CN116740654B - Substation operation prevention and control method based on image recognition technology - Google Patents
Substation operation prevention and control method based on image recognition technology Download PDFInfo
- Publication number
- CN116740654B CN116740654B CN202311015674.4A CN202311015674A CN116740654B CN 116740654 B CN116740654 B CN 116740654B CN 202311015674 A CN202311015674 A CN 202311015674A CN 116740654 B CN116740654 B CN 116740654B
- Authority
- CN
- China
- Prior art keywords
- image
- equipment
- texture feature
- training
- texture
- 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.)
- Active
Links
- 230000002265 prevention Effects 0.000 title claims abstract description 44
- 238000005516 engineering process Methods 0.000 title claims abstract description 40
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000010191 image analysis Methods 0.000 claims abstract description 23
- 239000011159 matrix material Substances 0.000 claims description 53
- 239000013598 vector Substances 0.000 claims description 48
- 238000012549 training Methods 0.000 claims description 46
- 238000011161 development Methods 0.000 claims description 20
- 230000007246 mechanism Effects 0.000 claims description 20
- 238000013527 convolutional neural network Methods 0.000 claims description 14
- 238000000605 extraction Methods 0.000 claims description 14
- 238000003062 neural network model Methods 0.000 claims description 7
- 230000001902 propagating effect Effects 0.000 claims 1
- 230000002159 abnormal effect Effects 0.000 abstract description 11
- 238000007689 inspection Methods 0.000 abstract description 10
- 238000001514 detection method Methods 0.000 description 19
- 230000006870 function Effects 0.000 description 11
- 230000018109 developmental process Effects 0.000 description 10
- 238000012423 maintenance Methods 0.000 description 10
- 230000005540 biological transmission Effects 0.000 description 8
- 238000012545 processing Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000012546 transfer Methods 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 239000000306 component Substances 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000004880 explosion Methods 0.000 description 3
- 238000007792 addition Methods 0.000 description 2
- 230000033228 biological regulation Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000005336 cracking Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000002427 irreversible effect Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 230000003278 mimic effect Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/54—Extraction of image or video features relating to texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Abstract
A substation operation prevention and control method based on an image recognition technology comprises the steps of collecting equipment images of equipment to be recognized through cameras deployed in a substation; performing image analysis on the equipment image of the equipment to be identified to obtain equipment image texture features; and determining whether the appearance of the device to be identified is damaged or not based on the image texture features of the device. Therefore, the interference of factors such as manual inspection, image noise and the like can be avoided, the accuracy of detecting the equipment breakage abnormal condition is improved, and an alarm is timely sent out, so that potential safety hazards are avoided, and the operation safety and efficiency are improved.
Description
Technical Field
The application relates to the technical field of intelligent substations, in particular to a substation operation prevention and control method based on an image recognition technology.
Background
With the development of a power system and the improvement of the intelligent degree of a transformer substation, the safety and the efficiency of transformer substation operation become important concerns, wherein operation prevention and control are important links for guaranteeing the normal operation of transformer substation equipment and the safety of operators.
The traditional substation operation prevention and control scheme mainly relies on manual inspection and periodic maintenance, and workers are required to check the appearance condition of equipment one by one. However, this solution is inefficient, time consuming, and prone to leakage and error. In addition, manual inspection relies on subjective judgment of staff to a great extent, is easily influenced by personal experience and emotion, causes misjudgment and omission, and especially for some tiny damages or hidden dangers, is easily ignored or misjudged as normal conditions, so that the operation prevention and control effect of the transformer substation is reduced, and the potential safety hazard is increased. The existing schemes also finish the prevention and control of the operation of the transformer substation based on the image recognition technology, but due to the complex environment of the transformer substation, the quality of the acquired image can be affected by the interference of various factors, so that the detection precision of some tiny damages or hidden dangers to equipment is reduced.
Therefore, an optimized substation operation prevention and control scheme based on image recognition technology is desired.
Disclosure of Invention
The present application has been made to solve the above-mentioned technical problems. The embodiment of the application provides a substation operation prevention and control method based on an image recognition technology, which comprises the steps of acquiring equipment images of equipment to be recognized through a camera arranged in a substation; performing image analysis on the equipment image of the equipment to be identified to obtain equipment image texture features; and determining whether the appearance of the device to be identified is damaged or not based on the image texture features of the device. Therefore, the interference of factors such as manual inspection, image noise and the like can be avoided, the accuracy of detecting the equipment breakage abnormal condition is improved, and an alarm is timely sent out, so that potential safety hazards are avoided, and the operation safety and efficiency are improved.
In a first aspect, a substation operation prevention and control method based on an image recognition technology is provided, which includes:
acquiring equipment images of equipment to be identified through cameras deployed in a transformer substation;
performing image analysis on the equipment image of the equipment to be identified to obtain equipment image texture features; and
and determining whether the appearance of the equipment to be identified is damaged or not based on the image texture characteristics of the equipment.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a substation operation prevention and control method based on an image recognition technology according to an embodiment of the present application.
Fig. 2 is a schematic diagram of a substation operation prevention and control method based on an image recognition technology according to an embodiment of the present application.
Fig. 3 is a flowchart of the sub-steps of step 120 in a substation operation prevention and control method based on image recognition technology according to an embodiment of the present application.
Fig. 4 is a block diagram of a substation operation prevention and control system based on an image recognition technology according to an embodiment of the present application.
Fig. 5 is a schematic view of a substation operation prevention and control method based on an image recognition technology according to an embodiment of the present application.
