Fan blade defect detection method and device, electronic equipment and storage medium

Abstract

The application provides a fan blade defect detection method and device, electronic equipment and a storage medium. The method comprises the following steps: projecting an orthogonal polarization encoded light field to a blade surface, acquiring an encoded polarization image stream and performing orthogonal decoding processing, and separating to obtain a target blade polarization image; based on posture data in illumination parameters and blade rotation parameters, performing feature enhancement processing on the target blade polarization image to generate an enhanced feature image; weighting and fusing phase feature parameters and texture morphological feature parameters to generate a blade surface fusion feature set; performing similarity matching operation on the blade surface fusion feature set and a defect standard feature library, determining a blade defect state, and outputting a detection result. The encoded polarization image is separated to obtain the target blade polarization image, and the enhanced feature image is generated by enhancement. The defect is detected based on the enhanced feature image, the quality of the image and the extracted feature is improved, and the accuracy of the defect detection is improved.

Description

Technical Field

[0001] This application relates to the field of wind power testing technology, and in particular to a method, device, electronic equipment and storage medium for detecting defects in wind turbine blades. Background Technology

[0002] As the core power component of a wind power generation system, the surface defects of wind turbine blades, such as cracks, corrosion, and wear, can directly affect the aerodynamic performance and operational safety of the unit, and may even lead to serious accidents such as blade breakage. Therefore, defect detection of wind turbine blades is a key link in ensuring the stable operation of wind power generation systems.

[0003] Currently, non-contact defect detection methods based on polarization imaging technology are being gradually applied in the field of wind turbine blade inspection. This method utilizes the difference in polarization characteristics between the normal area and the defect area on the blade surface to identify defects, and has the advantages of not requiring contact with the blade and being adaptable to outdoor inspection environments.

[0004] However, in actual inspection scenarios, wind turbine blades are constantly rotating at high speeds, and their real-time rotation angle, three-dimensional attitude, and other parameters are continuously and dynamically changing. Simultaneously, the outdoor ambient light parameters also fluctuate naturally over time. Traditional polarization imaging detection methods do not adapt to the dynamic changes in blade rotation attitude and the fluctuations in ambient light parameters; they only employ fixed polarization light field projection strategies and image acquisition parameters. This results in the acquired blade polarization images being easily affected by attitude changes and light fluctuations, leading to blurred blade surface features and insufficient differentiation between defective and normal areas in polarization characteristics. Consequently, these methods cannot provide a high-quality image foundation for subsequent defect feature extraction, ultimately resulting in defect detection accuracy that fails to meet practical engineering requirements. Summary of the Invention

[0005] This application aims to at least partially address one of the technical problems in the related art.

[0006] Therefore, this application proposes a method, apparatus, electronic device and storage medium for detecting defects in wind turbine blades.

[0007] One embodiment of this application proposes a method for detecting defects in wind turbine blades, including:

[0008] Step S1: Receive multi-source core parameters synchronously transmitted by the multi-source sensor module. The multi-source core parameters include blade rotation parameters, environmental parameters, illumination parameters, and imaging device parameters. Step S2: Generate an orthogonal polarization coded light field based on the blade rotation parameters and environmental parameters; send a control command to the projection module to project the orthogonal polarization coded light field onto the blade surface; send an acquisition command to the imaging device to obtain the coded polarization image stream; perform orthogonal decoding processing on the coded polarization image stream to separate and obtain the polarization image of the target blade. Step S3: Based on the attitude data in the illumination parameters and blade rotation parameters, a dynamic weight allocation mechanism is constructed to perform feature enhancement processing on the polarization image of the target blade and generate an enhanced feature image. Step S4: After preprocessing the enhanced feature image, extract phase feature parameters and texture morphology feature parameters, perform weighted fusion processing on the two types of feature parameters, and generate a blade surface fusion feature set. Step S5: Call the preset defect standard feature library, perform similarity matching operation between the fused feature set of the blade surface and the defect standard feature library, determine the defect status of the blade and output the detection result.

[0009] Optionally, the blade rotation parameters include rotation period, rotation speed, rotation fundamental frequency, real-time rotation angle, and three-dimensional attitude data; The environmental parameters include dust concentration, humidity, and salt spray content; The illumination parameters include the solar altitude angle and the solar azimuth angle; The imaging device parameters include the refractive index of the lens of the polarization imaging device, the initial angle of the polarizer, and the imaging resolution parameters; wherein, all received parameters are bound to the same timestamp.

[0010] Optionally, step S2 includes: Based on the rotating fundamental frequency, the modulation period of the orthogonal polarization coding is set; Combining the modulation period, environmental parameters, and initial angle of the polarizer, an orthogonal polarization coding rule is determined. Based on the orthogonal polarization coding rule, the coding polarization direction of the orthogonal polarization coded light field is optimized to generate an orthogonal polarization coded light field. Send control commands to the projection module to project the orthogonal polarization coded light field onto the blade surface; Based on the projection result and imaging resolution of the orthogonal polarization coded light field, a collection command is sent to the imaging device to control the imaging device to collect the coded polarization image stream at the corresponding resolution, and at the same time trigger the imaging device to record the real-time rotation angle of the blade corresponding to each frame of the image. Based on the orthogonal polarization coding rules, with the real-time rotation angle of the blade as the attitude calibration basis and the lens refractive index in the imaging equipment parameters as the optical correction parameter, the coded polarization image stream is subjected to frame-by-frame orthogonal decoding processing to obtain the orthogonal decoding result. Based on the orthogonal decoding results, a target blade polarization image containing only blade surface feature information is obtained.

[0011] Optionally, the method further includes: Local region segmentation is performed on the polarization image of the target blade to obtain multiple independent local regions; Calculate the average pixel gray value of each local region, and the deviation between the average pixel gray value of each local region and the average pixel gray value of the global pixel gray value of the polarization image of the target leaf; Based on the deviation value, the surface reflection coefficient of the corresponding local area is corrected in reverse, wherein the corrected surface reflection coefficient increases non-linearly with the increase of the deviation value; The corrected surface reflectance coefficients of each local region are substituted into the effective illumination intensity calculation process to calculate the effective illumination intensity of each local region.

[0012] Optionally, step S3 includes: Extract the three-dimensional attitude data and illumination parameters from the blade rotation parameters, and calculate the angle between the blade surface normal vector and the solar incident vector; Based on the included angle and the effective light intensity of each local area, the effective light intensity of the windward and leeward sides of the blade and the difference in light intensity between the two are calculated to obtain the spatial distribution data of the effective light intensity. Extract the rotational speed parameter from the blade rotation parameters, and calculate the comprehensive weight allocation factor based on the set illumination weight coefficient and rotational speed weight coefficient; Based on the difference in light intensity and the comprehensive weight allocation factor, the enhancement weights of the windward and leeward sides are adjusted, and pixel intensity adjustment is performed on the polarization image of the target blade to generate an enhanced feature image.

[0013] Optionally, adjusting the enhancement weights of the windward and leeward sides based on the light intensity difference and the comprehensive weight allocation factor, and performing pixel intensity adjustment on the polarization image of the target blade to generate an enhanced feature image, includes: Along the boundary line between the windward and leeward sides of the blade, a gradual adjustment zone is constructed based on the multiple independent local regions; Based on the spatial distribution data of the gradient adjustment zone and the effective light intensity, the continuous variation features of the effective light intensity are extracted, and a gradient weight sequence is generated based on the continuous variation features. The gradient weight sequence is used to smoothly adjust the pixel intensity of the target blade polarization image, thereby generating an enhanced feature image.

[0014] Optionally, step S4 includes: Preprocessing is performed on the enhanced feature image, including Gaussian filtering for noise reduction and histogram equalization. Phase reconstruction is performed on the preprocessed enhanced feature image to extract phase feature parameters, wherein the phase feature parameters include the coordinates of phase abrupt change points, phase fluctuation amplitude, and phase uniformity index. Edge detection and texture analysis are performed on the preprocessed enhanced feature image to extract texture morphology feature parameters, wherein the texture morphology feature parameters include the edge contour, texture roughness and area size of the defect region; Based on the dust concentration and salt spray content in the environmental parameters, the weight ratio of phase feature parameters and texture morphology feature parameters is dynamically allocated, and weighted fusion processing is performed on the dynamically allocated phase feature parameters and texture morphology feature parameters to generate a fused feature set of the blade surface.

[0015] Optionally, step S5 includes: The system invokes a preset defect standard feature library, which contains fused feature templates for crack, corrosion, wear, and dent types. Based on the fusion feature set of the blade surface and the fusion feature template in the defect standard feature library, a similarity matching operation is performed to obtain the similarity matching operation result; Based on the similarity matching calculation results and the preset feature difference threshold, it is determined whether the blade has defects and the type of defects, and the defect determination result is obtained. Based on the defect determination results, and combined with the rotation period and real-time rotation angle in the blade rotation parameters, the defect location coordinates are calibrated, and the detection results including the defect location coordinates, size, and severity level are output.

[0016] Another embodiment of this application proposes a wind turbine blade defect detection device, comprising: The data acquisition module is used to receive multi-source core parameters synchronously transmitted by the multi-source sensor module. The multi-source core parameters include blade rotation parameters, environmental parameters, illumination parameters, and imaging device parameters. The separation module is used to generate an orthogonal polarization coded light field based on the blade rotation parameters and environmental parameters, send control commands to the projection module to project the orthogonal polarization coded light field onto the blade surface, send acquisition commands to the imaging device to obtain the coded polarization image stream, perform orthogonal decoding processing on the coded polarization image stream, and separate the polarization image of the target blade. The enhancement module is used to construct a dynamic weight allocation mechanism based on the attitude data in the illumination parameters and blade rotation parameters, and to perform feature enhancement processing on the polarization image of the target blade to generate an enhanced feature image. The feature extraction module is used to preprocess the enhanced feature image, extract phase feature parameters and texture morphology feature parameters, perform weighted fusion processing on the two types of feature parameters, and generate a blade surface fusion feature set. The defect detection module is used to call a preset defect standard feature library, perform similarity matching calculation between the fused feature set of the blade surface and the defect standard feature library, determine the defect status of the blade, and output the detection result.

[0017] Another embodiment of this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements the method described in the foregoing aspect.

[0018] Another embodiment of this application proposes a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in the foregoing aspect.

[0019] Another embodiment of this application proposes a chip including processing circuitry configured to perform the method described in one aspect above.

[0020] Another embodiment of this application proposes a computer program product that, when executed by a processor, implements the method described in the foregoing aspect.

[0021] The wind turbine blade defect detection method, device, electronic equipment, chip, and storage medium proposed in this application can achieve the following beneficial effects: By receiving multi-source core parameters and generating orthogonal polarization-coded light fields based on blade rotation parameters and environmental parameters, the problem of traditional fixed polarization light field projection strategies being susceptible to attitude changes and illumination fluctuations is improved. This helps to enhance the stability of the acquired blade polarization images and provides a more reliable image basis for subsequent defect feature extraction.

[0022] A dynamic weight allocation mechanism is constructed based on attitude data in illumination parameters and blade rotation parameters. Feature enhancement processing is performed on the polarization image of the target blade, which can improve the problems of blurred blade surface features and insufficient differentiation of polarization characteristics between defective and normal areas, and help improve the identifiability of defect features.

[0023] By using multi-source parameter coordinated regulation, feature enhancement and weighted fusion processing, combined with similarity matching calculations of the defect standard feature library, the accuracy of blade defect detection can be improved, better meeting the actual engineering inspection needs.

[0024] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0025] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a schematic flowchart of a wind turbine blade defect detection method provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a wind turbine blade defect detection device provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of a chip proposed in an embodiment of this application. Detailed Implementation

[0026] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0027] The following description, with reference to the accompanying drawings, describes a method, apparatus, electronic device, chip, and storage medium for detecting defects in wind turbine blades according to embodiments of this application.

[0028] Figure 1 This is a flowchart illustrating a method for detecting defects in wind turbine blades provided in an embodiment of this application.

[0029] As one implementation, the wind turbine blade defect detection method of this application embodiment can be configured in a wind turbine blade defect detection device, which can be applied to any electronic device so that the electronic device can perform the wind turbine blade defect detection function.

[0030] Among them, electronic devices can be any device with computing capabilities, such as mobile terminals, which can be hardware devices with various operating systems, touch screens and / or displays, such as mobile phones, tablets, personal digital assistants, wearable devices, etc.

[0031] As another implementation, the wind turbine blade defect detection method of this application embodiment can also be executed by a chip with processing capabilities. The chip includes an image signal processing chip (ISP), a central processing unit (CPU), an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a field-programmable gate array (FPGA), a system on a chip (SOC), a reduced instruction set computer (RISC), etc., which will not be listed here.

[0032] It should be noted that all data collection operations related to users in this application are conducted with the user's authorization and in strict compliance with relevant laws and regulations such as privacy and security.

[0033] like Figure 1 As shown, the method may include the following steps: Step S1: Receive multi-source core parameters synchronously transmitted by the multi-source sensor module. The multi-source core parameters include blade rotation parameters, environmental parameters, illumination parameters, and imaging device parameters. In this embodiment, the blade rotation parameters include rotation period, rotation speed, rotation fundamental frequency, real-time rotation angle, and three-dimensional attitude data; The environmental parameters include dust concentration, humidity, and salt spray content; The illumination parameters include the solar altitude angle and the solar azimuth angle; The imaging device parameters include the refractive index of the lens of the polarization imaging device, the initial angle of the polarizer, and the imaging resolution parameters; wherein, all received parameters are bound to the same timestamp to ensure the spatiotemporal consistency of multi-source parameters.

[0034] In the specific implementation process, multi-source sensor modules are pre-deployed on the wind turbine nacelle, blade surface and detection equipment. Among them, blade rotation parameters are collected by attitude sensors and rotary encoders installed at the blade root, environmental parameters are collected by sand and dust sensors, humidity sensors and salt spray sensors installed outside the nacelle, illumination parameters are collected by solar azimuth sensors installed on the detection equipment, and imaging equipment parameters are obtained by the parameter acquisition module of the polarization imaging equipment itself.

[0035] The multi-source sensor module synchronously collects data according to a preset sampling frequency and transmits all parameters to the processor. After receiving the parameters transmitted by each sensor module, the processor binds the same timestamp to each parameter to ensure the consistency of parameters from different sources and of different types in time and space.

[0036] Step S2: Generate an orthogonal polarization coded light field based on the blade rotation parameters and environmental parameters; send a control command to the projection module to project the orthogonal polarization coded light field onto the blade surface; send an acquisition command to the imaging device to obtain the coded polarization image stream; perform orthogonal decoding processing on the coded polarization image stream to separate and obtain the polarization image of the target blade. In this embodiment, the orthogonal polarization coded light field refers to a polarized light field encoded using orthogonal polarization, whose polarization directions are orthogonally distributed according to a preset rule; the coded polarization image stream refers to a continuous image sequence acquired by the imaging device that contains orthogonal polarization coded information; the orthogonal decoding process refers to the process of inversely processing the coded polarization image stream according to the orthogonal polarization coding rules; the target leaf polarization image refers to a polarization image that, after orthogonal decoding processing, separates the background interference information and contains only the surface feature information of the leaf.

[0037] Traditional fixed-polarization light fields cannot adapt to blade rotation and environmental changes, resulting in severe image interference. This step generates an orthogonal polarization-coded light field based on blade rotation parameters and environmental parameters, enabling the light field characteristics to adapt to the blade rotation state and environmental conditions, reducing interference during light field projection. By acquiring the coded polarization image stream and performing orthogonal decoding, background interference and blade surface features can be effectively separated, resulting in a clean polarization image of the target blade.

[0038] Step S3: Based on the attitude data in the illumination parameters and blade rotation parameters, a dynamic weight allocation mechanism is constructed to perform feature enhancement processing on the polarization image of the target blade and generate an enhanced feature image. Step S4: After preprocessing the enhanced feature image, extract phase feature parameters and texture morphology feature parameters, perform weighted fusion processing on the two types of feature parameters, and generate a blade surface fusion feature set. Step S5: Call the preset defect standard feature library, perform similarity matching operation between the fused feature set of the blade surface and the defect standard feature library, determine the defect status of the blade and output the detection result.

[0039] Optionally, step S2 includes: Based on the rotating fundamental frequency, the modulation period of the orthogonal polarization coding is set; Combining the modulation period, environmental parameters, and initial angle of the polarizer, an orthogonal polarization coding rule is determined. Based on the orthogonal polarization coding rule, the coding polarization direction of the orthogonal polarization coded light field is optimized to generate an orthogonal polarization coded light field. Send control commands to the projection module to project the orthogonal polarization coded light field onto the blade surface; Based on the projection result and imaging resolution of the orthogonal polarization coded light field, a collection command is sent to the imaging device to control the imaging device to collect the coded polarization image stream at the corresponding resolution, and at the same time trigger the imaging device to record the real-time rotation angle of the blade corresponding to each frame of the image. Based on the orthogonal polarization coding rules, with the real-time rotation angle of the blade as the attitude calibration basis and the lens refractive index in the imaging equipment parameters as the optical correction parameter, the coded polarization image stream is subjected to frame-by-frame orthogonal decoding processing to obtain the orthogonal decoding result. Based on the orthogonal decoding results, a target blade polarization image containing only blade surface feature information is obtained.

[0040] In this embodiment, the rotating fundamental frequency refers to the basic frequency of the blade rotation, which is equal to the ratio of the rotational speed to 60; the modulation period refers to the time it takes for the coding rules of the orthogonal polarization coded light field to repeat once, which is the reciprocal of the rotating fundamental frequency.

[0041] The modulation period of the orthogonal polarization coded light field is synchronized with the blade rotation fundamental frequency, which enables the coding period of the light field to match the rotation period of the blade. This ensures that the light field can uniformly cover the blade surface during blade rotation, improves the effectiveness of light field projection, and reduces the problem of weak image signal caused by period asynchrony.

[0042] In the specific implementation process, the processor extracts the rotational fundamental frequency from the blade rotation parameters received in step S1. Based on the reciprocal relationship between the modulation period and the rotating fundamental frequency: the modulation period of the orthogonal polarization coding is set. The calculation formula is:

[0043] For example, the blade rotation fundamental frequency Substituting 0.25Hz into the formula, we can obtain the modulation period T = 1 / 0.25 = 4s. That is, the coding rule of the orthogonal polarization coded light field repeats once every 4s: synchronized with the blade rotation period (4s).

[0044] It should be noted that the orthogonal polarization coding rule refers to the rule used to define the polarization direction distribution of the orthogonal polarization coded light field; the coding polarization direction refers to the direction of polarized light in the orthogonal polarization coded light field.

[0045] Environmental parameters can affect the propagation characteristics of polarized light. For example, in high humidity, the polarization direction of polarized light is prone to shift. By combining the modulation period, environmental parameters, and the initial angle of the polarizer to determine the orthogonal polarization coding rule, the coding rule can be adapted to environmental conditions and equipment characteristics, optimizing the coded polarization direction, reducing polarization direction shift caused by environmental interference, and generating a high-quality orthogonal polarization coded light field.

[0046] In the specific implementation process, the processor calls the set modulation period, combines the received environmental parameters and the initial polarizer angle in the imaging device parameters, and determines the orthogonal polarization coding rule based on the preset coding rule algorithm. This rule defines the change of the polarization direction of the orthogonal polarization coded light field with time within the modulation period. After determining the coding rule, the coded polarization direction of the orthogonal polarization coded light field is optimized based on this rule to generate the orthogonal polarization coded light field.

[0047] For example, the modulation period is 4s, the ambient humidity is 60%, and the initial angle of the polarizer is 0°. The determined orthogonal polarization coding rule is: within 0-2s, the polarization direction is 0°; within 2-4s, the polarization direction is 90°, orthogonally distributed. Based on this rule, the coding polarization direction is optimized to generate an orthogonal polarization coded light field, whose polarization direction changes with time according to this rule.

[0048] A projection module is a device used to project orthogonally polarized coded light fields, and typically includes a polarization light source and a coding device.

[0049] In practice, after the processor generates the orthogonally polarized coded light field, it sends a control command to the projection module. The command includes parameters such as the projection intensity and projection time of the light field. Upon receiving the control command, the projection module projects the orthogonally polarized coded light field onto the surface of the rotating wind turbine blades according to the command requirements.

[0050] For example, the processor sends a control command to the projection module, which requests a projection intensity of 1000 lux. After receiving the command, the projection module continuously projects the orthogonal polarization encoded light field onto the surface of the rotating blade at an intensity of 1000 lux.

[0051] Imaging resolution refers to the number of pixels in an image captured by an imaging device, usually expressed as width × height. The real-time rotation angle of the blade refers to the rotation angle of the blade when the imaging device captures each frame of the image.

[0052] For example, with an imaging resolution of 1920×1080 and an acquisition frequency of 10Hz, the processor sends an acquisition command to the imaging device, requesting the acquisition of an coded polarization image stream at a resolution of 1920×1080 and a frequency of 10Hz. Upon receiving the command, the imaging device begins acquiring images: simultaneously recording the real-time rotation angle of the blade corresponding to each frame of the image; for example, the first frame corresponds to 30°, the second frame to 33°, and so on.

[0053] Attitude calibration basis refers to the basis used to calibrate the image attitude during the decoding process, namely the real-time rotation angle of the blade; optical correction parameters refer to the parameters used to correct optical errors during the decoding process, namely the refractive index of the lens of the imaging device; frame-by-frame orthogonal decoding processing refers to the process of performing orthogonal decoding processing on each frame of the encoded polarized image stream; orthogonal decoding result refers to the image sequence obtained after frame-by-frame orthogonal decoding processing.

[0054] The real-time rotation angle of the blades causes changes in image attitude, and the refractive index of the lens introduces optical errors, affecting decoding accuracy. By using orthogonal polarization coding rules as a benchmark, the real-time rotation angle of the blades as the basis for attitude calibration, and the refractive index of the lens as an optical correction parameter, frame-by-frame orthogonal decoding of the coded polarization image stream can effectively calibrate image attitude, correct optical errors, and improve decoding accuracy.

[0055] For example, the orthogonal polarization encoding rule is an orthogonal distribution of 0° and 90°. In a certain frame, the leaf's real-time rotation angle is 30°, and the lens refractive index is 1.5. The processor uses this encoding rule as a reference, calibrates the image pose based on the 30° rotation angle, corrects optical errors based on the 1.5 lens refractive index, and performs orthogonal decoding on that frame. After decoding all images frame by frame, the orthogonal decoding result is obtained.

[0056] For example, the orthogonal decoding result includes information such as blades, sky, and nacelle. The processor uses background subtraction to separate background information such as the sky and nacelle from the image, obtaining a target blade polarization image that only contains the surface features of the blade. The texture of the blade is clearly visible in the image, without background interference.

[0057] Optionally, the method further includes: Local region segmentation is performed on the polarization image of the target blade to obtain multiple independent local regions; Calculate the average pixel gray value of each local region, and the deviation between the average pixel gray value of each local region and the average pixel gray value of the global pixel gray value of the polarization image of the target leaf; Based on the deviation value, the surface reflection coefficient of the corresponding local area is corrected in reverse, wherein the corrected surface reflection coefficient increases non-linearly with the increase of the deviation value; The corrected surface reflectance coefficients of each local region are substituted into the effective illumination intensity calculation process to calculate the effective illumination intensity of each local region.

[0058] In this embodiment, considering that in outdoor wind farms, different areas of the wind turbine blade, such as the tip and root, have different surface roughness and coating thickness, resulting in different reflection characteristics in each area; the above embodiment calculates the effective light intensity based on global parameters and performs feature enhancement, without considering the differences in this area; in practical applications, it is easy to have biases in the light intensity assessment of local areas such as the tip and root of the blade, so this step is performed to reduce the bias in light intensity assessment.

[0059] Local region segmentation refers to the process of dividing the polarization image of a target leaf into multiple independent local regions; independent local regions refer to leaf regions with different reflective properties, such as the leaf tip region, leaf body region, and leaf root region.

[0060] For example, in the polarization image of the target leaf, the gray value of the leaf tip region is low, the gray value of the leaf body region is medium, and the gray value of the leaf root region is high. The processor uses a region growing algorithm, with the typical gray values ​​of the leaf tip, leaf body, and leaf root as seed points, to divide the image into three independent local regions: leaf tip, leaf body, and leaf root.

[0061] The processor retrieves multiple independent local regions, calculates the average grayscale value of all pixels within each local region, and obtains the average grayscale value of each local region. Simultaneously, it calculates the average grayscale value of all pixels in the polarization image of the target blade to obtain the global average grayscale value. Finally, it calculates the difference between the average grayscale value of each local region and the global average grayscale value to obtain the deviation value.

[0062] The surface reflectance coefficient refers to the reflectance coefficient of the blade surface to polarized light. Reverse correction refers to the process of adjusting the surface reflectance coefficient in the opposite direction based on the deviation value. That is, the larger the deviation value, the greater the correction range of the surface reflectance coefficient.

[0063] The deviation value of a local area characterizes the difference between the reflection characteristics of that area and the global area. By correcting the surface reflection coefficient based on the deviation value, the corrected surface reflection coefficient can more accurately reflect the actual reflection characteristics of the area, thereby improving the accuracy of subsequent calculations of effective light intensity.

[0064] In the specific implementation process, the processor calls the obtained deviation value and, based on a preset reflection coefficient correction algorithm, performs a reverse correction on the surface reflection coefficient of each local area. The corrected surface reflection coefficient increases non-linearly with the increase of the deviation value, and the calculation formula is as follows:

[0065] in, This is the corrected surface reflectance coefficient; The original surface reflectance coefficient; This is a correction factor; This is the deviation value; It is a non-linear exponent, typically ranging from 1.5 to 2.0.

[0066] The effective illumination intensity calculation process refers to the process of calculating the effective illumination intensity based on the surface reflectance and the incident illumination intensity. The effective illumination intensity of each local region refers to the effective illumination intensity of each local region after correcting for the surface reflectance.

[0067] The surface reflectance is a key parameter for calculating the effective illumination intensity. By substituting the corrected surface reflectance into the effective illumination intensity calculation process, the effective illumination intensity of each local area can be recalculated, making the calculation results more accurately reflect the actual illumination conditions of each area.

[0068] The recalculated effective light intensity of each local area more accurately reflects the actual light conditions of each area of ​​the leaf. Based on this data, a dynamic weight allocation mechanism is constructed and feature enhancement processing is performed, which can make the weight allocation more suitable for the actual situation of each local area and improve the targeting and effectiveness of feature enhancement.

[0069] Optionally, step S3 includes: Extract the three-dimensional attitude data and illumination parameters from the blade rotation parameters, and calculate the angle between the blade surface normal vector and the solar incident vector; Based on the included angle and the effective light intensity of each local area, the effective light intensity of the windward and leeward sides of the blade and the difference in light intensity between the two are calculated to obtain the spatial distribution data of the effective light intensity. Extract the rotational speed parameter from the blade rotation parameters, and calculate the comprehensive weight allocation factor based on the set illumination weight coefficient and rotational speed weight coefficient; Based on the difference in light intensity and the comprehensive weight allocation factor, the enhancement weights of the windward and leeward sides are adjusted, and pixel intensity adjustment is performed on the polarization image of the target blade to generate an enhanced feature image.

[0070] In this embodiment, the dynamic weight allocation mechanism refers to the mechanism of dynamically adjusting the weight allocation based on the attitude data in the illumination parameters and blade rotation parameters; the illumination intensity difference refers to the difference in effective illumination intensity between the windward and leeward sides of the blade; the spatial distribution data of effective illumination intensity refers to the distribution data of effective illumination intensity in different areas of the blade surface; the illumination weight coefficient refers to the weight coefficient set based on the illumination intensity difference; the rotation speed weight coefficient refers to the weight coefficient set based on the blade rotation speed parameter; the comprehensive weight allocation factor refers to the core factor used to adjust the enhancement weight, calculated by combining the illumination weight coefficient and the rotation speed weight coefficient; and the enhanced feature image refers to the image with improved clarity of blade surface features after feature enhancement processing.

[0071] The blade surface normal vector is a vector perpendicular to a point on the blade surface; the solar incident vector is a vector of sunlight incident on a point on the blade surface; and the included angle is the angle between the blade surface normal vector and the solar incident vector.

[0072] In the specific implementation process, the processor extracts the three-dimensional attitude data and illumination parameters from the blade rotation parameters received in step S1: Based on the three-dimensional attitude data, it calculates the normal vector of each point on the blade surface, and calculates the solar incidence vector based on the solar altitude angle and solar azimuth angle. Using the vector angle calculation formula, it calculates the angle between the normal vector of each point on the blade surface and the solar incidence vector. The vector angle calculation formula is as follows:

[0073] in, Angle Blade surface normal vector Let the solar incident vector be... For vector dot product, 、 These are the magnitudes of the two vectors, respectively.

[0074] Effective light intensity refers to the light intensity that actually shines on the surface of the blade and can be reflected; the windward side refers to the side of the blade facing the direction of the wind during rotation; the leeward side refers to the side of the blade away from the direction of the wind during rotation; the spatial distribution data of effective light intensity refers to the distribution data of effective light intensity on the windward side, leeward side and various local areas of the blade.

[0075] The difference in light intensity between the windward and leeward sides of the blade directly affects the clarity of image features. By calculating the effective light intensity on the windward and leeward sides and the difference between them based on the included angle and recalculated effective light intensity in each local area, the distribution of light intensity on the blade surface can be clearly identified.

[0076] In the specific implementation process, the processor calls the calculated included angle, combines it with the recalculated effective illumination intensity of each local area, and uses the illumination intensity calculation algorithm to solve the problem.

[0077] First, determine the baseline value for solar incident light intensity. The solar intensity can be directly collected by a light sensor or obtained by looking up tables using the solar altitude angle and azimuth angle. For each local area, the effective solar incident light intensity component of that area is calculated based on the included angle, using the following formula:

[0078] in, For the first iThe effective solar irradiance components of a local area For the first i The angle between the blade surface normal vector and the solar incident vector in a local region.

[0079] Combined with the corrected local surface reflectance coefficient The final effective illuminance of this local area is calculated using the following formula:

[0080] in, The effective illumination intensity is recalculated for the i-th local region.

[0081] Finally, based on the division of the windward and leeward sides of the blades, the effective light intensity of all local areas belonging to the windward side is averaged to obtain the effective light intensity of the windward side. The effective illuminance on the leeward side is obtained by averaging the effective illuminance of all local areas on the leeward side. The difference between the two is calculated to obtain the difference in light intensity. The formula is:

[0082] At the same time, integrate the effective light intensity data of all local areas. This yields spatial distribution data of effective light intensity.

[0083] The light intensity weighting coefficient is a coefficient set based on the difference in light intensity to characterize the degree of influence of the difference in light intensity on the weight allocation; the rotation speed weighting coefficient is a coefficient set based on the blade rotation speed parameter to characterize the degree of influence of the rotation speed on the weight allocation; the comprehensive weight allocation factor is a factor obtained by weighting the light intensity weighting coefficient and the rotation speed weighting coefficient.

[0084] Both light intensity differences and blade rotation speed affect the feature enhancement effect. The greater the light intensity difference, the more necessary it is to adjust the weights to improve the feature clarity in low-light areas; the faster the rotation speed, the more necessary it is to improve the dynamics of weight allocation. By setting light intensity weight coefficients and rotation speed weight coefficients, the comprehensive weight allocation factor can be calculated to incorporate the influence of both into the weight allocation mechanism, thereby improving the rationality and accuracy of weight allocation.

[0085] In the specific implementation process, the processor extracts the rotational speed parameter from the received blade rotation parameters and sets an illumination weighting coefficient based on the obtained light intensity difference. The greater the difference in light intensity, The larger the value, the higher the speed weighting coefficient is set based on the speed parameter. The faster the rotation speed, The larger the value, the better. After setting, a weighted calculation method is used to calculate the comprehensive weight allocation factor. W The calculation formula is:

[0086] in, , The preset weighting coefficients and proportions satisfy... .

[0087] Enhancement weight refers to the weight used to adjust the pixel intensity of the polarization image of the target blade; pixel intensity adjustment refers to the process of adjusting the grayscale value of the image pixels according to the enhancement weight; enhanced feature image refers to the image with improved clarity of the blade surface features after pixel intensity adjustment.

[0088] Adjusting the enhancement weights based on differences in light intensity and a comprehensive weighting allocation factor allows the weight allocation to be adapted to the actual light conditions and rotational speed of the blades. By adjusting the pixel intensity of the target blade's polarization image, the feature clarity of weakly lit areas can be improved, enhancing the distinction between defective and normal areas, resulting in an enhanced feature image.

[0089] In the specific implementation process, the processor calls upon the obtained light intensity difference and comprehensive weight allocation factor, and adjusts the enhancement weights of the windward and leeward sides of the blade according to the preset weight adjustment algorithm. The greater the light intensity difference, the greater the enhancement weight of the leeward side; the larger the comprehensive weight allocation factor, the greater the magnitude of the weight adjustment. After the weight adjustment is completed, the processor uses a pixel intensity adjustment algorithm to adjust the pixel intensity of the target blade polarization image according to the adjusted enhancement weights, increasing the pixel grayscale values ​​of weak light areas such as the leeward side and decreasing the pixel grayscale values ​​of strong light areas such as the windward side. After adjustment, an enhanced feature image is generated.

[0090] Optionally, adjusting the enhancement weights of the windward and leeward sides based on the light intensity difference and the comprehensive weight allocation factor, and performing pixel intensity adjustment on the polarization image of the target blade to generate an enhanced feature image, includes: Along the boundary line between the windward and leeward sides of the blade, a gradual adjustment zone is constructed based on the multiple independent local regions; Based on the spatial distribution data of the gradient adjustment zone and the effective light intensity, the continuous variation features of the effective light intensity are extracted, and a gradient weight sequence is generated based on the continuous variation features. The gradient weight sequence is used to smoothly adjust the pixel intensity of the target blade polarization image, thereby generating an enhanced feature image.

[0091] In this embodiment, in the outdoor blade rotation detection scenario, there is a significant gradient difference in light intensity between the windward and leeward sides of the blade. Although the feature enhancement processing in the above embodiment can improve feature clarity, it does not design a smooth transition strategy for this gradient difference. In practical applications, abrupt grayscale changes in the transition area between the windward and leeward sides are likely to occur, causing the texture continuity in this area to be disrupted and the abrupt grayscale changes to be mistakenly identified as defects. To solve the above problems, this embodiment is implemented.

[0092] The gradient adjustment zone refers to the area constructed at the boundary between the windward and leeward sides of the blade, used to achieve smooth adjustment of pixel intensity.

[0093] For example, the boundary between the windward and leeward sides of the blade is located in the middle of the image, and the local region segmentation result divides the blade into three regions: the tip, the blade body, and the root. The processor constructs a gradient adjustment area with a width of 10 pixels at the boundary line, which covers the transition area between the windward and leeward sides.

[0094] The continuous variation characteristic of effective illumination intensity refers to the continuous variation law of effective illumination intensity within the gradient adjustment zone; the gradient weight sequence refers to the weight sequence generated based on the continuous variation characteristic of effective illumination intensity, used to smoothly adjust pixel intensity.

[0095] For example, the effective illumination intensity within the gradient adjustment zone continuously changes from 800 lux on the windward side to 400 lux on the leeward side, with an enhancement weight of 0.4 on the windward side and 0.6 on the leeward side. The gradient weight sequence generated after extracting this continuous change feature is [0.4, 0.42, 0.44, 0.46, 0.48, 0.5, 0.52, 0.54, 0.56, 0.6], achieving a smooth transition from 0.4 to 0.6.

[0096] Pixel intensity smoothing refers to adjusting the pixel intensity within the gradient adjustment area using a gradient weight sequence: achieving a smooth transition from the windward side to the leeward side; enhanced feature image refers to an image with good texture continuity obtained after smoothing adjustment.

[0097] For example, the original grayscale value of the pixel in the gradient adjustment area changes continuously from 200 on the windward side to 100 on the leeward side, with a gradient weight sequence of [0.4, 0.42, ..., 0.6]. After adjusting the pixel intensity according to the weight sequence, the adjusted grayscale value smoothly changes from 200×0.4=80 to 100×0.6=60, without any abrupt grayscale changes, and the texture continuity is good.

[0098] This implementation refines the smoothing logic of pixel adjustment, constructs a gradient adjustment area and generates a gradient weight sequence, thereby achieving smooth adjustment of pixel intensity in the transition area. This effectively avoids abrupt changes in grayscale, ensures the continuity of the blade surface texture, reduces the possibility of misjudgment of defects, and further improves the reliability of defect detection results.

[0099] Optionally, step S4 includes: Preprocessing is performed on the enhanced feature image, including Gaussian filtering for noise reduction and histogram equalization. Phase reconstruction is performed on the preprocessed enhanced feature image to extract phase feature parameters, wherein the phase feature parameters include the coordinates of phase abrupt change points, phase fluctuation amplitude, and phase uniformity index. Edge detection and texture analysis are performed on the preprocessed enhanced feature image to extract texture morphology feature parameters, wherein the texture morphology feature parameters include the edge contour, texture roughness and area size of the defect region; Based on the dust concentration and salt spray content in the environmental parameters, the weight ratio of phase feature parameters and texture morphology feature parameters is dynamically allocated, and weighted fusion processing is performed on the dynamically allocated phase feature parameters and texture morphology feature parameters to generate a fused feature set of the blade surface.

[0100] In this embodiment, preprocessing refers to the preliminary processing of the enhanced feature image, including Gaussian filtering for noise reduction and histogram equalization; phase feature parameters refer to the set of parameters extracted through phase reconstruction to characterize the phase characteristics of the blade surface, including the coordinates of phase abrupt change points, phase fluctuation amplitude, and phase uniformity index; texture morphology feature parameters refer to the set of parameters extracted through edge detection and texture analysis to characterize the texture morphology characteristics of the blade surface, including the edge contour of the defect area, texture roughness, and area size; weighted fusion processing refers to the process of fusing the phase feature parameters and texture morphology feature parameters according to preset weights; the blade surface fusion feature set refers to the feature set obtained after weighted fusion processing, which includes the phase and texture morphology features of the blade surface.

[0101] In the specific implementation process, the processor calls the generated enhanced feature image and first uses a Gaussian filtering algorithm to denoise the image: the kernel size of the Gaussian filter is set to 3×3, the standard deviation is 1.0, and convolution operations are used to smooth the noise in the image. After denoising, a histogram equalization algorithm is used to process the image, calculate the gray-level histogram of the image, and improve the contrast of the image by adjusting the distribution of gray levels.

[0102] Phase reconstruction refers to the process of reconstructing phase information from a preprocessed enhanced feature image. Phase abrupt change point coordinates refer to the coordinates of points in the image where the phase changes abruptly; phase fluctuation amplitude refers to the range of phase fluctuation in the image; and phase uniformity index refers to an index used to characterize the uniformity of phase distribution in an image.

[0103] The phase characteristics of defective and normal areas on the blade surface differ significantly. Phase abrupt changes typically correspond to the edge of the defect, and the amplitude and uniformity of phase fluctuations can characterize the severity of the defect. By performing phase reconstruction on the preprocessed image and extracting phase feature parameters, phase feature information of the blade surface can be obtained.

[0104] For example, after phase reconstruction, the preprocessed enhanced feature image yields a phase distribution image. In the phase distribution image, phase abrupt change points with coordinates (100, 200), (105, 205), etc., are detected. These points correspond to the edge positions of cracks on the blade surface. The calculated phase fluctuation amplitude is 0.5 rad, and the phase uniformity index is 0.1, indicating that there is a defect in this area, and the defect is relatively serious.

[0105] The processor calls the preprocessed enhanced feature image and uses an edge detection algorithm, such as the Canny edge detection algorithm, to process the enhanced feature image and detect the edge contours of the defect region in the image. After edge detection, a texture analysis algorithm is used to analyze the defect region, calculate the texture roughness of the defect region, for example, by calculating the contrast of the gray-level co-occurrence matrix; calculate the number of pixels in the defect region, and combine it with the image resolution to obtain the area size of the defect region.

[0106] Dynamic weight allocation refers to the process of dynamically adjusting the weight ratio of phase feature parameters and texture morphology feature parameters according to changes in environmental parameters; the blade surface fusion feature set refers to a comprehensive feature set that includes phase feature parameters and texture morphology feature parameters.

[0107] Outdoor environments, such as dust and salt spray, can affect phase and texture features to varying degrees. For example, higher dust concentrations reduce the reliability of phase features while increasing the reliability of texture features. By dynamically assigning weights based on environmental parameters, the fused feature set can be better adapted to the current environment, improving its reliability. Weighted fusion processing combines the advantages of both types of features, providing a comprehensive characterization of the blade surface.

[0108] For example, the dust concentration in the environmental parameters is 0.01 mg / m³. 3 The salt spray content was 0.005 mg / m³. 3The weight of the dynamically assigned phase feature parameters is 0.4, and the weight of the texture morphology feature parameters is 0.6. The extracted phase feature parameters (such as phase abrupt change point coordinates, phase fluctuation amplitude, and phase uniformity index) and texture morphology feature parameters (such as edge contour, texture roughness, and area size) are weighted and fused according to the weights to generate a blade surface fusion feature set.

[0109] Optionally, step S5 includes: The system invokes a preset defect standard feature library, which contains fused feature templates for crack, corrosion, wear, and dent types. Based on the fusion feature set of the blade surface and the fusion feature template in the defect standard feature library, a similarity matching operation is performed to obtain the similarity matching operation result; Based on the similarity matching calculation results and the preset feature difference threshold, it is determined whether the blade has defects and the type of defects, and the defect determination result is obtained. Based on the defect determination results, and combined with the rotation period and real-time rotation angle in the blade rotation parameters, the defect location coordinates are calibrated, and the detection results including the defect location coordinates, size, and severity level are output.

[0110] In this embodiment, the defect standard feature library refers to a pre-set database containing various blade defect fusion feature templates, including cracks, corrosion, wear, and dents; similarity matching operation refers to the process of calculating the similarity between the blade surface fusion feature set and the fusion feature templates in the defect standard feature library; the detection result refers to the final detection result containing the defect location coordinates, size, and severity level.

[0111] For example, in the defect standard feature library, the fusion feature template for cracks is as follows: phase abrupt change points are linearly distributed, phase fluctuation amplitude is 0.3-0.6 rad, phase uniformity index is 0.05-0.15, edge contour is linear, texture roughness is 0.6-0.9, and area size depends on the crack size. The fusion feature template for corrosion is as follows: phase abrupt change points are planarly distributed, phase fluctuation amplitude is 0.2-0.4 rad, phase uniformity index is 0.1-0.2, edge contour is irregular planar, texture roughness is 0.4-0.7, and area size is relatively large.

[0112] For example, the cosine similarity algorithm is used to calculate the similarity value between the fused feature set of the blade surface and various defect templates. The similarity value with the crack template is 0.9, the similarity value with the corrosion template is 0.3, the similarity value with the wear template is 0.2, and the similarity value with the dent template is 0.1. The similarity matching operation results are [0.9, 0.3, 0.2, 0.1].

[0113] In the specific implementation process, the processor presets the feature difference threshold to 0.2, calls the obtained similarity matching operation result, and calculates the feature difference threshold corresponding to each similarity value, that is: feature difference threshold = 1 - similarity value.

[0114] The calculated feature difference thresholds are compared with preset thresholds. If the feature difference threshold for a certain type of defect is less than the preset threshold, the blade is determined to have that type of defect. If all feature difference thresholds are greater than the preset thresholds, the blade is determined to have no defects. After the determination is completed, the defect determination result is obtained.

[0115] For example, the preset feature difference threshold is 0.2, and the similarity matching operation result is [0.9, 0.3, 0.2, 0.1]. The calculated feature difference thresholds are 0.1, 0.7, 0.8, and 0.9, respectively. Among them, the feature difference threshold of 0.1 corresponding to the crack is less than the preset threshold of 0.2, so it is determined that the blade has a crack defect, and the defect determination result is that a crack defect exists.

[0116] Defect location coordinate calibration refers to the process of correcting the defect location coordinates by combining the blade rotation parameters; the detection result refers to the comprehensive result including the defect location coordinates, size and severity level; the severity level refers to the severity of the defect classified according to indicators such as the size of the defect and the amplitude of phase fluctuation, and is usually divided into three levels: slight, moderate and severe.

[0117] For example, the defect determination result is a crack defect, with the pixel coordinates of the defect location being (100, 200), the image resolution being 0.1 mm / pixel, the real-time blade rotation angle being 30°, and the rotation period being 4 seconds. Using a coordinate calibration algorithm, the pixel coordinates are converted to physical coordinates (10 mm, 20 mm), and combined with the rotation angle calibration, the actual location is determined to be in the blade body region, 100 mm from the blade root. The defect size is 10 mm × 0.5 mm, the phase fluctuation amplitude is 0.5 rad, and according to a preset standard, the severity level is determined to be medium. The final output detection result is: Defect type: crack; Defect location: blade body region, 100 mm from the blade root; Size: 10 mm × 0.5 mm; Severity level: medium.

[0118] In summary, this embodiment improves the interference of attitude changes and illumination fluctuations on image quality by receiving multi-source core parameters and coordinating the generation, acquisition, and decoding of polarized light fields, combined with feature enhancement through dynamic weight allocation and feature extraction through weighted fusion. This enhances the stability and clarity of leaf polarized images, improves the identifiability of defect features, and helps to improve the accuracy of defect detection, thus better meeting the needs of actual engineering inspection.

[0119] To achieve the above embodiments, this application also proposes a wind turbine blade defect detection device.

[0120] Figure 2 This is a schematic diagram of a wind turbine blade defect detection device provided in an embodiment of this application.

[0121] like Figure 2 As shown, the device may include: The data acquisition module 210 is used to receive multi-source core parameters synchronously transmitted by the multi-source sensor module. The multi-source core parameters include blade rotation parameters, environmental parameters, illumination parameters, and imaging device parameters. The separation module 220 is used to generate an orthogonal polarization coded light field based on the blade rotation parameters and environmental parameters, send control commands to the projection module to project the orthogonal polarization coded light field onto the blade surface, send acquisition commands to the imaging device to obtain a coded polarization image stream, perform orthogonal decoding processing on the coded polarization image stream, and separate the polarization image of the target blade. The enhancement module 230 is used to construct a dynamic weight allocation mechanism based on the attitude data in the illumination parameters and blade rotation parameters, and to perform feature enhancement processing on the polarization image of the target blade to generate an enhanced feature image. The feature extraction module 240 is used to preprocess the enhanced feature image, extract phase feature parameters and texture morphology feature parameters, perform weighted fusion processing on the two types of feature parameters, and generate a blade surface fusion feature set. The defect detection module 250 is used to call a preset defect standard feature library, perform similarity matching calculation between the fused feature set of the blade surface and the defect standard feature library, determine the defect status of the blade, and output the detection result.

[0122] It should be noted that the foregoing explanation of the method embodiments also applies to the apparatus of this embodiment, and will not be repeated here.

[0123] To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method described in the foregoing method embodiments.

[0124] To implement the above embodiments, this application also proposes a computer program product having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method described in the foregoing method embodiments.

[0125] To implement the above embodiments, this application also proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements the method described in the foregoing method embodiments.

[0126] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. For example, the electronic device 800 may be a mobile phone, computer, digital broadcasting terminal, messaging device, game console, tablet device, medical device, fitness equipment, personal digital assistant, etc.

[0127] Reference Figure 3 The electronic device 800 may include one or more of the following components: processing component 802, memory 804, power component 806, multimedia component 808, audio component 810, input / output (I / O) interface 812, sensor component 814, and communication component 816.

[0128] Processing component 802 typically controls the overall operation of electronic device 800, such as operations associated with display, telephone calls, data communication, camera operation, and recording operations. Processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the methods described above. Furthermore, processing component 802 may include one or more modules to facilitate interaction between processing component 802 and other components. For example, processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802.

[0129] Memory 804 is configured to store various types of data to support the operation of electronic device 800. Examples of such data include instructions for any application or method operating on electronic device 800, contact data, phonebook data, messages, pictures, videos, etc. Memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0130] Power component 806 provides power to various components of electronic device 800. Power component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 800.

[0131] Multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors may sense not only the boundaries of the touch or swipe action but also the duration and pressure associated with the touch or swipe operation. In some embodiments, multimedia component 808 includes a front-facing camera and / or a rear-facing camera. When the electronic device 800 is in an operating mode, such as a shooting mode or a video mode, the front-facing camera and / or the rear-facing camera may receive external multimedia data. Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

[0132] Audio component 810 is configured to output and / or input audio signals. For example, audio component 810 includes a microphone (MIC) configured to receive external audio signals when electronic device 800 is in an operating mode, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 804 or transmitted via communication component 816. In some embodiments, audio component 810 also includes a speaker for outputting audio signals.

[0133] I / O interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to, home buttons, volume buttons, power buttons, and lock buttons.

[0134] Sensor assembly 814 includes one or more sensors for providing state assessments of various aspects of electronic device 800. For example, sensor assembly 814 can detect the on / off state of electronic device 800, the relative positioning of components such as the display and keypad of electronic device 800, changes in position of electronic device 800 or a component of electronic device 800, the presence or absence of user contact with electronic device 800, orientation or acceleration / deceleration of electronic device 800, and temperature changes of electronic device 800. Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. Sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, sensor assembly 814 may also include an accelerometer, gyroscope, magnetometer, pressure sensor, or temperature sensor.

[0135] Communication component 816 is configured to facilitate wired or wireless communication between electronic device 800 and other devices. Electronic device 800 can access wireless networks based on communication standards, such as Wi-Fi, 4G, or 5G, or combinations thereof. In one exemplary embodiment, communication component 816 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, communication component 816 also includes a near-field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, Infrared Data Association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0136] In an exemplary embodiment, the electronic device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the methods described above.

[0137] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 804 including instructions, which can be executed by a processor 820 of an electronic device 800 to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0138] To implement the above embodiments, this application also proposes a chip, including: the chip includes a processing circuit configured to perform the methods provided in the foregoing embodiments.

[0139] Figure 4 This is a schematic diagram of the structure of a chip according to an embodiment of this application. See also... Figure 4 The diagram shown is a schematic representation of the structure of chip 1100, but it is not limited to this.

[0140] Chip 1100 includes processing circuitry 1101, which is configured to perform any of the above methods.

[0141] In some embodiments, chip 1100 further includes one or more interface circuits 1102. Optionally, the interface circuit 1102 is connected to memory 1103, and the interface circuit 1102 can be used to receive signals from memory 1103 or other devices, and the interface circuit 1102 can be used to send signals to memory 1103 or other devices. For example, the interface circuit 1102 can read instructions stored in memory 1103 and send the instructions to processing circuit 1101.

[0142] In some embodiments, the interface circuit 1102 performs at least one of the communication steps such as sending and / or receiving in the above method, while the processing circuit 1101 performs other steps.

[0143] In some embodiments, the terms interface circuit, interface, transceiver pin, transceiver, etc., can be used interchangeably.

[0144] In some embodiments, chip 1100 further includes one or more memories 1103 for storing instructions. Optionally, all or part of the memories 1103 may be located outside of chip 1100.

