A method for recognizing inspection images of an intelligent substation inspection robot

Abstract

The present invention provides a method for identifying inspection images of an intelligent inspection robot in a substation. The method includes: collecting visible light images and infrared light images of substation equipment in the current substation environment; matching to obtain a target image processing model based on the current humidity, current light intensity, and current visibility; inputting the visible light image and the infrared light image into the target image processing model to obtain a first target image and a second target image; performing deep feature extraction on the first target image and the second target image to obtain a first depth image and a second depth image, and fusing the first depth image and the second depth image to obtain a fused image; inputting the fused image into an image recognition model to obtain a recognition result, and determining the equipment status of the substation equipment based on the recognition result. The present invention improves the accuracy of image recognition in a complex and changeable substation environment, accurately identifies the equipment status of substation equipment, and ensures the safe and stable operation of the substation.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and particularly to a method for recognizing inspection images of intelligent inspection robots in substations. Background Art

[0002] With the continuous expansion of the scale of the power system and the increasing requirement for power supply reliability, the safe and stable operation of substations has become increasingly important. In order to timely and accurately detect potential fault hazards and abnormal conditions that may exist in substation equipment, intelligent inspection robots are widely used in the daily inspection work of substations.

[0003] During the inspection process, intelligent inspection robots will collect a large number of device images. Accurately identifying and analyzing these images is a key link for realizing efficient fault diagnosis. At present, the existing methods for recognizing inspection images of intelligent inspection robots in substations mainly include image recognition methods based on traditional machine learning, that is, for example, using artificial feature extraction combined with classifiers such as support vector machine (SVM), AdaBoost, etc. for recognition. Usually, features such as texture, shape, and color are designed manually first, and then the classifier is used to judge whether the device state in the image is normal. However, the features designed manually by this method often rely on expert experience, have poor adaptability to the complex and changeable substation environment, and it is difficult to comprehensively and accurately capture the features under various device abnormal conditions, resulting in low recognition accuracy. Especially when facing some subtle fault features or newly emerging fault types, misjudgment or missed judgment is likely to occur. Summary of the Invention

[0004] The present invention provides a method for recognizing inspection images of intelligent inspection robots in substations, so as to improve the accuracy of image recognition in a complex and changeable substation environment, accurately recognize the device state of substation equipment in the inspection images, and ensure the safe and stable operation of substations.

[0005] In a first aspect, the present invention provides a method for recognizing inspection images of intelligent inspection robots in substations, which is applied to intelligent inspection robots in substations. The recognition method includes:

[0006] Collect visible light images and infrared light images of substation equipment in the current substation environment; the current substation environment includes the current humidity, the current light intensity, and the current visibility;

[0007] Based on the current humidity, the current light intensity, and the current visibility, perform matching in a preset model library to obtain a target image processing model in the current substation environment;

[0008] Input the visible light image and the infrared light image into the target image processing model to obtain a first target image and a second target image respectively output by the target image processing model;

[0009] Perform depth feature extraction on the first target image and the second target image to obtain a first depth image and a second depth image respectively, and perform image fusion on the first depth image and the second depth image to obtain a fused image;

[0010] Input the fused image into a pre-trained image recognition model to obtain the recognition result input by the image recognition model, and determine the device state of the substation equipment based on the recognition result;

[0011] Wherein, the processing process of the target image processing model includes denoising processing, illumination uniformity processing, and defogging processing; the image recognition model is trained based on sample images and their corresponding state result labels.

[0012] In a second aspect, the present invention further provides an identification system for inspection images of a substation intelligent inspection robot, which is applied to the substation intelligent inspection robot and is used to implement the identification method for inspection images of the substation intelligent inspection robot as described in the first aspect; the identification system includes:

[0013] An acquisition module, configured to acquire visible light images and infrared light images of substation equipment in the current substation environment; the current substation environment includes the current humidity, the current illumination intensity, and the current visibility;

[0014] A matching module, configured to perform matching in a preset model library based on the current humidity, the current illumination intensity, and the current visibility to obtain a target image processing model in the current substation environment;

[0015] An image processing module, configured to input the visible light image and the infrared light image into the target image processing model to obtain a first target image and a second target image respectively output by the target image processing model;

[0016] An extraction and fusion module, configured to perform depth feature extraction on the first target image and the second target image to obtain a first depth image and a second depth image respectively, and perform image fusion on the first depth image and the second depth image to obtain a fused image;

[0017] A device state identification module, configured to input the fused image into a pre-trained image recognition model to obtain the recognition result input by the image recognition model, and determine the device state of the substation equipment based on the recognition result;

[0018] Wherein, the processing process of the target image processing model includes denoising processing, illumination uniformity processing, and defogging processing; the image recognition model is trained based on sample images and their corresponding state result labels.

[0019] In a third aspect, the present invention further provides an electronic device, including: a memory for storing a computer software program; and a processor for reading and executing the computer software program to implement the method for identifying inspection images of a substation intelligent inspection robot as described in any one of the above.

[0020] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium storing a computer software program, which when executed by a processor implements the method for identifying inspection images of a substation intelligent inspection robot as described in any one of the above.

[0021] In a fifth aspect, the present invention further provides a computer program product including a computer program, which when executed by a processor implements the method for identifying inspection images of a substation intelligent inspection robot as described in any one of the above.

[0022] The method for identifying inspection images of a substation intelligent inspection robot provided by the embodiments of the present invention matches an accurate image processing model according to humidity, light intensity, and visibility to perform denoising processing, light uniformity processing, and defogging processing on the collected images, so as to remove the influence of environmental factors in the substation environment on the image quality, improve the quality of the collected images, thereby improving the accuracy of image recognition in the complex and changeable substation environment. Further, deep feature extraction and image fusion are performed on the processed images, and the texture features and detail features of the images in different modes can be fused to obtain a more accurate fused image. Then, combined with the image recognition model to identify the fused image, the device state of the substation equipment in the inspection image can be accurately identified, ensuring the safe and stable operation of the substation. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a schematic flow chart of the method for identifying inspection images of a substation intelligent inspection robot provided by the embodiments of the present invention;

[0024] Figure 2 is a schematic structural diagram of the system for identifying inspection images of a substation intelligent inspection robot provided by the embodiments of the present invention;

[0025] Figure 3 is an embodiment diagram of the electronic device provided by the embodiments of the present invention;

[0026] Figure 4 is an embodiment diagram of the computer-readable storage medium provided by the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0028] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality of" means two or more, unless otherwise specifically defined.

[0029] In the description of the present invention, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described as "for example" in the present invention is not necessarily construed as being more preferred or advantageous than other embodiments. In order for any person skilled in the art to implement and use the present invention, the following description is given. In the following description, details are set forth for the purpose of explanation. It should be understood that those skilled in the art can recognize that the present invention can be implemented without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid unnecessary details from obscuring the description of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope that conforms to the principles and features disclosed in the present invention.

