CN109902647A - The portable online Bird's Nest intelligent identification Method of one kind and system - Google Patents
The portable online Bird's Nest intelligent identification Method of one kind and system Download PDFInfo
- Publication number
- CN109902647A CN109902647A CN201910176773.8A CN201910176773A CN109902647A CN 109902647 A CN109902647 A CN 109902647A CN 201910176773 A CN201910176773 A CN 201910176773A CN 109902647 A CN109902647 A CN 109902647A
- Authority
- CN
- China
- Prior art keywords
- bird
- nest
- intelligent
- identification method
- detection
- 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.)
- Granted
Links
Abstract
The invention discloses a kind of portable online Bird's Nest intelligent identification Method and systems, system, which uses, is based on Embedded design scheme of hardware and software, it is small and exquisite portable, only need temporary erection in locomotive driving platform, high-definition shooting can be carried out to the Bird's Nest of Along Railway catenary mast and realizes on-line intelligence inspection, simultaneously the Bird's Nest precise positioning that will test and and alarm, can effectively find the Bird's Nest on Along Railway catenary mast.
Description
Technical field
The present invention relates to contact net power supply safety detection field, especially a kind of portable online Bird's Nest intelligent identification Method
And system.
Background technique
Bird's Nest inspection is the important process content of contact net power supply safety detection, birds pillar along the railway or transmission of electricity
Nesting on overhead line structures can cause adverse effect to equipment such as contact nets, or even endanger the safe operation of train, for
Existing Bird's Nest, generally carries out Bird's Nest detection by way of manual inspection or off line data analysis, but these traditional approach
It not only takes time and effort, but also security risk cannot be excluded in time, very big puzzlement is brought to railroad operator, therefore, be badly in need of one
Online, timely, the intelligent Bird's Nest method for inspecting of kind and system.
Summary of the invention
For technical problem present in background technique, the present invention proposes a kind of portable online Bird's Nest intelligent recognition side
Method, the Bird's Nest intelligent identification Method specifically include:
S1: real time picture is obtained;
S2: deep learning network model is utilized, real time picture is recognized, and provide identification result;
S3: if identification result is thought to carry out early warning there are Bird's Nest;It is no to then follow the steps S1 or terminate the recognition methods.
It preferably, also include the secondary-confirmation of identification result, according to two in the step S2 of the Bird's Nest intelligent identification Method
Secondary confirmation is as a result, update identification result.
Preferably, the secondary-confirmation of the identification result, according to secondary-confirmation as a result, updating the specific method of identification result
Are as follows:
S21: the priori Characteristics Detection based on Bird's Nest structure feature is carried out;If priori Characteristics Detection result is judged as non-Bird's Nest,
Identification result is updated, and is jumped as step S3;Otherwise, step S22 is jumped directly to;
S22: the multi-target detection algorithm based on deep learning is used, the big target depended on to Bird's Nest detects, if not depositing
In big target, then identification result is updated, and the S3 that gos to step, otherwise, jumps directly to step S3.
Preferably, the priori Characteristics Detection rule of the S21 includes one or more of following rule:
1) judge whether the length-width ratio of Bird's Nest is in setting range, if so, being considered Bird's Nest;
2) judge whether height of the Bird's Nest apart from ground is in setting range, if so, being considered Bird's Nest;
3) judge whether the size of Bird's Nest is in setting range, if so, being considered Bird's Nest.
Preferably, the concrete methods of realizing of the S22 are as follows:
S221: establish training sample database: the image that these are contained Bird's Nest is classified simultaneously according to the big target that Bird's Nest is depended on
Label;The big target includes tree, signal transmitting tower, power circuit bar, catenary mast;
S222: deep learning network training sample training: is carried out using the target data marked;
S223: result detection: it is directed to real time picture, carries out big target detection;
If not detecting catenary mast in image, it is considered non-contact net pillar Bird's Nest, and updates identification knot
Fruit;
If detecting catenary mast, judge whether Bird's Nest position and shore position are in a certain range, if so, recognizing
For catenary mast, there are Bird's Nests.
Preferably, the concrete methods of realizing of early warning is carried out in the S3 are as follows:
In conjunction with one grade of basic database of GPS positioning information and a bar, the accurate location information where Bird's Nest is confirmed, and export report
Alert information.
