Method and system for removing background interference of spectral image of insulator

Abstract

The invention discloses an insulator spectral image background interference removal method and system, and belongs to the technical field of online detection of the external insulation state of a power transmission line, and the method comprises the steps: obtaining an insulator spectral image containing an insulator and background interference through a spectral image information collection device; repeating the above steps to obtain insulator spectral images of different angles, different wavebands and different background interferences, constructing an insulator spectral image background interference data set, and establishing an image segmentation network-based insulator spectral image background interference removal model; and for an insulator spectral image of which the background interference is to be removed, removing the image background interference by using the insulator spectral image background interference removal model, so that only an effective insulator region is reserved in the image. According to the method, the background interference of the spectral image of the insulator can be accurately removed, and the insulator detection efficiency can be conveniently improved by operation and maintenance personnel of a power system.

Description

Technical Field

[0001] The present invention belongs to the technical field of online detection of external insulation status of power transmission lines, and in particular relates to a method and system for removing background interference from an insulator spectral image. Background Art

[0002] Insulators are the most numerous and diverse electrical components in power transmission and transformation equipment. They are exposed to the air for a long time and are inevitably affected by agricultural pollution, industrial pollution and other factors, causing problems such as insulator contamination, aging, and damage, which significantly reduce their external insulation strength, and ultimately lead to large-scale, long-term power outages, which seriously threaten the safe and stable operation of the power system. Therefore, accurately detecting the state of insulators plays a very important role in ensuring the safe and stable operation of insulators.

[0003] Currently, the means of detecting the live status of insulators are relatively limited, mainly based on visible light, infrared and ultraviolet imaging. However, the accuracy of the diagnostic model of visible light images is low in practical applications. The detection results of infrared images and ultraviolet images are not only affected by the state of the insulator itself, but also by its operating voltage and operating load. Moreover, these means can only reflect abnormal conditions in the late stage of fault development, and it is difficult to timely discover potential fault hazards such as contamination, aging and damage.

[0004] Since there is a lot of irrelevant background information other than insulators in the images collected by the spectral imaging device, removing the background interference of the insulator spectral image has become an important basis and prerequisite for the spectral analysis of the line insulator status. However, most of the existing methods are applied to grayscale or color images, and the direct application to insulator spectral images has poor adaptability. It is necessary to use the insulator spectral characteristics to optimize the image background interference removal method.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the invention and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention aims to provide a method for removing background interference from an insulator spectral image, so as to solve the current technical problem of low efficiency and accuracy in removing background interference from an insulator spectral image.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] A first aspect of the present invention provides a method for removing background interference from an insulator spectral image, comprising the following steps:

[0009] Step S01, using a spectral image information acquisition device to obtain insulator spectral image data including insulators and background interference;

[0010] Step S02, repeating step S01 to obtain insulator spectral image data of different angles, different bands, and different background interferences, marking the effective insulator positions in the image as labels, and constructing an insulator spectral image background interference data set;

[0011] Step S03, constructing an insulator spectral image background interference removal model based on an image segmentation network;

[0012] Step S04: for the insulator spectrum image whose background interference is to be removed, the background interference of the image is removed by using the insulator spectrum image background interference removal model, so that only the effective insulator area is retained in the image.

[0013] Optionally, in step S01, the spectral image acquisition device includes an imaging unit and an angle adjustment unit, the imaging unit is used to image the insulator to capture the spectral characteristics of the insulator surface, and the angle adjustment unit is used to adjust the angle of the imaging unit to obtain the spectral characteristics of the insulator surface at different angles.

[0014] Optionally, two multi-spectral imaging sensors are deployed in the imaging unit, and the angles of the two multi-spectral imaging sensors are adjusted respectively to image the insulator at two different angles simultaneously, so as to capture the optical features of the same insulator at two viewing angles.

[0015] Optionally, a rotating table is deployed in the angle adjustment unit to control the angle at which the multi-spectral imaging sensor images the insulator, and a fixed arm is deployed to fix the multi-spectral imaging sensor to the rotating table so that it rotates synchronously with the rotating table.

