after taking pictures of the same tower from different perspectives; each first image group takes pictures of a different tower; The mask calculation module 320 is used to determine the grayscale mask of the target image region in each of the first images in the first image group; the target image region is the image region of the tower in the first image. The parameter acquisition module 330 is used to input the first image group into the first model to obtain the three-dimensional reconstruction model parameters; the first model is a Transformer model based on visual geometric constraints; the three-dimensional reconstruction model parameters include camera pose, depth map, three-dimensional point cloud and confidence information; The model building module 340 is used to weight the confidence information in the parameters of the 3D reconstruction model based on the grayscale mask, and to determine the 3D reconstruction model based on the weighting result and the parameters of the 3D reconstruction model. System building module 350 is used to generate a three-dimensional reconstructed digital twin system for power based on pre-determined power-themed vector semantic data and a three-dimensional reconstruction model.
[0055] In one embodiment, optionally, determining the grayscale mask for the target image region in each of the first images in the first image group includes: Determine the bounding box coordinates of each first image in the first image group; the bounding box is used to mark the image range where the tower is located in the first image; Based on the bounding box coordinates, determine the grayscale mask corresponding to the target image region in the first image.
[0056] In one embodiment, optionally, determining the grayscale mask corresponding to the target image region in the first image based on the bounding box coordinates includes: Based on the bounding box coordinates, the target image region in the first image is determined; Based on SAM2 segmentation, the target image region in the first image is segmented, and a fixed threshold binary classification process is performed to generate an initial mask. Perform connected component processing on the initial mask to determine the grayscale mask.
[0057] In one embodiment, optionally, determining the bounding box coordinates of each first image in the first image group includes: In the first image group, a first number of second images and a second number of third images are determined; the sum of the second and third numbers is the total number of images in the first image group. Label the bounding boxes of the towers in the second image and determine their coordinates; Boundary box coordinates for each third image are generated by projecting bounding box coordinates from the bounding box coordinates of the second image.
[0058] In one embodiment, optionally, determining the bounding box coordinates of each first image in the first image group includes: In the first image group, a first number of second images and a second number of third images are determined; the sum of the second and third numbers is the total number of images in the first image group. Label the bounding boxes of the towers in the second image and determine their coordinates; The bounding box coordinates in the second image are normalized according to the width and height of the first image and mapped to the third image to determine the bounding box coordinates of each third image.
[0059] In one embodiment, optionally, the confidence information includes 3D point cloud confidence and depth confidence; Among them, the confidence information in the 3D reconstruction model parameters is weighted based on a grayscale mask, including: Element-wise multiplication is used to weight the confidence scores of the grayscale mask and the 3D point cloud. Element-wise multiplication is used to weight the grayscale mask and the depth confidence score.
[0060] In one embodiment, optionally, determining a plurality of first image groups includes: Several first image sets were obtained by taking real-world photos of each tower from several angles using drones.
[0061] The three-dimensional reconstruction power digital twin system construction device provided in the embodiments of the present invention can execute the three-dimensional reconstruction power digital twin system construction method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method execution.
[0062] In one embodiment, Figure 5 This is a structural block diagram of an electronic device for a method of constructing a three-dimensional reconstructed power digital twin system provided in an embodiment of the present invention, as shown in the diagram. Figure 5 The diagram illustrates a schematic representation of an electronic device 10 that can be used to implement embodiments of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.
[0063] like Figure 5 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, which is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor 11, and the computer program is executed by the at least one processor 11 to enable the at least one processor 11 to execute the three-dimensional reconstruction power digital twin system construction method provided by the present invention.
[0064] The processor 11 can perform various appropriate actions and processes based on a computer program stored in the read-only memory (ROM) 12 or a computer program loaded from the storage unit 18 into the random access memory (RAM) 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.
[0065] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.
[0066] Processor 11 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the method for constructing a three-dimensional reconstruction power digital twin system provided by this invention.
[0067] In some embodiments, the methods provided herein may be implemented as computer programs tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the methods described above may be performed. Alternatively, in other embodiments, processor 11 may be configured by any other suitable means (e.g., by means of firmware) to perform a method for constructing a three-dimensional reconstruction power digital twin system.
[0068] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard parts (ASSPs), systems-on-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.
[0069] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.
[0070] In the context of this invention, a computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for constructing a three-dimensional reconstructed power digital twin system provided by this invention.
[0071] The present invention also provides a computer program product comprising a computer program that, when executed by a processor, implements the method provided according to embodiments of the present invention. A computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, the computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.
[0072] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).
[0073] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.
[0074] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.
[0075] This invention also provides a computer program product, including a computer program that, when executed by a processor, can implement the methods provided in any embodiment of this application.
[0076] In the implementation of the computer program product, computer program code for performing the operations of this application can be written in one or more programming languages or a combination thereof. Programming languages include object-oriented programming languages such as Java, Smalltalk, and C++, as well as conventional procedural programming languages such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0077] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.
[0078] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.
Claims
1. A method for constructing a three-dimensional reconstruction power digital twin system, characterized in that, include: Determine several first image groups; The first image group is a collection of images generated from the first images of the same tower taken from different perspectives; The towers photographed in each of the first image groups are different; Determine the grayscale mask of the target image region in each first image of the first image group; the target image region is the image region of the tower in the first image. The first image group is input into the first model to obtain the three-dimensional reconstruction model parameters; the first model is a Transformer model based on visual geometric constraints; the three-dimensional reconstruction model parameters include camera pose, depth map, three-dimensional point cloud and confidence information; The confidence information in the 3D reconstruction model parameters is weighted based on the grayscale mask, and the 3D reconstruction model is determined based on the weighting result and the 3D reconstruction model parameters. Based on the predetermined power-themed vector semantic data and the three-dimensional reconstruction model, a three-dimensional reconstructed power digital twin system is generated.
2. The method according to claim 1, characterized in that, Determining the grayscale mask for the target image region in each of the first images in the first image group includes: Determine the bounding box coordinates of each first image in the first image group; the bounding box is used to mark the image range where the tower is located in the first image; Based on the bounding box coordinates, the grayscale mask corresponding to the target image region in the first image is determined.
3. The method according to claim 2, characterized in that, Based on the bounding box coordinates, the grayscale mask corresponding to the target image region in the first image is determined, including: Based on the bounding box coordinates, the target image region in the first image is determined; Based on SAM2 segmentation, the target image region in the first image is segmented, and a fixed threshold binary classification process is performed to generate an initial mask. The initial mask is processed to determine the grayscale mask.
4. The method according to claim 2, characterized in that, Determining the bounding box coordinates of each first image in the first image group includes: In the first image group, a first number of second images and a second number of third images are determined; the sum of the second number and the third number is the total number of images in the first image group. The towers in the second image are labeled with bounding boxes, and the coordinates of the bounding boxes are determined; Based on the bounding box coordinates of the second image, bounding box projection is performed to generate the bounding box coordinates of each third image.
5. The method according to claim 2, characterized in that, Determining the bounding box coordinates of each first image in the first image group includes: In the first image group, a first number of second images and a second number of third images are determined; the sum of the second number and the third number is the total number of images in the first image group. The towers in the second image are labeled with bounding boxes, and the coordinates of the bounding boxes are determined; The bounding box coordinates in the second image are normalized according to the width and height of the first image and mapped to the third image to determine the bounding box coordinates of each third image.
6. The method according to claim 1, characterized in that, The confidence information includes 3D point cloud confidence and depth confidence; The weighting of confidence information in the 3D reconstruction model parameters based on the grayscale mask includes: The grayscale mask and the confidence level of the 3D point cloud are weighted by element-wise multiplication. The grayscale mask and the depth confidence score are weighted using element-wise multiplication.
7. The method according to claim 1, characterized in that, Determine several first image groups, including: Several first image sets were obtained by taking real-world photos of each tower from several angles using drones.
8. A device for constructing a three-dimensional reconstruction digital twin system for power systems, characterized in that, include: The image acquisition module is used to determine several first image groups; The first image group is a collection of images generated from the first images of the same tower taken from different perspectives; The towers photographed in each of the first image groups are different; The mask calculation module is used to determine the grayscale mask of the target image region in each of the first images in the first image group; the target image region is the image region of the tower in the first image. The parameter acquisition module is used to input the first image group into the first model to obtain the three-dimensional reconstruction model parameters; the first model is a Transformer model based on visual geometric constraints; the three-dimensional reconstruction model parameters include camera pose, depth map, three-dimensional point cloud and confidence information; The model building module is used to weight the confidence information in the parameters of the three-dimensional reconstruction model based on the grayscale mask, and to determine the three-dimensional reconstruction model based on the weighting result and the parameters of the three-dimensional reconstruction model. The system construction module is used to generate a three-dimensional reconstructed digital twin system for power based on pre-determined power-thematic vector semantic data and the three-dimensional reconstruction model.
9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for constructing a three-dimensional reconstructed power digital twin system as described in any one of claims 1-7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the method for constructing a three-dimensional reconstructed power digital twin system as described in any one of claims 1-7.