Detailed Description
The following description of the technical solutions according to the embodiments of the present application will be given with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Unless defined otherwise, all technical and scientific terms used in the embodiments of the application have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to limit the scope of the present application.
In describing embodiments of the present application, unless otherwise indicated and limited thereto, the term "connected" should be construed broadly, for example, it may be an electrical connection, or may be a communication between two elements, or may be a direct connection, or may be an indirect connection via an intermediate medium, and it will be understood by those skilled in the art that the specific meaning of the term may be interpreted according to circumstances.
It should be noted that, the term "first\second\third" related to the embodiment of the present application is merely to distinguish similar objects, and does not represent a specific order for the objects, and it is understood that "first\second\third" may interchange a specific order or sequence where allowed. It is to be understood that the "first\second\third" distinguishing objects may be interchanged where appropriate such that embodiments of the application described herein may be practiced in sequences other than those illustrated or described herein.
It should be understood that the transformer substation is an important component in the power system, and is mainly used for transmitting and distributing electric energy, and plays a role in converting electric energy of the high-voltage transmission line into low-voltage electric energy for use by users.
The general operation of the substation includes: 1. the power transmission line is connected in, the transformer substation receives the electric energy transmitted by the high-voltage power transmission line, and the electric energy is converted into low-voltage electric energy suitable for users through the transformer. The operator needs to access the transmission line to the main transformer or the distribution transformer of the substation as required. 2. Transformers are used for converting and distributing electric energy, and are core equipment of a transformer substation. The operation state of the transformer, including temperature, oil level, cooling system, etc. needs to be monitored by operators to ensure the normal operation of the transformer. Meanwhile, regular maintenance and overhaul, such as oil maintenance, insulation detection and the like of the oil immersed transformer, are also required. 3. Switching devices operate and there are various switching devices in substations, such as circuit breakers, load switches, etc., for controlling the transmission and distribution of electrical energy. Operators need to operate the switching devices as required to ensure the normal transmission and distribution of electrical energy. 4. The protection device detects and maintains, and the protection device in the transformer substation is used for monitoring the abnormal condition of the power system and taking corresponding protection measures to ensure the safety of equipment and personnel. The operator needs to detect and maintain the protection device regularly to ensure the reliability and accuracy. 5. Safety measures and accident handling, operators need to observe various safety regulations and operation rules, and ensure the safety of themselves and others. Meanwhile, when an accident or abnormal situation occurs, corresponding emergency measures need to be taken in time, and the accident is processed and investigated.
Substation operation involves the transmission, conversion and distribution of electrical energy, and the operation, maintenance and protection of equipment. The operators need to have relevant expertise and skills, strictly follow the safety regulations, and ensure the normal operation and the power supply reliability of the transformer substation.
Further, substations are an important component of electrical power systems, involving high voltage power equipment and complex electrical systems. There is a certain risk in the operation process, such as electric shock, fire, explosion, etc. Therefore, necessary precautions must be taken to ensure the safety of the operator.
Substation operations involve a large number of equipment operations and maintenance tasks including transmission line access, transformer operations and maintenance, switchgear operations, protection device detection and maintenance, and the like. Through reasonable prevention and control measures, the operation efficiency can be improved, and the operation time and the resource waste are reduced. Substation equipment is a core component of a power system, and normal operation of the equipment is crucial to stable operation of the power system. By means of the prevention and control measures, abnormal conditions of the equipment can be found in time, maintenance and maintenance can be performed, and the influence of equipment faults on a power system is avoided. Certain potential risks exist in the operation of the transformer substation, such as equipment faults, misoperation and the like, and accidents can be caused. Through strengthening prevention and control measures, the occurrence of accidents can be prevented, and the safe operation of the transformer substation is ensured.
Therefore, the necessity of the operation prevention and control of the transformer substation is embodied in the aspects of ensuring the safety of operators, improving the operation efficiency, protecting equipment, preventing accidents and the like. By adopting proper technology and management measures, the safe and efficient operation of the transformer substation can be effectively realized.
Fig. 1 is a flowchart of a substation operation prevention and control method based on an image recognition technology according to an embodiment of the present application. Fig. 2 is a schematic diagram of a substation operation prevention and control method based on an image recognition technology according to an embodiment of the present application. As shown in fig. 1 and fig. 2, the substation operation prevention and control method based on the image recognition technology includes: 110, collecting equipment images of equipment to be identified through cameras deployed in the transformer substation; 120, performing image analysis on the device image of the device to be identified to obtain device image texture features; and, 130, determining whether the appearance of the device to be identified is damaged based on the device image texture features.
In the step 110, the position of the camera should be selected to cover the equipment to be identified in a full area, ensure that the image quality is clear and visible, and the installation of the camera should be fixed and reliable, so as to avoid the blurring or the unrecognizable image caused by vibration or displacement. The camera is used for acquiring the equipment image, so that the appearance information of the equipment can be acquired in real time, manual inspection is not needed, and the operation efficiency is improved. The state of the equipment can be monitored all the time, the abnormal condition of the equipment can be found in time, and the accident risk caused by equipment damage is avoided.
In the step 120, the selection of the image analysis algorithm should be optimized according to specific equipment characteristics, so as to ensure that the texture features of the equipment image can be accurately extracted, and the image is preprocessed, such as denoising, enhancement, etc., so as to improve the image quality and the analysis accuracy. Through image analysis, texture features of the device image, such as colors, lines, etc., can be extracted for subsequent damage detection. And the device image analysis can be automatically performed, so that the manual intervention is reduced, and the detection accuracy and efficiency are improved.
In the step 130, a suitable algorithm model is established to classify the normal appearance and the damaged appearance of the device, so as to realize accurate damage detection, wherein the model needs to be designed and trained specifically for different types of devices, and the accuracy of detection is improved. By analyzing based on the texture characteristics, whether the appearance of the equipment to be identified is damaged or not can be judged, and an alarm can be sent out in time. The detection precision of equipment damage can also be improved, missing report and false report are reduced, and the operation safety and efficiency are improved.
The substation operation prevention and control method based on the image recognition technology collects equipment images through the cameras, performs image analysis and texture feature extraction, and achieves detection of equipment appearance damage, so that operation efficiency and safety are improved.
Specifically, in the step 110, a device image of the device to be identified is acquired by a camera disposed in the substation. Aiming at the technical problems, the technical concept of the application is that the camera arranged in the transformer substation is used for collecting the equipment image of the equipment to be identified, and the image analysis and identification technology is introduced into the rear end so as to automatically detect the appearance of the equipment to be identified after background interference is filtered and the image is enhanced, thereby judging whether the appearance of the equipment to be identified is damaged. Through the mode, the interference of factors such as manual inspection, image noise and the like can be avoided, so that the accuracy of detecting the equipment breakage abnormal condition is improved, an alarm is timely sent, potential safety hazards are avoided, and the operation safety and efficiency are improved.
Specifically, in the technical scheme of the application, firstly, equipment images of equipment to be identified, which are acquired by cameras deployed in a transformer substation, are acquired. The camera arranged in the transformer substation is used for collecting the equipment image of the equipment to be identified, and plays an important role in finally determining whether the appearance of the equipment to be identified is damaged.
The camera can acquire the image of the equipment in real time, manual inspection is not needed, and the appearance state of the equipment can be monitored all the time. Thus, abnormal conditions of the equipment such as damage, breakage, deformation and the like can be found in time, and the accident risk caused by the equipment damage is avoided.
The camera is used for collecting the equipment images, so that the operation efficiency can be greatly improved. Compared with the traditional manual inspection mode, the camera can monitor a plurality of devices simultaneously and can transmit image data in real time, so that the monitoring of the states of the devices is more timely and accurate.
The equipment images acquired by the camera can be stored and analyzed to form a history record of the equipment state. Through analysis of historical data, potential problems and trends of equipment can be found, maintenance and repair can be performed in advance, and irreversible loss caused by equipment damage is avoided.
Through image recognition and analysis techniques, automated damage detection can be performed on the device images. Based on the texture features of the equipment image, an algorithm model can be established to classify and identify equipment damage, so that whether the appearance of equipment to be identified is damaged or not is judged, and the accuracy and the efficiency of detection can be improved.
Specifically, in the step 120, an image analysis is performed on the device image of the device to be identified to obtain a texture feature of the device image. Fig. 3 is a flowchart of the substeps of step 120 in the substation operation prevention and control method based on the image recognition technology according to the embodiment of the present application, as shown in fig. 3, performing image analysis on the device image of the device to be recognized to obtain the device image texture feature, including: 121, extracting a device image foreground part from the device image of the device to be identified; 122, calculating a directional gradient histogram of the foreground portion of the device image; and 123, performing feature extraction on the directional gradient histogram through an image deep texture feature extractor based on a deep neural network model to obtain the device image texture feature.
The texture features of the equipment images can be quickly obtained through automatic image analysis and texture feature extraction, manual intervention and time consumption are not needed, the operation efficiency is greatly improved, and the requirement on human resources is reduced. By monitoring the equipment images in real time and extracting the texture features, the appearance damage condition of the equipment can be found in time, repair measures can be taken early, and safety accidents caused by equipment faults are avoided. The camera is deployed to collect the equipment image, and the real-time monitoring of the appearance of the equipment can be realized through image analysis and texture feature extraction, so that the state of the equipment can be mastered at any time, and abnormal conditions can be found in time. The image deep texture feature extractor based on the deep neural network model can extract features of the directional gradient histogram to obtain more accurate equipment image texture features, is beneficial to improving the accuracy of damage detection and reduces the false judgment rate.
It should be understood that, due to the complex environment in the transformer substation, various factors may interfere during the process of collecting the image of the device, and during the analysis of the image features, the interference of background features in the image may also easily cause weak attention to the feature information of the device to be identified, so as to affect the accuracy of appearance detection of the device to be identified. Therefore, in the technical scheme of the application, the foreground part of the device image is further extracted from the device image, so that the attention is concentrated on the device to be identified, the background interference is eliminated, and the accuracy of image analysis is improved.
Then, considering that if the appearance of the device to be identified is damaged, such as corrosion, cracking or loosening of the surface of the device, the texture end of the image is presented, therefore, in the technical scheme of the application, the direction gradient histogram of the foreground part of the image of the device is further calculated, and then the original RGB image is replaced by the direction gradient histogram as input data, so that the appearance texture characteristic information related to the device to be identified in the foreground part of the image of the device is more fully captured. It is worth mentioning that, here, the direction gradient histogram is a method for describing local texture features of an image, the algorithm divides the image into small-sized cell spaces, calculates gradients of pixels in the cells, generates cells (HOG (Histogram of Oriented Gradient, HOG) according to the gradient distribution, then counts the distribution of cells HOG in larger-sized block spaces, generates block spaces HOG, and describes local texture information in the image.
It should be appreciated that the directional gradient histogram may better describe the texture features of the image, and by calculating the directional gradient histogram of the foreground portion of the device image, the details of the texture features of the device appearance may be more fully captured, thereby improving the accuracy of the texture feature analysis of the device image.
When using the original RGB image as input data, the image may contain a lot of background information that is not related to the appearance of the device, which information may interfere with the analysis of the texture features of the device. By extracting the histogram of the directional gradient of the foreground portion of the device image, attention can be focused on the area related to the appearance of the device, and the influence of redundant information is reduced.
The direction gradient histogram is an efficient feature representation method, which requires less computation compared to the original RGB image. This means that the use of a directional gradient histogram can reduce the computational complexity and increase the execution efficiency of the algorithm when performing image analysis and texture feature extraction.
That is, using the directional gradient histogram of the foreground portion of the device image as input data can more fully capture the appearance texture feature information about the device to be identified in the foreground portion of the device image, thereby improving the accuracy and efficiency of image analysis and texture feature extraction.
Further, the feature extraction of the directional gradient histogram by an image deep texture feature extractor based on a deep neural network model to obtain the device image texture feature comprises: passing the directional gradient histogram through an image deep texture feature extractor based on a convolutional neural network model using a spatial attention mechanism to obtain a texture feature matrix; and fusing the directional gradient histogram and the texture feature matrix by utilizing a residual thought to obtain a multi-scale texture feature matrix as the texture feature of the equipment image.
Wherein the deep neural network model is based on a convolutional neural network model using a spatial attention mechanism
Further, it is also considered that, when appearance damage detection of the device to be identified is actually performed, attention should be paid to the appearance texture feature of the device to be identified at a spatial position in the image while neglecting disturbance feature information irrelevant to appearance quality detection. In view of the ability of the attention mechanism to select the focus position, a more resolved representation of the feature is produced, and the feature after addition to the attention module will change adaptively as the network deepens. Therefore, in the technical scheme of the application, the direction gradient histogram is processed by an image deep texture feature extractor based on a convolutional neural network model using a spatial attention mechanism so as to extract the implicit feature distribution information of the appearance quality of the equipment to be identified focused on the spatial position in the direction gradient histogram, thereby obtaining a texture feature matrix. It should be noted that, here, the image features extracted by the spatial attention reflect weights of the differences of the spatial dimension features, so as to suppress or strengthen the features of different spatial positions, thereby extracting the appearance quality feature information focused on the device to be identified in space.
The direction gradient histogram is processed by an image deep texture feature extractor based on a convolutional neural network model using a spatial attention mechanism, so that the appearance quality implicit feature distribution information focused on the spatial position in the direction gradient histogram can be extracted, and a texture feature matrix is obtained.
The convolutional neural network model using the spatial attention mechanism can weight features at different positions in the image when extracting texture features, and more attention is paid to important spatial positions. Therefore, the accuracy and the distinguishing degree of the texture features can be improved, and the appearance quality information of the equipment to be identified can be captured better. By the processing mode, the obtained texture feature matrix can be used for subsequent image analysis and damage detection tasks, and the effect and accuracy of operation prevention and control are improved.
Among other things, spatial attention mechanisms are a technique for image processing and computer vision tasks that mimic the attention mechanisms of the human visual system in processing visual information. The spatial attention mechanism can make the neural network pay more attention to important spatial positions when processing images, so that the accuracy and performance of tasks are improved.
In convolutional neural networks, the spatial attention mechanism is implemented by an attention-introducing module. And learning a weight matrix according to the input image feature map, wherein the weight matrix is used for guiding feature extraction and weighting of the neural network at different spatial positions. Thus, the network can be focused on the region with important information in the image, and the extraction capability of the key features is improved.
Through a spatial attention mechanism, the neural network can pay attention to important information in the image more accurately, and the performance of tasks such as target identification, positioning and segmentation is improved. In the deep texture feature extraction of the image, a spatial attention mechanism can help a network to better capture the appearance quality implicit feature distribution information of the equipment to be identified, and the accuracy and the distinguishing degree of the texture features are improved.
It should be appreciated that since the directional gradient histogram may describe edge and texture directional information in the image regarding the device to be identified, the texture feature matrix may capture appearance texture variations and detail feature information in the image regarding the device to be identified. Therefore, in the technical scheme of the application, the residual error idea is further utilized to fuse the direction gradient histogram and the texture feature matrix so as to obtain a multi-scale texture feature matrix. Therefore, the direction gradient histogram and the complementary information in the texture feature matrix can be combined to improve the diversity and richness of appearance feature expression of the equipment to be identified, so that subtle changes and differences which are not completely expressed in original features are captured, and the distinguishing capability of the features and the subsequent appearance damage detection accuracy of the equipment to be identified are improved.
It should be understood that the residual concept is an important concept in deep learning, and is used for solving the problems of gradient elimination and gradient explosion in the deep neural network training process. The core idea is to directly add the input and output of the network by introducing a jump connection, thus constructing a residual block.
In deep neural networks, each layer converts input into output by learning. However, when the number of network layers is large, information transfer may be hindered, resulting in gradients that are difficult to propagate efficiently, thereby affecting training and performance of the network. To solve this problem, the residual concept proposes the concept of a jump connection. The jump connection allows the network to pass information directly from layer to layer, adding the input features to the output features, so that the network can gradually adjust the input features closer to the desired output features by learning the residuals. This approach makes it easier for the network to learn the residual portion, thereby speeding up the training process of the network and improving the performance of the network.
In the image processing task, the residual thought can be utilized to fuse the characteristic information of different scales. For example, when the directional gradient histogram and the texture feature matrix are fused, a multi-scale texture feature matrix can be obtained by adding the directional gradient histogram and the texture feature matrix. Therefore, the network can utilize the information of the two characteristics simultaneously, and the judging capability of the appearance quality of the equipment to be identified is improved.
