An offshore wind farm remote monitoring method based on image processing

Abstract

The present application relates to the field of remote monitoring, especially to a kind of offshore wind farm remote monitoring method based on image processing, the present application is by controlling mobile inspection unit group to carry image acquisition equipment to the area of offshore wind farm is patrolled;When the first mobile unit moves to different area, the pre-acquisition image for the area of offshore wind farm collected is obtained, and the pre-acquisition image is analyzed for interference;Background interference characteristic parameter is calculated based on background interference feature, and the background interference category of each area is determined;Adaptively control the inspection parameter when the post mobile unit moves to target area.The present application improves the flexibility and environmental adaptability of inspection through first-aid-post cooperative mechanism;Focus on the interference source around the equipment, quantify the background interference intensity, improve the pertinence of interference identification;Dynamic adaptation interference intensity, reduce the influence of sea area interference source, improve image quality from the source;Multi-dimensional verification image quality, ensure the usability and reliability of inspection result.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of remote monitoring, in particular to a sea wind farm remote monitoring method based on image processing. BACKGROUND

[0002] With the continuous expansion of the scale of offshore wind farms, equipment operation and maintenance has become a key link to ensure stable energy supply. Among them, inspection as the core means of operation and maintenance, bears the important function of real-time monitoring of the state of related equipment, discovering potential faults in advance, which is directly related to the power generation efficiency of the wind farm, the service life of the equipment and the operation safety.

[0003] In particular, when night inspection is carried out, there is no shelter on the sea, and natural light sources such as moonlight, starlight, or distant ship lights, and coastal lighthouse lights, are easy to directly illuminate the flat sea surface. Because the surface of the sea is in a mirror state when there is no wind or light wind, the light will form a mirror reflection spot in a regular reflection form, and these spots may cover the key areas such as the wind turbine tower and blades in the inspection image, or directly enter the imaging field of view of the inspection equipment, resulting in a high-light reflection area in the local image, resulting in poor quality of the collected image, and unable to accurately analyze the subsequent operation of the related equipment state.

[0004] Therefore, a remote monitoring method for offshore wind farms is needed to accurately deal with the image quality problem caused by night sea mirror reflection, realize "interference identifiable, parameter adjustable, image verifiable" closed-loop management and control, and provide reliable data support for subsequent analysis of equipment state.

[0005] Chinese patent application publication No. CN120389521A discloses a multi-machine cooperative distribution network inspection method and system based on digital twinning, belonging to the field of digital twinning technology. The technical solution maps the power distribution network architecture and base station distribution to the GIS remote sensing map, constructs a comprehensive digital twinning model, and delimits the inspection area based on this. By analyzing the base station distribution and terrain characteristics in the area, the comprehensive inspection difficulty is determined, and then the inspection area is divided into multiple task blocks. The optimal target nest is selected from the deployed unmanned aerial vehicle nest to perform the inspection task using the contract network mechanism, and the optimal inspection route is generated according to the task block and nest position. Finally, the unmanned aerial vehicle controlled by the target nest completes all the inspection tasks according to the planned route. This scheme realizes intelligent scheduling of inspection resources and precise allocation of tasks, significantly improves the efficiency of power distribution network inspection, reduces the cost of inspection, and provides a strong guarantee for the safe and stable operation of the power grid.

[0006] However, the prior art still has the following problems,

[0007] Traditional inspection methods rely on fixed inspection paths to cover each point individually, failing to flexibly adjust the inspection sequence or skip meaningless areas based on actual interference, thus reducing inspection efficiency. Furthermore, the use of uniform parameter acquisition fails to focus on key observation areas, resulting in a large amount of low-quality data that directly affects the status assessment of related equipment, reducing the effectiveness and usability of the inspection. Summary of the Invention

[0008] To address these issues, this invention provides a remote monitoring method for offshore wind farms based on image processing. This method overcomes the limitations of existing technologies that rely on fixed inspection paths for point-to-point coverage, failing to flexibly adjust the inspection sequence or skip meaningless areas based on actual interference, thus reducing inspection efficiency. Furthermore, the use of a uniform parameter acquisition mode makes it impossible to focus on the main observation area, resulting in a large amount of low-quality data that directly affects the status assessment of related equipment, reducing the effectiveness and usability of the inspection.

