Defect detection method and system for power equipment
By acquiring infrared temperature images and visible light images, phase consistent edge feature detection and SIFT registration, the problem of difficulty in identifying power equipment during drone inspections is solved, and thermal analysis of power equipment with high accuracy and low complexity is achieved.
Patent Information
- Application Number
- CN202411041578.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-12-03
AI Technical Summary
During drone inspection, background interference of infrared images of power equipment makes it difficult to identify the equipment. The existing dual-light registration method has insufficient processing of size and viewing angle differences, and the calculation complexity is high, making it difficult to deploy at the edge of the drone .
By acquiring infrared temperature images and visible light images, performing phase-consistent edge feature detection, feature points are extracted using SIFT algorithm, and registering based on feature points to reduce background interference and improve the thermal analysis accuracy of power equipment.
It realizes accurate identification and defect detection of power equipment, reduces the complexity of the algorithm, is suitable for drone inspection environments, and improves the accuracy and environmental adaptability of infrared thermal analysis.
Smart Images

Figure CN119067918A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of operation and maintenance of overhead power transmission and distribution lines, and more specifically, to a defect detection method and system for power equipment. Background Art
[0002] As the number of power transmission lines continues to grow, traditional manual inspections can no longer meet the needs of operation and maintenance. Since 2013, State Grid and China Southern Grid have carried out intelligent inspections of power transmission lines using drones. Currently, drones have been used on a large scale and in a standardized manner. More than 80 million images are inspected each year, and more than 750,000 defects have been found. The types of defects found cover all types of towers, insulators, hardware, ground wires, foundations, channel environments, grounding devices, and auxiliary settings. Using deep learning to detect targets and efficiently and accurately identify defects in massive images is a must for saving manpower and material resources and improving inspection efficiency.
[0003] Infrared temperature measurement technology for power equipment is a vital monitoring tool that works based on the infrared radiation energy naturally emitted by objects. The intensity of this radiation energy is closely related to factors such as the temperature, material properties, and surface color of the power equipment. Infrared imaging technology cleverly uses infrared cameras and precision optical imaging lenses to capture the infrared radiation energy distribution map emitted by the tested power equipment in a non-contact manner, and converts it into a visual infrared thermal image, so that the temperature distribution data on the surface of the power equipment can be obtained quickly and accurately, thereby judging whether its working status is normal. Once abnormal hot spots are found on the surface of the equipment, it means that there may be potential faults or risks. Therefore, infrared temperature measurement technology plays an irreplaceable role in the early fault diagnosis, fire risk prevention, regular maintenance and overhaul of power equipment.
[0004] At present, there are two main methods for infrared fault diagnosis of power equipment in power transmission lines: 1. Directly identify power equipment faults in infrared images; 2. Identify power equipment through visible light images, align visible light and infrared images, locate power equipment in infrared images and perform temperature detection. In the first method, in the inspection scenario, drones often take top-down photos and infrared temperature measurements of power equipment. The surrounding background will interfere with the accuracy of equipment temperature measurement and identification, making it difficult to identify power equipment in infrared images. In comparison, the second method combines the different features of visible light images and infrared images to greatly improve the feature resolution of target power equipment, and locate power equipment defects in infrared images more accurately. Therefore, it has an increasingly broad application prospect in transmission line drone inspections.
[0005] In the second category of methods, visible light and infrared image registration (dual-light registration) is the key factor in infrared temperature measurement of power equipment in transmission lines.
[0006] At present, the common dual-light registration methods for power equipment include:
[0007] Feature matching is performed through image area features. This method is simple, intuitive and easy to implement, but it lacks the ability to deal with image deviations caused by size differences and viewing angle differences between visible light images and infrared images. It is mainly used for the identification of insulator strings, and lacks the ability to identify locations such as wire clamps and cable lead-in connectors. The algorithm is not expandable enough.
[0008] Generate dual-light fusion images for registration through GAN and other generation methods. This method has good registration and fusion effects on power equipment and strong scalability, but requires a large amount of hardware equipment calculation, which is not conducive to subsequent edge deployment on drones. Summary of the invention
[0009] In view of the above problems, the present invention proposes a defect detection method for electric power equipment, comprising:
[0010] For a target electric power device, an infrared temperature image and a visible light image of the target electric power device are acquired, and the saliency of the infrared temperature image is enhanced to obtain a target infrared temperature image;
[0011] Performing phase-consistent edge feature detection on the target infrared temperature image and the visible light image to obtain a detection result, and extracting feature points in the detection result using a SIFT detection algorithm;
[0012] Based on the feature points, the target infrared temperature image and the visible light image are registered, and based on the registration result, defects of the target electric power equipment are detected.
[0013] Optionally, the saliency of the infrared temperature image is enhanced, including:
[0014] The infrared temperature image is converted from the original color space to the CIELAB color space, and filtered by an adaptive median filter. The characteristics of color and intensity in the CIELAB color space are used to calculate the average value of the color vector, and the saliency value of each pixel of the infrared temperature image is determined based on the average value. The saliency of the infrared temperature image is enhanced by the saliency value.
[0015] Optionally, enhancing the saliency of the infrared temperature image further includes: normalizing the saliency values of pixels on each channel of the infrared temperature image and assigning a weight to each channel.
[0016] Optionally, the significance value is calculated as follows:
[0017]
[0018] Among them, I m (x,y) is the adaptive median vector, I u (x, y) is the average CIELAB vector, x0 and y0 are the offsets between the center surround filter and the pixel point, S i is the significance value of the original channel, S S is the significance value after the fusion of the three channels of LAB, ω i is the weight of the corresponding channel, x and y are the horizontal and vertical coordinates of the infrared image pixel, S(x,y) is the calculation method of the saliency value of the (x,y) region, S min is the minimum value in the region S(x,y), S max is the maximum value in the S(x,y) region, A is the significant value area calculation, and i is the channel count value of the image.
[0019] Optionally, phase-consistent edge feature detection is performed based on a two-dimensional Gabor filtering algorithm.
[0020] Optionally, using a SIFT detection algorithm to extract feature points in the detection result includes:
[0021] The image is Gaussian blurred to form a scale space, local extreme points are found in the scale space, the positions and scales of the local extreme points are determined, and one or more directions are assigned to the local extreme points to obtain feature points.
[0022] In another aspect, the present invention further provides a defect detection system for electric power equipment, comprising:
[0023] An image acquisition unit is used to acquire an infrared temperature image and a visible light image of a target power device, and enhance the significance of the infrared temperature image to obtain a target infrared temperature image;
[0024] A feature extraction unit is used to perform phase-consistent edge feature detection on the target infrared temperature image and the visible light image to obtain a detection result, and extract feature points in the detection result using a SIFT detection algorithm;
[0025] The detection unit is used to align the target infrared temperature image and the visible light image based on the feature points, and detect defects of the target power equipment based on the alignment result.
[0026] Optionally, the saliency of the infrared temperature image is enhanced, including:
[0027] The infrared temperature image is converted from the original color space to the CIELAB color space, and filtered by an adaptive median filter. The characteristics of color and intensity in the CIELAB color space are used to calculate the average value of the color vector, and the saliency value of each pixel of the infrared temperature image is determined based on the average value. The saliency of the infrared temperature image is enhanced by the saliency value.
[0028] Optionally, the method further includes: normalizing the significant values of pixels on each channel of the infrared temperature image and assigning a weight to each channel.
[0029] Optionally, the significance value is calculated as follows:
[0030]
[0031] Among them, I m (x,y) is the adaptive median vector, I u (x, y) is the average CIELAB vector, x0 and y0 are the offsets between the center surround filter and the pixel point, S i is the significance value of the original channel, S S is the significance value after the fusion of the three channels of LAB, ω i is the weight of the corresponding channel, x and y are the horizontal and vertical coordinates of the infrared image pixel, S(x,y) is the calculation method of the saliency value of the (x,y) region, S min is the minimum value in the region S(x,y), S max is the maximum value in the S(x,y) region, A is the significant value area calculation, and i is the channel count value of the image.
[0032] Optionally, phase-consistent edge feature detection is performed based on a two-dimensional Gabor filtering algorithm.
[0033] Optionally, using a SIFT detection algorithm to extract feature points in the detection result includes:
[0034] The image is Gaussian blurred to form a scale space, local extreme points are found in the scale space, the positions and scales of the local extreme points are determined, and one or more directions are assigned to the local extreme points to obtain feature points.
[0035] In yet another aspect, the present invention further provides a computing device, comprising: one or more processors;
[0036] a processor for executing one or more programs;
[0037] When the one or more programs are executed by the one or more processors, the above-described method is implemented.
[0038] In yet another aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed, the method described above is implemented.
[0039] Compared with the prior art, the present invention has the following beneficial effects:
[0040] The present invention provides a defect detection method for electric power equipment, comprising: for a target electric power equipment, obtaining an infrared temperature image and a visible light image of the target electric power equipment, and enhancing the significance of the infrared temperature image to obtain a target infrared temperature image; for the target infrared temperature image and the visible light image, performing phase-consistent edge feature detection to obtain a detection result, and using a SIFT detection algorithm to extract feature points in the detection result; based on the feature points, registering the target infrared temperature image and the visible light image, and based on the registration result, detecting defects in the target electric power equipment. The present invention makes full use of the advantages of rich information of visible light images and the characteristics of infrared images that can be used to carry out thermal analysis of equipment, reduces the interference of complex backgrounds on the thermal analysis effect of electric power equipment, has good environmental adaptability, improves the accuracy of the infrared thermal analysis algorithm, and reduces the complexity of the algorithm, providing an accurate diagnostic means for thermal analysis of electric power equipment. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 It is a schematic diagram of the process of the present invention;
[0042] Figure 2 It is the infrared image Gabor filtering enhancement image of the method of the present invention;
[0043] Figure 3 This is a comparison diagram of the edge structure filtering of visible light and infrared images by the method of the present invention;
[0044] Figure 4 This is a schematic diagram of the visible light and infrared dual light registration method of the present invention;
[0045] Figure 5 It is a structural diagram of the system of the present invention. DETAILED DESCRIPTION
[0046] Now, exemplary embodiments of the present invention are described with reference to the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely and to fully convey the scope of the present invention to those skilled in the art. The terms used in the exemplary embodiments shown in the accompanying drawings are not intended to limit the present invention. In the accompanying drawings, the same units / elements are marked with the same reference numerals.
[0047] Unless otherwise specified, the terms (including technical terms) used herein have the commonly understood meanings to those skilled in the art. In addition, it is understood that the terms defined in commonly used dictionaries should be understood to have the same meanings as those in the context of the relevant fields, and should not be understood as idealized or overly formal meanings.
[0048] Embodiment 1:
[0049] The present invention proposes a defect detection method for electric power equipment, such as Figure 1 As shown, including:
[0050] For a target electric power device, an infrared temperature image and a visible light image of the target electric power device are acquired, and the saliency of the infrared temperature image is enhanced to obtain a target infrared temperature image;
[0051] Performing phase-consistent edge feature detection on the target infrared temperature image and the visible light image to obtain a detection result, and extracting feature points in the detection result using a SIFT detection algorithm;
[0052] Based on the feature points, the target infrared temperature image and the visible light image are registered, and based on the registration result, defects of the target electric power equipment are detected.
[0053] Among them, the saliency of the infrared temperature image is enhanced, including:
[0054] The infrared temperature image is converted from the original color space to the CIELAB color space, and filtered by an adaptive median filter. The characteristics of color and intensity in the CIELAB color space are used to calculate the average value of the color vector, and the saliency value of each pixel of the infrared temperature image is determined based on the average value. The saliency of the infrared temperature image is enhanced by the saliency value.
[0055] The enhancement of the saliency of the infrared temperature image also includes: normalizing the saliency value of the pixel on each channel of the infrared temperature image and attaching a weight to each channel.
[0056] The calculation formula of the significant value is as follows:
[0057]
[0058]
[0059] Among them, I m (x,y) is the adaptive median vector, I u (x, y) is the average CIELAB vector, x0 and y0 are the offsets between the center surround filter and the pixel point, Si is the significance value of the original channel, S S is the significance value after the fusion of the three channels of LAB, ω i is the weight of the corresponding channel, x and y are the horizontal and vertical coordinates of the infrared image pixel, S(x,y) is the calculation method of the saliency value of the (x,y) region, S min is the minimum value in the region S(x,y), S max is the maximum value in the S(x,y) region, A is the significant value area calculation, and i is the channel count value of the image.
[0060] Among them, phase-consistent edge feature detection is performed based on a two-dimensional Gabor filtering algorithm.
[0061] The feature points in the detection result are extracted using the SIFT detection algorithm, including:
[0062] The image is Gaussian blurred to form a scale space, local extreme points are found in the scale space, the positions and scales of the local extreme points are determined, and one or more directions are assigned to the local extreme points to obtain feature points.
[0063] The above-mentioned significance enhancement method is as follows:
[0064] The algorithm first converts the image from the original color space to the CIELAB color space, and then processes it with an adaptive median filter to preserve most of the low-frequency content (corresponding to the overall structure of the image) and key high-frequency content (corresponding to edges and details). Subsequently, the algorithm uses the color and intensity characteristics in the CIELAB space to calculate the average value of the LAB color vector to obtain the saliency value of each pixel.
[0065] Since the brightness and color features of CIELAB vectors differ in value, direct saliency calculation may lead to an imbalance in the weights of different channels. Therefore, the improved algorithm in this paper normalizes the saliency values of each channel to ensure that they are on the same order of magnitude. At the same time, the algorithm assigns appropriate weights to each channel according to the speed of change of the channel characteristics to ensure that the saliency values of each channel can be fairly and effectively considered when calculating the Euclidean distance of pixels, avoiding the problem of suppressing smaller value components during the fusion process due to differences in order of magnitude. The algorithm is shown in Equation 1 and Equation 2:
[0066]
[0067]
[0068] Where: I m (x,y) is the adaptive median vector; I u(x,y) is the average CIELAB vector; x0 and y0 are the offsets between the center surround filter and the pixel; S i is the significance value of the original channel; S S is the significance value after the three channels of LAB are fused; ω i is the weight of the corresponding channel. By mapping the image pixel features to the CIELAB space and performing adaptive median filtering, the algorithm effectively utilizes the information of the three LAB channels, improves the execution efficiency of the algorithm, improves the edge details of the image, and makes the edge features of the segmented target and the background more obvious, and finally obtains the preprocessing results of the infrared image, such as Figure 2 shown.
[0069] The algorithm can now make full use of the information of the three channels L (brightness), A (red-green), and B (yellow-blue) in the CIELAB color space. This improvement not only improves the execution efficiency of the algorithm, but also significantly enhances the edge details of power equipment in infrared images while maintaining the overall structure of the image. Through fine processing, the algorithm can highlight the edge structure between the target and the background in the image, making these edges visually clearer and more obvious. Ultimately, the algorithm successfully provides high-quality preprocessing results for infrared images, laying a solid foundation for subsequent image analysis and processing.
[0070] The significant edge details of power equipment in infrared images can be enhanced through CIELAB space mapping and median filtering, but there are still inherent differences between infrared images and visible light images. In order to bridge this difference and further extract the edge structure features of power equipment in the two images, the present invention proposes an edge structure feature detection method based on local energy of phase consistency. It can highlight the key features of power equipment in the image, such as corners, edges, etc. The core idea of phase consistency is that these key features in the image will show a high degree of consistency in the phase of the frequency components after Fourier transform. This consistency is not affected by the image scale or lighting conditions, so the phase consistency feature has extremely high stability and wide applicability.
[0071] The calculation method of phase consistency mainly involves Fourier transform and local energy analysis. Specifically, the image is first converted from the spatial domain to the frequency domain through Fourier transform, and then the phase consistency of each frequency component in the local area is calculated. This process involves a detailed analysis of the amplitude and phase of the frequency components to determine which frequency components are consistent in phase, thereby identifying the key features in the image;
[0072] By applying the phase consistency method, the researchers were able to more effectively extract the edge structure features of power equipment in infrared images and visible light images and reduce the differences between them. This will help improve the applicability and effectiveness of the algorithm in different scenarios and lay a solid foundation for subsequent image fusion, target detection and other tasks.
[0073] In order to obtain the phase consistency metric of a two-dimensional image, it is necessary to use two-dimensional Gabor filtering to calculate local energy in multiple directions. Gabor filtering is a short-time windowed Fourier transform (Fourier transform in a specific time window), which is a special case when the window function in the short-time Fourier transform is a Gaussian function. Therefore, the Gabor filter can extract relevant features at different scales and directions in the frequency domain. In addition, the Gabor function is similar to the function of the human eye, so it is often used for texture recognition;
[0074] The two-dimensional Gabor filter function is defined as:
[0075]
[0076] In formula 3:
[0077] Wavelength (λ): represents the wavelength parameter of the cosine function in the Gabor kernel function. Its value is specified in pixels, usually greater than or equal to 2, but not greater than 1 / 5 of the input image size;
[0078] Direction (θ): Indicates the direction of the parallel stripes in the Gabor filter kernel. Valid values are real numbers from 0° to 360°.
[0079] Phase offset (φ): represents the phase parameter of the cosine function in the Gabor kernel function. Its value range is -180° to 180°. The equations corresponding to 0° and 180° are symmetric with the origin, and the equations corresponding to -90° and 90° are centrally symmetric with respect to the origin.
[0080] Aspect ratio (γ): The spatial aspect ratio determines the ellipticity of the Gabor function shape. When γ = 1, the shape is circular; when γ < 1, the shape is elongated along the parallel stripe direction, and this value is usually 0.5.
[0081] σ represents the standard deviation of the Gaussian factor of the Gabor function, and x and y represent the gradient values of the image in the horizontal and vertical dimensions respectively.
[0082] Since the two-dimensional Gabor belongs to the frequency domain filter, the inverse Fourier transform can be used to obtain its complex expression in the spatial domain. The real part is the even-symmetric Gabor wavelet, and the imaginary part is the odd-symmetric Log-Gabor wavelet. The odd-symmetric wavelet and the even-symmetric wavelet are used for convolution filtering in different directions and scales, respectively, and the amplitude and phase components of the image in different directions and scales are obtained as follows:
[0083]
[0084] In formula 4: and They are odd symmetric wavelets and even symmetric wavelets respectively. In order to obtain the edge structure features of the image, it is necessary to calculate the phase consistency values of the image pixels in different directions respectively, and obtain the phase consistency moment, so that the numerator is the sum of the weighted and noise-compensated local energy in all directions and scales, and the denominator is the sum of the amplitude components in all directions and scales. The calculation method of phase consistency is:
[0085]
[0086] In formula 6: W o is the filtering direction weight; T is the noise threshold; ε is the minimum constant; is a non-positive suppression function.
[0087] The phase consistency moment is calculated in each direction according to the phase consistency measurement value PC(X, Y) obtained by formula 6 using the moment analysis method. Through analysis, it is found that the axis corresponding to the minimum moment is the main axis representing the feature direction, and the axis of the maximum moment is perpendicular to the main axis, indicating the importance of the feature. When the value of the minimum moment at a certain point is also large, this point is the corner point of the image. Through the phase consistency method, a larger range of power equipment features in visible light and infrared images can be registered, and the local smooth illumination change features in the infrared image can be effectively kept unchanged. In this way, a stable edge structure feature map of infrared and visible light images is obtained, as shown in Figure 3 shown.
[0088] After the edge structure feature maps of the power equipment in the infrared image and the visible light image are obtained by the phase consistency algorithm, similar feature points can be extracted on the edge structure for matching. SIFT (Scale-Invariant Feature Transform) is an algorithm used for image processing, which aims to detect and describe local features in images. It performs image registration by finding extreme points in the image and extracting their position, scale, and rotation invariants.
[0089] The core steps include:
[0090] Extreme value detection in scale space: The scale space is formed by Gaussian blurring the image, and then local extreme points are found in this space. These points are called key points or feature points.
[0091] Keypoint localization: At each candidate location, a fine-tuned model is used to determine the location and scale of keypoints. Keypoints are selected based on their stability.
[0092] Direction determination: Each keypoint is assigned one or more directions based on the local gradient direction of the image. All subsequent image data operations are transformed relative to the direction, scale, and position of the keypoint, thereby providing invariance to these transformations.
[0093] Keypoint Description: In the neighborhood around each keypoint, the local gradients of the image are measured at a selected scale. These gradients are converted into a representation that allows for large local shape deformations and lighting changes.
[0094] The SIFT algorithm is characterized by its invariance to rotation, scaling, and brightness changes, and has a high tolerance for light, noise, and slight changes in viewing angle. This allows the SIFT feature to have a high detection rate even when some objects are obscured. By matching the features in the edge structure feature map of power equipment in visible light and infrared images with the SIFT algorithm, it is possible to quickly and accurately achieve the registration of power equipment in visible light and infrared images of transmission line drone inspections, such as Figure 4 shown.
[0095] The present invention makes full use of the advantages of rich information of visible light images and the characteristics of infrared images that can be used to carry out thermal analysis of equipment, reduces the interference of complex background on the thermal analysis effect of power equipment, and has good environmental adaptability.
[0096] The present invention realizes the registration of visible light image and infrared image of power equipment through infrared image saliency enhancement, Gabor filter edge structure feature extraction and SIFT edge feature matching algorithm, and further realizes infrared defect detection through visible light power equipment positioning and infrared temperature information.
[0097] The present invention proposes a registration of infrared and visible light images for transmission line drone inspection based on edge structure features of power equipment. The significance enhancement of infrared images includes but is not limited to feature enhancement algorithms such as adaptive median filtering algorithm and Gaussian filtering algorithm. The edge structure feature extraction method includes but is not limited to edge detection algorithms such as gabor filtering and log-gabor filtering. The edge feature matching algorithm includes but is not limited to SIFT feature matching, ORB feature matching and other algorithms. It realizes infrared temperature measurement and recognition of power equipment including but not limited to transformers, tension clamps, ice ditch clamps, pole-mounted disconnectors, insulators, lightning arresters, etc., realizes the recognition of multiple types of electrical equipment, has good scalability of temperature measurement and recognition types, and can be expanded to carry out visible light defect recognition, etc.
[0098] Embodiment 2:
[0099] The present invention also provides a defect detection system 200 for power equipment, such as Figure 5 As shown, including:
[0100] The image acquisition unit 201 is used to acquire an infrared temperature image and a visible light image of a target power device, and enhance the significance of the infrared temperature image to obtain a target infrared temperature image;
[0101] A feature extraction unit 202 is used to perform phase-consistent edge feature detection on the target infrared temperature image and the visible light image to obtain a detection result, and extract feature points in the detection result using a SIFT detection algorithm;
[0102] The detection unit 203 is used to align the target infrared temperature image and the visible light image based on the feature points, and detect defects of the target power equipment based on the alignment result.
[0103] Among them, the saliency of the infrared temperature image is enhanced, including:
[0104] The infrared temperature image is converted from the original color space to the CIELAB color space, and filtered by an adaptive median filter. The characteristics of color and intensity in the CIELAB color space are used to calculate the average value of the color vector, and the saliency value of each pixel of the infrared temperature image is determined based on the average value. The saliency of the infrared temperature image is enhanced by the saliency value.
[0105] The method further includes: normalizing the significant value of pixels on each channel of the infrared temperature image and attaching a weight to each channel.
[0106] The calculation formula of the significant value is as follows:
[0107]
[0108]
[0109] Among them, I m (x,y) is the adaptive median vector, I u (x, y) is the average CIELAB vector, x0 and y0 are the offsets between the center surround filter and the pixel point, S i is the significance value of the original channel, S S is the significance value after the fusion of the three channels of LAB, ω i is the weight of the corresponding channel, x and y are the horizontal and vertical coordinates of the infrared image pixel, S(x,y) is the calculation method of the saliency value of the (x,y) region, S min is the minimum value in the region S(x,y), S max is the maximum value in the S(x,y) region, A is the significant value area calculation, and i is the channel count value of the image.
[0110] Among them, phase-consistent edge feature detection is performed based on a two-dimensional Gabor filtering algorithm.
[0111] The feature points in the detection result are extracted using the SIFT detection algorithm, including:
[0112] The image is Gaussian blurred to form a scale space, local extreme points are found in the scale space, the positions and scales of the local extreme points are determined, and one or more directions are assigned to the local extreme points to obtain feature points.
[0113] The present invention makes full use of the advantages of rich information of visible light images and the characteristics of infrared images for conducting thermal analysis of equipment, reduces the interference of complex background on the thermal analysis effect of power equipment, has good environmental adaptability, improves the accuracy of infrared thermal analysis algorithm, and reduces the complexity of the algorithm, providing an accurate diagnostic means for thermal analysis of power equipment.
[0114] Embodiment 3:
[0115] Based on the same inventive concept, the present invention also provides a computer device, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding functions, so as to implement the steps of the method in the above embodiment.
[0116] Embodiment 4:
[0117] Based on the same inventive concept, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It is understandable that the computer-readable storage medium here can include both a built-in storage medium in a computer device and an extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the steps of the method in the above embodiment.
[0118] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.
[0119] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0120] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0121] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0122] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.
[0123] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. A defect detection method for electric power equipment, It is characterized in that include: For a target electric power device, an infrared temperature image and a visible light image of the target electric power device are acquired, and the saliency of the infrared temperature image is enhanced to obtain a target infrared temperature image; Performing phase-consistent edge feature detection on the target infrared temperature image and the visible light image to obtain a detection result, and extracting feature points in the detection result using a SIFT detection algorithm; Based on the feature points, the target infrared temperature image and the visible light image are registered, and based on the registration result, defects of the target electric power equipment are detected.
2. The defect detection method according to claim 1, It is characterized in that Enhance the saliency of infrared temperature images, including: The infrared temperature image is converted from the original color space to the CIELAB color space, and filtered by an adaptive median filter. The characteristics of color and intensity in the CIELAB color space are used to calculate the average value of the color vector, and the saliency value of each pixel of the infrared temperature image is determined based on the average value. The saliency of the infrared temperature image is enhanced by the saliency value.
3. The defect detection method according to claim 2, It is characterized in that The method further comprises: normalizing the significant value of the pixel on each channel of the infrared temperature image, and attaching a weight to each channel.
4. The defect detection method according to claim 2, It is characterized in that The calculation formula of the significant value is as follows: Among them, I m (x,y) is the adaptive median vector, I u (x,y) is the average CIELAB vector, x 0 and 0 is the offset between the center surround filter and the pixel point, S i is the significance value of the original channel, S S is the significance value after the fusion of the three channels of LAB, ω i is the weight of the corresponding channel, x and y are the horizontal and vertical coordinates of the infrared image pixel, S(x,y) is the calculation method of the saliency value of the (x,y) region, S min is the minimum value in the region S(x,y), S max is the maximum value in the S(x,y) region, A is the significant value area calculation, and i is the channel count value of the image.
5. The defect detection method according to claim 1, It is characterized in that Phase consistent edge feature detection is performed based on two-dimensional Gabor filtering algorithm.
6. The defect detection method according to claim 1, It is characterized in that The extracting feature points in the detection result using the SIFT detection algorithm includes: The image is Gaussian blurred to form a scale space, local extreme points are found in the scale space, the positions and scales of the local extreme points are determined, and one or more directions are assigned to the local extreme points to obtain feature points.
7. A defect detection system for electric power equipment, It is characterized in that include: An image acquisition unit, for acquiring an infrared temperature image and a visible light image of a target power device, and enhancing the significance of the infrared temperature image to obtain a target infrared temperature image; A feature extraction unit is used to perform phase-consistent edge feature detection on the target infrared temperature image and the visible light image to obtain a detection result, and extract feature points in the detection result using a SIFT detection algorithm; The detection unit is used to align the target infrared temperature image and the visible light image based on the feature points, and detect defects of the target power equipment based on the alignment result.
8. The defect detection system according to claim 7, It is characterized in that Enhance the saliency of infrared temperature images, including: The infrared temperature image is converted from the original color space to the CIELAB color space, and filtered by an adaptive median filter. The characteristics of color and intensity in the CIELAB color space are used to calculate the average value of the color vector, and the saliency value of each pixel of the infrared temperature image is determined based on the average value. The saliency of the infrared temperature image is enhanced by the saliency value.
9. The defect detection system according to claim 8, It is characterized in that The enhancement of the saliency of the infrared temperature image also includes: normalizing the saliency value of the pixels on each channel of the infrared temperature image and attaching a weight to each channel.
10. The defect detection system according to claim 8, It is characterized in that The calculation formula of the significant value is as follows: Among them, I m (x,y) is the adaptive median vector, I u (x,y) is the average CIELAB vector, x 0 and 0 is the offset between the center surround filter and the pixel point, S i is the significance value of the original channel, S S is the significance value after the fusion of the three channels of LAB, ω i is the weight of the corresponding channel, x and y are the horizontal and vertical coordinates of the infrared image pixel, S(x,y) is the calculation method of the saliency value of the (x,y) region, S min is the minimum value in the region S(x,y), S max is the maximum value in the S(x,y) region, A is the significant value area calculation, and i is the channel count value of the image.
11. The defect detection system according to claim 7, It is characterized in that Phase consistent edge feature detection is performed based on two-dimensional Gabor filtering algorithm.
12. The defect detection system according to claim 7, It is characterized in that The extracting feature points in the detection result using the SIFT detection algorithm includes: The image is Gaussian blurred to form a scale space, local extreme points are found in the scale space, the positions and scales of the local extreme points are determined, and one or more directions are assigned to the local extreme points to obtain feature points.
13. A computer device, It is characterized in that include: one or more processors; a processor for executing one or more programs; When the one or more programs are executed by the one or more processors, the method according to any one of claims 1 to 6 is implemented.
14. A computer-readable storage medium, It is characterized in that A computer program is stored thereon, and when the computer program is executed, the method according to any one of claims 1 to 6 is implemented.
Citation Information
Patent Citations
Image edge detection method and device, computer equipment and readable storage medium
CN111681256A
Infrared and visible light image registration method and system and storage medium
CN112950685A
Transformer substation power equipment fault detection method based on multi-source fusion
CN115661044A
Power transformation equipment temperature abnormity early warning method and system based on infrared visible light
CN116703986A
Cited By
Power construction risk assessment method and system
CN120913044A
A method and system for risk assessment in power construction
CN120913044B