Preferably, the warning message is by being wirelessly transmitted to mobile display device.
Preferably, in the S2 deep learning network specific construction method are as follows:
1) the deep learning network model, including convolutional layer, pond layer, full articulamentum and classifier layer are constructed;
2) the Bird's Nest local feature extracted in image is calculated by the convolutional layer;
3) sampling processing is carried out by local feature of the pond layer to image, farthest reduces the resolution ratio of image;
4) it is handled by the complete layer-by-layer feature extraction of articulamentum;
5) class probability by classifier layer prediction Bird's Nest and exact position, defining classification device loss function, described point
Class device loss function includes classification loss and position loss.
Meanwhile the present invention also proposes a kind of portable Bird's Nest cruising inspection system, including data acquisition module, central processing mould
Block, mobile display device;The data acquisition module, central processing module, mobile display device are sequentially connected;
Data acquisition module includes high-definition image acquisition unit and locating module;
Central processing module includes compression storing data unit, core control panel, Power Management Unit, acquisition module control list
Member, abnormal alarm unit, Bird's Nest intelligent detection unit, database;The compression storing data unit, is adopted at Power Management Unit
Collection module control unit, abnormal alarm unit, Bird's Nest intelligent detection unit, database are connected with core control panel;
It is connected between the abnormal alarm unit and mobile display device.
The invention has the benefit that replacing traditional manual inspection using portable online Bird's Nest intelligent inspection system
Mode carries out Bird's Nest defects detection, can find Bird's Nest and accurate positionin on catenary mast, in real time so as to railroad operator
Hidden danger removing is carried out in time, it is ensured that railway power supply safety saves manpower and time cost simultaneously, improves working efficiency.This system
It is so small and exquisite portable, only to need temporary erection in locomotive driving platform, so that it may to iron based on Embedded design scheme of hardware and software
The Bird's Nest of curb line pillar carries out high-definition shooting and intelligent patrol detection, while the Bird's Nest precise positioning and alarm that will test.
Detailed description of the invention
Fig. 1 is portable online Bird's Nest intelligent identifying system structural schematic diagram;
Fig. 2 is the structural schematic diagram of deep learning network in portable online Bird's Nest intelligent identification Method;
Fig. 3 is portable online Bird's Nest recognition methods flow chart.
Specific embodiment
For a clearer understanding of the technical characteristics, objects and effects of the present invention, this hair of Detailed description of the invention is now compareed
Bright specific embodiment.
As shown in Fig. 1, portable online Bird's Nest cruising inspection system includes data acquisition module, central processing module, movement
Show equipment.The data acquisition module, central processing module, mobile display device are sequentially connected.
Data acquisition module includes high-definition image acquisition unit and locating module.High-definition image acquisition unit uses a work
Industry camera carries out image real-time acquisition.The camera selects high-speed industrial camera and the conjunction of adequate resolution (2048*1024)
The camera lens of suitable focal length (25mm) carries out high frame per second shooting, this system to the catenary mast bar on railway both sides at an appropriate angle
In order to meet the requirement of complete covering shooting, acquisition frame rate cannot be below 40 frames/second.
Central processing module includes compression storing data unit, core control panel, Power Management Unit, acquisition module control
Unit, abnormal alarm unit, Bird's Nest intelligent detection unit, database store these image datas of camera acquisition to limited
Memory space in, in order to realize efficient compaction algorithms ability, compression storing data unit is using the JPEG figure based on FPGA
As compression chip, high compressed capability and power consumption Energy Efficiency Ratio can be obtained, positioning unit is real using the positioning of GPS/ Beidou bimodulus
Existing precise positioning, panoramic picture collecting part and framing unit match and may be implemented to determine image geographical location
Position, compression storing data unit dodge the storage sky that -3D editions 500G solid state hard disks of enlightening most distinguished high speed series constitute 2TB using four pieces
Between, the data and the corresponding position from framing unit for compressing storage unit output for storing data are believed
Breath, core control panel are used to control the data transmission in data acquisition module, are substantially carried out three work: 1) by JPEG compression list
The location information that the output data and framing unit of member obtain is delivered directly to data storage cell;2) JPEG compression is drawn
The location information synchronizing that the raw image data and framing unit for holding up input terminal acquisition obtain is transferred to central processing module;
3) by the raw image data synchronous transfer of JPEG compression engine input terminal acquisition to mobile display device.
