Method and system for monitoring disturbance range of environmental water conservation land of transformer substation
Image data acquired through the drone, combined with wavelet transformation and ResNet-50 feature extraction network, solves the problem of insufficient detection accuracy of land disturbance range in traditional monitoring methods, and achieves more accurate and efficient monitoring of land disturbance range.
Patent Information
- Application Number
- CN202411022660.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-07-29
AI Technical Summary
During the construction of traditional substations, there is insufficient detection accuracy of the land disturbance range monitoring method. Especially when complex change information modeling and feature overlap or statistical distribution modeling are inaccurate, large errors are easily generated and it is difficult to accurately monitor the land disturbance range.
The drone is used to obtain the front and back phase images of the substation engineering area, and the wavelet transformation decomposition and interpolation processing are used, combined with the ResNet-50 feature extraction network fused with the lightweight channel attention module, and the image is extracted and differentially processed. , generate a binary classification change prediction map and output a land disturbance range image.
It improves the recognition accuracy of image change areas, provides more reliable land disturbance range monitoring data, reduces information loss, enhances detection accuracy and efficiency, and can effectively monitor the land disturbance range in substation projects.
Smart Images

Figure CN118968409A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of substation environmental and water conservation, and in particular relates to a method and system for monitoring the land disturbance range of substation environmental and water conservation. Background Art
[0002] Substations are a key component of the power system, responsible for converting the high-voltage electricity generated by power stations into voltage levels suitable for long-distance transmission and distribution. Substations are also responsible for converting high-voltage electricity to lower voltages suitable for domestic and industrial use.
[0003] The construction and operation of substations are crucial to the stability and reliability of power supply.
[0004] With the rapid development of industrialization and urbanization, environmental protection and water resource protection have become increasingly important. The construction and operation of substations may have an impact on the surrounding environment, especially in terms of land disturbance.
[0005] Land disturbance can lead to a variety of environmental problems, such as soil erosion, water loss, ecosystem damage and water pollution, which not only affect the quality of life of local residents, but may also have long-term impacts on the entire ecosystem.
[0006] During the construction of the substation project, construction workers excavated the ground, causing differences between the surface of the construction area and the surrounding environment. This difference is mainly reflected in the difference between new soil and original soil, and between exposed surface and vegetation-covered surface. The boundary extraction result of the difference is the project construction disturbance area, thereby calculating the scope of land disturbance.
[0007] Traditional remote sensing image change detection mainly uses methods such as image difference, ratio, change vector analysis (CVA), etc., which respectively apply neighborhood information, structural features and contrast difference maps, mean difference maps and principal component analysis to change detection, and the detection accuracy has been improved to a certain extent. However, the above methods partially rely on manually constructed feature representations, and have limited modeling capabilities for complex high-level change information. In addition, when the features of change and non-change classes overlap or their statistical distribution modeling is inaccurate, the detection results will produce large errors and have certain limitations. Summary of the invention
[0008] The purpose of the embodiments of the present invention is to provide a method and system for monitoring the disturbance range of water conservation and land in a substation, aiming to solve the technical problems raised in the above-mentioned background technology.
[0009] To achieve the above object, the present invention provides the following technical solutions.
[0010] In a first aspect of the present invention, a method for monitoring the disturbance range of land for water conservation and environmental protection in a substation is provided, the method comprising the following steps:
[0011] Step S101: using a drone to obtain a pre-phase image and a post-phase image of a substation project area;
[0012] Step S102: The pre-phase image and the post-phase image are decomposed into high-frequency and low-frequency components by wavelet transform, the high-frequency components are processed by bilinear interpolation method, and the low-frequency components are processed by bicubic interpolation method, and then inverse wavelet transform is performed to output a high-resolution image;
[0013] Step S103: In the encoder, a ResNet-50 feature extraction network integrated with a lightweight channel attention module is used to perform feature extraction processing on the high-resolution image, and each feature layer extracted is subtracted and the absolute value is taken to obtain a difference feature map;
[0014] Step S104: The difference feature map is merged into the decoder part by upsampling after channel splicing. In the decoder, the feature map enhanced by feature difference is decoded through an MLP structure containing only one hidden layer to generate a binary classification change prediction map and output the final change result image.
[0015] In a further limitation of the embodiment of the present invention, before the high-frequency component is processed using the bilinear interpolation method, the high-frequency component is also processed using the Canny algorithm to extract edges to obtain new high-frequency components, and then the new high-frequency components are processed using the bilinear interpolation method, which can effectively retain edge information and improve the interpolation speed.
[0016] In a further limitation of the embodiment of the present invention, the ResNet-50 feature extraction network integrated with the lightweight channel attention module includes:
[0017] An input layer, the input layer is used to receive input images. In the embodiment of the present invention, the input images are high-resolution pre-phase images and post-phase images;
[0018] Residual convolution block: includes convolution, batch normalization and ReLU activation function;
[0019] Lightweight channel attention module: added after the residual block;
[0020] Average pooling layer: Average pooling is performed after the residual block and CAM;
[0021] Fully connected layer: maps the pooled features.
[0022] In a further limitation of the embodiment of the present invention, in the step of performing feature extraction processing on the high-resolution image using the ResNet-50 feature extraction network integrated with the lightweight channel attention module, the front phase image and the back phase image are feature extracted by two feature extraction modules with shared parameters;
[0023] After the image is processed by convolution and pooling, the residual convolution block is used to extract multi-scale semantic features;
[0024] The residual convolution block consists of a ResNeXt unit with 2x downsampling and a 1×1 convolution, followed by a unified upsampling module to obtain the same feature map size as the first layer residual block. In this way, output features that combine high-level semantic information and low-level spatial information of different scales can be obtained;
[0025] The transformed feature maps of the four layers are concatenated on the channel and input into a 3×3 convolutional layer to obtain the final feature map.
[0026] In a further limitation of the embodiment of the present invention, when information at different scales is fused, data at the same scale is required to be fused, and upsampling and downsampling methods are used to keep the image data at the same resolution. Specifically, the scale calculation method is as follows:
[0027]
[0028] In the formula, padding represents the padding of feature data; kernel_size represents the size of the convolution kernel; stride represents the stride of each convolution movement; H out Represents the height of the feature map; W out Indicates the width of the feature map.
[0029] In a further limitation of the embodiment of the present invention, the output of the cth convolutional layer in the nth residual convolutional block is expressed as: n,c =σ(W n,c [F n-1 , F n,1 , ..., F n,c-1 ]);
[0030] Where σ represents the ReLU activation function, W n,c Represents the weight value of the cth convolutional layer;
[0031] The output features of each layer of each residual convolution block are mapped and spliced, and each layer of the output features of the previous residual convolution block and the output features of the current residual convolution block are connected to the subsequent layers. Based on this form, local features can be well preserved.
[0032] In a further limitation of the embodiment of the present invention, three residual convolution blocks are used as the backbone of the feature extraction network. After obtaining the local information features of each residual convolution block, a global feature fusion and a global residual learning are performed; the fully connected layer fuses the output features of each residual convolution block, and performs a global residual learning with the initial input features, which can effectively reduce information loss;
[0033] The fused features are sent to a 3×3 convolution, and the obtained features are fused with themselves to obtain the final feature result. The final feature result F′ is expressed as:
[0034] F′=F cat +Conv(F cat );
[0035] In the formula, Conv represents a 3×3 convolution operation, and F cat It represents the result of global fusion and residual learning of features, which can effectively retain low-level features.
[0036] In a further limitation of the embodiment of the present invention, the step of fusing the change feature map into the decoder part by upsampling after channel splicing includes: among f1, f2, f3, f4 and f5 in the difference feature map, f3, f4 and f5 are upsampled twice respectively and spliced and fused with f2 to obtain p2, and then p2 is upsampled twice and spliced and fused with f1 to obtain p1.
[0037] In a further limitation of the embodiment of the present invention, in the step of decoding the feature map after feature difference enhancement to generate a binary change prediction map and outputting the final change result image, MLP decodes the upsampled feature map p1 to generate a change prediction map. The change prediction map is a two-dimensional image, in which the value of each pixel represents the possibility of a change at that position. The change prediction map is converted into a binary image through a threshold, in which pixels above the threshold are marked as 1, and pixels not above the threshold are marked as 0, 0 and 1 represent the labels of the unchanged area and the changed area, respectively. Correspondingly, the changed area is the land disturbance range that needs to be monitored in the substation project.
[0038] In a second aspect of the present invention, a system for monitoring the environmental and water conservation land disturbance range of a substation is provided, the system comprising:
[0039] An image acquisition module is used to acquire pre-phase images and post-phase images of the substation engineering area using a drone;
[0040] An image processing module is used to decompose the pre-phase image and the post-phase image into high-frequency and low-frequency components through wavelet transformation, process the high-frequency components using a bilinear interpolation method, process the low-frequency components using a bicubic interpolation method, and then perform an inverse wavelet transformation to output a high-resolution image;
[0041] The encoder module is used to perform feature extraction processing on high-resolution images using a ResNet-50 feature extraction network integrated with a lightweight channel attention module, subtract each feature layer extracted and take the absolute value to obtain a difference feature map;
[0042] The decoder module is used to fuse the difference feature map into the decoder part by upsampling the difference feature map after channel splicing. In the decoder, the feature map enhanced by feature difference is decoded through an MLP structure containing only one hidden layer to generate a binary classification change prediction map and output the final change result image.
[0043] Compared with the prior art, the beneficial effects of the method and system for monitoring the environmental and water conservation land disturbance range of a substation of the present invention are:
[0044] First, the pre-phase image and the post-phase image of the present invention are decomposed into high-frequency and low-frequency components by wavelet transformation, the high-frequency component is processed by bilinear interpolation method, and the low-frequency component is processed by bicubic interpolation method, and then the inverse wavelet transformation is performed, so that a high-resolution image can be output, so as to provide more reliable image data for subsequent models to identify image change areas;
[0045] Second, in the encoder of the present invention, a ResNet-50 feature extraction network integrated with a lightweight channel attention module is used to perform feature extraction processing on high-resolution images, and each extracted feature layer is subtracted and the absolute value is taken to obtain a difference feature map;
[0046] Third, in the decoder of the present invention, the difference feature map is merged into the decoder part by upsampling after channel splicing, and the feature map enhanced by feature difference is decoded through an MLP structure containing only one hidden layer to generate a binary classification change prediction map, and output the final change result image.
[0047] To sum up, the present invention is based on a feature extraction network of deep learning, takes the drone images in the substation construction process as the data source, establishes a substation construction land disturbance monitoring system, obtains the land disturbance information in the substation construction area through image information extraction, and judges whether the scope of land disturbance in the construction project is reasonable, so as to achieve the purpose of rapid monitoring of soil and water conservation in the construction process. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention.
[0049] Figure 1 This is a structural diagram of a method for monitoring the range of land disturbance in a substation for water conservation and environment protection according to the present invention;
[0050] Figure 2 It is a flow chart of the implementation of a method for monitoring the range of land disturbance in a substation for water conservation and environment protection according to the present invention;
[0051] Figure 3 A block diagram of the principle of high-resolution image processing provided by an embodiment of the present invention;
[0052] Figure 4 A structural block diagram of a substation environmental water conservation and land disturbance range monitoring system provided by the present invention;
[0053] Figure 5 The present invention provides a structural block diagram of a computer device. DETAILED DESCRIPTION
[0054] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0055] The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.