[0145] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0146] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0147] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0148] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0149] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0150] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0151] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0152] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for detecting defects in wind turbine blades, characterized in that, include: Step S1: Receive multi-source core parameters synchronously transmitted by the multi-source sensor module. The multi-source core parameters include blade rotation parameters, environmental parameters, illumination parameters, and imaging device parameters. Step S2: Generate an orthogonal polarization coded light field based on the blade rotation parameters and environmental parameters; send a control command to the projection module to project the orthogonal polarization coded light field onto the blade surface; send an acquisition command to the imaging device to obtain the coded polarization image stream; perform orthogonal decoding processing on the coded polarization image stream to separate and obtain the polarization image of the target blade. Step S3: Based on the attitude data in the illumination parameters and blade rotation parameters, a dynamic weight allocation mechanism is constructed to perform feature enhancement processing on the polarization image of the target blade and generate an enhanced feature image. Step S4: After preprocessing the enhanced feature image, extract phase feature parameters and texture morphology feature parameters, perform weighted fusion processing on the two types of feature parameters, and generate a blade surface fusion feature set. Step S5: Call the preset defect standard feature library, perform similarity matching operation between the fused feature set of the blade surface and the defect standard feature library, determine the defect status of the blade and output the detection result.

2. The method according to claim 1, characterized in that, The blade rotation parameters include rotation period, rotation speed, rotation fundamental frequency, real-time rotation angle, and three-dimensional attitude data; The environmental parameters include dust concentration, humidity, and salt spray content; The illumination parameters include the solar altitude angle and the solar azimuth angle; The imaging device parameters include the refractive index of the lens of the polarization imaging device, the initial angle of the polarizer, and the imaging resolution parameters; wherein, all received parameters are bound to the same timestamp.

3. The method according to claim 2, characterized in that, Step S2 includes: Based on the rotating fundamental frequency, the modulation period of the orthogonal polarization coding is set; Combining the modulation period, environmental parameters, and initial angle of the polarizer, an orthogonal polarization coding rule is determined. Based on the orthogonal polarization coding rule, the coding polarization direction of the orthogonal polarization coded light field is optimized to generate an orthogonal polarization coded light field. Send control commands to the projection module to project the orthogonal polarization coded light field onto the blade surface; Based on the projection result and imaging resolution of the orthogonal polarization coded light field, a collection command is sent to the imaging device to control the imaging device to collect the coded polarization image stream at the corresponding resolution, and at the same time trigger the imaging device to record the real-time rotation angle of the blade corresponding to each frame of the image. Based on the orthogonal polarization coding rules, with the real-time rotation angle of the blade as the attitude calibration basis and the lens refractive index in the imaging equipment parameters as the optical correction parameter, the coded polarization image stream is subjected to frame-by-frame orthogonal decoding processing to obtain the orthogonal decoding result. Based on the orthogonal decoding results, a target blade polarization image containing only blade surface feature information is obtained.

4. The method according to claim 3, characterized in that, The method further includes: Local region segmentation is performed on the polarization image of the target blade to obtain multiple independent local regions; Calculate the average pixel gray value of each local region, and the deviation between the average pixel gray value of each local region and the average pixel gray value of the global pixel gray value of the polarization image of the target leaf; Based on the deviation value, the surface reflection coefficient of the corresponding local area is corrected in reverse, wherein the corrected surface reflection coefficient increases non-linearly with the increase of the deviation value; The corrected surface reflectance coefficients of each local region are substituted into the effective illumination intensity calculation process to calculate the effective illumination intensity of each local region.

5. The method according to claim 4, characterized in that, Step S3 includes: Extract the three-dimensional attitude data and illumination parameters from the blade rotation parameters, and calculate the angle between the blade surface normal vector and the solar incident vector; Based on the included angle and the effective light intensity of each local area, the effective light intensity of the windward and leeward sides of the blade and the difference in light intensity between the two are calculated to obtain the spatial distribution data of the effective light intensity. Extract the rotational speed parameter from the blade rotation parameters, and calculate the comprehensive weight allocation factor based on the set illumination weight coefficient and rotational speed weight coefficient; Based on the difference in light intensity and the comprehensive weight allocation factor, the enhancement weights of the windward and leeward sides are adjusted, and pixel intensity adjustment is performed on the polarization image of the target blade to generate an enhanced feature image.

6. The method according to claim 5, characterized in that, The step of adjusting the enhancement weights of the windward and leeward sides based on the light intensity difference and the comprehensive weight allocation factor, performing pixel intensity adjustment on the polarization image of the target blade, and generating an enhanced feature image includes: A gradual adjustment zone is constructed based on the multiple independent local regions along the boundary line between the windward and leeward sides of the blade. Based on the spatial distribution data of the gradient adjustment zone and the effective light intensity, the continuous variation characteristics of the effective light intensity are extracted, and a gradient weight sequence is generated based on the continuous variation characteristics. The gradient weight sequence is used to smoothly adjust the pixel intensity of the target blade polarization image, thereby generating an enhanced feature image.

7. The method according to claim 6, characterized in that, Step S4 includes: Preprocessing is performed on the enhanced feature image, including Gaussian filtering for noise reduction and histogram equalization. Phase reconstruction is performed on the preprocessed enhanced feature image to extract phase feature parameters, wherein the phase feature parameters include the coordinates of phase abrupt change points, phase fluctuation amplitude, and phase uniformity index. Edge detection and texture analysis are performed on the preprocessed enhanced feature image to extract texture morphology feature parameters, wherein the texture morphology feature parameters include the edge contour, texture roughness and area size of the defect region; Based on the dust concentration and salt spray content in the environmental parameters, the weight ratio of phase feature parameters and texture morphology feature parameters is dynamically allocated, and weighted fusion processing is performed on the dynamically allocated phase feature parameters and texture morphology feature parameters to generate a fused feature set of the blade surface.

8. The method according to claim 7, characterized in that, Step S5 includes: The system invokes a preset defect standard feature library, which contains fused feature templates for crack, corrosion, wear, and dent types. Based on the fusion feature set of the blade surface and the fusion feature template in the defect standard feature library, a similarity matching operation is performed to obtain the similarity matching operation result; Based on the similarity matching calculation results and the preset feature difference threshold, it is determined whether the blade has defects and the type of defects, and the defect determination result is obtained. Based on the defect determination results, and combined with the rotation period and real-time rotation angle in the blade rotation parameters, the defect location coordinates are calibrated, and the detection results including the defect location coordinates, size, and severity level are output.

9. A device for detecting defects in wind turbine blades, characterized in that, include: The data acquisition module is used to receive multi-source core parameters synchronously transmitted by the multi-source sensor module. The multi-source core parameters include blade rotation parameters, environmental parameters, illumination parameters, and imaging device parameters. The separation module is used to generate an orthogonal polarization coded light field based on the blade rotation parameters and environmental parameters, send control commands to the projection module to project the orthogonal polarization coded light field onto the blade surface, send acquisition commands to the imaging device to obtain the coded polarization image stream, perform orthogonal decoding processing on the coded polarization image stream, and separate the polarization image of the target blade. The enhancement module is used to construct a dynamic weight allocation mechanism based on the attitude data in the illumination parameters and blade rotation parameters, and to perform feature enhancement processing on the polarization image of the target blade to generate an enhanced feature image. The feature extraction module is used to preprocess the enhanced feature image, extract phase feature parameters and texture morphology feature parameters, perform weighted fusion processing on the two types of feature parameters, and generate a blade surface fusion feature set. The defect detection module is used to call a preset defect standard feature library, perform similarity matching calculation between the fused feature set of the blade surface and the defect standard feature library, determine the defect status of the blade, and output the detection result.

10. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements the method as described in any one of claims 1-8.

Images