Central processing module uses NVIDIA TX2 nucleus module, built-in 6 core CPU, 256 core GPU are a stylobates in
The AI single module supercomputer of NVIDIA Pascal framework is capable of providing powerful operational capability, low-profile, energy conservation height
Effect can cast large-scale and complex deep neural network, and GPU can be with the real-time perfoming Bird's Nest intelligently inspection in acquisition inside Tx2
It surveys, central processing module is substantially carried out five work, is respectively as follows:
A. image is analyzed using deep learning method, obtains the specific location where Bird's Nest, the Bird's Nest intelligent recognition
Method specifically includes:
S1: real time picture is obtained;S2: utilizing deep learning network model, recognize to real time picture, and provides identification knot
Fruit;S3: if identification result is thought to carry out early warning there are Bird's Nest;It is no to then follow the steps S1 or terminate the recognition methods.It is preferred that
, in the step S2 of the Bird's Nest intelligent identification Method, also include identification result secondary-confirmation, according to secondary-confirmation as a result,
Update identification result.Preferably, the secondary-confirmation of the identification result, according to secondary-confirmation as a result, updating the tool of identification result
Body method are as follows: S21: the priori Characteristics Detection based on Bird's Nest structure feature is carried out;If priori Characteristics Detection result is judged as non-bird
Nest then updates identification result, and jumps as step S3;Otherwise, step S22 is jumped directly to;S22: using based on deep learning
Multi-target detection algorithm, the big target depended on to Bird's Nest detects, and big target, then update identification result if it does not exist,
And otherwise the S3 that gos to step jumps directly to step S3.Preferably, the priori Characteristics Detection rule of the S21 includes as follows
One or more of rule: 1) judging whether the length-width ratio of Bird's Nest is in setting range, if so, being considered Bird's Nest;2)
Judge whether height of the Bird's Nest apart from ground is in setting range, if so, being considered Bird's Nest;3) judge Bird's Nest size whether
In setting range, if so, being considered Bird's Nest.Preferably, the concrete methods of realizing of the S22 are as follows: S221: training sample is established
These: being contained the image of Bird's Nest by this library, is classified and is marked according to the big target that Bird's Nest is depended on;The big target includes
Tree, signal transmitting tower, power circuit bar, catenary mast;S222: it sample training: is carried out using the target data marked deep
Spend learning network training;S223: result detection: it is directed to real time picture, carries out big target detection;If do not detected in image
Catenary mast is then considered non-contact net pillar Bird's Nest, and updates identification result;If detecting catenary mast,
Then judge whether Bird's Nest position and shore position are in a certain range, there are Bird's Nests if so, thinking catenary mast.It is preferred that
, the concrete methods of realizing of early warning is carried out in the S3 are as follows: in conjunction with one grade of basic database of GPS positioning information and a bar, really
Recognize the accurate location information where Bird's Nest, and exports warning message.Preferably, the warning message is by being wirelessly transmitted to movement
Show equipment.Preferably, in the S2 deep learning network specific construction method as shown in Fig. 2, specifically: 1) building described in
Deep learning network model, including convolutional layer, pond layer, full articulamentum and classifier layer;2) pass through the convolutional layer meter
Calculate the Bird's Nest local feature extracted in image;3) sampling processing is carried out by local feature of the pond layer to image, it is maximum
Reduce to degree the resolution ratio of image;4) it is handled by the complete layer-by-layer feature extraction of articulamentum;5) pass through the classifier layer
Predict class probability and the exact position of Bird's Nest, defining classification device loss function, the classifier loss function includes classification damage
Become estranged position loss.
B., intelligent recognition is gone out to the image of bird's nest and corresponding location information is transferred to Bird's Nest staqtistical data base, Bird's Nest system
Library is counted for storing the image comprising bird's nest and location information corresponding thereto, and can be after inspection, root
Bird's Nest early warning report is generated according to the information stored, staff is showed by mobile display device, it is subsequent for staff
It is checked and is overhauled;
C., intelligent recognition is gone out to the image of bird's nest and corresponding location information is transferred to mobile display device, carries out abnormal report
It is alert;
D. the normal operation of data acquisition module is controlled;
E. the normal power supply of power supply module is controlled.
Power supply module uses lithium battery to be powered for the operation of system.
Mobile display device is substantially carried out quadrinomial job: 1) image information that real-time display data acquisition module obtains;2)
Real-time display intelligent recognition go out bird's nest image and corresponding location information, issue Bird's Nest warning message;3) in display interface
High-definition image acquisition unit control interface is set, and user can adjust the course of work and related ginseng of panorama camera by the interface
Related control information can be transferred to central processing module by number, mobile display device, realize the control for data acquisition module;
4) after inspection, Bird's Nest early warning report is shown for staff, is checked for staff.
In in high-intensitive cabinet, the cabinet is erected at erection dress for data acquisition module, central processing module integrated installation
It sets, stringer provides the decorating position and angle Selection of flexibility and changeability, holder using sucker, tripod cooperation holder
System can support level rotate angle: 0 °~360 °, vertical rotation angle: -75 °~+40 °, fully ensure that contact net branch
Mast and its ambient enviroment carry out having a good shooting angle when Image Acquisition, and guarantee that system is in when acquiring data
Stable state.
Mobile display device is realized by Wi-Fi to be transmitted with the data of data acquisition module and central processing module.
Based on Embedded portable Bird's Nest cruising inspection system use be based on Embedded design scheme of hardware and software, it is small and exquisite just
It takes, only needs temporary erection in locomotive driving platform, so that it may which high-definition shooting and intelligence are carried out to the Bird's Nest of Along Railway catenary mast
Can inspection, while the Bird's Nest precise positioning that will test and and alarm, can effectively find Along Railway catenary mast
On Bird's Nest.
As shown in figure 3, about Bird's Nest recognizer, first using deep learning algorithm to institute's acquired image data into
Row Bird's Nest Preliminary detection;Secondary-confirmation is carried out to the Bird's Nest that Preliminary detection arrives then in conjunction with interference filter algorithm, exports final bird
Nest testing result.The depth convolutional network that Bird's Nest detection algorithm based on deep learning uses is binaryzation neural network.That is: institute
Some network weights are approximate using binaryzation, and the convolutional neural networks with binaryzation weight are deep significantly less than the standard of non-two-value
Convolutional network is spent, also, when weighted value is two-value, convolution algorithm is only estimated by addition and subtraction, does not need multiplication, from
And operation is caused to accelerate, the convolutional neural networks of comparison with standard, the approximate convolutional neural networks of binaryzation weight can be preferable
The memory of small portable apparatus is adapted to, while keeping identical detection accuracy horizontal.Utilize the deep learning net after training
Network model carries out Bird's Nest detection to present system acquired image data, marks to detection containing Bird's Nest image data
Remember and record Bird's Nest location information (coordinate (x, y) of Bird's Nest central point in the picture, the formed rectangle frame of Bird's Nest range
It is long and wide).
Bird's Nest detection algorithm based on deep learning is capable of detecting when in input picture the position for whether having Bird's Nest and Bird's Nest
Confidence breath, still, can have least a portion of non-Bird's Nest target (erroneous detection) similar with Bird's Nest feature, and not in testing result
Be detected all Bird's Nests all be threaten railway contact line power supply safety, so must be in these Bird's Nests detected
Confirm the Bird's Nest on real catenary mast, filters out on branch, the non-railway power supply dress such as power cord bar, signal transmitting tower
The incoherent Bird's Nest set.So the present invention after based on deep learning algorithmic preliminaries detection Bird's Nest, devises a set of dry
Disturb filter algorithm.
The erroneous detection target of the Preliminary detection of the above-mentioned deep learning algorithm of the filter algorithm main filtration and it is incoherent just
Normal Bird's Nest retains the Bird's Nest defect on the catenary mast for really influencing railway power supply safety.
The filter algorithm is divided into 2 layers of filtering.
A, it is filtered using the priori characteristic based on Bird's Nest structure.Layer filtering, is mainly used for filtering out some non-Bird's Nest mesh
Standard type.
(1) overall ratio that Bird's Nest is set according to the length-width ratio of rectangle frame is more than the target of certain threshold value for length-width ratio,
Not think to be Bird's Nest, filter;(2) Bird's Nest does not appear in the position close to ground level generally, so, if the bird detected
Nest position is in certain threshold range of the lower edge of image, then it is assumed that he is jamming target, filtering;(3) size one of Bird's Nest
As in a fixed range, will not be too small, will not be too big.So the length of the affiliated rectangle frame for the Bird's Nest detected
A certain fixed threshold is respectively smaller than with width or is greater than a certain fixed threshold, then it is assumed that it is jamming target, filtering.(4) duplicate removal is grasped
Make, when occurring continuous n frame with the case where Bird's Nest target, algorithm design only retains the maximum frame detection knot of confidence level
Fruit.
B, the region in image where Bird's Nest is analyzed, treetop, signal transmitting tower, power circuit bar and contact are substantially at
Net pillar.For these four types of objects, feature is obvious, so the just multi-target detection based on deep learning in the present invention
Algorithm, while whether containing these four types of objects and the specific location of the object in detection image.
Training sample database is established: bird's nest image of the result that Bird's Nest detects by reservation after filter algorithm one.By this
Big target (tree, signal transmitting tower, power circuit bar and the contact net branch that the image containing Bird's Nest is depended on according to Bird's Nest a bit
Column) classified and is marked.
Sample training: carrying out deep learning network training using the 4 class target datas marked,
Testing result: if not detecting catenary mast in image, and other three classes are detected.It is considered that this Bird's Nest
The not defect Bird's Nest on catenary mast.It is filtered;If detecting catenary mast, also to continue this pillar
Position is corrected with the Bird's Nest position previously detected, if the position of Bird's Nest is recognized in a certain range of shore position
Bird's Nest is catenary mast Bird's Nest defect thus.Retain this frame image at this time, and according to longitude and latitude at this time and combines a bar
One profile database positions the accurate location of Bird's Nest, generates alarm.
The above disclosure is only the preferred embodiments of the present invention, cannot limit the right model of the present invention with this certainly
It encloses, therefore equivalent changes made in accordance with the claims of the present invention, is still within the scope of the present invention.
Claims (10)
1. a kind of portable online Bird's Nest intelligent identification Method, which is characterized in that the Bird's Nest intelligent identification Method specifically includes:
S1: real time picture is obtained;
S2: deep learning network model is utilized, real time picture is recognized, and provide identification result;
S3: if identification result is thought to carry out early warning there are Bird's Nest;It is no to then follow the steps S1 or terminate the recognition methods.
2. a kind of Bird's Nest intelligent identification Method described in claim 1, which is characterized in that the step of the Bird's Nest intelligent identification Method
It also include the secondary-confirmation of identification result, according to secondary-confirmation as a result, updating identification result in rapid S2.
3. a kind of portable online Bird's Nest intelligent identification Method as claimed in claim 2, which is characterized in that the identification result
Secondary-confirmation, according to secondary-confirmation as a result, updating identification result method particularly includes:
S21: the priori Characteristics Detection based on Bird's Nest structure feature is carried out;If priori Characteristics Detection result is judged as non-Bird's Nest,
Identification result is updated, and is jumped as step S3;Otherwise, step S22 is jumped directly to;
S22: the multi-target detection algorithm based on deep learning is used, the big target depended on to Bird's Nest detects, if not depositing
In big target, then identification result is updated, and the S3 that gos to step, otherwise, jumps directly to step S3.
4. a kind of portable online Bird's Nest intelligent identification Method as claimed in claim 3, which is characterized in that the priori of the S21
Characteristics Detection rule includes one or more of following rule:
1) judge whether the length-width ratio of Bird's Nest is in setting range, if so, being considered Bird's Nest;
2) judge whether height of the Bird's Nest apart from ground is in setting range, if so, being considered Bird's Nest;
3) judge whether the size of Bird's Nest is in setting range, if so, being considered Bird's Nest.
5. portable online Bird's Nest intelligent identification Method described in a kind of claim 3 or 4, which is characterized in that the tool of the S22
Body implementation method are as follows:
S221: establish training sample database: the image that these are contained Bird's Nest is classified simultaneously according to the big target that Bird's Nest is depended on
Label;The big target includes tree, signal transmitting tower, power circuit bar, catenary mast;
S222: deep learning network training sample training: is carried out using the target data marked;
S223: result detection: it is directed to real time picture, carries out big target detection;
If not detecting catenary mast in image, it is considered non-contact net pillar Bird's Nest, and updates identification knot
Fruit;
If detecting catenary mast, judge whether Bird's Nest position and shore position are in a certain range, if so, recognizing
For catenary mast, there are Bird's Nests.
6. portable online Bird's Nest intelligent identification Method described in a kind of claim 5, which is characterized in that carried out in the S3 pre-
Alert concrete methods of realizing are as follows:
In conjunction with one grade of basic database of GPS positioning information and a bar, the accurate location information where Bird's Nest is confirmed, and export report
Alert information.
7. a kind of portable online Bird's Nest intelligent identification Method as claimed in claim 6, which is characterized in that the warning message is logical
It crosses and is wirelessly transmitted to mobile display device.
8. portable online Bird's Nest intelligent identification Method according to claim 7, which is characterized in that the deep learning net
Network model is binaryzation weighting network.
9. one kind is based on Embedded portable online Bird's Nest intelligent identifying system, which is characterized in that described based on Embedded
Portable Bird's Nest cruising inspection system includes that portable online Bird's Nest cruising inspection system includes data acquisition module, central processing module, shifting
Dynamic display equipment;The data acquisition module, central processing module, mobile display device are sequentially connected;
Data acquisition module includes high-definition image acquisition unit and locating module;
Central processing module includes compression storing data unit, core control panel, Power Management Unit, acquisition module control list
Member, abnormal alarm unit, Bird's Nest intelligent detection unit, database;The compression storing data unit, is adopted at Power Management Unit
Collection module control unit, abnormal alarm unit, Bird's Nest intelligent detection unit, database are connected with core control panel;
It is connected between the abnormal alarm unit and mobile display device.
10. portable online Bird's Nest intelligent identifying system according to claim 8, which is characterized in that the central processing
Intelligent recognition is gone out the image of bird's nest to module and corresponding location information is transferred to Bird's Nest staqtistical data base, and can patrol
After inspection, Bird's Nest early warning report is generated according to the information stored in the Bird's Nest staqtistical data base, passes through the mobile display
Device shows staff, then checks and overhauls for staff;The central processing module goes out intelligent recognition
The image of bird's nest is transferred to the mobile display device with corresponding location information, carries out abnormal alarm;
In in high strength shell cabinet, the cabinet is erected at frame for the data acquisition module, central processing module integrated installation
If on device, the stringer provides the decorating position of flexibility and changeability using sucker, tripod cooperation holder and angle is selected
It selects.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910176773.8A CN109902647B (en) | 2019-03-08 | 2019-03-08 | Portable online bird nest intelligent identification method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910176773.8A CN109902647B (en) | 2019-03-08 | 2019-03-08 | Portable online bird nest intelligent identification method and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109902647A true CN109902647A (en) | 2019-06-18 |
CN109902647B CN109902647B (en) | 2023-04-18 |
Family
ID=66946855
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910176773.8A Active CN109902647B (en) | 2019-03-08 | 2019-03-08 | Portable online bird nest intelligent identification method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109902647B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112715427A (en) * | 2020-12-30 | 2021-04-30 | 青海保绿丰生态农林科技有限公司 | Automatic monitoring system for rules of hawk-leading nesting, number of spawning, brooding and predation habits |
CN114913473A (en) * | 2022-03-21 | 2022-08-16 | 中国科学院光电技术研究所 | Lightweight single-body imaging contact network safety patrol instrument |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101709954A (en) * | 2009-12-11 | 2010-05-19 | 北京航空航天大学 | Overhead wire icing on-line monitoring device based on machine vision |
CN103530620A (en) * | 2013-10-30 | 2014-01-22 | 华北电力大学 | Method for identifying bird nest on electric transmission line tower |
CN103745224A (en) * | 2013-12-24 | 2014-04-23 | 西南交通大学 | Image-based railway contact net bird-nest abnormal condition detection method |
CN105208347A (en) * | 2015-10-08 | 2015-12-30 | 成都时代星光科技有限公司 | Automatic unmanned aerial vehicle aerial patrolling and real-time image collecting and transmitting monitoring device for railway line |
CN105389555A (en) * | 2015-11-06 | 2016-03-09 | 天津航天中为数据系统科技有限公司 | Aerial-image-based hidden-danger target analysis method and system |
CN106250835A (en) * | 2016-07-26 | 2016-12-21 | 国网福建省电力有限公司 | Bird's Nest recognition methods on the transmission line of electricity of feature based identification |
CN106971152A (en) * | 2017-03-16 | 2017-07-21 | 天津大学 | A kind of method of Bird's Nest in detection transmission line of electricity based on Aerial Images |
CN106980827A (en) * | 2017-03-16 | 2017-07-25 | 天津大学 | A kind of method of Bird's Nest in identification transmission line of electricity based on Aerial Images |
CN107665336A (en) * | 2017-09-20 | 2018-02-06 | 厦门理工学院 | Multi-target detection method based on Faster RCNN in intelligent refrigerator |
CN108037133A (en) * | 2017-12-27 | 2018-05-15 | 武汉市智勤创亿信息技术股份有限公司 | A kind of power equipments defect intelligent identification Method and its system based on unmanned plane inspection image |
CN108537154A (en) * | 2018-03-28 | 2018-09-14 | 天津大学 | Transmission line of electricity Bird's Nest recognition methods based on HOG features and machine learning |
EP3376213A1 (en) * | 2017-03-15 | 2018-09-19 | Siemens Aktiengesellschaft | Method and assembly for monitoring the state of a system with operating elements |
CN108681733A (en) * | 2018-05-15 | 2018-10-19 | 苏州华眸智能科技有限公司 | A kind of pantograph abnormal state detecting system and detection method based on image procossing |
CN108764020A (en) * | 2018-03-30 | 2018-11-06 | 广东工业大学 | A kind of Bird's Nest recognition methods on high tension electric tower based on unmanned plane image |
CN109190545A (en) * | 2018-08-27 | 2019-01-11 | 国网天津市电力公司 | Bird's Nest automatic recognition system and its methods and applications in transmission line of electricity |
-
2019
- 2019-03-08 CN CN201910176773.8A patent/CN109902647B/en active Active
Patent Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101709954A (en) * | 2009-12-11 | 2010-05-19 | 北京航空航天大学 | Overhead wire icing on-line monitoring device based on machine vision |
CN103530620A (en) * | 2013-10-30 | 2014-01-22 | 华北电力大学 | Method for identifying bird nest on electric transmission line tower |
CN103745224A (en) * | 2013-12-24 | 2014-04-23 | 西南交通大学 | Image-based railway contact net bird-nest abnormal condition detection method |
CN105208347A (en) * | 2015-10-08 | 2015-12-30 | 成都时代星光科技有限公司 | Automatic unmanned aerial vehicle aerial patrolling and real-time image collecting and transmitting monitoring device for railway line |
CN105389555A (en) * | 2015-11-06 | 2016-03-09 | 天津航天中为数据系统科技有限公司 | Aerial-image-based hidden-danger target analysis method and system |
CN106250835A (en) * | 2016-07-26 | 2016-12-21 | 国网福建省电力有限公司 | Bird's Nest recognition methods on the transmission line of electricity of feature based identification |
EP3376213A1 (en) * | 2017-03-15 | 2018-09-19 | Siemens Aktiengesellschaft | Method and assembly for monitoring the state of a system with operating elements |
CN106971152A (en) * | 2017-03-16 | 2017-07-21 | 天津大学 | A kind of method of Bird's Nest in detection transmission line of electricity based on Aerial Images |
CN106980827A (en) * | 2017-03-16 | 2017-07-25 | 天津大学 | A kind of method of Bird's Nest in identification transmission line of electricity based on Aerial Images |
CN107665336A (en) * | 2017-09-20 | 2018-02-06 | 厦门理工学院 | Multi-target detection method based on Faster RCNN in intelligent refrigerator |
CN108037133A (en) * | 2017-12-27 | 2018-05-15 | 武汉市智勤创亿信息技术股份有限公司 | A kind of power equipments defect intelligent identification Method and its system based on unmanned plane inspection image |
CN108537154A (en) * | 2018-03-28 | 2018-09-14 | 天津大学 | Transmission line of electricity Bird's Nest recognition methods based on HOG features and machine learning |
CN108764020A (en) * | 2018-03-30 | 2018-11-06 | 广东工业大学 | A kind of Bird's Nest recognition methods on high tension electric tower based on unmanned plane image |
CN108681733A (en) * | 2018-05-15 | 2018-10-19 | 苏州华眸智能科技有限公司 | A kind of pantograph abnormal state detecting system and detection method based on image procossing |
CN109190545A (en) * | 2018-08-27 | 2019-01-11 | 国网天津市电力公司 | Bird's Nest automatic recognition system and its methods and applications in transmission line of electricity |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112715427A (en) * | 2020-12-30 | 2021-04-30 | 青海保绿丰生态农林科技有限公司 | Automatic monitoring system for rules of hawk-leading nesting, number of spawning, brooding and predation habits |
CN114913473A (en) * | 2022-03-21 | 2022-08-16 | 中国科学院光电技术研究所 | Lightweight single-body imaging contact network safety patrol instrument |
CN114913473B (en) * | 2022-03-21 | 2023-08-15 | 中国科学院光电技术研究所 | Lightweight monomer type imaging contact net safety inspection instrument |
Also Published As
Publication number | Publication date |
---|---|
CN109902647B (en) | 2023-04-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108037770B (en) | Unmanned aerial vehicle power transmission line inspection system and method based on artificial intelligence | |
CN110059631B (en) | Contact net non-contact type monitoring defect identification method | |
CN107808133B (en) | Unmanned aerial vehicle line patrol-based oil and gas pipeline safety monitoring method and system and software memory | |
CN111582234B (en) | Large-scale oil tea tree forest fruit intelligent detection and counting method based on UAV and deep learning | |
CN110348304A (en) | A kind of maritime affairs distress personnel search system being equipped on unmanned plane and target identification method | |
CN104581076A (en) | Mountain fire monitoring and recognizing method and device based on 360-degree panoramic infrared fisheye camera | |
CN110929592A (en) | Extraction method and system for outer boundary of mariculture area | |
CN109711348A (en) | Intelligent monitoring method and system based on the long-term real-time architecture against regulations in hollow panel | |
CN114743119B (en) | High-speed rail contact net hanger nut defect detection method based on unmanned aerial vehicle | |
CN109887343B (en) | Automatic acquisition and monitoring system and method for flight ground service support nodes | |
CN109801265A (en) | A kind of real-time transmission facility foreign matter detection system based on convolutional neural networks | |
CN112528979B (en) | Transformer substation inspection robot obstacle distinguishing method and system | |
CN105303162B (en) | A kind of Aerial Images insulator recognition methods based on target proposed algorithm | |
CN113759960A (en) | Unmanned aerial vehicle-based fan blade and tower barrel inspection identification system and method | |
CN107369291A (en) | The anti-external force damage alarm system and method for high-tension line based on deep learning | |
CN114240868A (en) | Unmanned aerial vehicle-based inspection analysis system and method | |
CN109902647A (en) | The portable online Bird's Nest intelligent identification Method of one kind and system | |
CN110163184A (en) | Intelligent video monitoring system and method | |
CN115761537A (en) | Power transmission line foreign matter intrusion identification method oriented to dynamic characteristic supplement mechanism | |
CN115909092A (en) | Light-weight power transmission channel hidden danger distance measuring method and hidden danger early warning device | |
CN112487894A (en) | Automatic inspection method and device for rail transit protection area based on artificial intelligence | |
CN112332541A (en) | Monitoring system and method for transformer substation | |
CN116310891A (en) | Cloud-edge cooperative transmission line defect intelligent detection system and method | |
CN109978847A (en) | Drag-line housing disease automatic identifying method based on transfer learning Yu drag-line robot | |
CN112529836A (en) | High-voltage line defect detection method and device, storage medium and electronic equipment |
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 |