[0056] like Figure 1 and Figure 2 As shown, in one embodiment provided by the present invention, a method for monitoring the disturbance range of land for water conservation and environmental protection in a substation comprises the following steps:
[0057] Step S101: using a drone to obtain a pre-phase image and a post-phase image of a substation project area;
[0058] The pre-phase image and post-phase image refer to two images acquired at different times. The pre-phase image is acquired earlier than the post-phase image. The purpose of the embodiment of the present invention is to obtain the change range of the post-phase image relative to the pre-phase image, that is, the land disturbance range caused by the substation project in the project area. In this disturbance range, the surface of the construction area and the surrounding environment are different. The boundary extraction result of the difference is the substation construction disturbance area, thereby monitoring the land disturbance range.
[0059] In the selection and preparation of drones, it is necessary to select a suitable drone based on the size of the substation project area and the complexity of the terrain, taking into account its flight altitude, endurance, load capacity, etc. The drone needs to be equipped with high-resolution photographic equipment. At the predetermined location, the drone automatically takes high-resolution surface images according to the set parameters.
[0060] Choose an appropriate time point after the project implementation to shoot the post-phase images so that the changes after the project implementation can be observed; specifically, try to shoot under the same lighting and climate conditions as the pre-phase images to reduce the impact of external factors on image comparison.
[0061] Specifically, the key to the task of monitoring the range of land disturbance for water conservation and environmental protection in substations is to distinguish the areas related to the changes from the areas not related to the changes in the image, among which the areas related to the changes are the parts that need to be paid attention to in the monitoring task.
[0062] For further information, please refer to Figure 1 and Figure 2 , the monitoring method of the present invention further comprises:
[0063] Step S102: The pre-phase image and the post-phase image are decomposed into high-frequency and low-frequency components by wavelet transform, the high-frequency components are processed by bilinear interpolation method, and the low-frequency components are processed by bicubic interpolation method, and then inverse wavelet transform is performed to output a high-resolution image;
[0064] Among them, most of the image information is stored in the low-frequency components, such as the image contour information; while some edge and detail information is stored in the high-frequency components. In order to achieve the best effect, different interpolation methods are needed for the two components.
[0065] For further information, please refer to Figure 1 and Figure 2 , the monitoring method of the present invention further comprises:
[0066] Step S103: In the encoder, a ResNet-50 feature extraction network integrated with a lightweight channel attention module is used to perform feature extraction processing on the high-resolution image, and each feature layer extracted is subtracted and the absolute value is taken to obtain a difference feature map;
[0067] Step S104: The difference feature map is merged into the decoder part by upsampling after channel splicing. In the decoder, the feature map enhanced by feature difference is decoded through an MLP structure containing only one hidden layer to generate a binary classification change prediction map and output the final change result image.
[0068] Among them, the binary change prediction mapping is 0 or 1, 0 and 1 represent the labels of the unchanged area and the changed area, respectively. Correspondingly, the changed area is the land disturbance range that needs to be monitored in the substation project.
[0069] Furthermore, before the embodiment of the present invention processes the high-frequency component using the bilinear interpolation method, it also includes performing edge extraction processing on the high-frequency component using the Canny algorithm to obtain a new high-frequency component, and then processing the new high-frequency component using the bilinear interpolation method, which can effectively retain edge information and improve the interpolation speed.
[0070] Furthermore, in a preferred embodiment of the present invention, a ResNet-50 feature extraction network integrated with a lightweight channel attention module includes:
[0071] An input layer, the input layer is used to receive input images. In the embodiment of the present invention, the input images are high-resolution pre-phase images and post-phase images;
[0072] Residual convolution block: includes convolution, batch normalization and ReLU activation function;
[0073] Lightweight Channel Attention Module (CAM): added after the residual block;
[0074] Average pooling layer: Average pooling is performed after the residual block and CAM;
[0075] Fully connected layer: maps the pooled features.