That is, the residual concept allows the network to directly transfer information from layer to layer by introducing a jump connection, thus solving the gradient problem in deep network training. In the image processing task, the residual thought can be used for fusing the characteristic information of different scales, so that the performance and accuracy of the network are improved.
The residual thought is utilized to fuse the direction gradient histogram and the texture feature matrix, on one hand, multi-scale feature fusion can be realized, and the direction gradient histogram and the texture feature matrix usually have different scales and resolutions. By utilizing the residual thought, the two features can be fused in networks of different layers, so that a multi-scale texture feature matrix is obtained. Thus, the texture features of the equipment to be identified can be captured better, and the accuracy of image analysis and texture feature extraction is improved.
On the other hand, the problems of gradient disappearance and gradient explosion, which are common problems in deep networks, can be alleviated, and the network is difficult to converge or overfit. By introducing a residual connection, gradients can be propagated more easily, avoiding gradual decrease or increase of gradients in the network. This helps to improve the training process of the network, improving the convergence and generalization ability of the model.
In yet another aspect, enhancement of information transfer may be performed, and the residual connection allows the network to directly transfer input features to output features, which may enhance transfer and retention of information. By combining the directional gradient histogram and the texture feature matrix with residual connection, the network can better utilize the information of the input features, and the extraction capability of the appearance texture features of the equipment is improved.
Specifically, in the step 130, determining whether the appearance of the device to be identified is damaged based on the device image texture features includes: and the multi-scale texture feature matrix passes through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the appearance of the equipment to be identified is damaged or not.
And then, the multi-scale texture feature matrix is passed through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the appearance of the equipment to be identified is damaged or not. That is, the classifying process is performed by using the appearance multi-scale characteristic information of the equipment to be identified, so that whether the appearance of the equipment to be identified is damaged or not is effectively detected, and an alarm is timely sent out, so that potential safety hazards are avoided.
Further, in the present application, the substation operation prevention and control method based on the image recognition technology further includes a training step: and training the deep texture feature extractor and the classifier based on the convolutional neural network model using a spatial attention mechanism. The training step comprises the following steps: acquiring training data, wherein the training data comprises a device image for training a device to be identified and a true value of whether the appearance of the device to be identified is damaged or not; extracting a training device image foreground portion from the training device image; calculating a training direction gradient histogram of the image foreground part of the training equipment; passing the training direction gradient histogram through the image deep texture feature extractor based on the convolutional neural network model using a spatial attention mechanism to obtain a training texture feature matrix; fusing the training direction gradient histogram and the training texture feature matrix by utilizing a residual thought to obtain a training multi-scale texture feature matrix; passing the training multi-scale texture feature matrix through the classifier to obtain a classification loss function value; expanding the training directional gradient histogram into a directional gradient histogram expanded pixel vector, and expanding the training texture feature matrix into a texture feature expanded vector; calculating a common manifold implicit similarity factor for the directional gradient histogram development pixel vector and the texture feature development vector to obtain a common manifold implicit similarity loss function value; and training the image deep texture feature extractor and the classifier based on the convolutional neural network model using a spatial attention mechanism with a weighted sum of the classification loss function value and the common manifold implicit similarity loss function value as a loss function value and traveling in a gradient descent direction.
Wherein passing the training multi-scale texture feature matrix through the classifier to obtain a classification loss function value comprises: expanding the training multi-scale texture feature matrix into classification feature vectors according to row vectors or column vectors; performing full-connection coding on the classification feature vectors by using a plurality of full-connection layers of the classifier to obtain coded classification feature vectors; and passing the coding classification feature vector through a Softmax classification function of the classifier to obtain the classification result.
In particular, in the technical solution of the present application, when the residual concept is used to fuse the directional gradient histogram and the texture feature matrix to obtain a multi-scale texture feature matrix, texture information contained in the directional gradient histogram is obtained by calculating gradients of respective pixels in the foreground portion of the device image, and texture information contained in the texture feature matrix is obtained by performing convolution kernel-based texture feature filtering on the directional gradient histogram, so that the texture feature matrix and the directional gradient histogram have differences in gradient and depth levels of feature expression, that is, the texture feature matrix and the directional gradient histogram have feature expression structural mismatch. If the directional gradient histogram and the texture feature matrix are directly fused in a position weighted sum manner to obtain a multi-scale texture feature matrix, structural imbalance of feature expression between the directional gradient histogram and the texture feature matrix can make geometric monotonicity of high-dimensional feature manifold expression in a high-dimensional feature space of the multi-scale texture feature matrix worse so as to influence accuracy of classification regression of the multi-dimensional feature manifold expression through a classifier.
In particular, the applicant of the present application has developed pixel vectors for the direction gradient histogram developed after the development of the direction gradient histogram, for example, denoted asAnd a texture feature expansion vector obtained after expansion of the texture feature matrix, for example, denoted +.>The common manifold implicit similarity factor is introduced as a loss function, specifically expressed as: calculating a common manifold implicit similarity factor for the directional gradient histogram development pixel vector and the texture feature development vector with a loss formula to obtain the common manifold implicit similarity loss function value; wherein, the loss formula is:
wherein,and->The direction gradient histogram development pixel vector and the texture feature development vector, ++>Is a transpose of the texture feature expansion vector, < >>Representing the two norms of the vector, and +.>Representing the square root of the Frobenius norm of the matrix, the direction gradient histogram development pixel vector and the texture feature development vector being both in the form of column vectors, +.>And->For the weight super parameter, ++>Representing vector multiplication, ++>Representing the multiplication by the position point,representing difference by position +.>Implicit similarity loss function values for the common manifold.
