CN113887466A - Method, device, equipment and medium for identifying hidden danger of water body in power transmission line corridor - Google Patents
Method, device, equipment and medium for identifying hidden danger of water body in power transmission line corridor Download PDFInfo
- Publication number
- CN113887466A CN113887466A CN202111197253.9A CN202111197253A CN113887466A CN 113887466 A CN113887466 A CN 113887466A CN 202111197253 A CN202111197253 A CN 202111197253A CN 113887466 A CN113887466 A CN 113887466A
- Authority
- CN
- China
- Prior art keywords
- model
- transmission line
- hrnet
- power transmission
- identifying
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 50
- 230000005540 biological transmission Effects 0.000 title claims abstract description 47
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 43
- 238000012549 training Methods 0.000 claims abstract description 41
- 238000012360 testing method Methods 0.000 claims abstract description 39
- 230000004927 fusion Effects 0.000 claims abstract description 33
- 230000002776 aggregation Effects 0.000 claims abstract description 15
- 238000004220 aggregation Methods 0.000 claims abstract description 15
- 238000005070 sampling Methods 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 8
- 238000010191 image analysis Methods 0.000 claims description 6
- 230000001965 increasing effect Effects 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 12
- 230000010339 dilation Effects 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 239000002904 solvent Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000002352 surface water Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Entrepreneurship & Innovation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Tourism & Hospitality (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Water Supply & Treatment (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Public Health (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method, a device, equipment and a medium for identifying potential water hazards of a power transmission line corridor, wherein the method comprises the steps of obtaining a satellite remote sensing image sample, and dividing the satellite remote sensing image sample into a training set and a testing set according to a preset proportion; constructing a first HRNet model, and training by using a training set to generate a second HRNet model; testing the second HRNet model by using the test set, and generating a target HRNet model when the test result meets a preset condition; and identifying the water hidden danger of the power transmission line by using the target HRNet model to generate an identification result. According to the invention, by introducing the characteristic pyramid hierarchical structure into the HRNet network, high-level and low-level semantic characteristics can be simultaneously explored, the capability of multilayer semantic representation of the model is enhanced, the problem of grid is avoided, and by adopting the fusion expansion convolution module and the multi-level characteristic data aggregation upsampling module, the diversity of receptive fields is increased, and the accuracy of the recognition result is improved.
Description
Technical Field
The invention relates to the technical field of power transmission line inspection, in particular to a method, a device, equipment and a medium for identifying hidden danger of a power transmission line corridor water body.
Background
The high-voltage transmission line is an important component in a power grid system, and the tower is one of carriers of the transmission line, so that the tower plays an important role in maintaining the safe and stable operation of the transmission line. When the extreme weather such as heavy rain is encountered, landslide, debris flow, reservoir dike breaking and the like often occur, and the tower close to the water body has great potential safety hazard and needs to be checked in time. Because the towers forming the transmission line are numerous, the distance between adjacent towers is long, the surface water resource environment is complex, and the manual on-site surveying method is not only low in efficiency, but also has certain dangers.
With the development of science and technology, the method for checking hidden dangers by combining the remote sensing technology with the deep learning technology becomes an effective method for acquiring the ground feature information. Among them, there are two common methods: one is a full convolution neural network DL-Unet improved based on U-net, which realizes effective segmentation of different types of ground objects in remote sensing images. The network improves the traditional convolution mode, introduces expansion convolution, increases the receptive field without increasing network parameters, however, the method easily causes the problem of 'grid', and a large number of expansion convolution layers are needed to obtain a sufficiently rich multi-scale characteristic diagram. The second mode is that the improved network based on GLNet and HRNet is used for semantic segmentation of high-resolution remote sensing images. The HRNet is used for replacing an original ResNet main body in a global branch, so that the characteristic diagram with stronger representation capability and higher resolution is obtained, but the traditional HRNet structure needs a large amount of repeated multi-scale fusion, so that the calculation parameter quantity is required to be larger, and the calculation complexity is greatly improved. Meanwhile, when the features at the final stage of the network are aggregated, if the feature map with smaller resolution is gradually enlarged only by simple bilinear upsampling, some detail features in the image are easily lost, and the accuracy of the identification result is further influenced.
Disclosure of Invention
The invention aims to provide a method, a device, equipment and a medium for identifying water hidden troubles of a power transmission line corridor, and aims to solve the problems of large calculation amount, high complexity, easy occurrence of gridding and low accuracy of identification results in the existing method for identifying the water hidden troubles of the power transmission line corridor.
In order to achieve the purpose, the invention provides a method for identifying hidden danger of a water body of a power transmission line, which comprises the following steps:
obtaining a satellite remote sensing image sample, and dividing the sample into a training set and a testing set according to a preset proportion;
constructing a first HRNet model, and training the first HRNet model by using the training set to generate a second HRNet model;
testing the second HRNet model by using the test set, and generating a target HRNet model when a test result meets a preset condition;
and identifying the water hidden danger of the power transmission line by using the target HRNet model to generate an identification result.
Further, preferably, the constructing the first HRNet model includes:
building a model frame by adopting a pyramid characteristic hierarchical structure, and using staggered convolution as a down-sampling method; the pyramid feature hierarchical structure is a four-layer structure, and multi-layer feature fusion can be performed between layers.
Further, preferably, the constructing the first HRNet model further includes:
and adopting a fusion expansion convolution module in the convolution layer, wherein the fusion expansion convolution module is integrated with a serial expansion convolution block and a parallel expansion convolution kernel.
Further, preferably, the constructing the first HRNet model further includes:
and constructing a multi-stage feature data aggregation up-sampling module to perform feature fusion of adjacent feature maps from the bottom layer to the top layer.
The invention also provides a device for identifying the hidden danger of the water body of the power transmission line, which comprises:
the system comprises a sample acquisition unit, a satellite remote sensing image analysis unit and a satellite remote sensing image analysis unit, wherein the sample acquisition unit is used for acquiring a satellite remote sensing image sample and dividing the sample into a training set and a testing set according to a preset proportion;
the training unit is used for constructing a first HRNet model, training the first HRNet model by using the training set and generating a second HRNet model;
the test unit is used for testing the second HRNet model by using the test set, and generating a target HRNet model when a test result meets a preset condition;
and the identification unit is used for identifying the water body hidden danger of the power transmission line by using the target HRNet model to generate an identification result.
Further, preferably, the training unit is further configured to:
building a model frame by adopting a pyramid characteristic hierarchical structure, and using staggered convolution as a down-sampling method; the pyramid feature hierarchical structure is a four-layer structure, and multi-layer feature fusion can be performed between layers.
Further, preferably, the training unit is further configured to:
and adopting a fusion expansion convolution module in the convolution layer, wherein the fusion expansion convolution module is integrated with a serial expansion convolution block and a parallel expansion convolution kernel.
Further, preferably, the training unit is further configured to:
and constructing a multi-stage feature data aggregation up-sampling module to perform feature fusion of adjacent feature maps from the bottom layer to the top layer.
The present invention also provides a terminal device, including:
one or more processors;
a memory coupled to the processor for storing one or more programs;
when the one or more programs are executed by the one or more processors, the one or more processors are enabled to implement the method for identifying the hidden danger of the transmission line water body.
The invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method for identifying the water body hidden danger of the power transmission line is realized.
Compared with the prior art, the invention has the beneficial effects that:
the invention discloses a method, a device, equipment and a medium for identifying hidden danger of a water body in a power transmission line corridor, wherein the method comprises the following steps: obtaining a satellite remote sensing image sample, and dividing the sample into a training set and a testing set according to a preset proportion; constructing a first HRNet model, and training the first HRNet model by using the training set to generate a second HRNet model; testing the second HRNet model by using the test set, and generating a target HRNet model when a test result meets a preset condition; and identifying the water hidden danger of the power transmission line by using the target HRNet model to generate an identification result.
According to the invention, by introducing the characteristic pyramid hierarchical structure into the HRNet network, high-level and low-level semantic characteristics can be simultaneously explored, the multilayer semantic representation capability of the model is enhanced, and rich multilayer semantic information is transmitted to the model, so that the problem of homogeneity among classes is solved, and the problem of 'gridding' is avoided. In addition, the diversity of the receptive field is increased by adopting the fusion expansion convolution module and the multi-stage characteristic data aggregation up-sampling module, so that the capability of a network model for finely identifying detailed characteristics is enhanced, and the accuracy of an identification result is improved.
Drawings
In order to more clearly illustrate the technical solution of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a method for identifying a water body hidden danger of a power transmission line according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an improved HRNet model network according to an embodiment of the present invention;
FIG. 3 is a block diagram of a fusion dilation convolution module according to an embodiment of the present invention;
FIG. 4 is a block diagram of a multi-level feature data aggregation upsampling module according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a device for identifying a water body hidden danger of a power transmission line according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of a terminal device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be understood that the step numbers used herein are for convenience of description only and are not intended as limitations on the order in which the steps are performed.
It is to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
The terms "comprises" and "comprising" indicate the presence of the described features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
The term "and/or" refers to and includes any and all possible combinations of one or more of the associated listed items.
Referring to fig. 1, an embodiment of the present invention provides a method for identifying a water body hidden danger of a power transmission line. As shown in fig. 1, the method for identifying the water body hidden danger of the power transmission line includes steps S10 to S40. The method comprises the following steps:
and S10, obtaining a satellite remote sensing image sample, and dividing the sample into a training set and a testing set according to a preset proportion.
In this step, firstly, the acquired satellite remote sensing image sample is divided into a training set and a testing set according to a preset proportion, and the training set and the testing set can be generally selected to respectively account for 70% and 30% of the total sample. Because the obtained initial satellite remote sensing image sample usually has more serious noise interference, in an optional embodiment, the initial sample can be removed by an image preprocessing mode to obtain a sample with higher quality. It should be noted that the image preprocessing method herein is a preprocessing method commonly used in the prior art, such as de-grayscale, image, etc. And will not be described in detail herein. The training set and the testing set in this embodiment respectively account for 70% and 30% of the total sample, which is only a preferred mode, and can be set according to the needs of the scene in practical application, and are not limited herein.
S20, constructing a first HRNet model, and training the first HRNet model by using the training set to generate a second HRNet model.
Specifically, in a specific embodiment, constructing the first HRNet model includes the following:
1) building a model frame by adopting a pyramid characteristic hierarchical structure, and using staggered convolution as a down-sampling method; the pyramid feature hierarchical structure is a four-layer structure, and multi-layer feature fusion can be performed between layers.
Specifically, a network structure diagram of the improved HRNet model is shown in fig. 2. On the basis of the traditional HRNet network, the advantages of the HRNet network are inherited, 4 feature graphs with different resolutions are generated by adopting a parallel method, high-resolution features in each layer are well reserved, and more accurate semantic prediction is generated potentially.
Further, in the embodiment, a pyramid feature hierarchical structure is introduced, the structure adopts a 4-layer top-down architecture, staggered convolution is applied as a down-sampling method, the step size is set to 2, after each down-sampling, the height and width of the feature map are reduced by half, and the number of corresponding channels is doubled compared with the previous one. The first-level feature diagram branch of the pyramid is directly added into the main high-resolution feature diagram, and multi-level feature fusion is carried out between the second-level feature diagram and the third-level feature diagram and corresponding outputs from various resolution branches.
2) And adopting a fusion expansion convolution module in the convolution layer, wherein the fusion expansion convolution module is integrated with a serial expansion convolution block and a parallel expansion convolution kernel.
It should be noted that, in order to reduce the complexity of the calculation and achieve a higher pixel-level accuracy based on a shallower semantic segmentation network, a new fusion expansion convolution module (FDC) is applied to the feature map extracted from each layer of the original HRNet model in this embodiment, so as to increase the diversity of the receptive field.
As shown in fig. 3, each block contains 3 dilation convolution kernels ( dilation rate 1, 2 and 5 respectively) applied in parallel to the input feature map, each dilation convolution block having a bypass connection that can further compensate for the detail features of the previous feature map. In addition, in the module, a serial expansion convolution block and a parallel expansion convolution kernel are integrated, so that the depth of a neural network can be minimized by using fewer expansion convolution layers, and a wider receptive field can be provided to realize multi-scale sampling. If the fusion dilation convolution module has m successive dilation blocks, each dilation block contains n dilation convolution kernels, and finally (n +1) × m different receptive fields can be obtained.
3) And constructing a multi-stage feature data aggregation up-sampling module to perform feature fusion of adjacent feature maps from the bottom layer to the top layer.
It should be noted that, a simple bilinear upsampling is adopted in the conventional feature aggregation method, the feature maps with smaller resolutions are sequentially enlarged, and the aggregation method between two adjacent layers is represented as follows:
in the formula (I), the compound is shown in the specification,
representing element summation, B representing bilinear upsampling, XiAnd Xi+1Respectively extracting characteristic graphs of the ith layer and the (i +1) th layer. If data dependency is not taken into account, the recovery of higher resolution prediction maps by purely using bilinear upsampling may result in the loss of fine detail.
Therefore, on the basis of generating strong feature map diversity, the embodiment provides an upsampling method based on multi-level feature data aggregation, the advantage of spatial correlation is fully utilized to improve the semantic reconstruction capability, and the model structure is as shown in fig. 4. In particular, the amount of the solvent to be used,
assuming that G is a true label with the same resolution as the input image, the loss function is minimized by:
in the formula, the multi-level feature data aggregation up-sampling is represented by a learnable transformation matrix T, DiLearnable upsampling, G, representing the ith layeri-1Is obtained by compressing G according to the image resolution of the (i-1) th layer. The feature fusion method between two adjacent layers is expressed as follows:
finally, all neighboring feature maps from the bottom layer to the top layer can be utilized to achieve more accurate pixel-level semantic prediction.
L(P,G)＝Lce(softmax(FD(Xp，Xp+1)),G)；
In the formula, XpAnd Xp+1Respectively representing the feature maps extracted from the two top levels, P being the final prediction map, LceRepresenting a cross entropy loss function.
And S30, testing the second HRNet model by using the test set, and generating a target HRNet model when the test result meets a preset condition. It can be understood that the purpose of step S30 is to test the training result, and when the preset condition is met, the model is proved to be trained well and to achieve satisfactory effect, the training result can be used for subsequent recognition; and when the preset condition is not met, proving that the model needs to be trained again until the preset condition is met. It should be noted that the preset condition is mainly defined according to the scene requirement, for example, "the error range of the output recognition result does not exceed 5%, and so on, so that different application environments may correspond to different preset conditions, and the content of the preset condition is not limited at all.
And S40, identifying the water body hidden danger of the power transmission line by using the target HRNet model, and generating an identification result.
In summary, according to the method for identifying the water body hidden danger of the power transmission line provided by the embodiment of the invention, the feature pyramid hierarchical structure is introduced into the HRNet network, so that high-level and low-level semantic features can be simultaneously explored, the capability of multilayer semantic representation of the model is enhanced, and rich multilayer semantic information is transmitted to the model, so that the problem of homogeneity among classes is solved, and the problem of 'grid' is avoided. In addition, the embodiment of the invention also adopts a fusion expansion convolution module and a multi-stage characteristic data aggregation up-sampling module to increase the diversity of the receptive field, thereby enhancing the capability of a network model for finely identifying the detailed characteristics and improving the accuracy of the identification result.
Referring to fig. 5, an embodiment of the present invention further provides a device for identifying hidden danger in a water body of a power transmission line, including:
the system comprises a sample acquisition unit 01, a satellite remote sensing image analysis unit and a satellite remote sensing image analysis unit, wherein the sample acquisition unit 01 is used for acquiring a satellite remote sensing image sample and dividing the sample into a training set and a test set according to a preset proportion;
the training unit 02 is used for constructing a first HRNet model, training the first HRNet model by using the training set and generating a second HRNet model;
the test unit 03 is configured to test the second HRNet model by using the test set, and generate a target HRNet model when a test result meets a preset condition;
and the identification unit 04 is used for identifying the water body hidden danger of the power transmission line by using the target HRNet model and generating an identification result.
In a specific embodiment, the training unit 02 is further configured to:
building a model frame by adopting a pyramid characteristic hierarchical structure, and using staggered convolution as a down-sampling method; the pyramid feature hierarchical structure is a four-layer structure, and multi-layer feature fusion can be performed between layers.
In a specific embodiment, the training unit 02 is further configured to:
and adopting a fusion expansion convolution module in the convolution layer, wherein the fusion expansion convolution module is integrated with a serial expansion convolution block and a parallel expansion convolution kernel.
In a specific embodiment, the training unit 02 is further configured to:
and constructing a multi-stage feature data aggregation up-sampling module to perform feature fusion of adjacent feature maps from the bottom layer to the top layer.
It can be understood that the device for identifying the water body hidden danger of the power transmission line provided by the embodiment of the invention is used for executing the method for identifying the water body hidden danger of the power transmission line according to any one of the embodiments. According to the embodiment of the invention, the characteristic pyramid hierarchical structure is introduced into the HRNet network, so that high-level and low-level semantic characteristics can be explored simultaneously, the multilayer semantic representation capability of the model is enhanced, rich multilayer semantic information is transmitted to the model, the problem of homogeneity among classes is solved, and the problem of 'gridding' is avoided. In addition, the embodiment of the invention also adopts a fusion expansion convolution module and a multi-stage characteristic data aggregation up-sampling module to increase the diversity of the receptive field, thereby enhancing the capability of a network model for finely identifying the detailed characteristics and improving the accuracy of the identification result.
Referring to fig. 6, an embodiment of the present invention provides a terminal device, including:
one or more processors;
a memory coupled to the processor for storing one or more programs;
when the one or more programs are executed by the one or more processors, the one or more processors implement the method for identifying the water body hidden danger of the power transmission line.
The processor is used for controlling the overall operation of the terminal equipment so as to complete all or part of the steps of the method for identifying the water body hidden danger of the power transmission line. The memory is used to store various types of data to support operation at the terminal device, and these data may include, for example, instructions for any application or method operating on the terminal device, as well as application-related data. The Memory may be implemented by any type of volatile or non-volatile Memory device or combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic Memory, flash Memory, magnetic disk, or optical disk.
In an exemplary embodiment, the terminal Device may be implemented by one or more Application Specific 1 integrated circuits (AS 1C), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a controller, a microcontroller, a microprocessor, or other electronic components, and is configured to perform the method for identifying the water body hazard of the power transmission line according to any one of the above embodiments, and achieve the technical effects consistent with the above method.
In another exemplary embodiment, a computer readable storage medium including a computer program is further provided, and the computer program is executed by a processor to implement the steps of the method for identifying a water hidden danger of a power transmission line according to any one of the above embodiments. For example, the computer-readable storage medium may be the above-mentioned memory including a computer program, and the above-mentioned computer program may be executed by a processor of a terminal device to complete the method for identifying a water body potential hazard of a power transmission line according to any one of the above-mentioned embodiments, and achieve the technical effects consistent with the above-mentioned method.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.
Claims (10)
1. A method for identifying hidden danger of a water body of a power transmission line is characterized by comprising the following steps:
obtaining a satellite remote sensing image sample, and dividing the sample into a training set and a testing set according to a preset proportion;
constructing a first HRNet model, and training the first HRNet model by using the training set to generate a second HRNet model;
testing the second HRNet model by using the test set, and generating a target HRNet model when a test result meets a preset condition;
and identifying the water hidden danger of the power transmission line by using the target HRNet model to generate an identification result.
2. The method for identifying the hidden danger of the water body of the power transmission line according to claim 1, wherein the constructing of the first HRNet model comprises the following steps:
building a model frame by adopting a pyramid characteristic hierarchical structure, and using staggered convolution as a down-sampling method; the pyramid feature hierarchical structure is a four-layer structure, and multi-layer feature fusion can be performed between layers.
3. The method for identifying the hidden danger of the water body of the power transmission line according to claim 1, wherein the constructing of the first HRNet model further comprises the following steps:
and adopting a fusion expansion convolution module in the convolution layer, wherein the fusion expansion convolution module is integrated with a serial expansion convolution block and a parallel expansion convolution kernel.
4. The method for identifying the hidden danger of the water body of the power transmission line according to claim 1, wherein the constructing of the first HRNet model further comprises the following steps:
and constructing a multi-stage feature data aggregation up-sampling module to perform feature fusion of adjacent feature maps from the bottom layer to the top layer.
5. The utility model provides a transmission line water hidden danger recognition device which characterized in that includes:
the system comprises a sample acquisition unit, a satellite remote sensing image analysis unit and a satellite remote sensing image analysis unit, wherein the sample acquisition unit is used for acquiring a satellite remote sensing image sample and dividing the sample into a training set and a testing set according to a preset proportion;
the training unit is used for constructing a first HRNet model, training the first HRNet model by using the training set and generating a second HRNet model;
the test unit is used for testing the second HRNet model by using the test set, and generating a target HRNet model when a test result meets a preset condition;
and the identification unit is used for identifying the water body hidden danger of the power transmission line by using the target HRNet model to generate an identification result.
6. The device for identifying the hidden danger of the water body of the power transmission line according to claim 5, wherein the training unit is further configured to:
building a model frame by adopting a pyramid characteristic hierarchical structure, and using staggered convolution as a down-sampling method; the pyramid feature hierarchical structure is a four-layer structure, and multi-layer feature fusion can be performed between layers.
7. The device for identifying the hidden danger of the water body of the power transmission line according to claim 5, wherein the training unit is further configured to:
and adopting a fusion expansion convolution module in the convolution layer, wherein the fusion expansion convolution module is integrated with a serial expansion convolution block and a parallel expansion convolution kernel.
8. The device for identifying the hidden danger of the water body of the power transmission line according to claim 5, wherein the training unit is further configured to:
and constructing a multi-stage feature data aggregation up-sampling module to perform feature fusion of adjacent feature maps from the bottom layer to the top layer.
9. A terminal device, comprising:
one or more processors;
a memory coupled to the processor for storing one or more programs;
when executed by the one or more processors, cause the one or more processors to implement the method of identifying a water body potential hazard of a power transmission line as recited in any one of claims 1-4.
10. A computer-readable storage medium, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the method for identifying a water body potential hazard of a power transmission line according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111197253.9A CN113887466A (en) | 2021-10-14 | 2021-10-14 | Method, device, equipment and medium for identifying hidden danger of water body in power transmission line corridor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111197253.9A CN113887466A (en) | 2021-10-14 | 2021-10-14 | Method, device, equipment and medium for identifying hidden danger of water body in power transmission line corridor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113887466A true CN113887466A (en) | 2022-01-04 |
Family
ID=79002777
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111197253.9A Pending CN113887466A (en) | 2021-10-14 | 2021-10-14 | Method, device, equipment and medium for identifying hidden danger of water body in power transmission line corridor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113887466A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111160276A (en) * | 2019-12-31 | 2020-05-15 | 重庆大学 | U-shaped cavity full-volume integral cutting network identification model based on remote sensing image |
| CN111738561A (en) * | 2020-05-29 | 2020-10-02 | 中国电建集团华东勘测设计研究院有限公司 | Regional water and soil loss dynamic monitoring method based on deep learning |
