CN112581604B - Substation equipment surface oil stain image data generation method and device - Google Patents
Substation equipment surface oil stain image data generation method and device Download PDFInfo
- Publication number
- CN112581604B CN112581604B CN202011518392.2A CN202011518392A CN112581604B CN 112581604 B CN112581604 B CN 112581604B CN 202011518392 A CN202011518392 A CN 202011518392A CN 112581604 B CN112581604 B CN 112581604B
- Authority
- CN
- China
- Prior art keywords
- equipment
- dimensional model
- oil stain
- substation
- transformer substation
- 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
- 238000000034 method Methods 0.000 title claims abstract description 76
- 230000011218 segmentation Effects 0.000 claims abstract description 19
- 238000002372 labelling Methods 0.000 claims abstract description 12
- 230000000007 visual effect Effects 0.000 claims abstract description 11
- 238000009877 rendering Methods 0.000 claims abstract description 9
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 8
- 238000003786 synthesis reaction Methods 0.000 claims abstract description 8
- 238000005094 computer simulation Methods 0.000 claims abstract description 7
- 238000013528 artificial neural network Methods 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 10
- 238000005286 illumination Methods 0.000 claims description 3
- 230000000717 retained effect Effects 0.000 claims description 3
- 238000007689 inspection Methods 0.000 description 12
- 238000001514 detection method Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 4
- 238000013135 deep learning Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000004519 grease Substances 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 230000002427 irreversible effect Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
- G06T19/20—Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computer Graphics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Architecture (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the invention provides a method and a device for generating oil stain image data on the surface of substation equipment, and belongs to the technical field of visual identification of substations. The method comprises the following steps: respectively carrying out high-simulation modeling according to the equipment and the images of the oil stain on the surface of the equipment to obtain a first three-dimensional model representing the transformer substation and a second three-dimensional model representing the oil stain on the surface of the equipment; adding the second three-dimensional model into the first three-dimensional model by adopting an object grid matching method to obtain a third three-dimensional model of the transformer substation with the greasy dirt on the surface of the equipment; selecting a plurality of target visual angles in a third three-dimensional model based on a point-plane matching method; acquiring an image part of oil stains on the surface of equipment according to a plurality of target visual angles; rendering different pixel values on each device and the greasy dirt on the surface of the device in the third three-dimensional model respectively, and representing the pixel values in the form of a semantic segmentation graph; and combining the semantic segmentation graph as a labeling part and an image part to synthesize equipment surface oil stain synthesis data of the transformer substation.
Description
Technical Field
The invention relates to the technical field of visual identification of substations, in particular to a method and a device for generating oil stain image data on the surface of substation equipment.
Background
The power system is a precondition and guarantee for the current stable development of economy and society, the transformer substation is inspected by adopting a manual inspection mode at present in China, the manual cost is high, the danger of the working environment is high, the working strength is high, and the working condition is hard. With the sharp decrease of the suitable population of physical laborers and the ageing of the population structure, the trend of 'robot replacement' is inevitably irreversible, and automatic equipment is inevitably introduced rapidly to replace the defect of labor force. With the gradual deepening of the construction of the intelligent power grid in China, substation inspection robots, high-definition video inspection and the like based on visual image data are widely applied, the contradiction that the scale of the power grid is continuously increased and the number of inspection workers cannot be synchronously increased is solved, the inspection quality and the working efficiency of equipment can be remarkably improved, the workload and the working intensity of operators are reduced, and the robot is used for replacing manual inspection and is a non-competing development trend.
The equipment inspection of the transformer substation is always the core work of the whole transformer substation in the operation inspection process. Most of electrical equipment in a transformer substation is oil-filled equipment, in a long-term operation process, oil leakage phenomena of different degrees can occur to the oil-filled equipment along with the change of an operation environment and a mode, the insulation level of the oil-filled equipment can be reduced by slight oil leakage, the power supply capacity of the oil-filled equipment is limited, the equipment safety is affected when the oil leakage is serious, and the stable operation of a power grid is endangered. Therefore, in daily inspection and maintenance of leaked oil, the inspection of leaked oil is particularly important.
