CN108109385A - A kind of vehicle identification of power transmission line external force damage prevention and hazardous act judgement system and method - Google Patents
A kind of vehicle identification of power transmission line external force damage prevention and hazardous act judgement system and method Download PDFInfo
- Publication number
- CN108109385A CN108109385A CN201810049610.9A CN201810049610A CN108109385A CN 108109385 A CN108109385 A CN 108109385A CN 201810049610 A CN201810049610 A CN 201810049610A CN 108109385 A CN108109385 A CN 108109385A
- Authority
- CN
- China
- Prior art keywords
- deep learning
- transmission line
- external force
- damage prevention
- vehicle
- 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
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
- G08G1/0175—Detecting movement of traffic to be counted or controlled identifying vehicles by photographing vehicles, e.g. when violating traffic rules
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of vehicle identifications of power transmission line external force damage prevention and hazardous act judgement system and method.The system includes front end power transmission line intelligent external force damage prevention prior-warning device, back-end central management platform and mobile client, the front-end and back-end of system are respectively provided with deep learning module, the deep learning module cooperating operation of front-end and back-end, the vehicle identification and vehicle risk behavior early warning for completing transmission line external force damage prevention differentiate.The model and parameter of the deep learning module of front end provide online after being trained by rear end, and central management platform determines the number of plies of deep learning network and convolution kernel size, and deep learning network is trained.The present invention also provides the method for discrimination of above-mentioned judgement system.The present invention had both avoided the training cost of front end deep learning network, reduces the complexity of front end system, and solves the problems, such as that transmission of video is costly low with real-time in the intellectual analysis of rear end.
Description
Technical field
The invention belongs to supply of electric power technology and its intelligent fields, and in particular to a kind of vehicle of power transmission line external force damage prevention is known
Not with hazardous act judgement system and method.
Background technology
Ultra-high-tension power transmission line is with coverage is big, distributed areas are wide, transmission range is long, nature and geographical conditions are complicated more
Become, destroy source it is more the features such as, they bring great challenge to the day-to-day operation of circuit, maintenance and maintenance, particularly shaft tower or
The potential outer broken behaviors such as the oversize vehicles such as pump truck, engineering truck near circuit pass through, the construction of neighbouring crane are to transmission line of electricity
Normal operation threatens.
Recent years, various new technology and equipments continually introduce to detect the outer broken behavior of electric power facility automatically
And early warning, the mode identification method based on machine vision are a kind of important methods.
At present, the method for the pattern-recognition applied to transmission line of electricity external force damage prevention early warning system of mainstream is that SIFT feature carries
It takes, based on svm classifier, for identifying Large Construction vehicle, color and angle of this method using SIFT methods extraction foreground area
Point reuses trained SVM classifier in advance and carries out vehicle judgement.But the defects of this method is the dimension mistake of feature vector
Height easily generates " dimension disaster ", and the feature of another SIFT extractions is by engineer, it is difficult to which the vehicle under describing complex scene is special
Sign, bad adaptability, generalization ability are poor, accuracy of identification is relatively low, early warning reliability deficiency.
The content of the invention
The purpose of the present invention is to propose to a kind of deep learning method based on joint front and back end, for the big of transmission of electricity thread environment
Type Construction traffic identifies and hazardous act analysis method, this method combination front end intelligence and back-end analysis, rear end in deep learning
Implementation model is trained, and discriminant classification is realized in front end.The present invention had not only reduced the complexity of front end training pattern, but also solved rear end intelligence
Can massive video data transmission problem, real time problems, accuracy rate problem in analysis, can real-time implementation be various dives to railway superstructures
Early warning, interception are carried out in the vehicle of harm and behavior.
To achieve the above object, the technical solution adopted by the present invention is a kind of vehicle identification of power transmission line external force damage prevention and danger
Behavior judgement system, the intelligent external force damage prevention prior-warning device of the system including transmission line of electricity front end, the central management platform of rear end and
Mobile client, central management platform are communicated by 3/4G networks with power transmission line intelligent external force damage prevention prior-warning device, mobile client
End is communicated by 3/4G networks with power transmission line intelligent external force damage prevention prior-warning device and central management platform, the intelligent external force damage prevention of front end
The central management platform of prior-warning device and rear end is respectively provided with deep learning module, the deep learning module cooperation of the front and back ends
Operation, the vehicle identification and vehicle risk behavior early warning for completing transmission line external force damage prevention differentiate.
Further, above-mentioned front end power transmission line intelligent external force damage prevention prior-warning device is by 3/4G intelligent cameras, solar cell
Plate, power-supply controller of electric, accumulator composition.