| CN112070069A (en) * | 2020-11-10 | 2020-12-11 | 支付宝(杭州)信息技术有限公司 | Method and device for identifying remote sensing image |
| CN112560706A (en) * | 2020-12-18 | 2021-03-26 | 广东电网有限责任公司电力科学研究院 | Method and device for identifying water body target of multi-source satellite image |
| CN113343883A (en) * | 2021-06-22 | 2021-09-03 | 长光卫星技术有限公司 | Port ore pile segmentation method based on improved HRNetV2 network |
-
2021
- 2021-10-14 CN CN202111197253.9A patent/CN113887466A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111160276A (en) * | 2019-12-31 | 2020-05-15 | 重庆大学 | U-shaped cavity full-volume integral cutting network identification model based on remote sensing image |
| CN111738561A (en) * | 2020-05-29 | 2020-10-02 | 中国电建集团华东勘测设计研究院有限公司 | Regional water and soil loss dynamic monitoring method based on deep learning |
| CN112070069A (en) * | 2020-11-10 | 2020-12-11 | 支付宝(杭州)信息技术有限公司 | Method and device for identifying remote sensing image |
| CN112560706A (en) * | 2020-12-18 | 2021-03-26 | 广东电网有限责任公司电力科学研究院 | Method and device for identifying water body target of multi-source satellite image |
| CN113343883A (en) * | 2021-06-22 | 2021-09-03 | 长光卫星技术有限公司 | Port ore pile segmentation method based on improved HRNetV2 network |
Non-Patent Citations (3)
| Title |
|---|
| FANG WANG 等: "CSE-HRNet: A Context and Semantic Enhanced High-Resolution Network for Semantic Segmentation of Aerial Imagery", 《IEEE ACCESS》, vol. 8, 16 October 2020 (2020-10-16), pages 182481 * |
| HUISI WU 等: "Optimized HRNet for image semantic segmentation", 《EXPERT SYSTEMS WITH APPLICATIONS》, 29 December 2020 (2020-12-29), pages 2 - 3 * |
| 王彤 等: "基于改进 U-net语义分割遥感影像的线路走廊隐患检测方法", 《南方电网技术》, vol. 13, no. 8, 31 August 2019 (2019-08-31), pages 67 - 73 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113298818B (en) | Remote sensing image building segmentation method based on attention mechanism and multi-scale features | |
| CN113160234B (en) | Unsupervised remote sensing image semantic segmentation method based on super-resolution and domain self-adaptation | |
| CN113780149B (en) | Remote sensing image building target efficient extraction method based on attention mechanism | |
| CN117078943B (en) | Remote sensing image road segmentation method integrating multi-scale features and double-attention mechanism | |
| CN116524361A (en) | Remote sensing image change detection network and detection method based on double twin branches | |
| CN112862774A (en) | Accurate segmentation method for remote sensing image building | |
| CN112287983B (en) | Remote sensing image target extraction system and method based on deep learning | |
| CN117237740B (en) | SAR image classification method based on CNN and Transformer | |
| CN115859094B (en) | Method for automatically identifying same-earthquake landslide by coupling convolution and entropy importance graph neural network | |
| CN113762396A (en) | Two-dimensional image semantic segmentation method | |
| Meng et al. | A block object detection method based on feature fusion networks for autonomous vehicles | |
| CN116863347A (en) | High-efficiency and high-precision remote sensing image semantic segmentation method and application | |
| Liu et al. | A new multi-channel deep convolutional neural network for semantic segmentation of remote sensing image | |
| CN113837931B (en) | Transformation detection method and device for remote sensing image, electronic equipment and storage medium | |
| Gu et al. | Advances in Rapid Damage Identification Methods for Post-Disaster Regional Buildings Based on Remote Sensing Images: A Survey | |
| CN113887466A (en) | Method, device, equipment and medium for identifying hidden danger of water body in power transmission line corridor | |
| CN111353441A (en) | Road extraction method and system based on position data fusion | |
| CN116611725A (en) | Land type identification method and device based on green ecological index | |
| CN116051850A (en) | Neural network target detection method, device, medium and embedded electronic equipment | |
| CN114723699A (en) | Intelligent identification method and system for detecting underground hidden danger of road by geological radar | |
| Wang et al. | Lightweight land cover classification via semantic segmentation of remote sensing imagery and analysis of influencing factors | |
| Wang | Remote sensing image semantic segmentation network based on ENet | |
| CN118135239B (en) | Fusion filtering multi-scale high-resolution remote sensing glacier extraction method | |
| CN118470333B (en) | Geographic environment semantic segmentation method and system based on remote sensing image | |
| Li et al. | Infrared Small Target Detection Algorithm Based on ISTD-CenterNet. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220104 |