CN108761237A - Unmanned plane electric inspection process image vital electrical component diagnoses automatically and labeling system - Google Patents
Unmanned plane electric inspection process image vital electrical component diagnoses automatically and labeling system Download PDFInfo
- Publication number
- CN108761237A CN108761237A CN201810526879.1A CN201810526879A CN108761237A CN 108761237 A CN108761237 A CN 108761237A CN 201810526879 A CN201810526879 A CN 201810526879A CN 108761237 A CN108761237 A CN 108761237A
- Authority
- CN
- China
- Prior art keywords
- power components
- electrical component
- inspection process
- image
- electric inspection
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to a kind of unmanned plane electric inspection process image vital electrical component automatically diagnosis and labeling system, include a power components module of target detection, an electrical component failures detection module, one provide human-computer interaction function human-computer interaction module, a labeling module and a unmanned plane electric inspection process Image Database.The present invention realizes unmanned plane electric inspection process image intelligentization and diagnoses, marks automatically, mitigates patrol officer's workload, promotes routing inspection efficiency and accuracy rate, provides data in the identification of electric inspection process vital electrical component and the application of field of fault detection for artificial intelligence and supports.Meanwhile the regulation and standardization for promoting the storage of electric inspection process image, marking.
Description
Technical field
It is special the present invention relates to artificial intelligence, computer technology, power grid fortune inspection maintenance technology and electric power monitoring technical field
It is not a kind of unmanned plane electric inspection process image vital electrical component diagnosis and labeling system automatically.
Background technology
With the development of economy and society, generated energy rises year by year with electricity consumption.Power grid scale constantly expands, transmission line of electricity fortune
The workload that inspection is safeguarded also increases therewith, and the human and material resources that power department every year will be costly carries out fortune inspection and safeguards
Work.In recent years with the development of unmanned air vehicle technique, in the way of unmanned plane carrying out electric inspection process becomes more and more general
Time, by analyzing the image data that unmanned plane obtains, to find the existing fault condition in inspection.It patrols at present
It is main still by being accomplished manually to examine image analysis, a large amount of inspection image brings huge choose to later data analyzing processing
War, electric inspection process personnel analyze electric inspection process image by eyes for a long time, not only inefficiency, and it is tired also easily to generate vision
Labor so as to cause serious misjudgement or is failed to judge.With the fast development of artificial intelligence technology, the inspection electricity based on deep learning
The identification of power image of component is possibly realized with fault detection technique.The present invention is by trained SSD target detections network, to a large amount of
Electric inspection process image handled, identify inspection image in vital electrical component, its malfunction is detected, and will
It, which is marked out, comes, and to mitigate the operating pressure of electric inspection process personnel, improves the efficiency and intelligent level of electric inspection process.
Invention content
Automatically diagnosis is a kind of unmanned plane electric inspection process image vital electrical component of being designed to provide of invention with mark
System, to overcome defect existing in the prior art.
To achieve the above object, the technical scheme is that：A kind of unmanned plane electric inspection process image vital electrical component
Automatic diagnosis and labeling system, including a power components module of target detection, an electrical component failures detection module, an offer people
The human-computer interaction module of machine interactive function, a labeling module and a unmanned plane electric inspection process Image Database；
The power components module of target detection obtains electric inspection process figure to be marked from the unmanned plane electric inspection process Image Database
Picture by the position of power components classification and power components in the power components target detection model inspection image, and is
The position coordinates of the classification information and the power components of electric inspection process image labeling power components to be marked in the picture, and pass through
The human-computer interaction module is shown；
The electrical component failures detection module obtains the image after power components module of target detection detection mark, leads to
It crosses an electrical component failures detection model and detects the corresponding malfunction of power components and abort situation in the image, and to wait marking
The position coordinates of the fault status information and failure of electric inspection process image labeling power components in the picture are noted, and pass through the people
Machine interactive module is shown；
The labeling module is used to generate the instruction of power components target detection model and electrical component failures detection model pre-training
Practice data and for detecting the mark stage automatically to power components target detection model and electrical component failures detection model
The detection mark image of processing is audited and is corrected, and there are mistakes by audit, and completes revised data as electric power
The training data that component target detection model or electrical component failures detection model are subsequently trained；
After the unmanned plane electric inspection process Image Database is for storing electric inspection process image to be marked, training data and mark processing
Image and its corresponding labeled data.
In an embodiment of the present invention, in the pre-training stage, the labeling module is from the unmanned plane electric inspection process image
Electric inspection process image data to be marked is obtained in library, and power components, power components event are selected in electric inspection process image center to be marked
Hinder region, marks classification information；Wherein, classification information includes：The position of power components classification and power components in the picture is sat
Mark, the position coordinates of electrical component failures state and failure in the picture；After completing mark, image and its labeled data are stored
To the unmanned plane electric inspection process Image Database, and as the training data in pre-training stage.
In an embodiment of the present invention, image, power components classification and the Electricity Department after the mark in training data are extracted
The position coordinates of part in the picture are trained through a SSD component target detection networks as target detection training set, establish institute
State power components target detection model.
In an embodiment of the present invention, image, electrical component failures state and event after the mark in training data are extracted
The position coordinates of barrier in the picture are trained through SSD fault targets detection network as fault detect training set, establish institute
State electrical component failures detection model.
In an embodiment of the present invention, automatic diagnosis and mark are achieved by the steps of：
Step S1：According to the unmanned plane electric inspection process image bank interface, obtained from the unmanned plane electric inspection process Image Database
Electric inspection process image to be marked；
Step S2：It is detected in electric inspection process image to be marked by power components module of target detection and frame selects power components
Region, and mark the position coordinates of power components classification information and power components in the picture；
Step S3：The result of step S2 detection marks is audited via labeling module and human-computer interaction circle module；If examining
It looks into errorless, is then transferred to step S4；Otherwise, it is modified processing, correction result is entered step into S4, and mark the image, as
The follow-up training data of power components target detection model；
Step S4：By the power components region of electrical component failures detection module read step S3 frame choosings, in the area
Simultaneously frame selects electrical component failures region for detection in domain, and marks electrical component failures status information and failure in the picture
Position coordinates；
Step S5：The result of step S4 detection marks is audited via labeling module and human-computer interaction circle module；If examining
It looks into errorless, then enters step S6；Otherwise, it is modified processing, correction result is entered step into S6, and mark the image, as
The follow-up training data of electrical component failures detection model；
Step S6：The handling result of fusion steps S3 and step S5, generate an xml document, be stored in together with image it is described nobody
Electro-mechanical force inspection Image Database；The xml document includes：The position of power components classification information and power components in the picture is sat
The position coordinates of mark and electrical component failures classification information and failure in the picture.
Compared to the prior art, the invention has the advantages that：A kind of unmanned plane electric inspection process provided by the invention
Diagnosis and labeling system can be to unmanned plane electric inspection process images based on depth learning technology automatically for image vital electrical component
In vital electrical component carry out the method and system for diagnosing, marking automatically, realize the crucial electricity of unmanned plane electric inspection process image
Intelligence Diagnosis, the automatic marking of power component are application of the artificial intelligence in electric inspection process critical component identification and fault detect
It provides data to support, promotes the storage of electric inspection process image, the standardization of mark, further increase image labeling and electric inspection process effect
Rate.
Description of the drawings
Fig. 1 is the system knot that unmanned plane electric inspection process image vital electrical component diagnoses automatically with labeling system in the present invention
Composition.
Fig. 2 is the pre-training that unmanned plane electric inspection process image vital electrical component diagnoses automatically with labeling system in the present invention
Flow chart.
Fig. 3 is the workflow that unmanned plane electric inspection process image vital electrical component diagnoses automatically with labeling system in the present invention
Cheng Tu.
Specific implementation mode
Below in conjunction with the accompanying drawings, technical scheme of the present invention is specifically described.
A kind of unmanned plane electric inspection process image vital electrical component of the present invention diagnoses automatically and labeling system, as shown in Figure 1,
The man-machine friendship of human-computer interaction function is provided including a power components module of target detection, an electrical component failures detection module, one
Mutual module, a labeling module and a unmanned plane electric inspection process Image Database.In the pre-training stage, which obtains nothing to be marked
Man-machine electric inspection process image is part unmanned plane electric inspection process image labeling power components classification to be marked and component locations letter
Breath and corresponding electrical component failures state and fault location information make power components target detection model pre-training number
According to collection and electrical component failures detection model pre-training data set, respectively to power components target detection model and power components event
Hinder detection model and carries out pre-training.Using trained power components target detection model treatment electric inspection process image to be marked,
Frame selects the power components position in image and marks power components classification and component position coordinates in the picture；Using training
Electrical component failures detection model handle the power components region that electric inspection process image center to be marked is selected, obtain corresponding electricity
Power unit failure state and abort situation coordinate.By merge in electric inspection process figure power components classification and component position information and
Electrical component failures classification and fault location information are generated comprising in electric inspection process figure where power components classification and power components
Position coordinates and power components fault category and the position coordinates in guilty culprit image xml document.
In the present embodiment, for labeling module, using electric power unmanned plane inspection image power components criteria classification term
Library and electric power unmanned plane inspection image electrical component failures criteria classification terminology bank are as power components in image and its failure shape
The mark Naming conventions of state.Labeling module obtains electric inspection process to be marked according to electric inspection process image bank interface from Image Database
Image data selects power components, electrical component failures region in electric inspection process image center to be marked, marks classification information.
Classification information includes the position coordinates of the classification and component of power components in the picture, and the malfunction and failure of power components exist
Position coordinates in image.
In the present embodiment, labeling module is interacted by human-computer interaction module with system operators, on the one hand, in system
Early period is run, operating personnel are labeled electric inspection process image using labeling module, make power components target detection model
With the pre-training data set of electrical component failures detection model；On the other hand, system during normal operation, operating personnel are to being
The automatic image for detecting, marking of system is audited, and is modified by labeling module to error label information, is power components mesh
The further training for marking detection model and electrical component failures detection model provides data.
In the present embodiment, for power components module of target detection, using electric power unmanned plane inspection image power components
Classification foundation of the standard categorization approach as power components in image, using electric power unmanned plane inspection image power components standard scores
Classification And Nomenclature specification of the class terminology bank as power components in image.
In the training stage, power components module of target detection uses SSD component target detection networks, utilizes labeling module mark
Note the classification of power components in processed image and image labeling, component position information is instructed as pre-training data set
Practice, obtains power components target detection model.And it will be detected mark processing in power components module of target detection, but handle
It is wrong, audit modification treated image and revised markup information are carried out as follow-up training number via labeling module
According to continuing to train to power components target detection model, further increase the accuracy of mark.In this process, by decimal
Power components target detection model according to collection pre-training is the SSD networks for target detection, and pre-training data set is smaller, but logical
It crosses system itself and is constantly produced from training dataset, it is accurate that the target identification of oneself will be continuously improved in power components target detection model
True rate.
In detection-phase, power components module of target detection, by obtaining electric inspection process to be marked from image database
Image data utilizes the vital electrical in the power components target detection model inspection image in advance Jing Guo small data set pre-training
Component categories and its position are that the position of the classification information and component of image labeling power components to be marked in figure is sat
Mark.
In the present embodiment, for electrical component failures detection module, using electric power unmanned plane inspection image power components
Classification foundation of the standard failure sorting technique as electrical component failures state in image, using electric power unmanned plane inspection image electricity
Classification And Nomenclature specification of the power unit failure criteria classification terminology bank as electrical component failures state in image.
In the training stage, electrical component failures detection module detects network using SSD fault targets, utilizes labeling module
It marks processed image artwork and image labeling information is trained as pre-training data set, obtain electrical component failures and examine
Disconnected model.And it will be detected mark processing in electrical component failures detection module, but processing is wrong, is carried out via labeling module
Audit modification treated image and revised markup information detect mould as follow-up training data to electrical component failures
Type continues to train, and further increases the accuracy of mark.In this process, by the electrical component failures of small data set pre-training
Detection model is the SSD networks for fault detect, and pre-training data set is smaller, but constantly generates self-training by system itself
The fault detect accuracy rate of oneself will be continuously improved in data set, the electrical component failures detection model mesh.
In detection-phase, the processed mark image of power components module of target detection and its corresponding mark letter are obtained
Breath utilizes the Electricity Department being detected in the electrical component failures detection model detection image in advance Jing Guo small data set pre-training
The corresponding malfunction of part and position, be image labeling power components to be marked fault status information and failure in figure
Coordinate position.
In the present embodiment, the pre-training training method of this system is as shown in Figure 2.Power components target detection model, in advance
Training set is made of the power components classification information of a certain number of electric inspection process image artworks for completing mark and image labeling,
Power components classification information refers to the position coordinates in the classification of power components and power components place artwork in image.Electricity Department
Part Fault Model, pre-training collection is by a certain number of electric inspection process image artworks for completing mark, the electric power of image labeling
Part classification information and the electrical component failures classification information of image labeling composition, electrical component failures classification information refer to image
Position coordinates in the malfunction and guilty culprit artwork of middle corresponding power components.
In the present embodiment, the workflow under this system normal operating condition is as shown in Figure 3.The power components target
Detection module and electrical component failures detection module, the detection carried out to electric inspection process image to be marked, annotation results are logical
Crossing human-computer interaction interface transfers to system operators to be audited.Wrong markup information is found for audit, by manually passing through
Labeling module is modified, and revised result is entered next work step, and by revised markup information and correspondence
The corresponding training set of image artwork income, corresponding detection model is further trained.Specifically comprise the following steps：
Step S1, according to electric inspection process image bank interface, unmanned plane electric inspection process image data is obtained from Image Database；
Step S2, using power components module of target detection, the unmanned plane electric inspection process image to be marked of acquisition is handled,
Detection, frame select power components region, and mark component categories and component in the electric inspection process image to be marked
Position coordinates in the picture.
Step S3, the region for selecting the step S2 power components classifications identified and frame via human-computer interaction interface into pedestrian
Work examines.Such as examine it is errorless, then enter next step；Otherwise to examining that wrong information carries out artificial correction, by correction result
Be included into target detection training set into next step, and by the image and its modified markup information, to target detection model into
Row further training.
Step S4, using electrical component failures detection module, the choosing electricity of frame described in step S3 in image to be marked is read
Power component area, detection, frame select electrical component failures region, and mark the event of corresponding power components in the region
The position coordinates of barrier state and failure in artwork.
Step S5, by the malfunction of the step S4 power components identified and frame favored area via human-computer interaction interface into
Row manual review.Such as examine it is errorless, then enter next step；Otherwise it to examining that wrong information carries out artificial correction, will correct
As a result enter next step, and the image and its modified markup information are included into fault detect training set, to fault detect mould
Type is further trained.
Step S6, the handling result of fusion steps S3 and step S5, it includes power components classification and electric power further to generate
Fault category and guilty culprit the area coordinate letter of location coordinate information and power components in electric inspection process figure where component
The xml document of breath.
In the present embodiment, the power components classification mark and electrical component failures classification mark of all progress, whether
It is manually realized by labeling module, or is realized by power components module of target detection or electrical component failures detection module
, defer to electric power unmanned plane inspection image power components standard categorization approach and electric power unmanned plane inspection image power components mark
Quasi- Fault Classification, using electric power unmanned plane inspection image power components criteria classification terminology bank and electric power unmanned plane inspection shadow
As electrical component failures criteria classification terminology bank marks Naming conventions as it.
In the present embodiment, the markup information of generation is stored with the file of xml formats.Xml formatted files it is interior
Hold and includes mainly：The size of electric inspection process picture；The vital electrical component categories information and component that image is included are in the picture
Location information；The location information of the malfunction and failure for the vital electrical component that image is included in the picture.Wherein, mesh
Tetra- values of target location information Xmin, Ymin, Xmax, Ymax indicate.
In the present embodiment, with the continuous expansion of training dataset and the continuous training to model, power components target
Detection module and electrical component failures detection module accurately can carry out intelligent dimension to electric inspection process image, gradually subtract
Light system is to artificial degree of dependence, from needing artificial mark to establish pre-training collection, to artificial only responsible audit, until final thorough
Bottom does not need manual intervention and automatic marking can be realized.
The above are preferred embodiments of the present invention, all any changes made according to the technical solution of the present invention, and generated function is made
When with range without departing from technical solution of the present invention, all belong to the scope of protection of the present invention.
Claims (5)
1. a kind of unmanned plane electric inspection process image vital electrical component diagnoses automatically and labeling system, which is characterized in that including one
Power components module of target detection, an electrical component failures detection module, one provide human-computer interaction function human-computer interaction module,
One labeling module and a unmanned plane electric inspection process Image Database；
The power components module of target detection obtains electric inspection process figure to be marked from the unmanned plane electric inspection process Image Database
Picture by the position of power components classification and power components in the power components target detection model inspection image, and is
The position coordinates of the classification information and the power components of electric inspection process image labeling power components to be marked in the picture, and pass through
The human-computer interaction module is shown；
The electrical component failures detection module obtains the image after power components module of target detection detection mark, leads to
It crosses an electrical component failures detection model and detects the corresponding malfunction of power components and abort situation in the image, and to wait marking
The position coordinates of the fault status information and failure of electric inspection process image labeling power components in the picture are noted, and pass through the people
Machine interactive module is shown；
The labeling module is used to generate the instruction of power components target detection model and electrical component failures detection model pre-training
Practice data and for detecting the mark stage automatically to power components target detection model and electrical component failures detection model
The detection mark image of processing is audited and is corrected, and there are mistakes by audit, and completes revised data as electric power
The training data that component target detection model or electrical component failures detection model are subsequently trained；
After the unmanned plane electric inspection process Image Database is for storing electric inspection process image to be marked, training data and mark processing
Electric inspection process image and its corresponding labeled data.
2. unmanned plane electric inspection process image vital electrical component according to claim 1 diagnoses automatically and labeling system,
It is characterized in that, in the pre-training stage, the labeling module obtains electric power to be marked from the unmanned plane electric inspection process Image Database
Inspection image data selects power components, electrical component failures region, mark classification letter in electric inspection process image center to be marked
Breath；Wherein, classification information includes：The position coordinates of power components classification and power components in the picture, electrical component failures shape
The position coordinates of state and failure in the picture；After completing mark, image and its labeled data are stored to the unmanned electro-mechanical force
Inspection Image Database, and as the training data in pre-training stage.
3. unmanned plane electric inspection process image vital electrical component according to claim 1 diagnoses automatically and labeling system,
It is characterized in that, extracts the image after the mark in training data, the position of power components classification and power components in the picture is sat
It is denoted as target detection training set, being trained through a SSD component target detection networks, establishing the power components target detection
Model.
4. unmanned plane electric inspection process image vital electrical component according to claim 1 diagnoses automatically and labeling system,
It is characterized in that, extracts the image after the mark in training data, the position of electrical component failures state and failure in the picture is sat
It is denoted as fault detect training set, being trained through SSD fault targets detection network, establishes the electrical component failures detection
Model.
5. unmanned plane electric inspection process image vital electrical component according to claim 1 diagnoses automatically and labeling system,
It is characterized in that, is achieved by the steps of automatic diagnosis and mark：
Step S1：According to the unmanned plane electric inspection process image bank interface, obtained from the unmanned plane electric inspection process Image Database
Electric inspection process image to be marked；
Step S2：It is detected in electric inspection process image to be marked by power components module of target detection and frame selects power components
Region, and mark the position coordinates of power components classification information and power components in the picture；
Step S3：The result of step S2 detection marks is audited via labeling module and human-computer interaction circle module；If examining
It looks into errorless, is then transferred to step S4；Otherwise, it is modified processing, correction result is entered step into S4, and mark the image, as
The follow-up training data of power components target detection model；
Step S4：By the power components region of electrical component failures detection module read step S3 frame choosings, in the area
Simultaneously frame selects electrical component failures region for detection in domain, and marks electrical component failures status information and failure in the picture
Position coordinates；
Step S5：The result of step S4 detection marks is audited via labeling module and human-computer interaction circle module；If examining
It looks into errorless, then enters step S6；Otherwise, it is modified processing, correction result is entered step into S6, and mark the image, as
The follow-up training data of electrical component failures detection model；
Step S6：The handling result of fusion steps S3 and step S5, generate an xml document, be stored in together with image it is described nobody
Electro-mechanical force inspection Image Database；The xml document includes：The position of power components classification information and power components in the picture is sat
The position coordinates of mark and electrical component failures classification information and failure in the picture.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810526879.1A CN108761237A (en) | 2018-05-29 | 2018-05-29 | Unmanned plane electric inspection process image vital electrical component diagnoses automatically and labeling system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810526879.1A CN108761237A (en) | 2018-05-29 | 2018-05-29 | Unmanned plane electric inspection process image vital electrical component diagnoses automatically and labeling system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108761237A true CN108761237A (en) | 2018-11-06 |
Family
ID=64003170
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810526879.1A Pending CN108761237A (en) | 2018-05-29 | 2018-05-29 | Unmanned plane electric inspection process image vital electrical component diagnoses automatically and labeling system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108761237A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109685930A (en) * | 2018-12-19 | 2019-04-26 | 深圳供电局有限公司 | Power components method for inspecting, device, computer equipment and storage medium |
| CN110059076A (en) * | 2019-04-19 | 2019-07-26 | 国网山西省电力公司电力科学研究院 | A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment |
| CN110298932A (en) * | 2019-07-26 | 2019-10-01 | 国网浙江省电力有限公司衢州供电公司 | A kind of UAV Intelligent patrol inspection integral control platform |
| CN110633702A (en) * | 2019-11-20 | 2019-12-31 | 江西博微新技术有限公司 | Unmanned aerial vehicle-based line maintenance charge calculation method, server and storage medium |
| CN111045452A (en) * | 2019-12-17 | 2020-04-21 | 昆明联诚科技股份有限公司 | Power line inspection method based on deep learning |
| CN111091543A (en) * | 2019-12-12 | 2020-05-01 | 哈尔滨市科佳通用机电股份有限公司 | Railway wagon swing bolster spring loss fault target detection method |
| CN111311967A (en) * | 2020-03-31 | 2020-06-19 | 普宙飞行器科技（深圳）有限公司 | Unmanned aerial vehicle-based power line inspection system and method |
| CN111815581A (en) * | 2020-06-27 | 2020-10-23 | 国网上海市电力公司 | Power isolating switch and automatic part identification method and device thereof |
| CN111860217A (en) * | 2020-06-30 | 2020-10-30 | 深圳优地科技有限公司 | Fault detection method and device, electronic equipment and storage medium |
| CN112102443A (en) * | 2020-09-15 | 2020-12-18 | 国网电力科学研究院武汉南瑞有限责任公司 | Marking system and marking method suitable for substation equipment inspection image |
| CN113239752A (en) * | 2021-04-27 | 2021-08-10 | 西安万飞控制科技有限公司 | Unmanned aerial vehicle aerial image automatic identification system |
| CN114061665A (en) * | 2021-11-19 | 2022-02-18 | 国网重庆市电力公司电力科学研究院 | Non-contact type inspection system and method applied to power distribution station room equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104361128A (en) * | 2014-12-05 | 2015-02-18 | 河海大学 | Data synchronization method of PC (Personnel Computer) end and mobile terminal based on hydraulic polling business |
| CN105141889A (en) * | 2015-07-28 | 2015-12-09 | 国家电网公司 | Power transmission line intelligent patrol system based on image vision |
| CN106326808A (en) * | 2015-06-23 | 2017-01-11 | 上海深邃智能科技有限公司 | Method for detecting bird nests in power transmission line poles based on unmanned plane images |
| CN106491216A (en) * | 2016-10-28 | 2017-03-15 | 苏州朗开医疗技术有限公司 | The internal destination object alignment system of one kind diagnosis and medical treatment alignment system |
| CN107607207A (en) * | 2017-08-22 | 2018-01-19 | 中国科学院深圳先进技术研究院 | A kind of power equipment thermal fault detection method, system and electronic equipment |
-
2018
- 2018-05-29 CN CN201810526879.1A patent/CN108761237A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104361128A (en) * | 2014-12-05 | 2015-02-18 | 河海大学 | Data synchronization method of PC (Personnel Computer) end and mobile terminal based on hydraulic polling business |
| CN106326808A (en) * | 2015-06-23 | 2017-01-11 | 上海深邃智能科技有限公司 | Method for detecting bird nests in power transmission line poles based on unmanned plane images |
| CN105141889A (en) * | 2015-07-28 | 2015-12-09 | 国家电网公司 | Power transmission line intelligent patrol system based on image vision |
| CN106491216A (en) * | 2016-10-28 | 2017-03-15 | 苏州朗开医疗技术有限公司 | The internal destination object alignment system of one kind diagnosis and medical treatment alignment system |
| CN107607207A (en) * | 2017-08-22 | 2018-01-19 | 中国科学院深圳先进技术研究院 | A kind of power equipment thermal fault detection method, system and electronic equipment |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109685930A (en) * | 2018-12-19 | 2019-04-26 | 深圳供电局有限公司 | Power components method for inspecting, device, computer equipment and storage medium |
| CN110059076A (en) * | 2019-04-19 | 2019-07-26 | 国网山西省电力公司电力科学研究院 | A kind of Mishap Database semi-automation method for building up of power transmission and transformation line equipment |
| CN110298932A (en) * | 2019-07-26 | 2019-10-01 | 国网浙江省电力有限公司衢州供电公司 | A kind of UAV Intelligent patrol inspection integral control platform |
| CN110633702A (en) * | 2019-11-20 | 2019-12-31 | 江西博微新技术有限公司 | Unmanned aerial vehicle-based line maintenance charge calculation method, server and storage medium |
| CN111091543B (en) * | 2019-12-12 | 2020-10-09 | 哈尔滨市科佳通用机电股份有限公司 | Railway wagon swing bolster spring loss fault target detection method |
| CN111091543A (en) * | 2019-12-12 | 2020-05-01 | 哈尔滨市科佳通用机电股份有限公司 | Railway wagon swing bolster spring loss fault target detection method |
| CN111045452A (en) * | 2019-12-17 | 2020-04-21 | 昆明联诚科技股份有限公司 | Power line inspection method based on deep learning |
| CN111311967A (en) * | 2020-03-31 | 2020-06-19 | 普宙飞行器科技（深圳）有限公司 | Unmanned aerial vehicle-based power line inspection system and method |
| CN111815581A (en) * | 2020-06-27 | 2020-10-23 | 国网上海市电力公司 | Power isolating switch and automatic part identification method and device thereof |
| CN111860217A (en) * | 2020-06-30 | 2020-10-30 | 深圳优地科技有限公司 | Fault detection method and device, electronic equipment and storage medium |
| CN112102443A (en) * | 2020-09-15 | 2020-12-18 | 国网电力科学研究院武汉南瑞有限责任公司 | Marking system and marking method suitable for substation equipment inspection image |
| CN113239752A (en) * | 2021-04-27 | 2021-08-10 | 西安万飞控制科技有限公司 | Unmanned aerial vehicle aerial image automatic identification system |
| CN113239752B (en) * | 2021-04-27 | 2024-06-18 | 西安万飞控制科技有限公司 | Unmanned aerial vehicle aerial image automatic identification system |
| CN114061665A (en) * | 2021-11-19 | 2022-02-18 | 国网重庆市电力公司电力科学研究院 | Non-contact type inspection system and method applied to power distribution station room equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108761237A (en) | Unmanned plane electric inspection process image vital electrical component diagnoses automatically and labeling system | |
| CN111582235B (en) | Alarm method, system and equipment for monitoring abnormal events in station in real time | |
| CN108537154A (en) | Transmission line of electricity Bird's Nest recognition methods based on HOG features and machine learning | |
| CN101576956B (en) | On-line character detection method based on machine vision and system thereof | |
| CN109389161A (en) | Rubbish identification evolutionary learning method, apparatus, system and medium based on deep learning | |
| CN104268588B (en) | Railway wagon brake shoe pricker loses the automatic testing method of failure | |
| CN111368690A (en) | Deep learning-based video image ship detection method and system under influence of sea waves | |
| CN105260734A (en) | Commercial oil surface laser code recognition method with self modeling function | |
| KR20200110979A (en) | Predictive maintenance system for defects in plant facilities | |
| CN106371013A (en) | Picture identification-based GIS switch fault automatic identification system | |
| CN108711148A (en) | A kind of wheel tyre defect intelligent detecting method based on deep learning | |
| CN113850799B (en) | YOLOv 5-based trace DNA extraction workstation workpiece detection method | |
| CN110931112A (en) | Brain medical image analysis method based on multi-dimensional information fusion and deep learning | |
| CN113205039A (en) | Power equipment fault image identification and disaster investigation system and method based on multiple DCNNs | |
| CN113485432A (en) | Photovoltaic power station electroluminescence intelligent diagnosis system and method based on unmanned aerial vehicle | |
| CN115171045A (en) | YOLO-based power grid operation field violation identification method and terminal | |
| CN115019294A (en) | Pointer instrument reading identification method and system | |
| CN117114420B (en) | Image recognition-based industrial and trade safety accident risk management and control system and method | |
| CN113487570A (en) | High-temperature continuous casting billet surface defect detection method based on improved yolov5x network model | |
| Zhou et al. | Wind turbine actual defects detection based on visible and infrared image fusion | |
| CN113420839B (en) | Semi-automatic labeling method and segmentation positioning system for stacking planar target objects | |
| CN116912872A (en) | Drawing identification method, device, equipment and readable storage medium | |
| CN112927222B (en) | Method for realizing multi-type photovoltaic array hot spot detection based on hybrid improved Faster R-CNN | |
| Liu et al. | Channel-Spatial attention convolutional neural networks trained with adaptive learning rates for surface damage detection of wind turbine blades | |
| Yang et al. | A Deep Learning-based Framework for Vehicle License Plate Detection. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181106 |
|
| RJ01 | Rejection of invention patent application after publication |