[0016] Optionally, the center wavelength range of the band of the insulator spectral image acquired by the multi-spectral imaging sensor is 400 nm to 900 nm.

[0017] Optionally, the continuous adjustment angle range of the angle adjustment unit is 0° to 180°.

[0018] Optionally, in step S02, constructing the insulator spectral image background interference dataset further includes:

[0019] An insulator spectral image dataset is constructed, which contains n images of the same size. For each image in the dataset, a binary label image matching its size is produced, in which the effective area of ​​the insulator is marked as 1, and the background and other non-target areas are marked as 0.

[0020] Optionally, the step S03 further includes:

[0021] The background interference removal model of insulator spectral image includes an encoder and a decoder;

[0022] The encoder includes 9 convolution layers with a convolution kernel of 3×3 and 4 maximum pooling layers with a field of view of 2×2; the insulator spectrum image is input into the encoder, and the convolution layer A is first used to extract features, and then the convolution layer B is used to further extract features, and then the maximum pooling layer A is downsampled to reduce the spatial dimension of the data, and then the convolution layer C and D are continued, and the maximum pooling layer B is further downsampled; this process is subsequently repeated twice, respectively through the convolution layers E, F and the pooling layer C, and the convolution layers G, H and the maximum pooling layer D, and finally the encoding process is completed through the convolution layer I;

[0023] The decoder includes 9 convolutional layers with 3×3 convolution kernels, 4 upsampling layers and 4 splicing layers; starting from the output of the convolutional layer I of the encoder, it first passes through the convolutional layer J, then uses the bilinear interpolation upsampling layer A to restore the spatial dimension of the image, and splices it with the corresponding level features of the encoder through the splicing layer A; subsequently, this process is repeated three times, the convolutional layers K and L are followed by the upsampling layer B and the splicing layer B, the convolutional layers M and N are followed by the upsampling layer C and the splicing layer C, the convolutional layers O and P are followed by the upsampling layer D and the splicing layer D, and finally the decoding process is completed through the convolutional layers Q and R, where the convolutional layer R outputs the final image segmentation result.

[0024] Optionally, the training of the insulator spectral image background interference removal model includes:

[0025] Set the number of network training iterations to 200, the batch training size to 32, the initial learning rate to 0.002, and the parameter optimization algorithm to stochastic gradient descent;

[0026] The insulator spectral image background interference dataset is randomly divided into training set, validation set and test set, where the data volume of the training set, validation set and test set is allocated in a ratio of 6:1:1;

[0027] The training set data is input into the model for training, and the validation set data is used to calculate the loss value of the network model and adjust the network parameters after each iteration. After the training is completed, the performance of the network model is evaluated using the test set data to obtain the insulator spectral image background interference removal model.

[0028] A second aspect of the present invention provides a system for removing background interference from an insulator spectral image, which is used to execute the above-mentioned method for removing background interference from an insulator spectral image. The system comprises:

[0029] Spectral image information acquisition equipment, used for acquiring spectral images of insulators including background interference;

[0030] An insulator background interference removal module is used to process the insulator spectrum image containing background interference to obtain the insulator spectrum image with the background interference removed;

[0031] A data storage module, used for storing the insulator spectral image collected by the spectral image information collection device and the insulator spectral image processed by the insulator background interference removal module;

[0032] The output module is used to output the insulator spectrum image with background interference removed.

[0033] The beneficial effects of the present invention are as follows:

[0034] The present invention designs a spectral image acquisition device optimized for the material characteristics of insulators, which can obtain spectral images of insulators at different angles, different bands, and different background interferences, and integrates multi-spectral imaging technology, so that the device can obtain high-quality spectral image data of insulators;

[0035] The present invention makes full use of the precision and efficiency advantages of deep convolutional neural networks to train parameter values ​​suitable for the insulator spectral image background interference removal model, thereby achieving accurate removal of insulator spectral image background interference, and facilitating power system operation and maintenance personnel to improve insulator detection efficiency.

[0036] The above description is only an overview of the technical solution of the present invention. In order to make the technical means of the present invention more clear and understandable, to the extent that those skilled in the art can implement it according to the contents of the specification, and to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the following is an example of the specific implementation of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0038] By reading the detailed description of the preferred specific embodiments below, various other advantages and benefits of the present invention will become clear to those of ordinary skill in the art. The drawings in the specification are only for the purpose of illustrating the preferred embodiments and are not considered to be limitations of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative work. Moreover, the same reference numerals are used to represent the same components throughout the drawings.

[0039] In the attached picture:

[0040] Figure 1This is a flow chart of the method for removing background interference from insulator spectral images;

[0041] Figure 2 The insulator spectral image data is obtained by using the spectral image information acquisition equipment;

[0042] Figure 3 This is the network structure diagram of the insulator spectral image background interference removal model based on the image segmentation network;

[0043] Figure 4 This is the result of removing background interference from the insulator spectrum image. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0045] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0047] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0048] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0049] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected or indirectly connected through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0050] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may include that the first and second features are in direct contact, or may include that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, a first feature being "above", "above" and "above" a second feature includes that the first feature is directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. A first feature being "below", "below" and "below" a second feature includes that the first feature is directly below and obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0051] In order to enable those skilled in the art to better understand the technical solution of the present invention, Figures 1 to 4 The present invention is further described in detail, and the accompanying drawings do not constitute a limitation on the embodiments of the present invention.

[0052] In embodiment 1, the present invention provides a method for removing background interference from an insulator spectrum image, comprising the following steps:

[0053] Step S01, using a spectral image information acquisition device to obtain insulator spectral image data including insulators and background interference;

[0054] Step S02, repeating step S01 to obtain insulator spectral image data of different angles, different bands, and different background interferences, marking the effective insulator positions in the image as labels, and constructing an insulator spectral image background interference data set;

[0055] Step S03, constructing an insulator spectral image background interference removal model based on an image segmentation network;

[0056] Step S04: for the insulator spectrum image whose background interference is to be removed, the background interference of the image is removed by using the insulator spectrum image background interference removal model, so that only the effective insulator area is retained in the image.

[0057] In a further preferred but non-limiting embodiment, in step S01, the spectral image acquisition device includes an imaging unit and an angle adjustment unit, the imaging unit is used to image the insulator to capture the spectral characteristics of the insulator surface, and the angle adjustment unit is used to adjust the angle of the imaging unit to obtain the spectral characteristics of the insulator surface at different angles.

[0058] In a further preferred but non-limiting embodiment, two multi-spectral imaging sensors are deployed in the imaging unit, and the angles of the two multi-spectral imaging sensors are adjusted respectively to simultaneously image the insulator at two different angles to capture the optical characteristics of the same insulator at two viewing angles.

[0059] In a further preferred but non-limiting embodiment, a rotating table is deployed in the angle adjustment unit to control the angle at which the multi-spectral imaging sensor images the insulator, and a fixed arm is deployed to fix the multi-spectral imaging sensor to the rotating table so that it rotates synchronously with the rotating table.

[0060] In a further preferred but non-limiting embodiment, the center wavelength of the band of the insulator spectral image acquired by the multi-spectral imaging sensor ranges from 400 nm to 900 nm.

[0061] In a further preferred but non-limiting embodiment, the continuous adjustment angle range of the angle adjustment unit is from 0° to 180°.

[0062] In a further preferred but non-limiting embodiment, in step S02, constructing the insulator spectral image background interference dataset further includes:

[0063] An insulator spectral image dataset is constructed, which contains n images of the same size. For each image in the dataset, a binary label image matching its size is produced, in which the effective area of ​​the insulator is marked as 1, and the background and other non-target areas are marked as 0.

[0064] In a further preferred but non-limiting embodiment, step S03 further includes:

[0065] like Figure 3 As shown, the background interference removal model of the insulator spectrum image includes an encoder and a decoder;

[0066] The encoder includes 9 convolution layers with a convolution kernel of 3×3 and 4 maximum pooling layers with a field of view of 2×2; the insulator spectrum image is input into the encoder, and the convolution layer A is first used to extract features, and then the convolution layer B is used to further extract features, and then the maximum pooling layer A is downsampled to reduce the spatial dimension of the data, and then the convolution layer C and D are continued, and the maximum pooling layer B is further downsampled; this process is subsequently repeated twice, respectively through the convolution layers E, F and the pooling layer C, and the convolution layers G, H and the maximum pooling layer D, and finally the encoding process is completed through the convolution layer I;

[0067] The decoder includes 9 convolutional layers with 3×3 convolution kernels, 4 upsampling layers and 4 Concat splicing layers; starting from the output of the convolutional layer I of the encoder, it first passes through the convolutional layer J, then uses the bilinear interpolation upsampling layer A to restore the spatial dimension of the image, and concat splices with the corresponding level features of the encoder through the Concat splicing layer A; subsequently, this process is repeated three times, the convolutional layers K and L are followed by the upsampling layer B and the Concat splicing layer B, the convolutional layers M and N are followed by the upsampling layer C and the Concat splicing layer C, the convolutional layers O and P are followed by the upsampling layer D and the Concat splicing layer D, and finally the decoding process is completed through the convolutional layers Q and R, where the convolutional layer R outputs the final image segmentation result.

[0068] In a further preferred but non-limiting embodiment, the training of the insulator spectral image background interference removal model includes:

[0069] Set the number of network training iterations to 200, the batch training size to 32, the initial learning rate to 0.002, and the parameter optimization algorithm to stochastic gradient descent;

[0070] The insulator spectral image background interference dataset is randomly divided into training set, validation set and test set, where the data volume of the training set, validation set and test set is allocated in a ratio of 6:1:1;

[0071] The training set data is input into the model for training, and the validation set data is used to calculate the loss value of the network model and adjust the network parameters after each iteration. After the training is completed, the performance of the network model is evaluated using the test set data to obtain the insulator spectral image background interference removal model.

[0072] In Embodiment 2, the present invention provides a system for removing background interference from an insulator spectrum image, which is used to execute the method for removing background interference from an insulator spectrum image described in Embodiment 1. The system includes:

[0073] Spectral image information acquisition equipment, used for acquiring spectral images of insulators including background interference;

[0074] An insulator background interference removal module is used to process the insulator spectrum image containing background interference to obtain the insulator spectrum image with the background interference removed;

[0075] A data storage module, used for storing the insulator spectral image collected by the spectral image information collection device and the insulator spectral image processed by the insulator background interference removal module;

[0076] The output module is used to output the insulator spectrum image with background interference removed.

[0077] In a preferred embodiment of the method, the insulator spectral image background interference removal model based on the image segmentation network includes three parts: an encoder, a decoder and a feature splicing.

Claims

1. A method for removing background interference from an insulator spectral image, characterized in that: The following steps are involved: Step S01, using a spectral image information acquisition device to obtain insulator spectral image data including insulators and background interference; Step S02, repeating step S01 to obtain insulator spectral image data of different angles, different bands, and different background interferences, marking the effective insulator positions in the image as labels, and constructing an insulator spectral image background interference data set; Step S03, constructing an insulator spectral image background interference removal model based on an image segmentation network; Step S04: for the insulator spectrum image whose background interference is to be removed, the background interference of the image is removed by using the insulator spectrum image background interference removal model, so that only the effective insulator area is retained in the image.

2. The method for removing background interference from an insulator spectrum image according to claim 1, characterized in that: In step S01, the spectral image acquisition device includes an imaging unit and an angle adjustment unit. The imaging unit is used to image the insulator to capture the spectral characteristics of the insulator surface, and the angle adjustment unit is used to adjust the angle of the imaging unit to obtain the spectral characteristics of the insulator surface at different angles.