Above-mentioned 3/4G intelligent cameras are by video imaging module, deep learning module, warning module, PTZ modules, compression mould
Block, memory module and communication module component composition, video imaging module realize Image Acquisition and pretreatment；Warning module is realized pre-
Alert information management and information reporting；PTZ modules are realized that video camera rotates up and down and are zoomed in or out with camera lens, adjust video camera
The visual field；Compression module realization video to the original yuv data boil down to JPEG picture comprising Large Construction vehicle and H.264/5；
Memory module realization stores compressed picture and video and based on early warning event and timestamp on SD storage mediums
Index.
Further, deep learning algorithm is respectively deployed in the front and back ends of system, and the central management platform of rear end realizes depth
The modeling of network model, mark, training in learning algorithm adjust the deep learning network of network model and parameter of rear end, and fixed
When newest deep learning network model and parameter are transmitted to front end online；Front end receives the deep learning network sent rear end
The deep learning network and parameter of front end intelligent camera are updated after model and parameter；The intelligent camera of front end captures monitoring point
The image of vehicular traffic carries out Classification and Identification and row to the image of acquisition using real-time update deep learning network model and parameter
For analysis, when determining dangerous situation, on-line early warning and by the vehicle of early warning event, picture and prediction occurring it is passing when it is front and rear short
Video carries out timestamp mark, compression, storage concurrently toward rear end central management platform.
The model and parameter configuration of deep learning module in the intelligent camera of front end are that dynamic is adjustable, after
The data and parameter that end central management platform is sent carry out model and parameter update.
Automatic detection vehicle in the deep learning real-time performance display foreground region of front end automatically extracts feature, at classification
Reason, behavioural analysis, are completed to the passing Large Construction vehicle identification of transmission line of electricity monitoring point and the differentiation early warning of hazardous act.
The central management platform of rear end determines the number of plies of deep learning network and convolution kernel size, according to different weather condition
And vehicle training deep learning network, obtain relevant parameter；Deep learning network uses the sub-sampling pixel through overcompression dimensionality reduction
Feature, Color Distribution Features, textural characteristics, structure feature, motion feature can recognize that all kinds of engineering trucks and its hazardous act.
Engineering truck recognition methods realization of the back-end central management platform based on deep learning includes background modeling, target inspection
Training and identification including survey, vehicle classification, vehicle behavior analysis.
In the deep learning network model of front end and parameter reproducting periods, front end record a video completely, is transferred to after mark
Back-end central management platform completes vehicle detection and recognition, it is ensured that not missing inspection by the deep learning network of rear end.
The present invention furthermore provides a kind of vehicle identification for realizing above-mentioned power transmission line external force damage prevention and differentiates system with hazardous act
The method of discrimination of system, specifically comprises the steps of：
Step 1, back-end central server determines the depth convolutional neural networks number of plies and convolution kernel size, can identify defeated
Engineering truck vehicle and process around electric wire tower stay and hazardous act；
Step 2, training depth convolutional neural networks, specific step are as follows：
Step 21, sample of the different automobile types picture as training under different weather, period is gathered, and be labeled,
Parametrization；
Step 22, sample image is pre-processed；
Step 23, training depth convolutional neural networks；
Step 24, the parameter of depth convolutional neural networks is obtained；
Step 25, parameter is passed into front end depth convolutional neural networks；
Step 26, the trained depth convolution of intelligent camera acquisition divides neural network model and grader；
Step 3, the intelligent camera of front end carries out Target Acquisition and identification, specific step are as follows：
Step 31, realtime graphic is obtained；
Step 32, the image got is pre-processed；
Step 33, carry out background modeling using hybrid neural networks and prospect identifies；
Step 34, trained depth convolutional neural networks carry out feature extraction to the prospect identified；
Step 35, the feature of depth convolutional neural networks extraction is classified by trained grader；
Step 36, whether output prospect is that engineering truck and vehicle risk behavior differentiate result.
Compared with prior art, the present invention has following advantageous effects：
1, in the present invention, the model and parameter of the deep learning module of front end are provided after being trained by rear end and online updating,
Back-end central management platform determines the number of plies of deep learning network and convolution kernel size, and deep learning network is trained,
Front end system complexity so can be reduced to avoid the training cost of front end deep learning network.
2, the present invention can solve massive video data transmission problem, real time problems, accuracy rate in the intellectual analysis of rear end
Problem, can real-time implementation be various carries out early warning, interception to the vehicle of railway superstructures potential hazard and behavior.
Description of the drawings
Fig. 1 is the system diagram of power transmission line intelligent external force damage prevention early warning system.
Fig. 2 is 3/4G intelligent camera functional block diagrams.
Fig. 3 is the training flow chart of Large Construction vehicle.
Fig. 4 is the identification process figure of Large Construction vehicle.
Fig. 5 is depth convolutional neural networks structure diagram.
Fig. 6 is deep learning dimensionality reduction modeling procedure figure.
Specific embodiment
In conjunction with attached drawing, the present invention is described in further detail.
The present invention provides a kind of oversize vehicle detection for transmission line of electricity external force damage prevention early warning and behavior judgement system, Fig. 1
It is system composition schematic diagram.The system is by the power transmission line intelligent external force damage prevention prior-warning device of front end, the central management platform of rear end
It is formed with mobile client；Central management platform is led to by 3/4G networks and power transmission line intelligent external force damage prevention early warning fore device
Letter, the power transmission line intelligent external force damage prevention prior-warning device of front end by 3/4G intelligent cameras, solar panel, power-supply controller of electric,
Accumulator forms.
Wherein, 3/4G intelligent cameras are made of components such as video imaging, deep learning, early warning, PTZ, compression, storages,
As shown in Figure 2.3/4G intelligent cameras are by the components group such as video imaging, deep learning, early warning, PTZ, compression, storage and communication
Into front end intelligent camera image-forming module realizes Image Acquisition and pretreatment, focusing, white balance, exposure, noise processed；Early warning
Module realizes warning information management and information reporting；PTZ realizes that video camera rotates up and down and is zoomed in or out with camera lens, adjusts
Camera coverage, compression are realized to the original yuv data boil down to JPEG picture comprising Large Construction vehicle and H.264/5 regarded
Frequently, storage is realized is managed storage to compressed picture and video on SD storage mediums.Central management platform includes
The components such as deep learning function module and information management module.
System is using the monitoring of deep learning real-time performance Large Construction vehicle and the differentiation of hazardous act, deep learning net
Network is real using the sub-sampling pixel characteristic through overcompression dimensionality reduction, Color Distribution Features, textural characteristics, structure feature, motion feature
Existing background modeling, target detection, vehicle classification, the training of vehicle behavior analysis, can recognize that wheeled digging machine, pump truck, crane, engineering truck
Wait vehicles and its hazardous act；
Establish deep learning module in the front and back end of system, deep learning network model and ginseng in the intelligent camera of front end
Number configuration is that dynamic is adjustable, and automatic detection vehicle in the deep learning real-time performance display foreground region of front end automatically extracts
Feature, classification processing, behavioural analysis complete that the passing Large Construction vehicle of transmission line monitoring point is identified and dangerous row
For differentiation early warning.
Back-end central management platform determines the number of plies of deep learning convolutional neural networks and convolution kernel size, completes front and back end
The modeling and training of deep learning algorithm, are labeled deep learning algorithm, training, percentage regulation learning model and parameter,
And deep learning network model and parameter are periodically transmitted to front end online；Back-end central management platform deep learning network is similary
Detect vehicle, extraction feature, classification processing, behavioural analysis, to the passing Large Construction vehicle of transmission line monitoring point video into
Row identification and the differentiation early warning of hazardous act.
The present invention furthermore provides a kind of oversize vehicle detection for transmission line of electricity external force damage prevention early warning and differentiates with behavior
Method, the early warning vehicle image and short-sighted frequency of back-end central management platform receiving front-end transmission and regarding for front end reproducting periods
Frequently, it is labeled, training adjusts the model and parameter of front and back end deep learning network according to actual conditions, periodically online will most
New deep learning model and parameter are transmitted to front end；Update front end is intelligently taken the photograph after front end receives the model sent rear end and parameter
The deep learning neutral net and parameter of camera；In front-end algorithm model and parameter reproducting periods, front end record a video completely, marks
Central management platform is transferred to after note and completes vehicle detection and recognition, it is ensured that not missing inspection.
It is the image of video capture monitoring point vehicular traffic in the power transmission line intelligent external force damage prevention prior-warning device of front end, right
The image of acquisition carries out Classification and Identification using deep learning method；Warning information is sent to related mobile phone and center when causing danger
Management platform, at the same by the vehicle of prediction occurring it is passing when front and rear short-sighted frequency to carry out timestamp mark, compression, storage concurrent backward
Hold central management platform.The oversize vehicle of the deep learning real-time performance transmission line of electricity external force damage prevention early warning of the front and back end cooperation
Detection and the key step of behavior method of discrimination：
Step 1, the depth convolutional neural networks number of plies and convolution kernel size are determined, can identify electric power pylon frame peripheral
The vehicles such as wheeled digging machine, pump truck, crane, engineering ladder truck pass through, stay and hazardous act；
Step 2, training depth convolutional neural networks, schematic diagram such as Fig. 3, step are as follows：
Step 21, the different automobile types picture under different weather (fine day, the bad weathers such as rain, snow, haze) is gathered, as
Trained sample, and be labeled, parameterize；
Step 22, sample image is pre-processed；
Step 23, training depth convolutional neural networks；
Step 24, the parameter of depth convolutional neural networks is obtained；
Step 25, parameter is passed into front end depth convolutional neural networks；
Step 26, trained depth volume integral neural network and grader are obtained.
Step 3, front end intelligent camera carries out Target Acquisition and identification, schematic diagram such as Fig. 4, step are as follows：
Step 31, realtime graphic is obtained；
Step 32, the image got is pre-processed；
Step 33, carry out background modeling using hybrid neural networks and prospect identifies；
Step 34, trained depth convolutional neural networks carry out feature extraction to the prospect identified；
Step 35, the feature of depth convolutional neural networks extraction is classified by trained grader；
Step 36, acquisition prospect whether be wheeled digging machine, pump truck, crane result.
Described depth convolutional neural networks (DCNN) its structure diagram as shown in figure 5, including input layer, convolutional layer,
Secondary sampling layer, full articulamentum, output layer.Input layer is vehicle picture and actual natural scene；Convolutional layer includes multiple characteristic patterns,
The local receptive field connection of each neuron and preceding layer in characteristic pattern, and carry out convolution with the convolution kernel with learning ability
Local feature is obtained, is exported to obtain the characteristic pattern of this layer by activation primitive, the calculation formula and activation primitive of convolutional layer are respectively：
Wherein l is the network number of plies, and k is convolution kernel, MjFor the receptive field of input layer, b is a biasing of each output figure
Value, e is natural Exponents, about takes 2.71828.
Secondary sampling layer is that input layer is sampled, and reduces the resolution ratio of input feature vector figure, reduces intrinsic dimensionality, secondary sampling
Layer formula be：
Wherein down () is pond function, and β is weight coefficient.
Convolutional layer is used to extract the feature of input picture with time sampling layer, and its whole is fed back to full articulamentum and carries out spy
Sign classification, final result is exported by output layer.
In order to improve operational efficiency and lower computational complexity, deep learning neutral net needs to carry out dimensionality reduction modeling,
Flow is as shown in Figure 6：
Step 1：Original image pixels are gathered as input data；
Step 2：Start node number, such as 2000 are set；
Step 3：Second layer neutral net start node, such as 1000 are set in textural characteristics according to COLOR COMPOSITION THROUGH DISTRIBUTION；
Step 4：Third layer neutral net start node, such as 500 are set according to movable information；
Step 5：Export the feature space data after dimensionality reduction；
Step 6：Convolution algorithm output characteristic figure；
Step 7：Correction characteristic pattern is exported by Nonlinear Processing；
Step 8：Maximum pond output pool characteristic；
Step 9：Into vehicle classification device, vehicle type recognition result is exported.
It should be noted that, although the present invention has been described by way of example and in terms of the preferred embodiments, but above-described embodiment content is not
For limiting the present invention's.Without departing from the spirit and scope of the invention, the similary category of any equivalence changes or retouching done
In the protection domain of the present invention.
Claims (10)
1. a kind of vehicle identification of power transmission line external force damage prevention and hazardous act judgement system, which is characterized in that the system includes transmission of electricity
The intelligent external force damage prevention prior-warning device of circuit front end, the central management platform and mobile client of rear end, central management platform pass through
3/4G networks communicate with power transmission line intelligent external force damage prevention prior-warning device, and mobile client passes through 3/4G networks and power transmission line intelligent
External force damage prevention prior-warning device and central management platform communication, the intelligent external force damage prevention prior-warning device of front end and the central management platform of rear end
Deep learning module, the deep learning module cooperating operation of the front and back ends are respectively provided with, the vehicle for completing transmission line external force damage prevention is known
Do not differentiate with vehicle risk behavior early warning.
2. the vehicle identification of power transmission line external force damage prevention according to claim 1 and hazardous act judgement system, which is characterized in that
The front end power transmission line intelligent external force damage prevention prior-warning device by 3/4G intelligent cameras, solar panel, power-supply controller of electric,
Accumulator forms.
3. the vehicle identification of power transmission line external force damage prevention according to claim 2 and hazardous act judgement system, which is characterized in that
The 3/4G intelligent cameras are by video imaging module, deep learning module, warning module, PTZ modules, compression module, storage
Module and communication module component composition, video imaging module realize Image Acquisition and pretreatment；Warning module realizes warning information
Management and information reporting；PTZ modules are realized that video camera rotates up and down and are zoomed in or out with camera lens, adjust camera coverage；
Compression module realization video to the original yuv data boil down to JPEG picture comprising Large Construction vehicle and H.264/5；Store mould
Compressed picture and video are stored on SD storage mediums for block realization and the index based on early warning event and timestamp.
4. the vehicle identification of power transmission line external force damage prevention according to claim 1 and hazardous act judgement system, which is characterized in that
Deep learning algorithm is respectively deployed in the front and back ends of system, and the central management platform of rear end realizes network in deep learning algorithm
Modeling, mark, the training of model adjust the deep learning network of network model and parameter of rear end, and timing online will be newest
Deep learning network model and parameter are transmitted to front end；Front end is received after the deep learning network model sent rear end and parameter more
The deep learning network and parameter of new front end intelligent camera；The intelligent camera of front end captures the figure of monitoring point vehicular traffic
Picture carries out Classification and Identification and behavioural analysis using real-time update deep learning network model and parameter to the image of acquisition, differentiates
When going out dangerous situation, on-line early warning and by the vehicle of early warning event, picture and prediction occurring it is passing when front and rear short-sighted frequency carry out when
Between stamp mark, compression, storage concurrently toward rear end central management platform.
5. the vehicle identification of power transmission line external force damage prevention according to claim 4 and hazardous act judgement system, which is characterized in that
The model and parameter configuration of deep learning module in the intelligent camera of front end are that dynamic is adjustable, according to back-end central pipe
The data and parameter that platform is sent carry out model and parameter update.
6. the vehicle identification of power transmission line external force damage prevention according to claim 4 and hazardous act judgement system, which is characterized in that
Automatic detection vehicle in the deep learning real-time performance display foreground region of front end automatically extracts feature, classification processing, behavior point
Analysis, is completed to the passing Large Construction vehicle identification of transmission line of electricity monitoring point and the differentiation early warning of hazardous act.
7. the vehicle identification of power transmission line external force damage prevention according to claim 4 and hazardous act judgement system, which is characterized in that
The central management platform of rear end determines the number of plies of deep learning network and convolution kernel size, is instructed according to different weather condition and vehicle
Practice deep learning network, obtain relevant parameter；Deep learning network uses sub-sampling pixel characteristic, color through overcompression dimensionality reduction
Distribution characteristics, textural characteristics, structure feature, motion feature can recognize that all kinds of engineering trucks and its hazardous act.
8. the vehicle identification of power transmission line external force damage prevention according to claim 4 and hazardous act judgement system, which is characterized in that
Engineering truck recognition methods realization of the back-end central management platform based on deep learning includes background modeling, target inspection
Training and identification including survey, vehicle classification, vehicle behavior analysis.
9. the vehicle identification of power transmission line external force damage prevention according to claim 4 and hazardous act judgement system, which is characterized in that
In the deep learning network model of front end and parameter reproducting periods, front end record a video completely, is transferred to back-end central after mark
Management platform completes vehicle detection and recognition, it is ensured that not missing inspection by the deep learning network of rear end.
10. a kind of vehicle identification for realizing power transmission line external force damage prevention as described in claim 1 is sentenced with hazardous act judgement system
Other method, it is characterised in that comprise the steps of：
Step 1, back-end central server determines the depth convolutional neural networks number of plies and convolution kernel size, can identify power transmission line
Engineering truck vehicle and process around pylon stay and hazardous act；
Step 2, training depth convolutional neural networks, specific step are as follows：
Step 21, sample of the different automobile types picture as training under different weather, period is gathered, and is labeled, parameter
Change；
Step 22, sample image is pre-processed；
Step 23, training depth convolutional neural networks；
Step 24, the parameter of depth convolutional neural networks is obtained；
Step 25, parameter is passed into front end depth convolutional neural networks；
Step 26, the trained depth convolution of intelligent camera acquisition divides neural network model and grader；
Step 3, the intelligent camera of front end carries out Target Acquisition and identification, specific step are as follows：
Step 31, realtime graphic is obtained；
Step 32, the image got is pre-processed；
Step 33, carry out background modeling using hybrid neural networks and prospect identifies；
Step 34, trained depth convolutional neural networks carry out feature extraction to the prospect identified；
Step 35, the feature of depth convolutional neural networks extraction is classified by trained grader；
Step 36, whether output prospect is that engineering truck and vehicle risk behavior differentiate result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810049610.9A CN108109385B (en) | 2018-01-18 | 2018-01-18 | System and method for identifying and judging dangerous behaviors of power transmission line anti-external damage vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810049610.9A CN108109385B (en) | 2018-01-18 | 2018-01-18 | System and method for identifying and judging dangerous behaviors of power transmission line anti-external damage vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108109385A true CN108109385A (en) | 2018-06-01 |
| CN108109385B CN108109385B (en) | 2022-06-14 |
Family
ID=62220117
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810049610.9A Active CN108109385B (en) | 2018-01-18 | 2018-01-18 | System and method for identifying and judging dangerous behaviors of power transmission line anti-external damage vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108109385B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108805207A (en) * | 2018-06-13 | 2018-11-13 | 江苏电力信息技术有限公司 | A kind of Large Construction vehicle raises arm detection algorithm |
| CN109873990A (en) * | 2019-03-13 | 2019-06-11 | 武汉大学 | A kind of illegal mining method for early warning in mine based on computer vision |
| CN110147757A (en) * | 2019-05-17 | 2019-08-20 | 国网山东省电力公司菏泽供电公司 | Passway for transmitting electricity engineering truck discrimination method and system based on convolutional neural networks |
| CN110428583A (en) * | 2019-07-18 | 2019-11-08 | 华东师范大学 | It is a kind of that early warning system and method are monitored in real time based on embedded development and the fork truck of deep learning |
| CN110689706A (en) * | 2019-10-11 | 2020-01-14 | 广东电网有限责任公司 | Method, device and system for monitoring large construction vehicle under overhead line |
| CN110795976A (en) * | 2018-08-03 | 2020-02-14 | 华为技术有限公司 | Method, device and equipment for training object detection model |
| CN110855932A (en) * | 2018-08-21 | 2020-02-28 | 杭州海康威视数字技术股份有限公司 | Alarm method and device based on video data, electronic equipment and storage medium |
| CN111881760A (en) * | 2020-06-30 | 2020-11-03 | 深圳金三立视频科技股份有限公司 | Transmission line external damage prevention identification method and terminal |
| CN112347947A (en) * | 2020-11-10 | 2021-02-09 | 厦门长江电子科技有限公司 | Image data processing system and method integrating intelligent detection and automatic test |
| CN112349056A (en) * | 2020-11-06 | 2021-02-09 | 暨南大学 | Cable channel prevents outer broken intelligent front end control and early warning system |
| CN113179389A (en) * | 2021-04-15 | 2021-07-27 | 江苏濠汉信息技术有限公司 | System and method for identifying crane jib of power transmission line dangerous vehicle |
| CN113255519A (en) * | 2021-05-25 | 2021-08-13 | 江苏濠汉信息技术有限公司 | Crane lifting arm identification system and multi-target tracking method for power transmission line dangerous vehicle |
| CN113780237A (en) * | 2021-09-27 | 2021-12-10 | 深圳供电局有限公司 | External damage prevention early warning method, device and system for underground pipeline |
| WO2022016573A1 (en) * | 2020-07-21 | 2022-01-27 | 南京智金科技创新服务中心 | Video monitoring analysis system and method |
| CN115099729A (en) * | 2022-08-25 | 2022-09-23 | 国网浙江慈溪市供电有限公司 | Power transmission line anti-external-damage supervision method and system based on navigation system |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104537788A (en) * | 2015-01-16 | 2015-04-22 | 南京理工大学 | High voltage electric transmission line external damage-proof intelligent image pre-warning system and method |
| CN105025263A (en) * | 2015-07-10 | 2015-11-04 | 国家电网公司 | Large-size vehicle positioning method for power transmission and distribution line channel |
| CN105744232A (en) * | 2016-03-25 | 2016-07-06 | 南京第五十五所技术开发有限公司 | Method for preventing power transmission line from being externally broken through video based on behaviour analysis technology |
| CN105894701A (en) * | 2016-04-05 | 2016-08-24 | 江苏电力信息技术有限公司 | Large construction vehicle identification and alarm method for preventing external damage to transmission lines |
| KR20170031895A (en) * | 2015-09-14 | 2017-03-22 | 대우조선해양 주식회사 | Virtual bridge system and control method thereof |
| CN106874840A (en) * | 2016-12-30 | 2017-06-20 | 东软集团股份有限公司 | Vehicle information recognition method and device |
| CN107038448A (en) * | 2017-03-01 | 2017-08-11 | 中国科学院自动化研究所 | Target detection model building method |
| CN107152966A (en) * | 2017-03-22 | 2017-09-12 | 广东工业大学 | A kind of high voltage transmission line based on deep learning waves early warning system |
| CN107247949A (en) * | 2017-08-02 | 2017-10-13 | 北京智慧眼科技股份有限公司 | Face identification method, device and electronic equipment based on deep learning |
| CN107369291A (en) * | 2017-07-13 | 2017-11-21 | 南京理工大学 | The anti-external force damage alarm system and method for high-tension line based on deep learning |
-
2018
- 2018-01-18 CN CN201810049610.9A patent/CN108109385B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104537788A (en) * | 2015-01-16 | 2015-04-22 | 南京理工大学 | High voltage electric transmission line external damage-proof intelligent image pre-warning system and method |
| CN105025263A (en) * | 2015-07-10 | 2015-11-04 | 国家电网公司 | Large-size vehicle positioning method for power transmission and distribution line channel |
| KR20170031895A (en) * | 2015-09-14 | 2017-03-22 | 대우조선해양 주식회사 | Virtual bridge system and control method thereof |
| CN105744232A (en) * | 2016-03-25 | 2016-07-06 | 南京第五十五所技术开发有限公司 | Method for preventing power transmission line from being externally broken through video based on behaviour analysis technology |
| CN105894701A (en) * | 2016-04-05 | 2016-08-24 | 江苏电力信息技术有限公司 | Large construction vehicle identification and alarm method for preventing external damage to transmission lines |
| CN106874840A (en) * | 2016-12-30 | 2017-06-20 | 东软集团股份有限公司 | Vehicle information recognition method and device |
| CN107038448A (en) * | 2017-03-01 | 2017-08-11 | 中国科学院自动化研究所 | Target detection model building method |
| CN107152966A (en) * | 2017-03-22 | 2017-09-12 | 广东工业大学 | A kind of high voltage transmission line based on deep learning waves early warning system |
| CN107369291A (en) * | 2017-07-13 | 2017-11-21 | 南京理工大学 | The anti-external force damage alarm system and method for high-tension line based on deep learning |
| CN107247949A (en) * | 2017-08-02 | 2017-10-13 | 北京智慧眼科技股份有限公司 | Face identification method, device and electronic equipment based on deep learning |
Non-Patent Citations (4)
| Title |
|---|
| 宋彪: "输电线路防施工外破的管控措施", 《宿州学院学报》 * |
| 宋彪: "输电线路防施工外破的管控措施", 《宿州学院学报》, 15 July 2013 (2013-07-15) * |
| 徐毅: "蚁群算法在电力巡检路线规划中的应用", 《计算机系统应用》 * |
| 徐毅: "蚁群算法在电力巡检路线规划中的应用", 《计算机系统应用》, 31 May 2015 (2015-05-31) * |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108805207A (en) * | 2018-06-13 | 2018-11-13 | 江苏电力信息技术有限公司 | A kind of Large Construction vehicle raises arm detection algorithm |
| CN108805207B (en) * | 2018-06-13 | 2020-08-04 | 江苏电力信息技术有限公司 | Method for detecting large construction vehicle lifting arm |
| US11423634B2 (en) | 2018-08-03 | 2022-08-23 | Huawei Cloud Computing Technologies Co., Ltd. | Object detection model training method, apparatus, and device |
| US11605211B2 (en) | 2018-08-03 | 2023-03-14 | Huawei Cloud Computing Technologies Co., Ltd. | Object detection model training method and apparatus, and device |
| CN110795976A (en) * | 2018-08-03 | 2020-02-14 | 华为技术有限公司 | Method, device and equipment for training object detection model |
| EP3624007A4 (en) * | 2018-08-03 | 2020-07-01 | Huawei Technologies Co., Ltd. | Method and apparatus for training object detection model, and device |
| CN110855932A (en) * | 2018-08-21 | 2020-02-28 | 杭州海康威视数字技术股份有限公司 | Alarm method and device based on video data, electronic equipment and storage medium |
| CN109873990A (en) * | 2019-03-13 | 2019-06-11 | 武汉大学 | A kind of illegal mining method for early warning in mine based on computer vision |
| CN110147757A (en) * | 2019-05-17 | 2019-08-20 | 国网山东省电力公司菏泽供电公司 | Passway for transmitting electricity engineering truck discrimination method and system based on convolutional neural networks |
| CN110428583B (en) * | 2019-07-18 | 2023-08-18 | 华东师范大学 | Forklift real-time monitoring and early warning system and method based on embedded development and deep learning |
| CN110428583A (en) * | 2019-07-18 | 2019-11-08 | 华东师范大学 | It is a kind of that early warning system and method are monitored in real time based on embedded development and the fork truck of deep learning |
| CN110689706A (en) * | 2019-10-11 | 2020-01-14 | 广东电网有限责任公司 | Method, device and system for monitoring large construction vehicle under overhead line |
| CN111881760A (en) * | 2020-06-30 | 2020-11-03 | 深圳金三立视频科技股份有限公司 | Transmission line external damage prevention identification method and terminal |
| CN111881760B (en) * | 2020-06-30 | 2021-10-08 | 深圳金三立视频科技股份有限公司 | Transmission line external damage prevention identification method and terminal |
| WO2022016573A1 (en) * | 2020-07-21 | 2022-01-27 | 南京智金科技创新服务中心 | Video monitoring analysis system and method |
| CN112349056A (en) * | 2020-11-06 | 2021-02-09 | 暨南大学 | Cable channel prevents outer broken intelligent front end control and early warning system |
| CN112347947A (en) * | 2020-11-10 | 2021-02-09 | 厦门长江电子科技有限公司 | Image data processing system and method integrating intelligent detection and automatic test |
| CN113179389A (en) * | 2021-04-15 | 2021-07-27 | 江苏濠汉信息技术有限公司 | System and method for identifying crane jib of power transmission line dangerous vehicle |
| CN113255519A (en) * | 2021-05-25 | 2021-08-13 | 江苏濠汉信息技术有限公司 | Crane lifting arm identification system and multi-target tracking method for power transmission line dangerous vehicle |
| CN113780237A (en) * | 2021-09-27 | 2021-12-10 | 深圳供电局有限公司 | External damage prevention early warning method, device and system for underground pipeline |
| CN115099729A (en) * | 2022-08-25 | 2022-09-23 | 国网浙江慈溪市供电有限公司 | Power transmission line anti-external-damage supervision method and system based on navigation system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108109385B (en) | 2022-06-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108109385A (en) | A kind of vehicle identification of power transmission line external force damage prevention and hazardous act judgement system and method | |
| CN112379231B (en) | Equipment detection method and device based on multispectral image | |
| WO2020199538A1 (en) | Bridge key component disease early-warning system and method based on image monitoring data | |
| CN112381784A (en) | Equipment detecting system based on multispectral image | |
| CN102867417B (en) | Taxi anti-forgery system and taxi anti-forgery method | |
| KR102122859B1 (en) | Method for tracking multi target in traffic image-monitoring-system | |
| CN103093249B (en) | A kind of taxi identification method based on HD video and system | |
| CN110728236B (en) | Vehicle loss assessment method and special equipment thereof | |
| CN104601956A (en) | Power transmission line online monitoring system and method based on fixed-wing unmanned aerial vehicle | |
| CN108918532A (en) | A kind of through street traffic sign breakage detection system and its detection method | |
| CN109326125B (en) | Picture quality diagnosis system and method based on embedded system | |
| CN113947555A (en) | Infrared and visible light fused visual system and method based on deep neural network | |
| CN110691224A (en) | Transformer substation perimeter video intelligent detection system | |
| CN116846059A (en) | Edge detection system for power grid inspection and monitoring | |
| CN112288682A (en) | Electric power equipment defect positioning method based on image registration | |
| CN114708532A (en) | Monitoring video quality evaluation method, system and storage medium | |
| WO2022104798A1 (en) | 5g-based unmanned electronic traffic police duty system | |
| CN112258402A (en) | Dense residual generation countermeasure network capable of rapidly removing rain | |
| CN117274967A (en) | Multi-mode fusion license plate recognition algorithm based on convolutional neural network | |
| CN107895365B (en) | Image matching method and monitoring system for power transmission channel external damage protection | |
| CN202887450U (en) | Taxi anti-fake system | |
| CN115641431A (en) | Power transmission line engineering vehicle detection method based on improved deep learning | |
| CN202650248U (en) | Intelligent transportation control device based on image processing | |
| CN114926456A (en) | Rail foreign matter detection method based on semi-automatic labeling and improved deep learning | |
| CN113705442A (en) | Outdoor large-board advertising picture monitoring and identifying system and method |
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 |