[0076] Specifically, in the step of performing feature extraction processing on high-resolution images using a ResNet-50 feature extraction network integrated with a lightweight channel attention module, the front-phase image and the back-phase image are feature extracted by two feature extraction modules with shared parameters;
[0077] After the image is processed by convolution and pooling, the residual convolution block is used to extract multi-scale semantic features;
[0078] The residual convolution block consists of a ResNeXt unit with 2x downsampling and a 1×1 convolution, followed by a unified upsampling module to obtain the same feature map size as the first layer residual block. In this way, output features that combine high-level semantic information and low-level spatial information of different scales can be obtained;
[0079] When information at different scales is fused, data at the same scale is required to be fused, and upsampling and downsampling methods are used to keep the image data at the same resolution. Specifically, the scale calculation method is as follows:
[0080]
[0081] In the formula, padding represents the padding of feature data; kernel_size represents the size of the convolution kernel; stride represents the stride of each convolution movement; H out Represents the height of the feature map; W out Indicates the width of the feature map.
[0082] Furthermore, the transformed feature maps of the four layers are concatenated on the channel and input into a 3×3 convolutional layer to obtain the final feature map.
[0083] Further, in the embodiment of the present invention, the output of the cth convolutional layer in the nth residual convolutional block is expressed as:
[0084] F n,c =σ(W n,c [F n-1 , F n,1 , ..., F n,c-1 ]);
[0085] Where σ represents the ReLU activation function, W n,c Represents the weight value of the cth convolutional layer;
[0086] The output features of each layer of each residual convolution block are mapped and spliced. The output features of the previous residual convolution block and each layer of the output features of the current residual convolution block are connected to the subsequent layer. Based on this form, the local features can be well preserved.
[0087] Preferably, the embodiment of the present invention adopts three residual convolution blocks as the backbone of the feature extraction network. After obtaining the local information features of each residual convolution block, a global feature fusion and global residual learning are performed; the fully connected layer fuses the output features of each residual convolution block, and performs a global residual learning with the initial input features, which can effectively reduce information loss.
[0088] Furthermore, the embodiment of the present invention sends the fused features into a 3×3 convolution, and fuses the obtained features with themselves to obtain a final feature result. The final feature result F′ is expressed as:
[0089] F′=F cat +Conv(F cat );
[0090] In the formula, Conv represents a 3×3 convolution operation, and F cat It represents the result of global fusion and residual learning of features, which can effectively retain low-level features.
[0091] Furthermore, the step of fusing the change feature map into the decoder part by upsampling after channel splicing of the difference feature map of the present invention includes: the step of fusing the change feature map into the decoder part by upsampling after channel splicing of the difference feature map includes: among f1, f2, f3, f4 and f5 in the difference feature map, f3, f4 and f5 are upsampled twice respectively and spliced and fused with f2 to obtain p2, and then p2 is upsampled twice and spliced and fused with f1 to obtain p1.
[0092] Furthermore, in the step of decoding the feature map enhanced by feature differences, generating a binary change prediction map, and outputting the final change result image of the present invention, MLP decodes the upsampled feature map p1 to generate a change prediction map. The change prediction map is a two-dimensional image, in which the value of each pixel represents the possibility of a change at that position. The change prediction map is converted into a binary image through a threshold, in which pixels above the threshold are marked as 1, and pixels not above the threshold are marked as 0, 0 and 1 represent the labels of the unchanged area and the changed area, respectively. Correspondingly, the changed area is the land disturbance range that needs to be monitored in the substation project.
[0093] It should be noted that the above figures are only schematic illustrations of the processes included in the method according to an exemplary embodiment of the present invention, and are not intended to be limiting. It is easy to understand that the processes shown in the above figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.
[0094] It should be understood that, although described in a certain order, these steps are not necessarily performed in sequence in the above order. Unless there is clear explanation in this article, the execution of these steps does not have strict order restriction, and these steps can be performed in other orders. Moreover, a part of the steps of the present embodiment may include a plurality of steps or a plurality of stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a part of the steps or stages in other steps or other steps.
[0095] In a second aspect of the embodiments of the present invention, the present invention further provides a system for monitoring the environmental and water conservation land disturbance range of a substation, comprising:
[0096] An image acquisition module 201 is used to acquire a pre-phase image and a post-phase image of a substation engineering area using a drone;
[0097] The image processing module 202 is used to decompose the pre-phase image and the post-phase image into high-frequency and low-frequency components through wavelet transformation, process the high-frequency components using a bilinear interpolation method, process the low-frequency components using a bicubic interpolation method, and then perform an inverse wavelet transformation to output a high-resolution image;
[0098] The encoder module 203 is used to perform feature extraction processing on the high-resolution image using a ResNet-50 feature extraction network integrated with a lightweight channel attention module, and to perform difference and take absolute values of each extracted feature layer to obtain a difference feature map;
[0099] The decoder module 204 is used to fuse the difference feature map into the decoder part by upsampling the difference feature map after channel splicing. In the decoder, the feature map enhanced by feature difference is decoded through an MLP structure containing only one hidden layer to generate a binary classification change prediction map and output the final change result image.
[0100] The substation environmental water conservation and land disturbance range monitoring system of the present invention has broad application prospects in the field of substation engineering management, can significantly improve the efficiency and reliability of engineering safety construction management, reduce production interruptions and losses, and provide a higher level of support for substation projects.
[0101] According to a third aspect of an embodiment of the present invention, a computer device 300 is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, a method for monitoring the land disturbance range of environmental water conservation in a substation according to any one of the above embodiments is implemented.
[0102] The computer device includes a processor and a memory, and may also include an input system and an output system. The processor, memory, input system and output system may be connected via a bus or other means, and the input system may receive input digital or character information, and generate signal input related to the migration of the monitoring range of land disturbance for environmental protection and water conservation of substations. The output system may include display devices such as display screens.
[0103] As a non-volatile computer-readable storage medium, the memory can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as the program instructions / modules corresponding to the substation environmental water conservation land disturbance range monitoring method in the embodiment of the present application. The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required by at least one function; the data storage area may store data created by the use of the substation environmental water conservation land disturbance range monitoring method, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the local module via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
[0104] The processor may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip in some embodiments. The processor is generally used to control the overall operation of the computer device. In this embodiment, the processor is used to run the program code stored in the memory or process data. The processors of the multiple computer devices of the computer device of this embodiment execute various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory, that is, implementing the steps of the substation environmental water conservation land disturbance range monitoring method of the above method embodiment.
[0105] It will also be appreciated by those skilled in the art that various exemplary logic blocks, modules, circuits and algorithm steps described in conjunction with the disclosure herein can be implemented as electronic hardware, computer software or a combination of the two. In order to clearly illustrate this interchangeability of hardware and software, a general description has been given to the functions of various schematic components, blocks, modules, circuits and steps. Whether this function is implemented as software or hardware depends on specific applications and the design constraints imposed on the entire system. Those skilled in the art can implement the function in various ways for each specific application, but this implementation decision should not be interpreted as causing a departure from the disclosed scope of the embodiments of the present invention.
[0106] Finally, it should be noted that the computer-readable storage medium (e.g., memory) herein may be a volatile memory or a nonvolatile memory, or may include both a volatile memory and a nonvolatile memory. As an example and not by way of limitation, a nonvolatile memory may include a read-only memory (ROM), a programmable ROM (PROM), an electrically programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), or a flash memory. A volatile memory may include a random access memory (RAM), which may act as an external cache memory. As an example and not by way of limitation, RAM may be obtained in a variety of forms, such as synchronous RAM (DRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct Rambus RAM (DRRAM). The storage devices of the disclosed aspects are intended to include, but are not limited to, these and other suitable types of memory.
[0107] The various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure herein may be implemented or executed using the following components designed to perform the functions herein: a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination of these components. A general purpose processor may be a microprocessor, but alternatively, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP, and / or any other such configuration.
[0108] The above are exemplary embodiments disclosed in the present invention, but it should be noted that various changes and modifications may be made without departing from the scope disclosed in the embodiments of the present invention as defined in the claims. The functions, steps and / or actions of the method claims according to the disclosed embodiments described herein do not need to be performed in any particular order. In addition, although the elements disclosed in the embodiments of the present invention may be described or required in individual form, they may also be understood as multiple unless explicitly limited to the singular.
[0109] It should be understood that, as used herein, the singular form "a" or "an" is intended to include the plural form as well, unless the context clearly supports an exception. It should also be understood that, as used herein, "and / or" refers to any and all possible combinations of one or more of the items listed in association. The serial numbers of the embodiments disclosed in the above embodiments of the present invention are for description only and do not represent the advantages and disadvantages of the embodiments.
[0110] A person skilled in the art should understand that the discussion of any of the above embodiments is only exemplary and is not intended to imply that the scope of the disclosure of the embodiments of the present invention (including the claims) is limited to these examples; under the concept of the embodiments of the present invention, the technical features in the above embodiments or different embodiments can also be combined, and there are many other changes in different aspects of the embodiments of the present invention as above, which are not provided in detail for the sake of simplicity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present invention should be included in the protection scope of the embodiments of the present invention.
Claims
1. A method for monitoring the disturbance range of land for water conservation in a substation, characterized in that: The method comprises the following steps: Step S101: using a drone to obtain a pre-phase image and a post-phase image of a substation project area; Step S102: The pre-phase image and the post-phase image are decomposed into high-frequency and low-frequency components by wavelet transform, the high-frequency components are processed by bilinear interpolation method, and the low-frequency components are processed by bicubic interpolation method, and then inverse wavelet transform is performed to output a high-resolution image; Step S103: In the encoder, a ResNet-50 feature extraction network integrated with a lightweight channel attention module is used to perform feature extraction processing on the high-resolution image, and each feature layer extracted is subtracted and the absolute value is taken to obtain a difference feature map; Step S104: The difference feature map is merged into the decoder part by upsampling after channel splicing. In the decoder, the feature map enhanced by feature difference is decoded through an MLP structure containing only one hidden layer to generate a binary classification change prediction map and output the final change result image.
2. The method for monitoring the disturbance range of substation water conservation and environment according to claim 1 is characterized in that: Before the high frequency component is processed by the bilinear interpolation method, the high frequency component is also processed by edge extraction using the Canny algorithm to obtain a new high frequency component, and then the new high frequency component is processed by the bilinear interpolation method.
3. The method for monitoring the disturbance range of substation water conservation and environment according to claim 2 is characterized in that: The ResNet-50 feature extraction network with a lightweight channel attention module includes: An input layer, the input layer is used to receive input images. In the embodiment of the present invention, the input images are high-resolution pre-phase images and post-phase images; Residual convolution block: includes convolution, batch normalization and ReLU activation function; Lightweight channel attention module: added after the residual block; Average pooling layer: Average pooling is performed after the residual block and CAM; Fully connected layer: maps the pooled features.
4. The method for monitoring the disturbance range of substation water conservation and environment according to claim 3 is characterized in that: In the step of extracting features from high-resolution images using a ResNet-50 feature extraction network fused with a lightweight channel attention module, features of the pre-phase image and the post-phase image are extracted by two feature extraction modules with shared parameters; After the image is processed by convolution and pooling, the residual convolution block is used to extract multi-scale semantic features; The residual convolution block consists of a ResNeXt unit with 2x downsampling and a 1×1 convolution, followed by a unified upsampling module to obtain the same feature map size as the first layer residual block, and obtain output features that combine high-level semantic information and low-level spatial information of different scales; The transformed feature maps of the four layers are concatenated on the channel and input into a 3×3 convolutional layer to obtain the final feature map.
5. The method for monitoring the disturbance range of land for water conservation and environmental protection in a substation according to claim 4 is characterized in that: When fusing information at different scales, it is required to fuse data at the same scale, and use upsampling and downsampling methods to keep the image data at the same resolution. Specifically, the scale calculation method is as follows: In the formula, padding represents the padding of feature data; kernel_size represents the size of the convolution kernel; Stride represents the stride of each convolution movement; H out Represents the height of the feature map; W out Indicates the width of the feature map.
6. The method for monitoring the disturbance range of land for water conservation and environmental protection in a substation according to claim 5 is characterized in that: The output of the cth convolutional layer in the nth residual convolutional block is expressed as: F n,c =σ(W n,c [F n-1 ,F n,1 ,...,F n,c-1 ]); Where σ represents the ReLU activation function, W n,c Represents the weight value of the cth convolutional layer; The output features of each layer of each residual convolution block are mapped and concatenated, and each layer of the output features of the previous residual convolution block and the output features of the current residual convolution block are connected to the subsequent layers.
7. The method for monitoring the disturbance range of land for water conservation and environmental protection in a substation according to claim 6 is characterized in that: Three residual convolution blocks are used as the backbone of the feature extraction network. After obtaining the local information features of each residual convolution block, a global feature fusion and global residual learning are performed; the fully connected layer fuses the output features of each residual convolution block and performs a global residual learning with the initial input features; The fused features are sent to a 3×3 convolution, and the obtained features are fused with themselves to obtain the final feature result. The final feature result F′ is expressed as: F′=F cat +ConV(F cat ); In the formula, Conv represents a 3×3 convolution operation, and F cat It represents the result of global fusion and residual learning of features.
8. The method for monitoring the disturbance range of land for environmental protection and water conservation in a substation according to claim 7 is characterized in that: The step of fusing the change feature map into the decoder part by upsampling the difference feature map after channel splicing includes: among f1, f2, f3, f4 and f3 in the difference feature map, f3, f4 and f5 are upsampled twice respectively and spliced and fused with f2 to obtain p2, and then p2 is upsampled twice and spliced and fused with f1 to obtain p1.
9. The method for monitoring the disturbance range of land for environmental protection and water conservation in a substation according to claim 2 is characterized in that: In the step of decoding the feature map enhanced by feature difference, generating a binary change prediction map, and outputting the final change result image, MLP decodes the upsampled feature map p1 to generate a change prediction map. The change prediction map is a two-dimensional image, in which the value of each pixel represents the possibility of change at that position. The change prediction map is converted into a binary image through a threshold, in which pixels above the threshold are marked as 1, and pixels not higher than the threshold are marked as 0.
10. A monitoring system for implementing the method for monitoring the disturbance range of land for environmental protection and water conservation in a substation as claimed in any one of claims 1 to 9, characterized in that: The monitoring system includes: An image acquisition module is used to acquire pre-phase images and post-phase images of the substation engineering area using a drone; An image processing module is used to decompose the pre-phase image and the post-phase image into high-frequency and low-frequency components through wavelet transformation, process the high-frequency components using a bilinear interpolation method, process the low-frequency components using a bicubic interpolation method, and then perform an inverse wavelet transformation to output a high-resolution image; The encoder module is used to perform feature extraction processing on high-resolution images using a ResNet-50 feature extraction network integrated with a lightweight channel attention module, subtract each feature layer extracted and take the absolute value to obtain a difference feature map; The decoder module is used to fuse the change feature map into the decoder part by upsampling the difference feature map after channel splicing. In the decoder, the feature map enhanced by feature difference is decoded through an MLP structure containing only one hidden layer to generate a binary classification change prediction map and output the final change result image.
Citation Information
Patent Citations
Lightweight remote sensing image super-resolution reconstruction method, device and equipment
CN115393191A
Traffic sign image segmentation algorithm based on improved space-time diagram convolution
CN115457509A
Remote sensing image change detection method fusing twinborn coding and decoding and attention mechanism
CN117953369A
Vortex spinning broken yarn intelligent detection method and system based on image analysis
CN117994236A
Remote-sensing picture interpolation method based on small wave fractal
CN1892698A
Cited By
Power transmission line tower footing land disturbance automatic extraction method based on full convolutional neural network
CN121259348A