3. The method for removing background interference from an insulator spectrum image according to claim 2, characterized in that: Two multi-spectral imaging sensors are deployed in the imaging unit, and the angles of the two multi-spectral imaging sensors are adjusted respectively to image the insulator at two different angles simultaneously, so as to capture the optical features of the same insulator at two viewing angles.

4. The method for removing background interference from an insulator spectrum image according to claim 2 or 3, characterized in that: A rotating table is deployed in the angle adjustment unit to control the angle at which the multi-spectral imaging sensor images the insulator, and a fixed arm is deployed to fix the multi-spectral imaging sensor on the rotating table so that the sensor rotates synchronously with the rotating table.

5. The method for removing background interference from insulator spectrum images according to claim 4, characterized in that: The center wavelength of the band of the insulator spectral image acquired by the multi-spectral imaging sensor ranges from 400nm to 900nm.

6. The method for removing background interference from an insulator spectrum image according to claim 4, characterized in that: The continuous adjustment angle range of the angle adjustment unit is from 0° to 180°.

7. The method for removing background interference from an insulator spectrum image according to claim 1, characterized in that: In the step S02, constructing the insulator spectral image background interference data set also includes: An insulator spectral image dataset is constructed, which contains n images of the same size. For each image in the dataset, a binary label image matching its size is produced, in which the effective area of ​​the insulator is marked as 1, and the background and other non-target areas are marked as 0.

8. The method for removing background interference from an insulator spectrum image according to claim 1, characterized in that: The step S03 further includes: The background interference removal model of insulator spectral image includes an encoder and a decoder; The encoder includes 9 convolution layers with a convolution kernel of 3×3 and 4 maximum pooling layers with a field of view of 2×2; the insulator spectrum image is input into the encoder, and the convolution layer A is first used to extract features, and then the convolution layer B is used to further extract features, and then the maximum pooling layer A is downsampled to reduce the spatial dimension of the data, and then the convolution layer C and D are continued, and the maximum pooling layer B is further downsampled; this process is subsequently repeated twice, respectively through the convolution layers E, F and the pooling layer C, and the convolution layers G, H and the maximum pooling layer D, and finally the encoding process is completed through the convolution layer I; The decoder includes 9 convolutional layers with 3×3 convolution kernels, 4 upsampling layers and 4 splicing layers; starting from the output of the convolutional layer I of the encoder, it first passes through the convolutional layer J, then uses the bilinear interpolation upsampling layer A to restore the spatial dimension of the image, and splices it with the corresponding level features of the encoder through the splicing layer A; subsequently, this process is repeated three times, the convolutional layers K and L are followed by the upsampling layer B and the splicing layer B, the convolutional layers M and N are followed by the upsampling layer C and the splicing layer C, the convolutional layers O and P are followed by the upsampling layer D and the splicing layer D, and finally the decoding process is completed through the convolutional layers Q and R, where the convolutional layer R outputs the final image segmentation result.

9. The method for removing background interference from an insulator spectrum image according to claim 7 or 8, characterized in that: The training of the insulator spectral image background interference removal model includes: Set the number of network training iterations to 200, the batch training size to 32, the initial learning rate to 0.002, and the parameter optimization algorithm to stochastic gradient descent; The insulator spectral image background interference dataset is randomly divided into training set, validation set and test set, where the data volume of the training set, validation set and test set is allocated in a ratio of 6:1:1; The training set data is input into the model for training, and the validation set data is used to calculate the loss value of the network model and adjust the network parameters after each iteration. After the training is completed, the performance of the network model is evaluated using the test set data to obtain the insulator spectral image background interference removal model.

10. An insulator spectral image background interference removal system, used to execute the method according to any one of claims 1 to 9, characterized in that: The system comprises: Spectral image information acquisition equipment, used for acquiring spectral images of insulators including background interference; An insulator background interference removal module is used to process the insulator spectrum image containing background interference to obtain the insulator spectrum image with the background interference removed; A data storage module, used for storing the insulator spectral image collected by the spectral image information collection device and the insulator spectral image processed by the insulator background interference removal module; The output module is used to output the insulator spectrum image with background interference removed.

Images