Here, the common manifold implicit similarity factor may spread out pixel vectors with the directional gradient histogramAnd the texture feature expansion vector +.>The structured association between them represents the common manifold of the respective feature manifolds in the cross dimension and shares the direction gradient histogram expanded pixel vector +_ with the same factorization weight>And the texture feature expansion vector +.>And (3) common constraints of manifold structural factors such as variability, correspondence, relevance and the like, so as to measure the distribution similarity of geometric derivative structure representations depending on common manifold to realize nonlinear geometric monotonicity of feature distribution transfer, and promote the geometric monotonicity of the directional gradient histogram relative to the high-dimensional feature manifold expression of the texture feature matrix to improve the expression effect of fusion features of the directional gradient histogram and the texture feature matrix. In this way, the appearance of the device to be identified can be madeThe quality detects to judge whether the outward appearance of waiting to discern equipment takes place to damage, so avoid artifical inspection and the interference of factors such as image noise, thereby improve the accuracy to the broken abnormal conditions detection of equipment, and in time send out the alarm, in order to avoid potential safety hazard, improve operation security and efficiency.
In summary, the substation operation prevention and control method 100 based on the image recognition technology according to the embodiment of the present application is illustrated, which collects the device image of the device to be recognized through the camera disposed in the substation, and introduces the image analysis and recognition technology at the back end to automatically perform the appearance detection of the device to be recognized after filtering the background interference and enhancing the image, so as to determine whether the appearance of the device to be recognized is damaged. Through the mode, the interference of factors such as manual inspection, image noise and the like can be avoided, so that the accuracy of detecting the equipment breakage abnormal condition is improved, an alarm is timely sent, potential safety hazards are avoided, and the operation safety and efficiency are improved.
In one embodiment of the present application, fig. 4 is a block diagram of a substation operation prevention and control system based on an image recognition technology according to an embodiment of the present application. As shown in fig. 4, a substation operation prevention and control system 200 based on an image recognition technology according to an embodiment of the present application includes: the image acquisition module 210 is configured to acquire an equipment image of equipment to be identified through a camera disposed in the substation; the image analysis module 220 is configured to perform image analysis on the device image of the device to be identified to obtain a texture feature of the device image; and an appearance determining module 230, configured to determine whether the appearance of the device to be identified is damaged based on the image texture features of the device.
Here, it will be understood by those skilled in the art that the specific functions and operations of the respective units and modules in the above-described image recognition technology-based substation operation prevention and control system have been described in detail in the above description of the image recognition technology-based substation operation prevention and control method with reference to fig. 1 to 3, and thus, repetitive descriptions thereof will be omitted.
As described above, the substation operation prevention and control system 200 based on the image recognition technology according to the embodiment of the present application may be implemented in various terminal devices, for example, a server or the like for substation operation prevention and control based on the image recognition technology. In one example, the substation operation prevention and control system 200 based on the image recognition technology according to the embodiment of the present application may be integrated into the terminal device as one software module and/or hardware module. For example, the substation operation prevention and control system 200 based on the image recognition technology may be a software module in the operating system of the terminal device, or may be an application program developed for the terminal device; of course, the substation operation prevention and control system 200 based on the image recognition technology may also be one of numerous hardware modules of the terminal device.
Alternatively, in another example, the substation operation prevention and control system 200 based on the image recognition technology and the terminal device may be separate devices, and the substation operation prevention and control system 200 based on the image recognition technology may be connected to the terminal device through a wired and/or wireless network and transmit the interactive information according to the agreed data format.
Fig. 5 is a schematic view of a substation operation prevention and control method based on an image recognition technology according to an embodiment of the present application. As shown in fig. 5, in this application scenario, first, a device image of a device to be identified is acquired by a camera disposed within a substation (e.g., C as illustrated in fig. 5); then, the acquired device image of the device to be identified is input into a server (e.g., S as illustrated in fig. 5) deployed with a substation operation prevention and control algorithm based on an image recognition technology, wherein the server is capable of processing the device image of the device to be identified based on the substation operation prevention and control algorithm of the image recognition technology to determine whether the appearance of the device to be identified is damaged.
It is also noted that in the apparatus, devices and methods of the present application, the components or steps may be disassembled and/or assembled. Such decomposition and/or recombination should be considered as equivalent aspects of the present application.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, this description is not intended to limit embodiments of the application to the form disclosed herein. Although a number of example aspects and embodiments have been discussed above, a person of ordinary skill in the art will recognize certain variations, modifications, alterations, additions, and subcombinations thereof.
Claims (2)
1. The substation operation prevention and control method based on the image recognition technology is characterized by comprising the following steps of:
acquiring equipment images of equipment to be identified through cameras deployed in a transformer substation;
performing image analysis on the equipment image of the equipment to be identified to obtain equipment image texture features; and
determining whether the appearance of the equipment to be identified is damaged or not based on the image texture characteristics of the equipment;
the image analysis of the device image of the device to be identified to obtain the texture features of the device image comprises the following steps:
extracting an equipment image foreground part from the equipment image of the equipment to be identified;
calculating a directional gradient histogram of the foreground part of the device image; and
performing feature extraction on the direction gradient histogram through an image deep texture feature extractor based on a deep neural network model to obtain the equipment image texture features;
the deep neural network model is based on a convolutional neural network model using a spatial attention mechanism;
the feature extraction of the directional gradient histogram by an image deep texture feature extractor based on a deep neural network model to obtain the equipment image texture features comprises the following steps:
passing the directional gradient histogram through an image deep texture feature extractor based on a convolutional neural network model using a spatial attention mechanism to obtain a texture feature matrix; and
fusing the direction gradient histogram and the texture feature matrix by utilizing a residual thought to obtain a multi-scale texture feature matrix as the texture feature of the equipment image;
wherein, based on the device image texture characteristics, determining whether the appearance of the device to be identified is damaged comprises: the multi-scale texture feature matrix is passed through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the appearance of equipment to be identified is damaged or not;
the substation operation prevention and control method based on the image recognition technology further comprises the training steps of: training the image deep texture feature extractor and the classifier based on a convolutional neural network model using a spatial attention mechanism;
wherein the training step comprises:
acquiring training data, wherein the training data comprises a device image for training a device to be identified and a true value of whether the appearance of the device to be identified is damaged or not;
extracting a training device image foreground part from the device image of the device to be recognized;
calculating a training direction gradient histogram of the image foreground part of the training equipment;
passing the training direction gradient histogram through the image deep texture feature extractor based on the convolutional neural network model using a spatial attention mechanism to obtain a training texture feature matrix;
fusing the training direction gradient histogram and the training texture feature matrix by utilizing a residual thought to obtain a training multi-scale texture feature matrix; and
passing the training multi-scale texture feature matrix through the classifier to obtain a classification loss function value;
expanding the training directional gradient histogram into a directional gradient histogram expanded pixel vector, and expanding the training texture feature matrix into a texture feature expanded vector;
calculating a common manifold implicit similarity factor for the directional gradient histogram development pixel vector and the texture feature development vector to obtain a common manifold implicit similarity loss function value; and
training the image deep texture feature extractor and the classifier based on the convolutional neural network model using a spatial attention mechanism by taking a weighted sum of the classification loss function value and the common manifold implicit similarity loss function value as a loss function value and propagating in a gradient descent direction;
wherein calculating a common manifold implicit similarity factor for the direction gradient histogram development pixel vector and the texture feature development vector to obtain a common manifold implicit similarity loss function value comprises:
calculating a common manifold implicit similarity factor for the directional gradient histogram development pixel vector and the texture feature development vector with a loss formula to obtain the common manifold implicit similarity loss function value;
wherein, the loss formula is:
，
wherein,and->The direction gradient histogram development pixel vector and the texture feature development vector, ++>Is the texture feature expansion vectorIs the transposed vector of>Representing the two norms of the vector, and +.>Representing the square root of the Frobenius norm of the matrix, the direction gradient histogram development pixel vector and the texture feature development vector being both in the form of column vectors, +.>And->For the weight super parameter, ++>Representing vector multiplication, ++>Representing multiplication by location +.>Representing difference by position +.>Implicit similarity loss function values for the common manifold.
2. The method for preventing and controlling operation of a transformer substation based on the image recognition technology according to claim 1, wherein the step of passing the training multi-scale texture feature matrix through the classifier to obtain a classification loss function value comprises the steps of:
expanding the training multi-scale texture feature matrix into classification feature vectors according to row vectors or column vectors;
performing full-connection coding on the classification feature vectors by using a plurality of full-connection layers of the classifier to obtain coded classification feature vectors; and
and the coding classification feature vector is passed through a Softmax classification function of the classifier to obtain the classification result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311015674.4A CN116740654B (en) | 2023-08-14 | 2023-08-14 | Substation operation prevention and control method based on image recognition technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311015674.4A CN116740654B (en) | 2023-08-14 | 2023-08-14 | Substation operation prevention and control method based on image recognition technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116740654A CN116740654A (en) | 2023-09-12 |
| CN116740654B true CN116740654B (en) | 2023-11-07 |
Family
ID=87910007
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311015674.4A Active CN116740654B (en) | 2023-08-14 | 2023-08-14 | Substation operation prevention and control method based on image recognition technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116740654B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117475304A (en) * | 2023-10-24 | 2024-01-30 | 北京佳格天地科技有限公司 | Agricultural pest identification method based on artificial intelligence |