[0009] To achieve the above objectives, the present invention provides a remote monitoring method for offshore wind farms based on image processing, comprising:

[0010] The mobile inspection unit group is equipped with image acquisition equipment to inspect the offshore wind farm area, including advance mobile unit and follow-up mobile unit moving and inspecting along the same moving path at different time intervals.

[0011] Acquire preliminary images of the offshore wind farm area collected when the advance mobile unit moves to different areas, and perform interference analysis on the preliminary images, including determining a background interference box based on the sea background in the preliminary images and the area where the target equipment is located, and extracting background interference features within the background interference box;

[0012] Based on the background interference features, background interference characterization parameters are calculated to determine the background interference category of each region;

[0013] The inspection parameters for controlling the movement of the rear moving unit to the target area based on the background interference category include:

[0014] The inspection height and image acquisition angle of the rear moving unit are continuously adjusted until the acquired image meets the sky background conditions. The main observation frame is determined based on the sky background and the area where the target device is located in the acquired image. The image features are analyzed to verify the image, and the verified image is used as the inspection image.

[0015] Alternatively, the rear moving unit can be controlled to maintain the inspection parameters of the advance moving unit and acquire inspection images.

[0016] Furthermore, a background interference box is determined based on the marine background in the pre-acquired image and the area where the target device is located, including:

[0017] Determine the edge contour corresponding to the target device;

[0018] The edge contour is enlarged by a first predetermined ratio to obtain a first reference frame;

[0019] The first reference box, which is in the background of the sea area, is determined as the background interference box.

[0020] Furthermore, the background interference features include texture density and the proportion of overexposed areas.

[0021] Furthermore, the process of calculating the background interference characterization parameters based on the background interference features includes:

[0022] The ratio of texture density threshold to texture density is used as the first background interference feature;

[0023] The ratio of the overexposed area percentage to the overexposed area percentage threshold is used as the second background interference feature;

[0024] The sum of the first background interference feature and the second background interference feature is used as the background interference characterization parameter.

[0025] Furthermore, the background interference categories for each region are determined, including:

[0026] If the background interferometric characterization parameter is greater than or equal to the background interferometric characterization parameter threshold, then the background interferometric category of the region is determined to be the strong interferometric category.

[0027] If the background interferometric characterization parameter is less than the background interferometric characterization parameter threshold, then the background interferometric category of the determined region is the weak interferometric category.

[0028] Furthermore, the inspection parameters for the rear-mounted moving unit when it moves to the target area based on the background interference category include:

[0029] If the background interference category of the area is a strong interference category, the inspection height and image acquisition angle of the rear moving unit are continuously adjusted until the acquired image meets the sky background conditions. Based on the sky background and the area where the target device is located in the acquired image, the main observation box is determined, the image features are analyzed to verify the image, and the verified image is used as the inspection image.

[0030] If the background interference category of the region is weak interference, then the rear moving unit is controlled to maintain the inspection parameters of the advance moving unit and acquire inspection images.

[0031] Furthermore, the sky background condition includes the target device being in a sky background at a proportion greater than a proportion threshold.

[0032] Furthermore, the process of determining the main observation box includes:

[0033] Determine the edge contour corresponding to the target device;

[0034] The edge contour is enlarged by a second predetermined ratio to obtain a second reference frame;

[0035] The second reference box, which is in the background of the sky, is designated as the primary view box.

[0036] Further, the image features are analyzed to verify the image, including:

[0037] If the sharpness of the acquired image is greater than the sharpness threshold and the spot coverage is less than the spot coverage threshold, then the verification is passed.

[0038] The image features include sharpness and light spot coverage.

[0039] Furthermore, the inspection parameters include inspection height and image acquisition angle.