[0030] Optionally, refer to Figure 1 , Figure 1 is a schematic flowchart of the method for identifying inspection images of a substation intelligent inspection robot provided by the present invention. In the embodiments of the present invention, the execution subject of the method for identifying inspection images of a substation intelligent inspection robot is the substation intelligent inspection robot. Therefore, the method for identifying inspection images of a substation intelligent inspection robot includes:

[0031] Step 10, collect visible light images and infrared light images of substation equipment in the current substation environment.

[0032] Optionally, the substation intelligent inspection robot in the embodiments of the present invention can be deployed on the substation or at a preset position in the substation area to collect image information of substation equipment in the substation in real time.

[0033] Further, a visible light camera and an infrared thermal imaging camera are installed on the intelligent substation inspection robot, and the time of the visible light camera and the infrared thermal imaging camera is the same, ensuring that the image information collected by the visible light camera and the infrared thermal imaging camera is aligned in time. Therefore, after receiving the instruction of equipment status recognition, the intelligent substation inspection robot calls the visible light camera and the infrared thermal imaging camera to collect the visible light image and the infrared light image of the substation equipment in the current substation environment, where the current substation environment includes the current humidity, the current light intensity, and the current visibility.

[0034] Step 20: Match based on the current humidity, the current light intensity, and the current visibility in the preset model library to obtain the target image processing model in the current substation environment.

[0035] Optionally, a preset model library is stored in the intelligent substation inspection robot. The preset model library includes multiple pre-trained image processing models. The processing capabilities of each image processing model for images in different environments are different, that is, each image processing model corresponds to different humidity, light intensity, and visibility. Among them, the image recognition model is trained based on the sample image and its corresponding status result label, and the training process is as described in steps A1 to A6.

[0036] Since the image quality of the collected visible light image and infrared light image is different under different humidity, light intensity, and visibility, in order to obtain the best-quality image, it is necessary to match the image processing model adapted to the current substation environment in the preset model library. That is, the intelligent substation inspection robot matches in the preset model library according to the current humidity, the current light intensity, and the current visibility to obtain the target image processing model adapted to the current substation environment, as specifically described in steps 201 to 204.

[0037] Step 30: Input the visible light image and the infrared light image into the target image processing model to obtain the first target image and the second target image respectively output by the target image processing model.

[0038] Further, the intelligent substation inspection robot inputs the visible light image and the infrared light image into the target image processing model, and the target image processing model processes the visible light image and the infrared light image to obtain the first target image and the second target image respectively output by the target image processing model. Among them, the processing process of the target image processing model includes denoising processing, light uniformity processing, and defogging processing. Therefore, it can be understood that the target image processing model performs denoising processing, light uniformity processing, and defogging processing on the visible light image and the infrared light image. The denoising processing process is as described in steps 301 to 303, the light uniformity processing is as described in steps 304 to 308, and the defogging processing is as described in steps 309 to 313.

[0039] Step 40: Extract the depth features of the first target image and the second target image respectively to obtain the first depth image and the second depth image, and fuse the first depth image and the second depth image to obtain the fused image.

[0040] Further, the intelligent substation inspection robot normalizes the first target image and the second target image to map the pixel value range of the image to the interval.

[0041] Therefore, the intelligent substation inspection robot normalizes each pixel point in the first target image , where and represent the horizontal and vertical coordinates of the image respectively. The specific normalization formula is as follows:

[0042] .

[0043] Among them, represents the number of pixel points in the first target image, represents the pixel value of the pixel point at in the first target image after normalization, represents the pixel value of the pixel point at in the first target image, represents the minimum pixel value in the first target image, represents the maximum pixel value in the first target image.

[0044] The intelligent substation inspection robot normalizes each pixel point in the second target image , and the specific normalization formula is as follows:

[0045] .

[0046] Among them, represents the number of pixel points in the second target image, represents the pixel value of the pixel point at in the second target image after normalization, represents the pixel value of the pixel point at in the second target image, represents the minimum pixel value in the second target image, represents the maximum pixel value in the second target image.

[0047] Furthermore, the embodiment of the present invention performs deep feature extraction based on a feature extraction algorithm with multi-scale convolutional kernels and adaptive weighting. In one embodiment, the multi-scale convolutional kernels include a set of convolutional kernels with different scales , where represents the scale identifier of the th convolutional kernel . The size of each convolutional kernel is , where is an odd number, ensuring that the convolutional kernel has a central element, which is convenient for symmetric convolution operations on the image. And the element values in the convolutional kernel are determined by the following method of random initialization and adaptive adjustment:

[0048] .

[0049] Among them, is a random number within the range of , and is a weight adaptively determined according to the local texture complexity of the image. The calculation formula of is as follows:

[0050] .

[0051] Among them, and respectively represent the gradient amplitudes of the pixel points at the point of the normalized first target image and the second target image, and represents the upper left corner coordinate of the local area of the image where the current convolutional kernel acts.

[0052] Furthermore, through the different convolutional kernels with different scales in the multi-scale convolutional kernel , convolution operations are respectively performed on the normalized image to obtain feature maps at different scales. For the first target image

[0053] after normalization processing, the calculation formula of its feature map at the th scale is as follows:

[0054] .

[0055] Among them, represents the pixel value of the pixel point at the th scale of the first feature map obtained after convolution operation on the normalized first target image at the point.

[0056] Similarly, for the second target image after normalization , the calculation formula for the feature map at the -th scale is as follows:

[0057] .

[0058] Among them, represents the second target image after normalization at the -th scale, and the second feature map obtained after convolution operation at the pixel value of the pixel at point .

[0059] Furthermore, non-linear activation is performed on the feature maps obtained at each scale. The first target image feature map after activation is denoted as , and the second target image feature map after activation is denoted as . Among them, the specific formula of the activation function is as follows: .

[0060] Among them, represents the pixel value of the feature map obtained by convolution, represents the adjustable parameter.

[0061] Therefore, for each pixel point in the first target image feature map , the pixel value can be expressed as . For each pixel point in the second target image feature map , the pixel value can be expressed as .

[0062] Furthermore, the feature maps after activation at different scales are fused to obtain the first depth image and the second depth image . Among them, the fusion method in the embodiments of the present invention adopts the method of weighted summation, and the weights are determined according to the contribution degree of each scale feature map to the overall feature expression.

[0063] For the first depth image , the pixel value of each pixel point in the first depth image can be expressed as: , where represents the preset weight, and the calculation formula of the preset weight is:

[0064] 。

[0065] Among them, and respectively represent the width and height of the first target image . is the starting coordinate of the first target image (usually ).

[0066] For the second depth image , the pixel value of each pixel point in the second depth image can be expressed as: , among which, represents a preset weight.

[0067] Optionally, in order to better fuse the first depth image and the second depth image , a feature alignment operation is first performed. The embodiment of the present invention adopts a method based on feature point matching and local deformation, specifically: by detecting the significant feature points in the first depth image and the second depth image , the feature point coordinates of the first depth image are detected as , and the feature point coordinates of the second depth image are detected as , among which, represents the coordinates of the th feature point in the first depth image ( is the abscissa, is the ordinate), represents the coordinates of the th feature point in the second depth image ( is the abscissa, is the ordinate).

[0068] Furthermore, calculate the matching relationship between feature points. The specific calculation formula is as follows:

[0069] Furthermore, calculate the matching relationship between feature points. The specific calculation formula is as follows:

[0070] .

[0071] Among them, represents a parameter for controlling the sensitivity of similarity measurement.

[0072] Further, according to the matched feature points, the first depth image after deformation is obtained by adjusting the image through the local deformation field and the second depth image after deformation . The local deformation field at the coordinate The displacement vector can be obtained by interpolation and other operations based on the matched feature points. The specific formula is as follows:

[0073] .

[0074] .

[0075] Among them, represents the dimension of the feature point coordinates of the first depth image , represents the dimension of the feature point coordinates of the second depth image ; represents the interpolation kernel function, such as the Gaussian kernel function.

[0076] Further, the embodiment of the present invention fuses the first depth image after deformation and the second depth image after deformation based on the algorithm of region adaptive weighted fusion to obtain the fused image , specifically: the image is divided into multiple non-overlapping small regions. For each region , calculate the complexity metric index of each region . The specific calculation formula is as follows:

[0077] .

[0078] Among them, represents the complexity metric index of each region , represents the gradient magnitude of the pixel point of the first depth image after deformation at the point , represents the gradient magnitude of the pixel point of the second depth image after deformation at the point , represents the number of pixel points in each region .

[0079] Further, according to the complexity metric index of each region , determine the fusion weights of the first depth image after deformation and the second depth image after deformation in each region . The specific formula is as follows:

[0080] ;

[0081] 。

[0082] Among them, represents the first depth image after deformation in each region of the fusion weight, represents the second depth image after deformation in each region of the fusion weight, representing a preset balance parameter.

[0083] The pixel value of the pixel point of the fused image at the coordinate is calculated as follows:

[0084] 。

[0085] Among them, represents the number of regions, represents the pixel value of the pixel point of the first depth image after deformation at the point, represents the pixel value of the pixel point of the second depth image after deformation at the point; represents an indicator function. When is within the region , , when is not within the region , 。

[0086] Step 50, input the fused image into a pre-trained image recognition model to obtain the recognition result input by the image recognition model, and determine the device state of the substation equipment based on the recognition result.

[0087] Optionally, input the fused image into a pre-trained image recognition model. The image recognition model performs enhancement processing on the fused image. The image enhancement method based on multi-modal transformation in the embodiments of the present invention includes two transformation modes: a spatial transformation mode and a pixel value transformation mode. Specifically:

[0088] For the spatial transformation mode, by randomly generating affine transformation parameters such as the translation amount and , the rotation angle , and the scaling factor to transform the image. The new coordinate The relationship with the original coordinates is as follows:

[0089] .

[0090] Among them, the transformation parameters need to meet the range constraints. For example, (in pixel units), , . When , the image is enlarged, and when , the image is reduced.

[0091] For the pixel value transformation mode, through the brightness-contrast adaptive adjustment algorithm. First, calculate the average brightness and the standard deviation of brightness :

[0092] .

[0093] .

[0094] Among them, and are the width and height of the image respectively.

[0095] Furthermore, adjust the pixel values of the image. The specific formula is as follows:

[0096] .

[0097] Among them, the enhanced fused image represents the pixel value of the pixel point at the coordinate , represents the gradient magnitude of the pixel point of the fused image at the coordinate .

[0098] After the above spatial transformation and pixel value transformation, the enhanced fused image is obtained.

[0099] Furthermore, normalize the enhanced fused image again so that the pixel value range of the enhanced fused image adapts to the input requirements of the image recognition model, such as normalizing to the interval . The specific normalization formula is as follows:

[0100] .

[0101] Among them, represents the minimum pixel value in the enhanced fused image , Represents the enhanced fused image The maximum pixel value in

[0102] Meanwhile, the enhanced fused image is cropped or padded according to the size requirements of the model input (if the sizes are inconsistent), and adjusted to a suitable format (for example, adjusted to the tensor format expected by the model, etc.) to obtain the processed input image , and is ready to be input into a pre-trained image recognition model.

[0103] Optionally, the image recognition model in the embodiments of the present invention includes multiple convolutional layers, attention layers, and fully connected layers. Therefore, after the processed input image is input into the image recognition model, in the convolutional layer stage, for the th convolutional layer and the th convolutional kernel , where represents the coordinates of the elements in the convolutional kernel, the size of the convolutional kernel is , and its parameters are updated based on the current input feature map , that is, the output of the previous layer or the feature map of the input image after previous processing and the prediction error feedback of the model are used for dynamic adjustment. The update formula is as follows:

[0104] .

[0105] Among them, represents the learning rate parameter, and respectively represent the width and height of the th layer feature map, represents the th layer prediction error; represents the adaptive weight, and the calculation formula of is as follows:

[0106] .

[0107] After operations such as multiple convolutional layers and attention layers, image features are continuously extracted and fused, and finally a prediction result vector is output through the fully connected layer, where is the number of categories, such as the number of different substation equipment status categories, represents the prediction probability of belonging to the th class status.

[0108] Furthermore, an uncertainty assessment is performed on the prediction probability vector output by the model. In the embodiments of the present invention, the uncertainty is measured by an uncertainty metric based on the entropy of the probability distribution. The specific formula of the uncertainty metric is as follows:

[0109] 。

[0110] Meanwhile, calculate the class index corresponding to the maximum probability value of the probability vector , and the specific formula is as follows:

[0111] 。

[0112] Furthermore, according to the uncertainty measurement index and the maximum probability class index determine the final substation equipment status.

[0113] Optionally, in the embodiment of the present invention, an uncertainty threshold is set. If , then directly take the class corresponding to the maximum probability as the equipment status, that is, the equipment status is the th class status.

[0114] If , it means that the model prediction uncertainty is relatively large. At this time, it is necessary to further refer to the local feature information in the image. Therefore, perform regional division on the processed input image . For each region , calculate its relevance score with various equipment statuses, and the specific formula is as follows:

[0115] 。

[0116] Among them, represents the relevance score of region with the th class equipment status, represents the pixel value of the pixel point of the processed input image at , represents the feature template of the th class equipment status at the coordinate .

[0117] Furthermore, calculate the comprehensive score of each class based on the region relevance score, and the specific formula is as follows: . Among them, represents the number of divided regions, represents the weight of region , and the calculation formula of weight is as follows:

[0118] 。

[0119] Indicates the processed input image At the coordinate The gradient magnitude of the pixel at the location.

[0120] Finally, determine the device status according to the comprehensive score. The device status is the status corresponding to the category with the highest comprehensive score, that is, the device status is the status of the category, where:

[0121] .

[0122] In the embodiment of the present invention, an accurate image processing model is matched according to humidity, light intensity, and visibility to perform denoising processing, light uniformity processing, and defogging processing on the collected image, so as to be able to remove the influence of environmental factors in the substation environment on the image quality, improve the quality of the collected image, thereby improving the accuracy of image recognition in the complex and changeable substation environment. Perform deep feature extraction and image fusion on the processed image, which can fuse the texture features and detail features of images in different modes to obtain a more accurate fused image, and then combine the image recognition model to recognize the fused image, which can accurately recognize the device status of the substation equipment in the inspection image, ensuring the safe and stable operation of the substation.

[0123] In one embodiment, the descriptions of steps 201 to 204 are as follows:

[0124] Step 201, obtain the optimal humidity, optimal light intensity, and optimal visibility adapted to each image processing model in the preset model library.

[0125] Optionally, each image processing model in the preset model library of the embodiment of the present invention is adapted to its corresponding optimal humidity, optimal light intensity, and optimal visibility, that is, the quality of the image obtained by processing the image under the optimal humidity, optimal light intensity, and optimal visibility is the best.

[0126] Therefore, the substation intelligent inspection robot obtains the optimal humidity, optimal light intensity, and optimal visibility adapted to each image processing model.

[0127] Step 202, based on the current humidity, current light intensity, and current visibility, and the optimal humidity, optimal light intensity, and optimal visibility of each image processing model, respectively determine the humidity matching value, light intensity matching value, and visibility matching value of each image processing model.

[0128] Further, the substation intelligent inspection robot calculates the humidity matching value of each image processing model according to the optimal humidity and current humidity of each image processing model. The specific formula is as follows:

[0129] .

[0130] Among them, represents the humidity matching value of the th image processing model, represents the current humidity, represents the th optimal humidity of the image processing model.

[0131] Furthermore, the intelligent substation inspection robot calculates the light intensity matching value with each image processing model according to the optimal light intensity and the current light intensity of each image processing model. The specific formula is as follows:

[0132] .

[0133] Among them, represents the light intensity matching value of the th image processing model, represents the current light intensity, represents the th optimal light intensity of the image processing model.

[0134] Furthermore, the intelligent substation inspection robot calculates the visibility matching value with each image processing model according to the optimal visibility and the current visibility of each image processing model. The specific formula is as follows:

[0135] .

[0136] Among them, represents the visibility matching value of the th image processing model, represents the current visibility, represents the th optimal visibility of the image processing model.

[0137] Step 203: Determine the comprehensive matching value with each image processing model based on the humidity matching value, light intensity matching value, and visibility matching value with each image processing model.

[0138] Furthermore, the intelligent substation inspection robot calculates the comprehensive matching value with each image processing model according to the humidity matching value, light intensity matching value, and visibility matching value with each image processing model. The specific calculation formula is as follows:

[0139] .

[0140] Among them, represents the comprehensive matching value of the th image processing model, , and Represents a preset weight, .

[0141] Step 204: Traverse all the comprehensive matching values, and determine the image processing model corresponding to the maximum comprehensive matching value as the target image processing model.

[0142] Optionally, the substation intelligent inspection robot traverses all the comprehensive matching values, and determines the image processing model corresponding to the maximum comprehensive matching value as the target image processing model.

[0143] In the embodiment of the present invention, an accurate image processing model is matched according to humidity, light intensity, and visibility to perform denoising processing, light uniformity processing, and defogging processing on the collected images, so as to remove the influence of environmental factors in the substation environment on the image quality, improve the quality of the collected images, and thus improve the accuracy of image recognition in the complex and changeable substation environment.

[0144] Optionally, the processing processes of the visible light image and the infrared light image in the target image processing model in the embodiment of the present invention are the same. Therefore, the processing process of the visible light image in the target image processing model is taken as an example in the embodiment of the present invention, and the processing process of the infrared light image is the same by analogy.

[0145] In one embodiment, the denoising process is as described in steps 301 to 303:

[0146] Step 301: Obtain local pixel points in the visible light image based on a local sliding window of a preset size.

[0147] Optionally, the input visible light image in the embodiment of the present invention is , where, for the visible light image at the pixel value of the pixel point can be expressed as , where and respectively represent the abscissa and ordinate of the visible light image , the image size of the visible light image is , is the number of pixels of the visible light image in the height direction, is the number of pixels of the visible light image in the width direction.

[0148] Furthermore, the preset size of the local sliding window in the embodiment of the present invention is , where is an odd number, such as 3, 5, etc. Therefore, through the local sliding window of the preset size to obtain the visible light image Local pixel points in it.

[0149] Step 302: For each pixel point in the local pixel points, calculate the pixel mean and pixel standard deviation of each pixel point within the local sliding window.

[0150] Further, for each pixel point in the local pixel points , calculate the pixel mean and pixel standard deviation of each pixel point within the local sliding window. The specific formulas are as follows:

[0151] .

[0152] Among them, represents the pixel mean of each pixel point within the local sliding window.

[0153] .

[0154] Among them, represents the pixel standard deviation of each pixel point within the local sliding window.

[0155] Step 303: Based on the pixel mean and pixel standard deviation of each pixel point within the local sliding window, perform denoising processing on each pixel point until all pixel points in the visible light image are denoised, obtaining the denoised image corresponding to the visible light image.

[0156] Further, the preset noise threshold in the embodiments of the present invention can determine a suitable value according to the actual characteristics and experience of the image. Therefore, compare the pixel standard deviation of each pixel point within the local sliding window with the preset noise threshold. If , it is determined that the pixel point may be affected by noise and is corrected. Specifically, perform denoising processing on each pixel point through the pixel mean and pixel standard deviation of each pixel point within the local sliding window. The formula for denoising processing of each pixel point is as follows:

[0157] .

[0158] Among them, represents the pixel value of the denoised image at .

[0159] Further, perform denoising processing on each pixel point according to the above process until all pixel points in the visible light image are denoised, obtaining the denoised image corresponding to the visible light image .

[0160] In the embodiment of the present invention, the acquired image is denoised by the target image processing model, so that the environmental factors in the substation environment can be removed from affecting the image quality, the quality of the acquired image is improved, and thus the accuracy of image recognition in the complex and changeable substation environment is improved.

[0161] In one embodiment, the process of uniform illumination processing is as described in steps 304 to 308:

[0162] Step 304: Divide the denoised image into image blocks of a preset size.

[0163] Optionally, the preset size in the embodiment of the present invention is , is a suitable positive integer, for example = 8, 16, etc. Therefore, the denoised image is divided into non-overlapping image blocks of size . Therefore, the denoised image can be divided into image blocks.

[0164] Step 305: Calculate the average brightness value of each image block based on the pixel values obtained after denoising each pixel point in each image block.

[0165] Further, for each image block , obtain the pixel values obtained after denoising each pixel point in each image block . According to the pixel values obtained after denoising each pixel point in each image block , calculate the average brightness value of each image block . The specific formula is as follows:

[0166] .

[0167] Among them, represents represents the th image block, represents the th average brightness value of the image block.

[0168] Step 306: Calculate the image brightness value of the denoised image based on the average brightness value of each image block.

[0169] Further, calculate the image brightness value of the denoised image according to the average brightness value of each image block . The specific calculation formula is as follows:

[0170] .

[0171] Step 307: Calculate the brightness adjustment coefficient for each image block based on the average brightness value of each image block and the image brightness value.

[0172] Further, calculate the brightness adjustment coefficient for each image block according to the average brightness value of each image block and the image brightness value. The specific formula is as follows:

[0173] 。

[0174] Wherein, represents the brightness adjustment coefficient of the th image block.

[0175] Step 308: Update the pixel values obtained after denoising each pixel point in each image block based on the brightness adjustment coefficient of each image block to obtain an image with uniform illumination processing.

[0176] Further, update the pixel values obtained after denoising each pixel point in each image block according to the brightness adjustment coefficient of each image block to obtain the pixel value of the pixel point at in the image after uniform illumination processing. All updated pixel points result in an image with uniform illumination processing , 。

[0177] In the embodiment of the present invention, the acquired image is subjected to uniform illumination processing through the target image processing model, so as to be able to remove the influence of environmental factors in the substation environment on the image quality, improve the quality of the acquired image, and thus improve the accuracy of image recognition in the complex and changeable substation environment.

[0178] In one embodiment, the defogging process is as described in steps 309 to 313:

[0179] Step 309: Obtain the pixel values of each pixel point in each color channel of the image after uniform illumination processing.

[0180] Optionally, obtain the pixel values of each pixel point in each color channel of the image after uniform illumination processing. For the pixel point in the image after uniform illumination processing ,its pixel value in each color channel can be expressed as , , represents the color channel.

[0181] Step 310: Based on the pixel values of each pixel in each color channel, determine the minimum value of the pixel values in the color channels within a preset neighborhood centered on each pixel, and obtain the dark channel of each pixel.

[0182] Optionally, the dark channel image in the embodiments of the present invention at the pixel of is the minimum value of the pixel values in the three color channels (such as the RGB color space) within a neighborhood centered on (the size is set to , being an appropriate value). Therefore, based on the pixel values of each pixel in each color channel, determine the minimum value of the pixel values in the color channels within a preset neighborhood centered on each pixel, and obtain the dark channel of each pixel. Among them, the dark channel of the pixel of the dark channel image at is as follows: The formula is as follows:

[0183] .

[0184] Among them, represents the RGB color space, represents the side length size of the preset neighborhood.

[0185] Step 311: Perform guided filtering calculation based on the dark channel of each pixel to obtain the transmittance estimation value.

[0186] Furthermore, perform guided filtering calculation according to the dark channel of each pixel to obtain the transmittance estimation value , and the specific formula of the transmittance estimation value is as follows:

[0187] .

[0188] Among them, represents the guided filtering function of the image after illumination uniformity processing.

[0189] For the guided filtering calculation, the specific analysis process is as follows:

[0190] The core idea of guided filtering is to perform a filtering operation on the image to be processed (i.e., the dark channel image in the embodiments of the present invention ) based on the guidance image (i.e., the image after illumination uniformity processing in the embodiments of the present invention), so that the filtered result can not only maintain the important detail information such as the edges of the image to be processed, but also be affected by the structural characteristics of the guidance image, achieving a better smoothing and estimation effect.

[0191] Optionally, assume the estimated value of the filtered transmittance in a local window centered on the pixel point with the window size set to , which is an odd number, and it satisfies a linear relationship with the guidance image and the dark channel image to be filtered , that is:

[0192] .

[0193] Among them, represents the th local window divided from the image. In the embodiments of the present invention, the image is divided into multiple such local windows for filtering processing respectively; represents the set of pixel points covered by the th local window, and are the linear coefficients to be determined within each local window.

[0194] Furthermore, in order to determine the coefficients and within each local window, the embodiments of the present invention define an objective function, and the goal is to make the filtered result as close as possible to the true transmittance, and at the same time require the coefficients and within the local window to change as gently as possible to ensure the smoothness of filtering. Among them, the objective function can be expressed as:

[0195] .

[0196] Among them, measures the error between the estimated transmittance and the dark channel image, represents the regularization term, represents the regularization parameter.

[0197] Optionally, in order to find the coefficients and that minimize the objective function , the embodiments of the present invention respectively take the partial derivatives of and and set the partial derivatives equal to 0.

[0198] Take the partial derivative of and set it to 0:

[0199] .

[0200] After arrangement, it can be obtained:

[0201] .

[0202] Among them, represents the number of pixel points within the local window, that is . .

[0203] Take the partial derivative of and set it to 0:

[0204] .

[0205] Substitute the expression of obtained above into this formula, and after simplification, we can get:

[0206] .

[0207] Through such calculations, the expressions of the coefficients and within each local window are obtained, so that the estimated transmittance value after filtering for each pixel point within the corresponding window can be calculated according to the linear filtering model defined above.

[0208] Furthermore, traverse all the pixel points of the entire image. For each pixel point , determine the local window it belongs to, and then according to the coefficients and corresponding to this window that have been calculated, calculate the estimated transmittance value after filtering according to the linear filtering model.

[0209] Step 312: Determine the maximum value of the pixel values corresponding to the pixel points with a preset percentage of the pixel values in the dark channel in the image after illumination uniformity processing as the estimated value of the atmospheric light intensity.

[0210] Furthermore, obtain the pixel values corresponding to the pixel points with a preset percentage (such as ) of the pixel values in the dark channel in the image after illumination uniformity processing, and determine the maximum value of the pixel values corresponding to the pixel points with a preset percentage of the pixel values in the dark channel in the image after illumination uniformity processing as the estimated value of the atmospheric light intensity . Therefore, the specific formula for the estimated value of the atmospheric light intensity is:

[0211] Among them, represents the pixel value of the pixel point at in the image after illumination uniformity processing, represents the estimated value of the atmospheric light intensity, represents the estimated transmittance value, represents the previous preset percentage, represents the dark channel image.

[0212] Step 313, based on the estimated transmittance value and the estimated atmospheric light intensity value, update the pixel value of each pixel in the image after illumination uniformity processing to obtain the first target image.

[0213] Further, based on the estimated transmittance value and the estimated atmospheric light intensity value, update the pixel value of each pixel in the image after illumination uniformity processing to obtain the first target image. The specific formula is as follows:

[0214] .

[0215] Among them, that is represents the image after defogging processing, represents the image after defogging processing at the pixel value of the pixel at, represents the preset coefficient.

[0216] In the embodiment of the present invention, the collected image is defogged through the target image processing model, so that the influence of environmental factors in the substation environment on the image quality can be removed, the quality of the collected image can be improved, and thus the accuracy of image recognition in the complex and changeable substation environment can be improved.

[0217] In one embodiment, the descriptions of steps A1 to A6 are as follows:

[0218] Step A1, input each sample image into the initial model to obtain the first state prediction result of each sample image output by the initial model.

[0219] Optionally, the training constraint conditions for the model training of the image recognition model in the embodiment of the present invention are: the loss function value is less than or equal to the first preset threshold, and the output result of the target optimization function is less than or equal to the second preset threshold, where the first preset threshold and the second preset threshold are set according to the actual situation. The target optimization function is constructed based on the model parameters of the initial model, and the model parameters include the weight matrices and bias terms of the hidden layer and the output layer. The initial model is trained based on a neural network.

[0220] Therefore, input each sample image into the initial model to obtain the first state prediction result of each sample image output by the initial model.

[0221] Step A2, input the first state prediction result and the state result label of each sample image into the loss function and the target optimization function of the initial model to obtain the first loss function value and the first output result.

[0222] Further, input the first - state prediction result and the state - result label of each sample image into the loss function and the target optimization function of the initial model to obtain the first loss - function value and the first output result. Among them, the formula of the loss function of the initial model can be expressed as:

[0223] 。

[0224] The formula of the target optimization function of the initial model can be expressed as:

[0225] 。

[0226] 。

[0227] 。

[0228] 。

[0229] Among them, represents the loss function, represents the number of sample images, represents the state prediction result of the th sample image, represents the state - result label of the th sample image, represents the exponential function, represents the value of the model parameters at the th step of training, represents the value of the model parameters at the th step of training, represents the learning rate, represents the loss function with respect to the model parameters gradient, represents the first - gradient contribution value of a single sample image to the loss function with respect to the model parameters , represents the second - gradient contribution value of a single sample image to the loss function with respect to the model parameters ; represents the sign function. When , ; when , ; when , ; represents the state prediction result of the th sample image with respect to the model parameters Partial derivative

[0230] Step A3. If the first loss function value is greater than the first preset threshold, and / or the first output result is greater than the second preset threshold, then update the parameters in the objective optimization function based on the first state prediction results and state result labels of each sample image.

[0231] Furthermore, if it is determined that the first loss function value is less than or equal to the first preset threshold, and the first output result of the objective optimization function is less than or equal to the second preset threshold, it indicates that the model parameters of the initial model and the parameters in the objective optimization function are the optimal parameters, and the initial model at this time is determined as the image recognition model.

[0232] Furthermore, if it is determined that the first loss function value is greater than the first preset threshold, and / or the first output result is greater than the second preset threshold, it indicates that the model parameters of the initial model or / and the parameters in the objective optimization function are not yet the optimal parameters. At this time, it is necessary to update the parameters in the objective optimization function and , adjust the parameters such that the weight matrices and bias terms of the hidden layer and the output layer are trained to be optimal parameters, and the learning rate is trained to be an optimal parameter. Because if the learning rate is too large, it may cause the model to "stride over" the optimal solution during training, making it difficult to converge or even increasing the loss function value; if the learning rate is too small, the update of the model parameters will be very slow, which will increase the time and computing resources required for training and it will take a long time to achieve a good training effect. Step A4. Input each sample image into the initial model with optimized parameters, and obtain the second state prediction results of each sample image output by the optimized initial model.

[0233] Step A4. Input each sample image into the initial model with optimized parameters, and obtain the second state prediction results of each sample image output by the optimized initial model.

[0234] Step A5. Input the second state prediction results and state result labels of each sample image into the loss function and the objective optimization function with optimized parameters, and obtain the second loss function value and the second output result.

[0235] Furthermore, input each sample image into the initial model with optimized parameters, and obtain the second state prediction results of each sample image output by the optimized initial model.

[0236] Furthermore, input the second state prediction results and state result labels of each sample image into the loss function and the objective optimization function with optimized parameters, and obtain the second loss function value and the second output result.

[0237] Step A6: Based on the training constraints of the image recognition model, make conditional judgments on the second loss function value and the second output result until the training constraints are met, and obtain the image recognition model.

[0238] Furthermore, if it is determined that the second loss function value is less than or equal to the first preset threshold, and the second output result of the target optimization function is less than or equal to the second preset threshold, it indicates that the model parameters of the initial model and the parameters in the target optimization function are the optimal parameters, and the initial model at this time is determined as the image recognition model.

[0239] Furthermore, if it is determined that the second loss function value is greater than the first preset threshold, or / and the second output result is greater than the second preset threshold, it indicates that the model parameters of the initial model or / and the parameters in the target optimization function are not yet the optimal parameters. At this time, the parameters in the target optimization function need to be updated according to the second state prediction result and the state result label of each sample image and until the second loss function value and the second output result meet the training constraints of the image recognition model, and the image recognition model is obtained.

[0240] In the embodiment of the present invention, an image processing model is trained. Therefore, the image collected is denoised, the illumination is made uniform, and the fog is removed through the image processing model, so that the environmental factors in the substation environment that affect the image quality can be removed, the quality of the collected image can be improved, and thus the accuracy of image recognition in the complex and changeable substation environment can be improved.

[0241] Furthermore, the image recognition system of the substation intelligent inspection robot provided by the present invention is described below. The image recognition system of the substation intelligent inspection robot described below can be mutually corresponding and referred to with the image recognition method of the substation intelligent inspection robot described above.

[0242] Optionally, referring to Figure 2 , Figure 2 is a schematic structural diagram of the image recognition system of the substation intelligent inspection robot provided by the present invention, which is applied to the substation intelligent inspection robot. The image recognition system of the substation intelligent inspection robot includes:

[0243] An acquisition module 210, configured to acquire visible light images and infrared light images of substation equipment in the current substation environment; the current substation environment includes the current humidity, the current illumination intensity, and the current visibility;

[0244] A matching module 220, configured to perform matching in a preset model library based on the current humidity, the current illumination intensity, and the current visibility to obtain a target image processing model in the current substation environment;

[0245] An image processing module 230, configured to input a visible light image and an infrared light image into a target image processing model, and obtain a first target image and a second target image respectively output by the target image processing model;

[0246] An extraction and fusion module 240, configured to perform depth feature extraction on the first target image and the second target image to obtain a first depth image and a second depth image respectively, and perform image fusion on the first depth image and the second depth image to obtain a fused image;

[0247] A device status recognition module 250, configured to input the fused image into a pre-trained image recognition model, obtain a recognition result input by the image recognition model, and determine the device status of the substation device based on the recognition result.

[0248] In the embodiment of the present invention, an accurate image processing model is matched according to humidity, light intensity, and visibility to perform denoising processing, light uniformity processing, and defogging processing on the collected images, so as to remove the influence of environmental factors in the substation environment on the image quality, improve the quality of the collected images, thereby improving the accuracy of image recognition in the complex and changeable substation environment. Performing depth feature extraction and image fusion on the processed images can fuse the texture features and detail features of the images in different modes to obtain a more accurate fused image. Then, combining the image recognition model to recognize the fused image can accurately recognize the device status of the substation device in the inspection image, ensuring the safe and stable operation of the substation.

[0249] Please refer to Figure 3 , Figure 3 which is an embodiment diagram of the electronic device provided by the embodiment of the present invention. As Figure 3 shown, the embodiment of the present invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, the following steps are implemented:

[0250] Collect a visible light image and an infrared light image of the substation device in the current substation environment; the current substation environment includes the current humidity, the current light intensity, and the current visibility;

[0251] Match in a preset model library based on the current humidity, the current light intensity, and the current visibility to obtain a target image processing model in the current substation environment;

[0252] Input the visible light image and the infrared light image into the target image processing model to obtain a first target image and a second target image respectively output by the target image processing model;

[0253] Extract depth features from the first target image and the second target image to obtain a first depth image and a second depth image respectively, and fuse the first depth image and the second depth image to obtain a fused image;

[0254] Input the fused image into a pre-trained image recognition model to obtain a recognition result input by the image recognition model, and determine the device state of the substation equipment based on the recognition result.

[0255] Please refer to Figure 4 , Figure 4 which is the embodiment diagram of the computer-readable storage medium provided by the embodiment of the present invention. As Figure 4 shown, this embodiment provides a computer-readable storage medium 400, on which a computer program 311 is stored. When the computer program 311 is executed by a processor, the following steps are implemented:

[0256] Collect visible light images and infrared light images of substation equipment in the current substation environment; the current substation environment includes the current humidity, the current light intensity, and the current visibility;

[0257] Match based on the current humidity, the current light intensity, and the current visibility in a preset model library to obtain a target image processing model in the current substation environment;

[0258] Input the visible light image and the infrared light image into the target image processing model to obtain a first target image and a second target image respectively output by the target image processing model;

[0259] Extract depth features from the first target image and the second target image to obtain a first depth image and a second depth image respectively, and fuse the first depth image and the second depth image to obtain a fused image;

[0260] Input the fused image into a pre-trained image recognition model to obtain a recognition result input by the image recognition model, and determine the device state of the substation equipment based on the recognition result.

[0261] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the recognition method for inspection images of a substation intelligent inspection robot provided by the above-mentioned various methods. The method includes:

[0262] Collect visible light images and infrared light images of substation equipment in the current substation environment; the current substation environment includes the current humidity, the current light intensity, and the current visibility;

[0263] Match based on the current humidity, current light intensity, and current visibility in a preset model library to obtain a target image processing model for the current substation environment;

[0264] Input the visible light image and the infrared light image into the target image processing model to obtain a first target image and a second target image respectively output by the target image processing model;

[0265] Extract deep features from the first target image and the second target image to obtain a first depth image and a second depth image respectively, and fuse the first depth image and the second depth image to obtain a fused image;

[0266] Input the fused image into a pre-trained image recognition model to obtain a recognition result input by the image recognition model, and determine the device status of the substation equipment based on the recognition result.

[0267] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0268] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0269] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for recognizing inspection images of an intelligent inspection robot in a substation, characterized in that, Applied to a substation intelligent inspection robot, the recognition method includes: Collect visible light images and infrared light images of substation equipment in the current substation environment; the current substation environment includes the current humidity, the current light intensity, and the current visibility; Match based on the current humidity, the current light intensity, and the current visibility in a preset model library to obtain a target image processing model in the current substation environment; Input the visible light image and the infrared light image into the target image processing model to obtain a first target image and a second target image respectively output by the target image processing model; Perform deep feature extraction on the first target image and the second target image to obtain a first depth image and a second depth image respectively, and fuse the first depth image and the second depth image to obtain a fused image; Input the fused image into a pre-trained image recognition model to obtain a recognition result input by the image recognition model, and determine the equipment status of the substation equipment based on the recognition result; Among them, the processing process of the target image processing model includes denoising processing, light intensity uniformity processing, and defogging processing; the image recognition model is trained based on sample images and their corresponding status result labels; The matching based on the current humidity, the current light intensity, and the current visibility in the preset model library to obtain a target image processing model in the current substation environment includes: Obtain the optimal humidity, the optimal light intensity, and the optimal visibility adapted to each image processing model in the preset model library; Based on the current humidity, the current light intensity, and the current visibility, as well as the optimal humidity, the optimal light intensity, and the optimal visibility of each image processing model, determine the humidity matching value, the light intensity matching value, and the visibility matching value for each image processing model respectively; Based on the humidity matching value, the light intensity matching value, and the visibility matching value for each image processing model, determine the comprehensive matching value for each image processing model; Traverse all the comprehensive matching values, and determine the image processing model corresponding to the largest comprehensive matching value as the target image processing model.

2. The recognition method of the inspection image of the intelligent substation inspection robot according to claim 1, characterized in that, The calculation formula of the comprehensive matching value is as follows: ; ； ； ； ； Among them, represents the comprehensive matching value of the th image processing model, represents the humidity matching value of the th image processing model, represents the light intensity matching value of the th image processing model, represents the visibility matching value of the th image processing model, represents the current humidity, represents the th optimal humidity of the image processing model, represents the current light intensity, represents the th optimal light intensity of the image processing model, represents the current visibility, represents the th optimal visibility of the image processing model, , and represent the preset weights.

3. The recognition method of the inspection image of the intelligent inspection robot for a substation according to claim 1, characterized in that, The process of the target image processing model performing denoising processing on the visible light image includes: Obtain local pixel points in the visible light image based on a locally sliding window of a preset size; For each pixel point in the local pixel points, calculate the pixel mean and pixel standard deviation of each pixel point within the local sliding window; Perform denoising processing on each pixel point based on the pixel mean and pixel standard deviation of each pixel point within the local sliding window until denoising processing is performed on all pixel points in the visible light image to obtain the denoised image corresponding to the visible light image; Among them, the formula for performing denoising processing on each pixel point based on the pixel mean and pixel standard deviation of each pixel point within the local sliding window is as follows: ； Among them, represents the pixel value of the denoised image at . represents the visible light image, represents the pixel value of the visible light image at . represents the abscissa and ordinate in the image, represents the pixel mean of each pixel point within the local sliding window, represents the pixel standard deviation of each pixel point within the local sliding window, represents the preset noise threshold.

4. The method for recognizing inspection images of a substation intelligent inspection robot according to claim 3, wherein, The process of the target image processing model performing light intensity uniformity processing on the visible light image includes: Divide the denoised image into image patches of a preset size; each image patch does not overlap with others; Based on the pixel values obtained after denoising for each pixel in each image patch, calculate the average brightness value of each image patch; Calculate the image brightness value of the denoised image based on the average brightness value of each image patch; Based on the average brightness value of each image patch and the image brightness value, calculate the brightness adjustment coefficient of each image patch; Update the pixel values obtained after denoising for each pixel in each image patch based on the brightness adjustment coefficient of each image patch to obtain the image after illumination uniformity processing; Among them, the calculation formula for calculating the brightness adjustment coefficient of each image patch based on the average brightness value of each image patch and the image brightness value is as follows: ； Among them, represents the brightness adjustment coefficient of the th image block, represents the average brightness value of the th image block, represents the image brightness value, represents the number of pixels of the visible light image in the height direction, represents the number of pixels of the visible light image in the width direction, represents the preset size.

5. The recognition method of the inspection image of the intelligent inspection robot for a substation according to claim 4, characterized in that, The process of the target image processing model for defogging the visible light image includes: Obtain the pixel values of each pixel in each color channel in the image after illumination uniformity processing; Based on the pixel values of each pixel in each color channel, determine the minimum value of the pixel values in the color channel within a preset neighborhood centered on each pixel to obtain the dark channel of each pixel; Perform guided filtering calculation based on the dark channel of each pixel to obtain the estimated transmission rate; Determine the maximum value of the pixel values corresponding to the pixels in the preset percentage of the pixel values in the dark channel in the image after illumination uniformity processing as the estimated atmospheric light intensity value; Update the pixel values of each pixel in the image after illumination uniformity processing based on the estimated transmission rate and the estimated atmospheric light intensity value to obtain the first target image; Among them, the specific formula for updating the pixel values of each pixel in the image after illumination uniformity processing based on the estimated transmission rate and the estimated atmospheric light intensity value is as follows: ； ； ； Among them, represents the pixel value of the dehazed image at . represents the pixel value of the image with uniform illumination at . represents the estimated value of the atmospheric light intensity. represents the estimated value of the transmittance. represents a preset coefficient. represents a previous preset percentage. represents the dark channel image. represents the guided filter function of the image after uniform illumination processing . represents the color channel. represents the RGB color space. represents the side length of a preset neighborhood. represents the pixel point at each pixel value on the color channel.

6. The recognition method of the inspection image of the intelligent substation inspection robot according to claim 1, characterized in that The training constraint conditions of the image recognition model are: the loss function value is less than or equal to the first preset threshold, and the output result of the target optimization function is less than or equal to the second preset threshold; The target optimization function is constructed based on the model parameters of the initial model, and the model parameters include the weight matrices and bias terms of the hidden layer and the output layer. The initial model is obtained by training based on a neural network; The specific process of training the image recognition model includes: Input each sample image into the initial model to obtain the first state prediction result of each sample image output by the initial model; Input the first state prediction result and the state result label of each sample image into the loss function and the target optimization function of the initial model to obtain the first loss function value and the first output result; If the first loss function value is greater than the first preset threshold, or / and, the first output result is greater than the second preset threshold, then update the parameters in the target optimization function based on the first state prediction result and the state result label of each sample image; Input each sample image into the initial model with optimized parameters to obtain the second state prediction result of each sample image output by the optimized initial model; Input the second state prediction result and the state result label of each sample image into the loss function and the target optimization function with optimized parameters to obtain a second loss function value and a second output result; Based on the training constraint conditions of the image recognition model, perform conditional determination on the second loss function value and the second output result until the training constraint conditions of the image recognition model are satisfied, and obtain the image recognition model; Among them, the formula of the loss function of the initial model is expressed as: ； The formula of the target optimization function is expressed as: ； ； ； ； Among them, represents the loss function, represents the number of sample images, represents the th state prediction result of the sample image, represents the th state result label of the sample image, represents the exponential function, represents the value of the model parameters at the th step of training, represents the value of the model parameters at the th step of training, represents the learning rate, represents the loss function with respect to the model parameters gradient, represents a single sample image with respect to the loss function regarding the model parameters first gradient contribution value, represents a single sample image with respect to the loss function regarding the model parameters second gradient contribution value; represents the sign function, when is the case, ; when is the case, ; when is the case, ; represents the th state prediction result of the sample image with respect to the model parameters partial derivative.

7. An identification system for inspection images of an intelligent inspection robot in a substation, characterized in that, Applied to the intelligent substation inspection robot, used to implement the method for recognizing inspection images of the intelligent substation inspection robot according to any one of claims 1 to 6; the recognition system includes: An acquisition module, configured to acquire visible light images and infrared light images of substation equipment in the current substation environment; the current substation environment includes the current humidity, the current light intensity, and the current visibility; A matching module, configured to perform matching in a preset model library based on the current humidity, the current light intensity, and the current visibility to obtain a target image processing model in the current substation environment; An image processing module, configured to input the visible light image and the infrared light image into the target image processing model to obtain a first target image and a second target image respectively output by the target image processing model; An extraction and fusion module, configured to perform deep feature extraction on the first target image and the second target image to obtain a first depth image and a second depth image respectively, and perform image fusion on the first depth image and the second depth image to obtain a fused image; A device state recognition module, configured to input the fused image into a pre-trained image recognition model to obtain a recognition result input by the image recognition model, and determine the device state of the substation equipment based on the recognition result; Among them, the processing process of the target image processing model includes denoising processing, light intensity uniform processing, and defogging processing; the image recognition model is trained based on sample images and their corresponding state result labels; The performing matching in a preset model library based on the current humidity, the current light intensity, and the current visibility to obtain a target image processing model in the current substation environment includes: Obtain the optimal humidity, optimal light intensity, and optimal visibility adapted to each image processing model in the preset model library; Based on the current humidity, the current light intensity, and the current visibility, and the optimal humidity, optimal light intensity, and optimal visibility of each image processing model, respectively determine the humidity matching value, light intensity matching value, and visibility matching value of each image processing model; Based on the humidity matching value, light intensity matching value, and visibility matching value of each image processing model, determine the comprehensive matching value of each image processing model; Traverse all the comprehensive matching values, and determine the image processing model corresponding to the largest comprehensive matching value as the target image processing model.

8. An electronic device, comprising: A memory, configured to store computer software programs; A processor for reading and executing the computer software program, characterized in that when the processor executes the computer software program, it implements the recognition method for the inspection images of the intelligent substation inspection robot according to any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium storing a computer software program, characterized in that, When the computer software program is executed by the processor, it implements the recognition method for the inspection images of the intelligent substation inspection robot according to any one of claims 1 to 6.

Images