| CN117636045A (en) * | 2023-12-07 | 2024-03-01 | 湖州练市漆宝木业有限公司 | Wood defect detection system based on image processing |
| CN117479127A (en) * | 2023-12-26 | 2024-01-30 | 深圳市掌锐电子有限公司 | 5G-based intelligent terminal and method for Internet of vehicles |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104809732A (en) * | 2015-05-07 | 2015-07-29 | 山东鲁能智能技术有限公司 | Electrical equipment appearance abnormity detection method based on image comparison |
| CN107945179A (en) * | 2017-12-21 | 2018-04-20 | 王华锋 | A kind of good pernicious detection method of Lung neoplasm of the convolutional neural networks of feature based fusion |
| CN108573256A (en) * | 2017-03-14 | 2018-09-25 | 山东鲁能智能技术有限公司 | A kind of substation's plate pressing equipment state identification method and device |
| CN111598861A (en) * | 2020-05-13 | 2020-08-28 | 河北工业大学 | Improved Faster R-CNN model-based non-uniform texture small defect detection method |
| CN112200145A (en) * | 2020-11-03 | 2021-01-08 | 上海万位科技有限公司 | Remote identification method for public security |
| CN113792631A (en) * | 2021-08-31 | 2021-12-14 | 电子科技大学 | Aircraft detection and tracking method based on multi-scale self-adaption and side-domain attention |
| WO2022089166A1 (en) * | 2020-11-02 | 2022-05-05 | 腾讯科技（深圳）有限公司 | Facial image processing method and apparatus, facial image display method and apparatus, and device |
| CN114494106A (en) * | 2021-11-04 | 2022-05-13 | 西安工业大学 | Deep learning multi-feature fusion-based oral mucosal disease identification method |
| CN115171045A (en) * | 2022-07-15 | 2022-10-11 | 国网江苏省电力有限公司 | YOLO-based power grid operation field violation identification method and terminal |
| CN115619685A (en) * | 2022-11-08 | 2023-01-17 | 广州大学 | Transformer method for tracking structure for image restoration |
| CN116309590A (en) * | 2023-05-22 | 2023-06-23 | 四川新迎顺信息技术股份有限公司 | Visual computing method, system, electronic equipment and medium based on artificial intelligence |
| CN116385707A (en) * | 2023-04-04 | 2023-07-04 | 河海大学 | Deep learning scene recognition method based on multi-scale features and feature enhancement |
-
2023
- 2023-08-14 CN CN202311015674.4A patent/CN116740654B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104809732A (en) * | 2015-05-07 | 2015-07-29 | 山东鲁能智能技术有限公司 | Electrical equipment appearance abnormity detection method based on image comparison |
| CN108573256A (en) * | 2017-03-14 | 2018-09-25 | 山东鲁能智能技术有限公司 | A kind of substation's plate pressing equipment state identification method and device |
| CN107945179A (en) * | 2017-12-21 | 2018-04-20 | 王华锋 | A kind of good pernicious detection method of Lung neoplasm of the convolutional neural networks of feature based fusion |
| CN111598861A (en) * | 2020-05-13 | 2020-08-28 | 河北工业大学 | Improved Faster R-CNN model-based non-uniform texture small defect detection method |
| WO2022089166A1 (en) * | 2020-11-02 | 2022-05-05 | 腾讯科技（深圳）有限公司 | Facial image processing method and apparatus, facial image display method and apparatus, and device |
| CN112200145A (en) * | 2020-11-03 | 2021-01-08 | 上海万位科技有限公司 | Remote identification method for public security |
| CN113792631A (en) * | 2021-08-31 | 2021-12-14 | 电子科技大学 | Aircraft detection and tracking method based on multi-scale self-adaption and side-domain attention |
| CN114494106A (en) * | 2021-11-04 | 2022-05-13 | 西安工业大学 | Deep learning multi-feature fusion-based oral mucosal disease identification method |
| CN115171045A (en) * | 2022-07-15 | 2022-10-11 | 国网江苏省电力有限公司 | YOLO-based power grid operation field violation identification method and terminal |
| CN115619685A (en) * | 2022-11-08 | 2023-01-17 | 广州大学 | Transformer method for tracking structure for image restoration |
| CN116385707A (en) * | 2023-04-04 | 2023-07-04 | 河海大学 | Deep learning scene recognition method based on multi-scale features and feature enhancement |
| CN116309590A (en) * | 2023-05-22 | 2023-06-23 | 四川新迎顺信息技术股份有限公司 | Visual computing method, system, electronic equipment and medium based on artificial intelligence |
Non-Patent Citations (2)
| Title |
|---|
| RA-MISNET:multilevel information sharing network based on residual attention for crowd counting;Jun Sang 等;《SPIE》;全文 * |
| 基于多注意力机制级联LSTM 模型的猪脸表情识别;温长吉 等;《农业工程 报》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116740654A (en) | 2023-09-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116740654B (en) | Substation operation prevention and control method based on image recognition technology | |
| CN108537154B (en) | Power transmission line bird nest identification method based on HOG characteristics and machine learning | |
| CN112734692A (en) | Transformer equipment defect identification method and device | |
| CN111898669B (en) | Abnormal event early warning system of direct-current submerged arc furnace based on machine learning | |
| CN109886232A (en) | A kind of power grid image identification system neural network based | |
| CN116824517B (en) | Substation operation and maintenance safety control system based on visualization | |
| CN117590223B (en) | Online monitoring system and method for circuit breaker | |
| CN116629465B (en) | Smart power grids video monitoring and risk prediction response system | |
| Zhang | Safety management of civil engineering construction based on artificial intelligence and machine vision technology | |
| CN116012762A (en) | Traffic intersection video image analysis method and system for power equipment | |
| CN106355187A (en) | Application of visual information to electrical equipment monitoring | |
| CN114037873A (en) | Transformer substation secondary circuit protection pressing plate detection system and method based on artificial intelligence | |
| Lu et al. | Anomaly detection method for substation equipment based on feature matching and multi-Semantic classification | |
| Changfu et al. | Research of substation equipment abnormity identification based on image processing | |
| Wang et al. | Substation Equipment Defect Detection based on Temporal-spatial Similarity Calculation | |
| JP4162067B2 (en) | Transmission line accident cause determination method | |
| Gu et al. | Object detection of overhead transmission lines based on improved YOLOv5s | |
| CN117686823B (en) | Automatic acceptance checking system for relay protection of transformer substation | |
| AU2021103901A4 (en) | Method for identifying abnormity of overhead lines with unbalanced samples | |
| Rongrong et al. | Fault Diagnosis Method of Distribution Equipment Based on Hybrid Model of Robot and Deep Learning | |
| CN113838049B (en) | Intelligent checking method for hard pressing plate of transformer substation suitable for portable equipment | |
| LU505201B1 (en) | Intelligent safety early warning method and system of electric power plant based monitoring data analysis | |
| CN112798979B (en) | Substation grounding wire state detection system and method based on deep learning technology | |
| CN118038069A (en) | Switch state detection method and device, storage medium and processor | |
| CN117686823A (en) | Automatic acceptance checking system for relay protection of transformer substation |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |