Offshore wind power plant identification method, device and equipment and storage medium

Abstract

The embodiment of the invention provides an offshore wind plant identification method and device, equipment and a storage medium, and is applied to the technical field of machine learning. The method comprises the following steps: acquiring a to-be-identified offshore remote sensing image, and further acquiring a slice image of a to-be-identified offshore wind power plant; the slice image of the offshore wind power plant to be identified is obtained by slicing the offshore remote sensing image to be identified; inputting the slice image of the offshore wind plant to be identified into a pre-trained offshore wind plant identification model to obtain an image coordinate of an identification frame corresponding to the slice image of the offshore wind plant to be identified; the identification frame is used for identifying the identified offshore wind plant; the offshore wind power plant identification model is constructed based on a YOLOv8 model and a C3K2 model; and converting the image coordinates into geographic coordinates to obtain position information and quantity information of the offshore wind power plant corresponding to the offshore remote sensing image to be identified. In this way, high-precision and high-efficiency full-automatic identification of the offshore wind plant can be realized.

Description

Technical Field

[0001] The present disclosure relates to the field of machine learning technology, and in particular to an offshore wind farm identification method, apparatus, device, and storage medium. Background Art

[0002] As a representative of new energy sources, offshore wind power holds significant significance for the nation's strategic transformation. It can reduce carbon emissions and increase the proportion of renewable energy in the traditional energy system. It can also drive the healthy development of related industrial chains, promoting technological innovation and industrial upgrading. Accurately understanding the spatiotemporal variations of offshore wind farms is crucial for estimating energy output.

[0003] Currently, offshore wind farm statistics still rely on manual, layer-by-layer reporting, a time-consuming and labor-intensive process, with uncertain accuracy. High-resolution satellite-based data capture of offshore wind farm locations and numbers is an effective and cost-effective method. However, with the increasing popularity and development of offshore wind power, relying solely on manual visual interpretation becomes prohibitively expensive. Therefore, there is an urgent need to develop a fully automated, high-precision offshore wind farm detection method to provide a key technical reference for accurately estimating the number of offshore wind farms.

[0004] Based on high-resolution optical remote sensing images to locate offshore wind farms, the currently commonly used target detection models include traditional methods and deep learning methods.

[0005] Target detection models based on traditional methods mainly include template detection and feature matching. This method is mainly based on target templates or manually designed related features, combined with image processing and other means to achieve automatic extraction of targets of interest. This type of method is effective for simple contextual features, but for complex scenes, the universality and robustness of this method are low.

[0006] Deep learning-based object detection methods mainly include one-stage and two-stage methods. One-stage methods, with their simple and elegant structure, are currently the mainstream object detection methods. Examples include the Yolo series (You Only Look Once algorithm is an end-to-end real-time object detection method based on deep learning. Its core idea is to transform object detection into a regression problem, directly predicting bounding box positions and class probabilities through a single forward propagation), the CenterNet series (a single-stage object detection algorithm based on keypoint detection that achieves efficient detection by predicting the center point position and related attributes such as size and orientation), and the RetinalNet series. Two-stage models include Fast-RCNN (Fast Regional Convolutional Network), Faster-RCNN, and Mask-RCNN (Regional Convolutional Neural Network). While two-stage methods offer higher accuracy, they suffer from high computational efficiency and resource consumption, making them difficult to deploy on edge devices.

[0007] Currently, YoloV8 is the most widely used model for target detection in satellite imagery, balancing efficiency and computational accuracy. This model consists of three components: backbone feature extraction, neck network, and head network. All three components are implemented based on a convolutional structure, resulting in very high computational efficiency and relatively high accuracy. It is widely used in typical target detection fields such as aircraft and oil tank inspections. However, there is currently a lack of an improved YoloV8 model specifically for offshore wind power detection to improve the efficiency and reduce the cost of offshore wind farm identification and detection. Summary of the Invention

[0008] The present disclosure provides an offshore wind farm identification method, apparatus, device, and storage medium.

[0009] According to a first aspect of the present disclosure, a method for identifying an offshore wind farm is provided. The method comprises:

[0010] Acquire an offshore remote sensing image to be identified, and then acquire a slice image of the offshore wind farm to be identified; the slice image of the offshore wind farm to be identified is obtained by slicing the offshore remote sensing image to be identified;

[0011] Inputting the slice image of the offshore wind farm to be identified into a pre-trained offshore wind farm identification model to obtain image coordinates of an identification frame corresponding to the slice image of the offshore wind farm to be identified; the identification frame is used to identify the identified offshore wind farm; the offshore wind farm identification model is constructed based on the YOLOv8 model and the C3K2 model;

[0012] The image coordinates are converted into geographic coordinates to obtain the location information and quantity information of the offshore wind farm corresponding to the offshore remote sensing image to be identified.

[0013] According to the above aspects and any possible implementation, an implementation is further provided.

[0014] Slicing the marine remote sensing image to be identified includes:

[0015] The marine remote sensing image to be identified is sliced according to a preset slice size and overlap size.

[0016] According to the above aspects and any possible implementation, an implementation is further provided.

[0017] The process of constructing the offshore wind farm identification model based on the YOLOv8 model and the C3K2 model includes the following steps:

[0018] Introducing the C3K2 model into the Backbone network of the YOLOv8 model;

[0019] Introduce the C3K2 model into the Neck network of the YOLOv8 model.

[0020] According to the above aspects and any possible implementation, an implementation is further provided.

[0021] The process of pre-training the offshore wind farm identification model includes the following steps:

[0022] Obtain images of labeled offshore wind farms as a sample set;

[0023] Training the offshore wind farm recognition model using the sample set as input and the image and image coordinates of the recognition frame as output;

[0024] The output recognition frame image is compared with the labeled offshore wind farm image, and the model parameters are optimized according to the comparison results until convergence, thereby obtaining a trained offshore wind farm recognition model.

[0025] According to the above aspects and any possible implementation, an implementation is further provided.

[0026] The identification frame carries confidence information;

[0027] Before converting the image coordinates into geographic coordinates to obtain the location information and quantity information of the offshore wind farm corresponding to the offshore remote sensing image to be identified, the method includes:

[0028] The recognition frames are sorted and screened according to the confidence information, and overlapping frames are removed from the recognition frames.

[0029] According to the above aspects and any possible implementation, an implementation is further provided.

[0030] The confidence sorting and screening of the recognition frames according to the confidence information includes:

[0031] The recognition frames are sorted in descending order according to the confidence level, and recognition frames with a confidence level less than a preset confidence threshold are eliminated.

[0032] According to the above aspects and any possible implementation, an implementation is further provided.

[0033] The performing overlapping frame removal processing on the identification frame includes:

[0034] Calculate the overlapping area between any two recognition frames, and remove the recognition frame with the larger overlapping area than the preset overlap threshold.

[0035] According to a second aspect of the present disclosure, a device for identifying an offshore wind farm is provided. The device comprises:

[0036] An acquisition module is used to acquire an offshore remote sensing image to be identified, and then acquire a slice image of the offshore wind farm to be identified; the slice image of the offshore wind farm to be identified is obtained by slicing the offshore remote sensing image to be identified;

[0037] a processing module, configured to input the slice image of the offshore wind farm to be identified into a pre-trained offshore wind farm identification model to obtain image coordinates of an identification frame corresponding to the slice image of the offshore wind farm to be identified; the identification frame is used to identify the identified offshore wind farm; the offshore wind farm identification model is constructed based on the YOLOv8 model and the C3K2 model;

[0038] The generating module is used to convert the image coordinates into geographic coordinates to obtain the location information and quantity information of the offshore wind farm corresponding to the offshore remote sensing image to be identified.

[0039] According to a third aspect of the present disclosure, an electronic device is provided, comprising: a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the above method when executing the program.

[0040] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the method according to the first aspect of the present disclosure is implemented.

[0041] The offshore wind farm identification method, apparatus, device, and storage medium provided by the embodiments of the present disclosure train an offshore wind farm identification model constructed based on the YOLOv8 model and the C3K2 model to obtain a model of an offshore wind farm that can automatically identify an image to be identified, and output the image coordinates of the offshore wind farm. Then, the longitude and latitude of the offshore wind farm are determined based on the image coordinates to obtain specific location information and quantity information of the offshore wind farm, thereby improving identification efficiency and reducing identification costs.

[0042] It should be understood that the contents described in the Summary of the Invention section are not intended to limit the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are provided for a better understanding of the present disclosure and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, among which:

[0044] Figure 1 A flow chart of an offshore wind farm identification method according to an embodiment of the present disclosure is shown;

[0045] Figure 2 shows a structural block diagram of an offshore wind farm identification model according to an embodiment of the present disclosure;

[0046] Figure 3 The figure shows the overall structural block diagram of the C3K2 model in the offshore wind farm identification model according to an embodiment of the present disclosure; wherein (a) is the structural block diagram of the C3K2 model; (b) is the structural block diagram of the C3K network in the C3K2 model;

[0047] Figure 4 A schematic diagram of offshore wind farm identification results according to an embodiment of the present disclosure is shown;

[0048] Figure 5 A block diagram of an offshore wind farm identification device according to an embodiment of the present disclosure is shown;

[0049] Figure 6 A schematic block diagram of an exemplary electronic device capable of implementing embodiments of the present disclosure is shown;

[0050] Figure 7 A structural block diagram of a windmill convolutional network according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0051] To make the purpose, technical solutions, and advantages of the embodiments of the present disclosure more clear, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present disclosure.

[0052] In this document, the term "and / or" simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document generally indicates that the related objects are in an "or" relationship.

[0053] In this paper, a high-precision offshore wind power intelligent detection model is designed based on the convolutional structure. The model is improved based on YoloV8, and the backbone feature extraction module and detection head part are optimized, with higher accuracy and lower calculation parameters. It can be subsequently deployed on drones or on-orbit satellites to realize fully automatic real-time detection of offshore wind farms.

[0054] Figure 1 A flow chart of an offshore wind farm identification method 100 according to an embodiment of the present disclosure is shown. The method 100 includes:

[0055] Step 110 : Acquire the offshore remote sensing image to be identified, and then acquire the slice image of the offshore wind farm to be identified.

[0056] In some embodiments, to address memory overflow issues in models running on very large images, a target detection method using overlapping slices is proposed. Within an entire very large image, offshore wind turbine identification is performed on any slice using overlapping slices. Specifically, slicing the offshore remote sensing image to be identified includes slicing the offshore remote sensing image to be identified according to a preset slice size and overlap.

[0057] In some embodiments, a marine remote sensing image to be identified is first acquired, for example, a high-resolution, large-scale marine image. This large-scale image is then dynamically sliced, ensuring that each slice maintains a certain degree of overlap with adjacent slices. This is done by slicing according to a preset slice size and overlap. For example, a slice size of 512 pixels, 1024 pixels, and so on, with an overlap of 64 pixels, is sufficient to cover the wind farm target for sub-meter remote sensing images. This results in multiple slice images.

[0058] Step 120 : Inputting the slice image of the offshore wind farm to be identified into a pre-trained offshore wind farm identification model to obtain image coordinates of an identification frame corresponding to the slice image of the offshore wind farm to be identified.

[0059] In some embodiments, offshore wind farm identification is performed on each slice image in sequence.

[0060] In some embodiments, the process of constructing the offshore wind farm identification model based on the YOLOv8 model and the C3K2 model includes the following steps: introducing the C3K2 model into the Backbone network of the YOLOv8 model; introducing the C3K2 model into the Neck network of the YOLOv8 model. Figure 2 and Figure 3 As shown in the figure, the C3K2 model is introduced into the Backbone network and Neck network of the YOLOv8 model to improve the computational efficiency and recognition accuracy of the model. Through actual experiments, the accuracy of wind farm recognition is improved by about 1%. Among them, Conv is responsible for extracting the features of the input image; Concat is responsible for combining feature maps of different levels; Detect is responsible for target detection tasks, including predicting bounding boxes, categories, and confidence levels; Split is responsible for dividing the image into multiple parts for processing; Bottleneck is used for feature extraction and enhancement, and its main function is to process the input feature map through convolution operations, and apply residual connections when appropriate to effectively propagate information. That is, based on the YoloV8 target detection model, the backbone feature module and the detection head are further optimized, and a C3K2 improvement module is proposed to improve the target detection accuracy, and multiple convolution superposition is used to output more effective target coordinate information. Among them, the improvement of C3K2 is to replace the Conv network in C3K2 with Figure 7 The pinwheel convolutional network shown.

[0061] In some embodiments, the process of pre-training the offshore wind farm recognition model includes the following steps: obtaining images of labeled offshore wind farms as a sample set; training the offshore wind farm recognition model with the sample set as input and the image and image coordinates of the recognition frame as output; comparing the output image of the recognition frame with the image of the labeled offshore wind farm, optimizing the model parameters according to the comparison results until convergence, and obtaining a trained offshore wind farm recognition model.

[0062] In some embodiments, offshore wind farms are manually identified and labeled on images of different resolutions to construct an offshore wind farm identification dataset as a sample set.

[0063] In some embodiments, the image of the area marked by the identification box in the recognition result is compared with the image of the corresponding labeled offshore wind farm. In response to the error between the image of the area marked by the identification box in the recognition result and the image of the corresponding labeled offshore wind farm being greater than a preset error threshold, the parameters of each layer in the pre-constructed offshore wind farm recognition model are adjusted; the above process is repeated until the error between the image of the area marked by the identification box in the recognition result and the image of the corresponding labeled offshore wind farm is less than the preset error threshold, thereby completing the training of the offshore wind farm recognition model.

[0064] In some embodiments, comparing the difference between the recognition result image and the original image, that is, the manually annotated offshore wind farm image, can be performed in the following ways: ① Mean square error MSE: calculating the average of the square of the difference between the pixels of the two images; ② Peak signal-to-noise ratio PSNR: comparing the dynamic range and mean square error of the two images; ③ Structural similarity index SSIM: comparing the similarity of the two images in terms of brightness, contrast and structural information; ④ Mutual information MI: measuring the degree of information overlap between the two images; ⑤ Pearson correlation coefficient: measuring the linear correlation between the two images; ⑥ Poisson noise ratio PNR: estimating the ratio between the noise in the image and the estimated signal, etc. In addition, image errors can also be measured based on neural networks, for example, the structural similarity index CNN-SSIM and perceptual loss PerceptualLoss based on convolutional neural network CNN.

[0065] In some embodiments, for each wind farm, the final output is in the form of [Xmin, Ymin, Xmax, Ymax], where the above four coordinates represent the image coordinates of the upper left corner and the lower right corner of the wind farm in a sample slice.

[0066] Step 130 : converting the image coordinates into geographic coordinates to obtain the location information and quantity information of the offshore wind farm corresponding to the offshore remote sensing image to be identified.

[0067] In some embodiments, the identification frame carries confidence information; before converting the image coordinates into geographic coordinates to obtain the location information and quantity information of the offshore wind farm corresponding to the offshore remote sensing image to be identified, it includes: sorting and screening the identification frames according to the confidence information, and removing overlapping frames from the identification frames. The confidence sorting and screening of the identification frames according to the confidence information includes: sorting the identification frames in descending order according to the confidence level, and removing identification frames with confidence levels less than a preset confidence threshold. The overlapping frame removal of the identification frames includes: calculating the overlapping area between any two identification frames, and removing the identification frame with the larger area among the two identification frames corresponding to the preset overlapping threshold. That is, the offshore wind farm recognition model outputs the recognition frame of the identified offshore wind farm and the image coordinates of the recognition frame, and also outputs the probability value of the recognition result being an offshore wind farm. The higher the probability value, the greater the possibility of identifying an offshore wind farm. Among them, by setting a confidence threshold, those recognition results with low probability values are eliminated, for example, the recognition results below 0.5 are eliminated. In addition, after all slices are detected, global non-maximum suppression is used to eliminate overlapping frames. That is, the overlapping area of any two recognition frames is calculated, and the recognition frame with a larger area among the recognition frames whose overlapping area is greater than the preset overlap threshold is eliminated, and only the recognition frame with a smaller area is retained to eliminate redundant overlapping frames, realize global detection of targets on ultra-large images, and ensure the accuracy of the recognition results.

[0068] In some embodiments, the remaining recognition frames after the confidence elimination and overlapping frame elimination processes are used as the final recognition results to determine the locations and number of the identified offshore wind farms.

[0069] In some embodiments, the image coordinates can be converted to geographic coordinates using tools such as GDAL to obtain the latitude and longitude information of the identified offshore wind farms. The identified offshore wind farms are then counted to determine the total number of offshore wind farms in the offshore remote sensing image to be identified.

[0070] It should be noted that for the aforementioned method embodiments, for simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present disclosure is not limited by the order of the actions described, because according to the present disclosure, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present disclosure.

[0071] The above is an introduction to the method embodiment. The following is a further explanation of the solution disclosed in the present disclosure through an apparatus embodiment.

[0072] Figure 5 FIG. 5 shows a block diagram of an offshore wind farm identification device 500 according to an embodiment of the present disclosure. Figure 5 As shown, the device 500 includes:

[0073] The acquisition module 510 is used to acquire the offshore remote sensing image to be identified, and then acquire a slice image of the offshore wind farm to be identified; the slice image of the offshore wind farm to be identified is obtained by slicing the offshore remote sensing image to be identified;

[0074] Processing module 520, configured to input the slice image of the offshore wind farm to be identified into a pre-trained offshore wind farm identification model to obtain image coordinates of an identification frame corresponding to the slice image of the offshore wind farm to be identified; the identification frame is used to identify the identified offshore wind farm; the offshore wind farm identification model is constructed based on the YOLOv8 model and the C3K2 model;

[0075] The generating module 530 is configured to convert the image coordinates into geographic coordinates to obtain the location information and quantity information of the offshore wind farm corresponding to the offshore remote sensing image to be identified.

[0076] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0077] According to an embodiment of the present disclosure, the present disclosure further provides an electronic device and a readable storage medium.

[0078] Figure 6 A schematic block diagram of an electronic device 600 that can be used to implement an embodiment of the present disclosure is shown. 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 assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0079] The electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a ROM 602 or a computer program loaded from a storage unit 608 into a RAM 603. The RAM 603 may also store various programs and data required for the operation of the electronic device 600. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An I / O interface 605 is also connected to the bus 604.

[0080] Multiple components in the electronic device 600 are connected to the I / O interface 605, including an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a magnetic disk, an optical disk, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the electronic device 600 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0081] The computing unit 601 can be a variety of general-purpose and / or specialized processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 601 performs the various methods and processes described above, such as method 100. For example, in some embodiments, method 100 can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as a storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by the computing unit 601, one or more steps of the method 100 described above can be performed. Alternatively, in other embodiments, the computing unit 601 can be configured to perform method 100 in any other appropriate manner (e.g., by means of firmware).

[0082] Various embodiments of the systems and techniques described 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 products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system comprising at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0083] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0084] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0085] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer 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 can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the 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 acoustic input, voice input, or tactile input).

[0086] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0087] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact through a communication network. The client-server relationship arises through computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, a server in a distributed system, or a server integrated with a blockchain.

[0088] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not limited herein.

[0089] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A method for identifying an offshore wind farm, characterized in that: include: Acquire offshore remote sensing images to be identified, and then obtain slice images of the offshore wind farm to be identified; The slice image of the offshore wind farm to be identified is obtained by slicing the offshore remote sensing image to be identified; Inputting the slice image of the offshore wind farm to be identified into a pre-trained offshore wind farm identification model to obtain image coordinates of an identification frame corresponding to the slice image of the offshore wind farm to be identified; the identification frame is used to identify the identified offshore wind farm; the offshore wind farm identification model is constructed based on the YOLOv8 model and the C3K2 model; The image coordinates are converted into geographic coordinates to obtain the location information and quantity information of the offshore wind farm corresponding to the offshore remote sensing image to be identified.

2. The method according to claim 1, characterized in that Slicing the marine remote sensing image to be identified includes: The marine remote sensing image to be identified is sliced according to a preset slice size and overlap size.

3. The method according to claim 1, characterized in that The process of constructing the offshore wind farm identification model based on the YOLOv8 model and the C3K2 model includes the following steps: Introducing the C3K2 model into the Backbone network of the YOLOv8 model; Introduce the C3K2 model into the Neck network of the YOLOv8 model.

4. The method according to claim 3, characterized in that The process of pre-training the offshore wind farm identification model includes the following steps: Obtain images of labeled offshore wind farms as a sample set; Training the offshore wind farm recognition model using the sample set as input and the image and image coordinates of the recognition frame as output; The output recognition frame image is compared with the labeled offshore wind farm image, and the model parameters are optimized according to the comparison results until convergence, thereby obtaining a trained offshore wind farm recognition model.

5. The method according to claim 1, wherein The identification frame carries confidence information; Before converting the image coordinates into geographic coordinates to obtain the location information and quantity information of the offshore wind farm corresponding to the offshore remote sensing image to be identified, the method includes: The recognition frames are sorted and screened according to the confidence information, and overlapping frames are removed from the recognition frames.

6. The method according to claim 5, characterized in that The confidence sorting and screening of the recognition frames according to the confidence information includes: The recognition frames are sorted in descending order according to the confidence level, and recognition frames with a confidence level less than a preset confidence threshold are eliminated.

7. The method according to claim 1, characterized in that The performing overlapping frame removal processing on the identification frame includes: Calculate the overlapping area between any two recognition frames, and remove the recognition frame with the larger overlapping area than the preset overlap threshold.

8. An offshore wind farm identification device, characterized in that: include: An acquisition module is used to acquire offshore remote sensing images to be identified, and then obtain slice images of the offshore wind farm to be identified; The slice image of the offshore wind farm to be identified is obtained by slicing the offshore remote sensing image to be identified; a processing module, configured to input the slice image of the offshore wind farm to be identified into a pre-trained offshore wind farm identification model to obtain image coordinates of an identification frame corresponding to the slice image of the offshore wind farm to be identified; the identification frame is used to identify the identified offshore wind farm; the offshore wind farm identification model is constructed based on the YOLOv8 model and the C3K2 model; The generating module is used to convert the image coordinates into geographic coordinates to obtain the location information and quantity information of the offshore wind farm corresponding to the offshore remote sensing image to be identified.

9. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 7.

Images