[0040] Compared with existing technologies, this invention controls a mobile inspection unit group equipped with image acquisition equipment to inspect offshore wind farm areas; acquires preliminary images of the offshore wind farm area collected by the advance mobile unit as it moves to different areas, and performs interference analysis on the preliminary images; calculates background interference characterization parameters based on background interference features to determine the background interference category of each area; and adaptively controls the inspection parameters of the follow-up mobile unit as it moves to the target area based on the background interference category. This invention improves the flexibility and environmental adaptability of inspection through an advance-follow-up collaborative mechanism; focuses on interference sources around the equipment, quantifies background interference intensity, and improves the targeting of interference identification; dynamically adapts to interference intensity to reduce the impact of marine interference sources, improving image quality from the source; and verifies image quality from multiple dimensions to ensure the usability and reliability of the inspection results.

[0041] In particular, this invention establishes a "precursor-rear" dual-unit collaborative inspection mechanism. The advance mobile unit is responsible for acquiring preceding images and performing background interference analysis, completing the interference analysis before the arrival of the rear mobile unit. This allows time for the rear mobile unit to adjust inspection parameters, avoiding the generation of numerous low-quality images due to untimely adjustments in the "acquisition-interference-judgment" mode. It anticipates interference in advance, avoiding acquisition lag. The rear mobile unit, based on the interference analysis conclusions of the advance mobile unit, adjusts the inspection parameters accordingly and acquires the final inspection images. This achieves a collaborative division of labor between "detection and acquisition," improving the efficiency of inspection tasks and enhancing the overall efficiency of large-scale offshore wind farm inspections.

[0042] In particular, this invention focuses on quantifying background interference around the equipment, eliminating interference from irrelevant backgrounds. This eliminates the need to process redundant information across the entire image, reducing computational load while accurately locating the core problem areas in nighttime inspections where "sea glare and specular reflection affect equipment imaging," thus improving the accuracy of interference identification. Furthermore, a differentiated dynamic adjustment strategy for inspection parameters based on background interference categories is employed. This includes: for areas with strong interference, adjusting parameters to convert the sea background to a sky background, avoiding core interference sources such as glare, specular reflection, and water defocusing at night, fundamentally improving image quality; for areas with weak interference, maintaining inspection parameters for image acquisition, especially in scenarios with limited nighttime inspection time windows, avoiding the time wasted by "forcibly adjusting parameters regardless of interference strength," significantly improving overall inspection speed while ensuring image quality meets standards. Attached Figure Description

[0043] Figure 1 This is a schematic diagram illustrating the steps of an image processing-based remote monitoring method for offshore wind farms according to an embodiment of the invention.

[0044] Figure 2 A logic decision diagram for determining the background interference category of each region in an embodiment of the invention;

[0045] Figure 3 This is a logic determination diagram of the inspection parameters when the rear moving unit moves to the target area based on the background interference category control in an embodiment of the invention.

[0046] Figure 4 This is a logical decision diagram for analyzing image features and verifying images in an embodiment of the invention. Detailed Implementation

[0047] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0048] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0049] It should be noted that in the description of this invention, terms such as "inner" and "outer" indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings. This is merely for ease of description and does not indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0050] Please see Figure 1The diagram illustrates the steps of an image-processing-based remote monitoring method for offshore wind farms according to an embodiment of the present invention. The image-processing-based remote monitoring method for offshore wind farms according to an embodiment of the present invention includes:

[0051] Step S1: Control the mobile inspection unit group equipped with image acquisition equipment to inspect the offshore wind farm area, including the advance mobile unit and the rear mobile unit moving and inspecting along the same moving path at different time intervals.

[0052] Step S2: Obtain the preliminary acquisition images of the offshore wind farm area collected when the advance mobile unit moves to different areas, and perform interference analysis on the preliminary acquisition images, including determining the background interference box based on the sea background in the preliminary acquisition image and the area where the target equipment is located, and extracting the background interference features within the background interference box.

[0053] Step S3: Calculate background interference characterization parameters based on the background interference features to determine the background interference category of each region;

[0054] Step S4, based on the background interference category, controls the inspection parameters when the rear moving unit moves to the target area, including:

[0055] The inspection height and image acquisition angle of the rear moving unit are continuously adjusted until the acquired image meets the sky background conditions. The main observation frame is determined based on the sky background and the area where the target device is located in the acquired image. The image features are analyzed to verify the image, and the verified image is used as the inspection image.

[0056] Alternatively, the rear moving unit can be controlled to maintain the inspection parameters of the advance moving unit and acquire inspection images.

[0057] Specifically, the mobile inspection unit refers to a drone, and the image acquisition device can be a high-definition camera. The mobile inspection unit includes a preliminary mobile unit and a rear mobile unit.

[0058] Specifically, the acquisition methods for texture density, overexposed area ratio, sharpness, and spot coverage can be carried out using any existing technology capable of recognizing the above features, which will not be elaborated further here.

[0059] Specifically, the background interference box is determined based on the marine background in the pre-acquired image and the area where the target device is located, including:

[0060] Determine the edge contour corresponding to the target device;

[0061] The edge contour is enlarged by a first predetermined ratio to obtain a first reference frame;

[0062] The first reference box, which is in the background of the sea area, is determined as the background interference box.

[0063] It is understandable that if the acquired image only covers a small area very close to the edge of the device, the background interference features (texture density, proportion of overexposed areas) may be incompletely extracted, making it impossible to accurately determine the true degree of interference. Therefore, based on the edge contour, the analysis is expanded outward to cover the high-incidence area of ​​interference at the boundary between the device and the sea, while not extending to distant and unrelated sea areas, ensuring that the interference analysis is focused and targeted. Therefore, in this embodiment, the first predetermined ratio is set to 1.2, that is, the edge contour boundary is expanded outward by 20%, which will not be elaborated further.

[0064] Specifically, the background interference features include texture density and the proportion of overexposed areas.

[0065] Specifically, the process of calculating background interference characterization parameters based on the background interference features includes:

[0066] The ratio of texture density threshold to texture density is used as the first background interference feature;

[0067] The ratio of the overexposed area percentage to the overexposed area percentage threshold is used as the second background interference feature;

[0068] The sum of the first background interference feature and the second background interference feature is used as the background interference characterization parameter.

[0069] In this embodiment, the purpose of setting the texture density threshold and the overexposed area proportion threshold is to characterize the situation where the background interference is relatively serious, resulting in poor image quality. By acquiring historical pre-collected images of several offshore wind farm area inspections completed by the advance mobile unit, the historical texture density and overexposed area proportion data are called to solve for the mean texture density and the mean overexposed area proportion, and the corresponding values ​​are used as the benchmark values ​​under normal conditions. Based on the purpose of setting the above two thresholds, the texture density threshold is determined as the product of the mean texture density and the density deviation coefficient, and the overexposed area proportion threshold is determined as the product of the mean overexposed area proportion and the area deviation coefficient. The density deviation coefficient is selected in the range [0.9, 0.95], preferably 0.9 in the implementation, and the area deviation coefficient is selected in the range [1.1, 1.2], preferably 1.1 in the implementation.

[0070] Specifically, this invention establishes a "precursor-rear" dual-unit collaborative inspection mechanism. The advance mobile unit is responsible for acquiring preceding images and performing background interference analysis, completing the interference analysis before the arrival of the rear mobile unit. This allows time for the rear mobile unit to adjust inspection parameters, avoiding the generation of numerous low-quality images due to untimely adjustments in the "acquisition-interference-judgment" mode. It anticipates interference in advance, mitigating acquisition lag. The rear mobile unit, based on the interference analysis conclusions of the advance mobile unit, specifically adjusts inspection parameters and acquires the final inspection images. This achieves a collaborative division of labor between "detection and acquisition," improving the efficiency of inspection tasks and overall enhancing the efficiency of large-scale offshore wind farm inspections.

[0071] Specifically, please refer to Figure 2 As shown, this is a logic decision diagram for determining the background interference category of each region according to an embodiment of the present invention. Determining the background interference category of each region includes:

[0072] If the background interferometric characterization parameter is greater than or equal to the background interferometric characterization parameter threshold, then the background interferometric category of the region is determined to be the strong interferometric category.

[0073] If the background interferometric characterization parameter is less than the background interferometric characterization parameter threshold, then the background interferometric category of the determined region is the weak interferometric category.

[0074] The background interference characterization parameter threshold is predetermined. The background interference characterization parameter threshold is determined by calculating the background interference characterization parameter when the texture density threshold is equal to the texture density and the overexposed area proportion is equal to the overexposed area proportion threshold.

[0075] Specifically, please refer to Figure 3 As shown, this is a logic determination diagram of the inspection parameters when the rear moving unit moves to the target area based on the background interference category control according to an embodiment of the present invention. The inspection parameters when the rear moving unit moves to the target area based on the background interference category control include:

[0076] If the background interference category of the area is a strong interference category, the inspection height and image acquisition angle of the rear moving unit are continuously adjusted until the acquired image meets the sky background conditions. Based on the sky background and the area where the target device is located in the acquired image, the main observation box is determined, the image features are analyzed to verify the image, and the verified image is used as the inspection image.

[0077] If the background interference category of the region is weak interference, then the rear moving unit is controlled to maintain the inspection parameters of the advance moving unit and acquire inspection images.

[0078] Specifically, the sky background condition includes the target device being in a sky background at a proportion greater than a proportion threshold.

[0079] Specifically, the process of determining the main observation box includes:

[0080] Determine the edge contour corresponding to the target device;

[0081] The edge contour is enlarged by a second predetermined ratio to obtain a second reference frame;

[0082] The second reference box, which is in the background of the sky, is designated as the primary view box.

[0083] It is understandable that the main observation frame is the core carrier for subsequent verification of image quality. It needs to include a small amount of sky background related to the equipment imaging, while reducing the impact of light and shadow changes (such as cloud shadows and moonlight reflections) in irrelevant areas (such as distant clouds and cluttered sky edges) on image acquisition, and avoiding interference with the verification results from including the aforementioned irrelevant areas. If the magnification ratio is too small, key parts of the equipment (such as the tips of wind turbine blades, tower connection points, and edges of small monitoring devices) will not be included in the frame, making it impossible to fully evaluate the equipment and directly affecting the equipment status detection (such as the inability to observe details such as blade cracks and loose parts). If the magnification ratio is too large, irrelevant areas included in the main observation frame will interfere with and affect image quality, leading to misjudgments in image quality verification. For example, dark areas in distant clouds may be misjudged as blurred equipment, and moonlight reflections in clouds may be misjudged as light spots. Therefore, in this embodiment, the second predetermined ratio is set to 1.3 times, that is, the edge contour boundary is extended outward by 30%, which will not be elaborated further.

[0084] Specifically, please refer to Figure 4 As shown, it is a logical decision diagram for analyzing image features and verifying images according to an embodiment of the present invention. The analysis of image features and verification of images includes:

[0085] If the sharpness of the acquired image is greater than the sharpness threshold and the spot coverage is less than the spot coverage threshold, then the verification is passed.

[0086] The image features include sharpness and light spot coverage.

[0087] In this embodiment, the purpose of setting the sharpness threshold and the spot coverage threshold is to characterize the poor quality of the acquired images. By acquiring historical images collected from several inspections of the offshore wind farm area by the rear-mounted mobile unit, historical sharpness data and historical spot coverage data are called to solve for the mean sharpness and mean spot coverage, which are then used as the benchmark values ​​under normal conditions. Based on the purpose of setting the above two thresholds, the sharpness threshold is determined to be the product of the mean sharpness and the sharpness deviation coefficient, and the spot coverage threshold is determined to be the product of the mean spot coverage and the coverage deviation coefficient. The sharpness deviation coefficient is selected in the interval [1.2, 1.3], preferably 1.2 in practice, and the coverage deviation coefficient is selected in the interval [0.9, 0.95], preferably 0.9 in practice.

[0088] Specifically, the verification logic is determined based on actual needs to ensure image availability. Clarity and spot coverage are used as verification indicators to address the "out-of-focus and glare" problems caused by the nighttime sea background. This ensures that verified images can clearly present equipment details and meet the actual needs of subsequent equipment status detection (such as component damage and oil stain detection), thus preventing invalid images from flowing into subsequent analysis stages.

[0089] Specifically, the inspection parameters include inspection height and image acquisition angle.

[0090] Specifically, this invention focuses on quantifying background interference around the equipment, eliminating interference from irrelevant backgrounds (such as distant sea surfaces and sky edges). It eliminates the need to process redundant information across the entire image, reducing computational load while accurately locating the core problem areas in nighttime inspections where "sea glare and specular reflections affect equipment imaging," thus improving the accuracy of interference identification. Furthermore, it employs a differentiated dynamic adjustment strategy for inspection parameters based on background interference categories. This includes: for areas with strong interference, adjusting parameters to convert the sea background to a sky background, avoiding core interference sources such as glare, specular reflections, and water blurring at night, fundamentally improving image quality; for areas with weak interference, maintaining inspection parameters for image acquisition, especially in scenarios with limited nighttime inspection time windows, avoiding the time wasted by "forcibly adjusting parameters regardless of interference strength," significantly improving overall inspection speed while ensuring image quality meets standards.

[0091] If the image processing-based remote monitoring method for offshore wind farms of the present invention is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0092] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A remote monitoring method for offshore wind farms based on image processing, characterized in that, include: The mobile inspection unit group is equipped with image acquisition equipment to inspect the offshore wind farm area, including advance mobile unit and follow-up mobile unit moving and inspecting along the same moving path at different time intervals. Acquire preliminary images of the offshore wind farm area collected when the advance mobile unit moves to different areas, and perform interference analysis on the preliminary images, including determining a background interference box based on the sea background in the preliminary images and the area where the target equipment is located, and extracting background interference features within the background interference box; Based on the background interference features, background interference characterization parameters are calculated to determine the background interference category of each region; The inspection parameters for controlling the movement of the rear moving unit to the target area based on the background interference category include: If the background interference category of the area is a strong interference category, the inspection height and image acquisition angle of the rear moving unit are continuously adjusted until the acquired image meets the sky background conditions. Based on the sky background and the area where the target device is located in the acquired image, the main observation box is determined, the image features are analyzed to verify the image, and the verified image is used as the inspection image. If the background interference category of the region is weak interference, then the rear moving unit is controlled to maintain the inspection parameters of the advance moving unit and acquire inspection images. The process of calculating background interference characterization parameters based on the aforementioned background interference features includes: The ratio of texture density threshold to texture density is used as the first background interference feature; The ratio of the overexposed area percentage to the overexposed area percentage threshold is used as the second background interference feature; The sum of the first background interference feature and the second background interference feature is used as the background interference characterization parameter; Determine the background interference category for each region, including: If the background interferometric characterization parameter is greater than or equal to the background interferometric characterization parameter threshold, then the background interferometric category of the determined region is the strong interferometric category; If the background interferometric characterization parameter is less than the background interferometric characterization parameter threshold, then the background interferometric category of the region is determined to be the weak interferometric category; The background interference features include texture density and the proportion of overexposed areas, and the sky background condition includes the proportion of the target device in the sky background being greater than a proportion threshold.

2. The remote monitoring method for offshore wind farms based on image processing according to claim 1, characterized in that, The background interferometer is determined based on the marine background in the pre-acquired image and the area where the target device is located, including: Determine the edge contour corresponding to the target device; The edge contour is enlarged by a first predetermined ratio to obtain a first reference frame; The first reference box, which is in the background of the sea area, is determined as the background interference box.

3. The remote monitoring method for offshore wind farms based on image processing according to claim 1, characterized in that, The process of determining the main observation box includes: Determine the edge contour corresponding to the target device; The edge contour is enlarged by a second predetermined ratio to obtain a second reference frame; The second reference box, which is in the background of the sky, is designated as the primary view box.

4. The remote monitoring method for offshore wind farms based on image processing according to claim 1, characterized in that, Image feature analysis and verification, including: If the sharpness of the acquired image is greater than the sharpness threshold and the spot coverage is less than the spot coverage threshold, then the verification is passed. The image features include sharpness and light spot coverage.

5. The remote monitoring method for offshore wind farms based on image processing according to claim 1, characterized in that, The inspection parameters include inspection height and image acquisition angle.

Images