In general, developing a leaked oil inspection algorithm for a patrol robot requires a large amount of oil stain image data on the surface of equipment with labels. However, the electric power industry has numerous devices, and is complex and variable. Meanwhile, most of the acquired data are taken from field shooting, and certain environmental limitations exist, so that a great challenge is brought to the construction of an intelligent recognition model of the surface greasy dirt.
Along with the continuous breakthrough of the computer vision technology, a plurality of methods based on deep learning are applied to intelligent inspection tasks of the transformer substation. Research and practice show that the performance of the deep learning algorithm is severely dependent on the quality and scale of the training data, so that deep learning-based detection of oil stains on the surface of equipment requires a large amount of high-quality image data. Considering the requirements of detection and identification of greasy dirt on the surface of equipment, the scale of the greasy dirt image data on the surface of the equipment is generally required to satisfy the following requirements besides being as large as possible: (1) The greasy dirt data on the surface of the equipment is required to be covered comprehensively, and the greasy dirt data is required to be covered as much as possible under the possible conditions; (2) The greasy dirt data on the surface of the equipment should contain data under various extreme conditions, such as heavy rain, heavy snow and the like, and are used for improving the detection capability of the algorithm under extreme weather; (3) The oil stain data on the surface of the equipment is accurately marked, and the marking content is as rich as possible, such as the refined oil stain marking area, serious conditions and the like, so as to support the design of oil stain detection algorithms on the surface of various equipment.
The method for constructing the oil stain data set on the surface of the equipment is based on a manual collection labeling method, namely, the real oil stain data on the surface of the equipment is obtained through a manual, robot, monitoring camera and the like. In a transformer substation, fixed acquisition cameras are generally used for acquiring equipment and judging abnormal conditions, or equipment replaced by surface greasy dirt is directly acquired through a handheld camera, so that the obtained data is real and effective, but cannot be obtained in a large amount, the coverage range is narrow, the acquisition efficiency is low, meanwhile, manual labeling is still needed, and time and labor are wasted.
Disclosure of Invention
The invention aims to provide a method and a device for generating oil stain image data on the surface of substation equipment, which can overcome the technical defect that the oil stain image data of the equipment are collected on the spot by people in the prior art.
In order to achieve the above purpose, an embodiment of the present invention provides a method for generating oil stain image data on a surface of substation equipment, the method including:
respectively carrying out high-simulation modeling according to the equipment acquired in the transformer substation and the image of the greasy dirt on the surface of the equipment so as to obtain a first three-dimensional model representing the transformer substation and a second three-dimensional model representing the greasy dirt on the surface of the equipment;
adding the second three-dimensional model into the first three-dimensional model by adopting an object grid matching method to obtain a third three-dimensional model which represents a transformer substation with oil stains on the surface of equipment;
selecting a plurality of target viewing angles in the third three-dimensional model based on a point-plane matching method;
acquiring an image part of the oil stain on the surface of the equipment according to the target visual angles;
rendering different pixel values on the equipment and the oil stain on the surface of the equipment in the third three-dimensional model respectively, and representing the pixel values in the form of a semantic segmentation graph;
and combining the semantic segmentation graph serving as a labeling part and the image part to synthesize equipment surface oil stain synthesis data of the transformer substation.
Optionally, the method further comprises:
and when the third three-dimensional model is generated, rendering the third three-dimensional model aiming at different preset weather types.
Optionally, the adding the second three-dimensional model to the first three-dimensional model by adopting the method of object grid matching to obtain a third three-dimensional model representing the transformer substation with the greasy dirt on the equipment surface specifically comprises the following steps:
determining a first grid area of an oil stain map of oil stains on the surface of each device in the second three-dimensional model;
determining a second mesh area of an exterior surface of each device in the first three-dimensional model;
searching the oil stain on the surface of the equipment according to the size relation between the first grid area and the second grid area to obtain an external surface on which the oil stain on the surface of the equipment possibly appears;
for any external surface, in the case that the second grid area is larger than the first grid area, determining that oil stains on the surface of the device corresponding to the first grid area may appear on the external surface corresponding to the second grid area.
Optionally, the method based on the point-plane matching selects a plurality of target perspectives in the third three-dimensional model specifically includes:
selecting a plurality of perspectives for the external surface in the third three-dimensional model to take a picture of the external surface;
determining whether other devices are included in the picture;
deleting the corresponding view angle under the condition that other devices are included in the picture;
in the case that the picture does not include other devices, reserving the corresponding view angle;
all of the retained views are combined to form the plurality of target views.
Optionally, the method further comprises:
and when the semantic segmentation map is generated, encoding the type of the oil stain on the surface of the equipment, the labeling frame, the type of the equipment, the current visual angle, the distance, the illumination and the weather, and adding the encoded data into the semantic segmentation map.
In a second aspect, the present invention also provides a substation equipment surface oil stain image data generating apparatus, the apparatus comprising a processor for being read by a machine to cause the machine to perform a method as described above.
In a third aspect, the present invention further provides a machine vision identification method for substation equipment, where the identification method includes:
generating equipment surface oil stain synthetic data of the transformer substation by adopting the method;
training a preset neural network based on the equipment surface oil stain synthesis data;
and identifying the equipment of the transformer substation by using the trained neural network.
In a fourth aspect, the present invention also provides a machine vision recognition device of substation equipment, the recognition device comprising a processor for performing the recognition method as described above.
In a fifth aspect, the invention also provides a storage medium storing instructions for reading by a machine to cause the machine to perform a method as described in any one of the above.
Through the technical scheme, the method and the device for generating the oil stain image data on the surface of the transformer substation equipment can overcome the technical defect of high difficulty in acquiring the data set due to the fact that the oil stain image on the surface of the transformer substation equipment is acquired on the spot by people in the prior art, and can acquire a large number of data sets in a three-dimensional modeling mode under the condition that only a small amount of acquisition work is input. According to the machine vision recognition method and device for the transformer substation equipment, the equipment surface oil stain synthetic data generated by the method and device are adopted, so that training efficiency and subsequent recognition accuracy of the neural network are improved under the condition that the training set and the testing set are huge enough.
Additional features and advantages of embodiments of the invention will be set forth in the detailed description which follows.
Drawings
The accompanying drawings are included to provide a further understanding of embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain, without limitation, the embodiments of the invention. In the drawings:
FIG. 1 is a flow chart of a substation equipment surface oil stain image data generation method according to one embodiment of the present invention;
FIG. 2 is a flow chart of a method of object grid matching according to one embodiment of the invention;
FIG. 3 is a flow chart of a method of point-to-face matching according to one embodiment of the invention; and
FIG. 4 is an exemplary diagram of a semantic segmentation map and a three-dimensional map according to one example of the invention.
Detailed Description
The following describes specific embodiments of the present invention in detail with reference to the drawings. It should be understood that the detailed description and specific examples, while indicating and illustrating the invention, are not intended to limit the invention.
In the embodiments of the present invention, unless otherwise indicated, terms of orientation such as "upper, lower, top, bottom" are used generally with respect to the orientation shown in the drawings or with respect to the positional relationship of the various components with respect to one another in the vertical, vertical or gravitational directions.
In addition, if there is a description of "first", "second", etc. in the embodiments of the present invention, the description of "first", "second", etc. is for descriptive purposes only and is not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions of the embodiments may be combined with each other, but it is necessary to base that the technical solutions can be realized by those skilled in the art, and when the technical solutions are contradictory or cannot be realized, the combination of the technical solutions should be considered to be absent and not within the scope of protection claimed in the present invention.
Fig. 1 is a flowchart of a substation equipment surface oil stain image data generating method according to an embodiment of the present invention. In fig. 1, the method may include:
in step S10, high-simulation modeling is performed according to the images of the equipment and the oil stain on the surface of the equipment acquired in the transformer substation, so as to obtain a first three-dimensional model representing the transformer substation and a second three-dimensional model representing the oil stain on the surface of the equipment. The first three-dimensional model can be a standard drawing of equipment of the transformer substation, and high-simulation modeling is conducted on each equipment of the transformer substation. The individual devices are then combined to form the first three-dimensional model according to their location distribution at the substation. The second three-dimensional model can be a second three-dimensional model which combines the penetration principle and visual special effects of the oil stain and aims at various states of the oil stain on the surface of the equipment so as to obtain various vivid oil stain patterns, namely the oil stain on the surface of the equipment.
In step S11, the second three-dimensional model is added to the first three-dimensional model by using the object grid matching method to obtain a third three-dimensional model representing the transformer substation with the greasy dirt on the surface of the equipment. For the object grid matching method, there may be steps such as those shown in fig. 2. In fig. 2, the object grid matching method may include:
in step S20, a first grid area of the greasy dirt map of each device surface greasy dirt in the second three-dimensional model is determined.
In step S21, a second mesh area of the exterior surface of each device in the first three-dimensional model is determined.
In step S22, the device surface oil stain is searched according to the size relation between the first grid area and the second grid area to obtain an external surface where the device surface oil stain may occur.
In step S23, for any external surface, in the case where the second mesh area is larger than the first mesh area, it is determined that the oil stain on the surface of the device corresponding to the first mesh area may appear on the external surface corresponding to the second mesh area.
In the method illustrated in fig. 2, each flat external surface on the equipment of the substation may be greasy on the equipment surface due to the particular environment of the substation. However, in rendering the third three-dimensional model, it is contemplated that some of the exterior surfaces may be less than the surface of the device. In this case, the greasy dirt on the surface of the device is forced to be rendered on the external surface, which obviously causes that the generated third three-dimensional model is difficult to meet the requirement. Therefore, in steps S20 to S23, each external surface is searched for a mesh area of oil stains on each device surface, respectively, thereby avoiding a case where the second mesh area is larger than the first mesh area.
Further, external conditions are different due to different devices of the substation. For some devices, only a portion of the device surface may be greasy. In such a context, if the device surface is continuously filled with oil stains in a random distribution manner, more meaningless third three-dimensional models are generated. Thus, the object grid matching method as shown in fig. 2 may also further comprise the step of selecting a corresponding device surface oil stain for different devices. Specifically, it may be that, for example, the corresponding possible occurrence of the greasy dirt on the surface of the device is determined according to the condition of the different devices, and then the method as shown in fig. 2 is performed. And searching the corresponding external surface for the greasy dirt on the surface of each device to judge the size relation between the first grid area and the second grid area. The type and pattern of the equipment surface grease stains that may occur on each external surface are ultimately determined.
In step S12, a plurality of target perspectives are selected in the third three-dimensional model based on the method of point-plane matching. The technical problem to be solved by the invention is that in the prior art, the acquisition difficulty of the data set is high because of relying on artificial field acquisition of oil stain images on the surface of equipment of the transformer substation. By the method as shown in fig. 1, a large number of data sets can be obtained by means of computer three-dimensional modeling with only a small number of acquisition tasks. However, the difference from on-site artificial collection is that: when the image is manually collected on site, the image can be manually judged in real time, so that other equipment or interference factors in the collected image are avoided; however, in the method of the present invention, since computer modeling is adopted, the acquisition of the image can naturally be completed only by a computer, so that it is difficult to determine whether other devices or interference factors are present in the captured image. Therefore, in this step S12, a plurality of target perspectives may be selected for photographing based on the method of point-plane matching. In particular, the method of the point-face matching may be, for example, as shown in fig. 3. In fig. 3, the method of the point-to-surface matching may include:
in step S30, for the external surface in the third three-dimensional model, a plurality of perspectives are selected to take a picture (image) of the external surface.
In step S31, it is determined whether other devices are included in the picture.
In step S32, when another device is included in the picture, the corresponding view is deleted.
In step S33, in the case where no other device is included in the picture, the corresponding view angle is reserved.
In step S34, all the retained viewing angles are combined to constitute a plurality of target viewing angles.
By the method shown in fig. 3, the photographed viewing angles are screened in advance, so that the condition that the photographed images do not meet the requirements is avoided, and the technical problem that the computer modeling photographed images replace the manual photographing in the prior art is solved.
In step S13, an image portion of the surface oil stain of the apparatus is acquired according to a plurality of target viewing angles.
In step S14, each device and device surface oil stain in the third three-dimensional model renders a different pixel value, respectively, and is represented in the form of a semantic segmentation map. In addition, when the semantic segmentation map is generated, the type of oil stain on the surface of the device, the labeling frame, the type of the device, the current visual angle, the distance, the illumination and the weather are encoded and added into the semantic segmentation map, so that the neural network can be conveniently and quickly identified in the subsequent process of training the neural network, and the overall training efficiency is improved. In one example of the present invention, the inventors employed an AirSim plug-in a virtual engine to render an object mask in a scene image. And rendering and storing different colors of the sky, the ground, the transformer substation equipment and oil stain areas of different types on the equipment in the scene when oil stains on the surface of one piece of equipment are generated, and representing the oil stain areas by using a semantic segmentation graph. And then analyzing the obtained semantic segmentation graph, generating corresponding labeling frames and oil stain pattern types and other information according to the connected domain and the pixel value of each part in the graph, wherein the generated labeling effect is shown in figure 4. The right side of the figure is rendered equipment surface greasy dirt image data and a corresponding semantic segmentation diagram, different pixel values are rendered for different models, information such as greasy dirt pattern types and the like is marked through pixel values of small color blocks at the lower left corner, and labeling contents for tasks such as detection and classification can be obtained based on the semantic segmentation diagram according to the numerical value of the pixel values and connected domain analysis.
In step S15, the semantic segmentation map is combined as a labeling part and an image part to synthesize device surface oil stain synthesis data of the substation.
Considering that the subsequent machine identification algorithm may encounter various weather effects of the substation during operation. Thus, in one embodiment of the invention, when generating a third three-dimensional model representing oil dirt with the device surface, the third three-dimensional model may be rendered for a different weather type that is preset. Among other things, the weather types may be different lighting conditions, such as day, noon and night, and different weather types in rainy, sunny, snowy, foggy and sand storm. The specific rendering mode may be rendering by using a virtual engine.
In a second aspect, the present invention also provides a substation equipment surface oil stain image data generating apparatus, which may comprise a processor which may be used to be read by a machine to cause the machine to perform a method as described above.
In a third aspect, the present invention also provides a machine vision recognition method of substation equipment, where the recognition method may be, for example, that first, the method described in any one of the above is used to generate equipment surface oil stain synthetic data of the substation; training a preset neural network based on the oil stain synthesis data on the surface of the equipment; and finally, identifying the equipment of the transformer substation by using the trained neural network.
In a fourth aspect, the present invention also provides a machine vision identification device for substation equipment, which may comprise a processor, which may be used to perform the identification method as described above.
In a fifth aspect, the invention also provides a storage medium having stored thereon instructions which can be used to be read by a machine to cause the machine to perform a method as described in any of the above.
Through the technical scheme, the method and the device for generating the oil stain image data on the surface of the transformer substation equipment can overcome the technical defect of high difficulty in acquiring the data set due to the fact that the oil stain image on the surface of the transformer substation equipment is acquired on the spot by people in the prior art, and can acquire a large number of data sets in a three-dimensional modeling mode under the condition that only a small amount of acquisition work is input. According to the machine vision recognition method and device for the transformer substation equipment, the equipment surface oil stain synthetic data generated by the method and device are adopted, so that training efficiency and subsequent recognition accuracy of the neural network are improved under the condition that the training set and the testing set are huge enough.
The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the embodiments of the present invention are not limited to the specific details of the foregoing embodiments, and various simple modifications may be made to the technical solutions of the embodiments of the present invention within the scope of the technical concept of the embodiments of the present invention, and all the simple modifications belong to the protection scope of the embodiments of the present invention.
In addition, the specific features described in the above embodiments may be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the various possible combinations of embodiments of the invention are not described in detail.
Those skilled in the art will appreciate that all or part of the steps in implementing the methods of the embodiments described above may be implemented by a program stored in a storage medium, including instructions for causing a single-chip microcomputer, chip or the like or processor (processor) to perform all or part of the steps of the methods of the embodiments described herein. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In addition, any combination of the various embodiments of the present invention may be made between the various embodiments, and should also be regarded as disclosed in the embodiments of the present invention as long as it does not deviate from the idea of the embodiments of the present invention.
Claims (9)
1. The method for generating the oil stain image data on the surface of the substation equipment is characterized by comprising the following steps of:
respectively carrying out high-simulation modeling according to the equipment acquired in the transformer substation and the image of the greasy dirt on the surface of the equipment so as to obtain a first three-dimensional model representing the transformer substation and a second three-dimensional model representing the greasy dirt on the surface of the equipment;
adding the second three-dimensional model into the first three-dimensional model by adopting an object grid matching method to obtain a third three-dimensional model which represents a transformer substation with oil stains on the surface of equipment;
selecting a plurality of target viewing angles in the third three-dimensional model based on a point-plane matching method;
acquiring an image part of the oil stain on the surface of the equipment according to the target visual angles;
rendering different pixel values on the equipment and the oil stain on the surface of the equipment in the third three-dimensional model respectively, and representing the pixel values in the form of a semantic segmentation graph;
and combining the semantic segmentation graph serving as a labeling part and the image part to synthesize equipment surface oil stain synthesis data of the transformer substation.
2. The method according to claim 1, wherein the method further comprises:
and when the third three-dimensional model is generated, rendering the third three-dimensional model aiming at different preset weather types.
3. The method according to claim 1, wherein adding the second three-dimensional model to the first three-dimensional model by using the object grid matching method to obtain a third three-dimensional model representing a substation with equipment surface oil stains comprises:
determining a first grid area of an oil stain map of oil stains on the surface of each device in the second three-dimensional model;
determining a second mesh area of an exterior surface of each device in the first three-dimensional model;
searching the oil stain on the surface of the equipment according to the size relation between the first grid area and the second grid area to obtain an external surface on which the oil stain on the surface of the equipment possibly appears;
for any external surface, in the case that the second grid area is larger than the first grid area, determining that oil stains on the surface of the device corresponding to the first grid area may appear on the external surface corresponding to the second grid area.
4. The method according to claim 1, wherein the method based on point-plane matching selects a plurality of target perspectives in the third three-dimensional model specifically comprises:
selecting a plurality of perspectives for an external surface in the third three-dimensional model to take a picture of the external surface;
determining whether other devices are included in the picture;
deleting the corresponding view angle under the condition that other devices are included in the picture;
in the case that the picture does not include other devices, reserving the corresponding view angle;
all of the retained views are combined to form the plurality of target views.
5. The method according to claim 1, wherein the method further comprises:
and when the semantic segmentation map is generated, encoding the type of the oil stain on the surface of the equipment, the labeling frame, the type of the equipment, the current visual angle, the distance, the illumination and the weather, and adding the encoded data into the semantic segmentation map.
6. A substation equipment surface oil stain image data generating device, characterized in that the device comprises a processor for being read by a machine to cause the machine to perform the method of any of claims 1 to 5.
7. A machine vision identification method of substation equipment, characterized in that the identification method comprises:
generating equipment surface oil stain synthesis data of a transformer substation by adopting the method as set forth in any one of claims 1 to 5;
training a preset neural network based on the equipment surface oil stain synthesis data;
and identifying the equipment of the transformer substation by using the trained neural network.
8. A machine vision recognition device of substation equipment, characterized in that the recognition device comprises a processor for performing the recognition method according to claim 7.
9. A storage medium storing instructions for reading by a machine to cause the machine to perform the method of any one of claims 1 to 5 and 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011518392.2A CN112581604B (en) | 2020-12-21 | 2020-12-21 | Substation equipment surface oil stain image data generation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011518392.2A CN112581604B (en) | 2020-12-21 | 2020-12-21 | Substation equipment surface oil stain image data generation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112581604A CN112581604A (en) | 2021-03-30 |
| CN112581604B true CN112581604B (en) | 2024-02-02 |
Family
ID=75136693
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011518392.2A Active CN112581604B (en) | 2020-12-21 | 2020-12-21 | Substation equipment surface oil stain image data generation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112581604B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107690672A (en) * | 2017-07-25 | 2018-02-13 | 深圳前海达闼云端智能科技有限公司 | Training data generation method, generating means and its image, semantic dividing method |
| CN110428388A (en) * | 2019-07-11 | 2019-11-08 | 阿里巴巴集团控股有限公司 | A kind of image-data generating method and device |
| CN111145136A (en) * | 2020-01-02 | 2020-05-12 | 国网安徽省电力有限公司检修分公司 | Synthesis method, system and storage medium for transformer substation meter image data set |
| CN112083000A (en) * | 2020-08-12 | 2020-12-15 | 中国电力科学研究院有限公司 | Intelligent identification method and system for appearance defects of substation equipment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11132826B2 (en) * | 2019-05-16 | 2021-09-28 | Caterpillar Inc. | Artificial image generation for training an object detection system |
-
2020
- 2020-12-21 CN CN202011518392.2A patent/CN112581604B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107690672A (en) * | 2017-07-25 | 2018-02-13 | 深圳前海达闼云端智能科技有限公司 | Training data generation method, generating means and its image, semantic dividing method |
| CN110428388A (en) * | 2019-07-11 | 2019-11-08 | 阿里巴巴集团控股有限公司 | A kind of image-data generating method and device |
| CN111145136A (en) * | 2020-01-02 | 2020-05-12 | 国网安徽省电力有限公司检修分公司 | Synthesis method, system and storage medium for transformer substation meter image data set |
| CN112083000A (en) * | 2020-08-12 | 2020-12-15 | 中国电力科学研究院有限公司 | Intelligent identification method and system for appearance defects of substation equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112581604A (en) | 2021-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111724439B (en) | Visual positioning method and device under dynamic scene | |
| CN110400387A (en) | A kind of joint method for inspecting, system and storage medium based on substation | |
| CN110188835B (en) | Data-enhanced pedestrian re-identification method based on generative confrontation network model | |
| CN111145136B (en) | Synthesis method, system and storage medium for transformer substation meter image data set | |
| CN108363951B (en) | Automatic acquisition method of deep learning sample library corresponding to remote sensing image land type identification | |
| CN101625723A (en) | Rapid image-recognizing method of power line profile | |
| CN115205264A (en) | High-resolution remote sensing ship detection method based on improved YOLOv4 | |
| CN110619623B (en) | Automatic identification method for heating of joint of power transformation equipment | |
| CN106096603A (en) | A kind of dynamic flame detection method merging multiple features and device | |
| CN111160261A (en) | Sample image labeling method and device for automatic sales counter and storage medium | |
| CN114240868A (en) | Unmanned aerial vehicle-based inspection analysis system and method | |
| CN112613454A (en) | Electric power infrastructure construction site violation identification method and system | |
| CN109146880A (en) | A kind of electric device maintenance method based on deep learning | |
| CN113344852A (en) | Target detection method and device for power scene general-purpose article and storage medium | |
| CN113327255A (en) | Power transmission line inspection image processing method based on YOLOv3 detection, positioning and cutting and fine-tune | |
| CN115082254A (en) | Lean control digital twin system of transformer substation | |
| CN114972177A (en) | Road disease identification management method and device and intelligent terminal | |
| CN110610485A (en) | Ultra-high voltage transmission line channel hidden danger early warning method based on SSIM algorithm | |
| CN113435452A (en) | Electrical equipment nameplate text detection method based on improved CTPN algorithm | |
| CN111507398A (en) | Transformer substation metal instrument corrosion identification method based on target detection | |
| CN112581604B (en) | Substation equipment surface oil stain image data generation method and device | |
| CN115439763A (en) | Water engineering surface crack detection method based on YOLOv5 model | |
| CN115239621A (en) | Transformer substation visual intelligent inspection method based on virtual-real matching | |
| CN111307267A (en) | Conductor galloping monitoring method based on concentric circle detection | |
| CN116452943A (en) | Water level identification method for image identification based on convolutional neural network |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: No. 397, Tongcheng South Road, Baohe District, Hefei City, Anhui Province 230061 Applicant after: Super high voltage branch of State Grid Anhui Electric Power Co.,Ltd. Applicant after: HEFEI ZHONGKE LEINAO INTELLIGENT TECHNOLOGY Co.,Ltd. Address before: No. 397, Tongcheng South Road, Baohe District, Hefei City, Anhui Province 230061 Applicant before: STATE GRID ANHUI POWER SUPPLY COMPANY OVERHAUL BRANCH Applicant before: HEFEI ZHONGKE LEINAO INTELLIGENT TECHNOLOGY Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |