CN104331710B - On off state identifying system - Google Patents
On off state identifying system Download PDFInfo
- Publication number
- CN104331710B CN104331710B CN201410668236.2A CN201410668236A CN104331710B CN 104331710 B CN104331710 B CN 104331710B CN 201410668236 A CN201410668236 A CN 201410668236A CN 104331710 B CN104331710 B CN 104331710B
- Authority
- CN
- China
- Prior art keywords
- image
- module
- sample
- storehouse
- state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- 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
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to remote digital video monitoring and image identification technical field, a kind of on off state identifying system, including image capture module, communication module, image pre-processing module, characteristics of image processing identification module and template matches module, the present invention choose Euclidean distance as grader.The discrimination highest of Euclidean distance, illustrate that it can effectively classify on off state feature, suitable for twisting type characteristic of switch sorting algorithm, and its average classification time is most short, meets algorithm real-time requirement.The system of the present invention establishes module by training sample database and automatically obtains training sample database from all original sample storehouses of acquisition so that the population size of whole training sample set is moderate, can effective identification switch state, be unlikely to have influence on the speed of system again.
Description
Technical field
The present invention relates to remote digital video monitoring and image identification technical field, and in particular to a kind of on off state identification
System.
Background technology
Main Basiss breaker auxiliary node judges state and the position of switch during power scheduling.But due to
The reasons such as corrosion, wear, aging, cause auxiliary switch switching not in place, can not correctly judge the actual position of switch sometimes, be
Command scheduling provides error message.The observation for manually going to scene to carry out on the spot is now needed, and discharges failure and generally requires
Longer time, unsafe factor is brought to power network power supply, or even need interruption maintenance.
Remote digital video monitoring and image identification system are exactly by remote digital video monitoring and image recognition technology knot
Altogether, the digital video signal collected is passed through by biography by equipment such as camera, decoder and video servers first
Defeated passage passes Surveillance center back in real time in a manner of video flowing, remote video monitoring is carried out to scene in Surveillance center, from video
Interception monitoring Target Photo, is analyzed digital video image, handled and is known by corresponding Preprocessing Technique in stream
Not.Therefore apply monitoring remote video and image recognition technology, can realize to power high voltage circuit breaker switch state oneself
Dynamic identification and fault warning.This technology is ensures that it is a kind of new that electric power enterprise production safety and quick diagnosis failure provide
Means directly perceived and accurate.
But traditional switch is mostly by colour recognition or by Text region, for example, to display " ON " or
Identification of the font of " OFF " etc., the identification of switch etc. is redirected for some, then can not accurately identify, cause the utilization of system
Scope is more limited to.
The content of the invention
Solve above-mentioned technical problem, the invention provides a kind of on off state identifying system, the shape suitable for redirecting switch
State identifies that recognition effect is good, and the degree of accuracy is high.
In order to achieve the above object, the technical solution adopted in the present invention is a kind of on off state identifying system, including figure
As acquisition module, communication module, image pre-processing module, characteristics of image processing identification module and template matches module, the figure
As acquisition module and communication module are arranged on motor computer room, the communication module is connected with image capture module telecommunications, the figure
As pretreatment module, characteristics of image identification module and template matches module are arranged at Surveillance center, and image pre-processing module
Data interaction is carried out by communication module and image capture module,
Described image acquisition module gathers in real time to disk rotary transition on off state on switch cubicle,
The realtime image data that the communication module collects image capture module is transferred to image pre-processing module,
Described image pretreatment module receives the realtime image data of communication module transmission, and carries out figure to the view data
As obtaining pretreatment image after the processing of gray processing, image smoothing, image sharpening and image binaryzation,
Described image feature recognition module is connected with image pre-processing module, is extracted pretreatment image and is determined current circle
Spiral transition on off state feature,
The template matches module measured switch state feature and the on off state image progress in given Sample Storehouse
Match somebody with somebody, similarity measure values are calculated by similarity measurement, obtain the state conclusion of current disc rotary switch.
Further, the template matches module uses Euclidean distance computational methods, calculate switch samples X to be identified with
ωiSample in class Sample StorehouseDistance d, whereinFor ωiQ in class Sample Storehouse
Individual sample, its calculation formula are as follows：
①
The distance of switch samples X more to be identified and each sample of all class Sample Storehouse kinds again, and carry out with
Lower judgement：
②
If meeting the switch samples X more to be identified of 2. formula, it is determined as X ∈ ωi, i.e., switch samples X shapes to be identified
State is ωiThe on off state of class Sample Storehouse, is otherwise determined as
Further, module also is established including training sample database in the template matches module, the module is former by designing
Beginning Sample Storehouse, redundant samples storehouse and training sample database, each sample in original sample storehouse is trained, and to training result
Classified, and be respectively put into redundant samples storehouse or training sample database, so as to automatically obtain training sample database.This mode causes
The population size of training sample set is moderate, can effectively distinguish the classification of traffic sign, is unlikely to have influence on the speed of system again
Degree.
Further, described image pretreatment module includes being converted to gray-scale map unit and image denoising unit.Described turn
Turn to gray-scale map unit and the color switching image collected is converted into gray level image, described image denoising unit is to gray level image
Medium filtering is carried out, noise point present in gray scale is removed, avoids the interference that noise spot is brought to image recognition.
Further, described image feature recognition module includes image smoothing unit, image sharpening unit, image binaryzation
Unit and feature identification unit, image is smoothed described image smooth unit so that image display effect is more clear
Wash, switch and background differentiation effect are more obvious, and gray-scale map is carried out binaryzation, switch and background by described image binarization unit
Further separation, feature identification unit are used for judging on off state.
The present invention compared with prior art, has the following advantages that by using above-mentioned technical proposal：
The present invention chooses Euclidean distance as grader, the discrimination highest of Euclidean distance, illustrates that it can effectively classify
Switch shape facility is redirected, is relatively specific for redirecting switch Shape Classification, and its average classification time is most short, meets to calculate
Method requirement of real-time.It is automatic to establish module from all original sample storehouses of acquisition by training sample database for the system of the present invention
Obtaining training sample database so that the population size of whole training sample set is moderate, can effectively distinguish the classification of switch shape,
It is unlikely to have influence on the speed of system again.The present invention can not only reach Real time identification for redirecting switch, at the same discrimination compared with
It is high.
Brief description of the drawings
Fig. 1 is the structural representation of embodiments of the invention；
Fig. 2 is the image pre-processing module of embodiments of the invention, characteristics of image processing identification module and template matches mould
The functional status schematic diagram of block；
Fig. 3 (a) is the on off state video artwork of the interception of embodiments of the invention；
Fig. 3 (b) is the image after the on off state gray processing of embodiments of the invention；
Fig. 3 (c) is the image after the on off state image smoothing of embodiments of the invention；
Fig. 3 (d) is the image after the on off state image sharpening of embodiments of the invention；
Fig. 3 (e) is the image after the on off state image binaryzation of embodiments of the invention；
Fig. 3 (f) is the image for the switch sections that the on off state feature recognition of embodiments of the invention is irised out.
Fig. 3 (g) is the measured switch binary image that the on off state feature of embodiments of the invention intercepts out.
Fig. 4 is the training sample database generation block diagram of embodiments of the invention.
Fig. 5 is the on off state OPEN state diagrams of embodiments of the invention.
Fig. 6 is the on off state CLOSE state diagrams of embodiments of the invention.
Fig. 7 is the state diagram of other states of the on off state of embodiments of the invention.
Embodiment
In conjunction with the drawings and specific embodiments, the present invention is further described.
As a specific embodiment, as depicted in figs. 1 and 2, a kind of on off state identifying system of the invention, including
Image capture module, communication module, image pre-processing module, characteristics of image processing identification module and template matches module, it is described
Image capture module and communication module are arranged on motor computer room, and the communication module is connected with image capture module telecommunications, described
Image pre-processing module, characteristics of image identification module and template matches module are arranged at Surveillance center, and image preprocessing mould
Block carries out data interaction by communication module and image capture module,
Described image acquisition module gathers in real time to disk rotary transition on off state on switch cubicle, with reference to shown in figure 3 (a),
The video artwork of the interception of the on off state intercepted by the present embodiment.
The realtime image data that the communication module collects image capture module is transferred to image pre-processing module,
Described image pretreatment module receives the realtime image data of communication module transmission, and carries out figure to the view data
As obtaining pretreatment image after the processing of gray processing, image smoothing, image sharpening and image binaryzation, with reference to figure 3 (b), Fig. 3 (c),
Fig. 3 (d), Fig. 3 (e) are respectively the image that image gray processing, image smoothing, image sharpening and image binaryzation handle to obtain.
Described image feature recognition module is connected with image pre-processing module, is extracted pretreatment image and is determined current circle
Spiral transition on off state feature, with reference to shown in figure 3 (f) and Fig. 3 (g), made the transition on off state spy to determine current disk rotary
The image of sign and to the image carry out image binaryzation processing after on off state figure.
The template matches module measured switch state feature and the on off state image progress in given Sample Storehouse
Match somebody with somebody, similarity measure values are calculated by similarity measurement, obtain the state conclusion of current disc rotary switch.
The template matches module uses Euclidean distance computational methods, it is known that on off state has two classes of OPEN and CLOSE
Not, can be set to:ω1And ω2.There is N per classiIndividual sample image, then ωiClass is expressed as
For switch samples to be identified, X=(x1,x2,...xn), calculate itself and OPEN, the distance of CLOSE state samples.Calculating is treated
The switch samples X and ω of identificationiSample in class Sample StorehouseDistance d, whereinFor
ωiQ-th of sample in class Sample Storehouse, its calculation formula are as follows：
①
The distance of switch samples X more to be identified and each sample of all class Sample Storehouse kinds again, and carry out with
Lower judgement：
②
If meeting the switch samples X more to be identified of 2. formula, it is determined as X ∈ ωi, i.e., switch samples X shapes to be identified
State is ωiThe on off state of class Sample Storehouse, is otherwise determined as
Also establish module including training sample database in the template matches module, the module by design original sample storehouse,
Redundant samples storehouse and training sample database, each sample in original sample storehouse is trained, and training result is classified,
And it is respectively put into redundant samples storehouse or training sample database.This division so that the population size of training sample set is moderate, Ji Nengyou
The differentiation of effect redirects the state of switch, is unlikely to have influence on the speed of system again.
For obtain multi-light, angle, position switch image as sample set, substantial amounts of experiment has been carried out, in room
Under the different illumination condition such as outer fine day, cloudy day, cloudy, fixing camera, change the state of switch, position, distance etc., shooting
Etc. big switch samples image, on off state identification storehouse is made, the research for the detection and identification of on off state.On off state
The method that detection is largely all based on statistical learning with the recognizer used in identifying system, discrimination to training set according to
Bad property is very strong.Sample for training is more, and pattern is more (illumination, weather, light, angle, position etc.), grader it is extensive
Ability is stronger, and obtained recognition effect is better.But training sample can excessively cause the room and time complexity of system
Can be very high.
Therefore, with reference to shown in figure 4, the system establishes mould from all original sample storehouses of acquisition by training sample database
Block automatically obtains training sample database, comprises the following steps：
1. obtaining original sample storehouse O by switching binary image detection, and original sample storehouse is divided into OPEN, CLOSE
Two class O of two states1,O2；
2. the original state of training sample and redundant samples is sky；
3. per a kind of O in pair original sample storehousei, the extraction of sample is trained respectively, takes certain in original sample storehouse a kind of
Sample OiIn the 1stPut into training sample database, and delete in original sample storehouse
4. all samples in pair training sample database are trained, characteristic vector storehouse is obtained；
5. compare all samples in original sample storehouse compared with characteristic vector storehouse, obtain each sample and feature to
The minimum range D (j) (j=1,2) in storehouse is measured, wherein, N OiIn current total sample number.
6. the maximum D (j) in 4 is found, i.e. sample corresponding to max (D (j))See whether it meets threshold condition
max(D(j))≥Thmin, wherein ThminFor sampleThe minimum threshold of training sample database can be entered, if meeting threshold condition,
By sampleSample Storehouse is put into, otherwise, then it is not put into；
7. finding out all samples that minimum range is less than threshold condition, that is, meet D (j)<Thmax, wherein, ThmaxFor sampleThe max-thresholds in sample redundancy storehouse can be entered, be put into redundancy storehouse；
8. the process of repeat step 4~6, until original sample storehouse OiFor sky；
9. the process of repeat step 2~8, until Oi(i=1,2) training sample database is all obtained.
With it is artificial select storehouse compared with, the advantages of selecting storehouse system automatically, is:(1) overcome the artificial randomness for selecting storehouse, can save
Substantial amounts of manpower and time；(2) the automatic sample for selecting storehouse system to select maximum inter- object distance so that sample is more accurate, more
Representative, redundancy is seldom, reduces the quantity of Sample Storehouse.
Described image pretreatment module includes being converted to gray-scale map unit and image denoising unit.It is described to be converted into gray-scale map
The color switching image collected is converted into gray level image by unit, and described image denoising unit carries out intermediate value filter to gray level image
Ripple, noise point present in gray scale is removed, avoid the interference that noise spot is brought to image recognition.
Described image feature recognition module includes image smoothing unit, image sharpening unit, image binaryzation unit and spy
Levy recognition unit, image is smoothed described image smooth unit cause image display effect more cleans, switch and
Background differentiation effect is more obvious, and gray-scale map is carried out binaryzation by described image binarization unit, and switch further divides with background
From feature identification unit is used for judging on off state.
Finally, by above-mentioned judgement, obtained switch has come to three kinds of states：OPEN states, CLOSE states and other
State is (such as：Switch is covered by other articles, and camera goes wrong).
When system judges that measured switch image is OPEN state.The lamp of OPEN on interface will be shown as green,
CLOSE lamp shows yellowly.As shown in figure 5, OPEN states.
When system judges that measured switch image is CLOSE state.The lamp of CLOSE on interface will be shown as green,
OPEN lamp shows yellowly.As shown in fig. 6, CLOSE states.
When it is other states that system, which judges measured switch image, two lamps will show red simultaneously, that is, mistake occur.Such as
Shown in Fig. 7, it is determined as other states.
Although specifically showing and describing the present invention with reference to preferred embodiment, those skilled in the art should be bright
In vain, do not departing from the spirit and scope of the present invention that appended claims are limited, in the form and details can be right
The present invention makes a variety of changes, and is protection scope of the present invention.
Claims (3)
- A kind of 1. on off state identifying system, it is characterised in that：Including image capture module, communication module, image preprocessing mould Block, characteristics of image processing identification module and template matches module, described image acquisition module and communication module are arranged on motor machine Room, the communication module are connected with image capture module telecommunications, described image pretreatment module, characteristics of image identification module and mould Plate matching module is arranged at Surveillance center, and image pre-processing module carries out data by communication module and image capture module Interaction,Described image acquisition module gathers in real time to disk rotary transition on off state on switch cubicle,The realtime image data that the communication module collects image capture module is transferred to image pre-processing module,Described image pretreatment module receives the realtime image data of communication module transmission, and image ash is carried out to the view data Pretreatment image is obtained after degreeization, image smoothing, image sharpening and image binaryzation processing,Described image feature recognition module is connected with image pre-processing module, is extracted pretreatment image and is determined current disk rotary Make the transition on off state feature,The template matches module measured switch state feature is matched with the on off state image in given Sample Storehouse, is led to Cross similarity measurement and similarity measure values are calculated, obtain the state conclusion of current disc rotary switch；Wherein, also establish module including training sample database in the template matches module, the module by design original sample storehouse, Redundant samples storehouse and training sample database, each sample in original sample storehouse is trained, and training result is classified, And redundant samples storehouse or training sample database are respectively put into, so as to automatically obtain training sample database；Specifically comprise the following steps：(1) obtains original sample storehouse O by switching binary image detection, and original sample storehouse is divided into OPEN, CLOSE two Two class O of kind state1,O2；(2) original state of training samples and redundant samples is sky；(3) is to every a kind of O in original sample storehousei, the extraction of sample is trained respectively, takes a kind of sample of certain in original sample storehouse OiIn the 1stPut into training sample database, and delete in original sample storehouse(4) is trained to all samples in training sample database, obtains characteristic vector storehouse；(5) compares all samples in original sample storehouse compared with characteristic vector storehouse, obtains each sample and characteristic vector The minimum range D (j) (j=1,2) in storehouse, wherein, N OiIn current total sample number；(6) finds the maximum D (j) in step 4, i.e. sample corresponding to max (D (j))See whether it meets threshold value bar Part max (D (j)) >=Thmin, wherein ThminFor sampleThe minimum threshold of training sample database can be entered, if meeting threshold condition, Then by sampleSample Storehouse is put into, otherwise, then it is not put into；(7) finds out all samples that minimum range is less than threshold condition, that is, meets D (j)<Thmax, wherein, ThmaxFor sample The max-thresholds in sample redundancy storehouse can be entered, be put into redundancy storehouse；(8) process of repeat steps 4~6, until original sample storehouse OiFor sky；(9) process of repeat steps 2~8, until Oi(i=1,2) training sample database is all obtained.
- A kind of 2. on off state identifying system according to claim 1, it is characterised in that：Described image pretreatment module bag Include and be converted to gray-scale map unit and image denoising unit, the gray-scale map unit that is converted to turns the color switching image collected Gray level image is turned to, described image denoising unit carries out medium filtering to gray level image, noise point present in gray scale is removed, Avoid the interference that noise spot is brought to image recognition.
- A kind of 3. on off state identifying system according to claim 1, it is characterised in that：Described image feature recognition module Including image smoothing unit, image sharpening unit, image binaryzation unit and feature identification unit, described image smooth unit pair Image is smoothed so that image display effect becomes apparent from, and switch and background differentiation effect are more obvious, described image Gray-scale map is carried out binaryzation by binarization unit, and switch further separates with background.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410668236.2A CN104331710B (en) | 2014-11-19 | 2014-11-19 | On off state identifying system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410668236.2A CN104331710B (en) | 2014-11-19 | 2014-11-19 | On off state identifying system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104331710A CN104331710A (en) | 2015-02-04 |
| CN104331710B true CN104331710B (en) | 2018-01-02 |
Family
ID=52406431
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410668236.2A Active CN104331710B (en) | 2014-11-19 | 2014-11-19 | On off state identifying system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104331710B (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106250902A (en) * | 2016-07-29 | 2016-12-21 | 武汉大学 | Power system on off state detection method based on characteristics of image template matching |
| CN106339722A (en) * | 2016-08-25 | 2017-01-18 | 国网浙江省电力公司杭州供电公司 | Line knife switch state monitoring method and device |
| CN106570865A (en) * | 2016-11-08 | 2017-04-19 | 国家电网公司 | Digital-image-processing-based switch state detecting system of power equipment |
| CN108068817A (en) * | 2017-12-06 | 2018-05-25 | 张家港天筑基业仪器设备有限公司 | A kind of automatic lane change device and method of pilotless automobile |
| CN108334815A (en) * | 2018-01-11 | 2018-07-27 | 深圳供电局有限公司 | Inspection method of power secondary equipment, and switch state identification method and system |
| CN108334824B (en) * | 2018-01-19 | 2022-05-06 | 国网电力科学研究院武汉南瑞有限责任公司 | High-voltage isolating switch state identification method based on background difference and iterative search |
| CN109409395A (en) * | 2018-07-29 | 2019-03-01 | 国网上海市电力公司 | Using the method for template matching method identification target object region electrical symbol in power monitoring |
| CN109100760B (en) * | 2018-08-16 | 2019-12-24 | 集美大学 | Big dipper and satellite communication bimodulus high accuracy location thing allies oneself with terminal |
| CN111382673A (en) * | 2020-01-09 | 2020-07-07 | 南京艾拓维讯信息技术有限公司 | KVM system and method for monitoring windmill power generation state |
| CN112178706B (en) * | 2020-10-14 | 2021-11-05 | 宁波方太厨具有限公司 | Method and system for identifying fire gear of stove and method and system for linking smoke stove |
| CN113901964A (en) * | 2021-12-07 | 2022-01-07 | 北京惠朗时代科技有限公司 | Staff face excitement detection method and system for intelligent company management |
| CN114202731A (en) * | 2022-02-15 | 2022-03-18 | 南京天创电子技术有限公司 | Multi-state knob switch identification method |
| CN116907349B (en) * | 2023-09-12 | 2023-12-08 | 北京宝隆泓瑞科技有限公司 | Universal switch state identification method based on image processing |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101335465A (en) * | 2008-07-24 | 2008-12-31 | 华中科技大学 | Round disk rotation type switch state image recognition apparatus for electric switch cabinet |
| CN101833673A (en) * | 2010-05-18 | 2010-09-15 | 华中科技大学 | Electric power switchgear switch state image recognition system |
| CN102938055A (en) * | 2012-10-09 | 2013-02-20 | 哈尔滨工程大学 | Hand bone identification system |
| CN103077376A (en) * | 2012-12-30 | 2013-05-01 | 信帧电子技术（北京）有限公司 | Method for re-identifying human body image based on video image |
-
2014
- 2014-11-19 CN CN201410668236.2A patent/CN104331710B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101335465A (en) * | 2008-07-24 | 2008-12-31 | 华中科技大学 | Round disk rotation type switch state image recognition apparatus for electric switch cabinet |
| CN101833673A (en) * | 2010-05-18 | 2010-09-15 | 华中科技大学 | Electric power switchgear switch state image recognition system |
| CN102938055A (en) * | 2012-10-09 | 2013-02-20 | 哈尔滨工程大学 | Hand bone identification system |
| CN103077376A (en) * | 2012-12-30 | 2013-05-01 | 信帧电子技术（北京）有限公司 | Method for re-identifying human body image based on video image |
Non-Patent Citations (1)
| Title |
|---|
| 一种简单的数字图像识别方法;韩帅;《电子质量》;20100930(第9期);第6-8页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104331710A (en) | 2015-02-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104331710B (en) | On off state identifying system | |
| CN113139521B (en) | Pedestrian boundary crossing monitoring method for electric power monitoring | |
| CN115082683A (en) | Injection molding defect detection method based on image processing | |
| CN111814686A (en) | Vision-based power transmission line identification and foreign matter invasion online detection method | |
| CN108537154A (en) | Transmission line of electricity Bird's Nest recognition methods based on HOG features and machine learning | |
| CN107179479B (en) | Visible light image-based transmission conductor strand scattering defect detection method | |
| CN105044122A (en) | Copper part surface defect visual inspection system and inspection method based on semi-supervised learning model | |
| CN109946304A (en) | Surface defects of parts on-line detecting system and detection method based on characteristic matching | |
| CN108268832A (en) | Electric operating monitoring method, device, storage medium and computer equipment | |
| CN104598916B (en) | A kind of construction method and train knowledge method for distinguishing of train identifying system | |
| CN103824092A (en) | Image classification method for monitoring state of electric transmission and transformation equipment on line | |
| CN108335294A (en) | The power distribution room abnormality image-recognizing method of complex condition | |
| CN105740844A (en) | Insulator cracking fault detection method based on image identification technology | |
| CN106504247B (en) | A kind of quick-fried automatic testing method of the insulator chain based on visible images and device | |
| CN112819784A (en) | Method and system for detecting broken strands and scattered strands of wires of distribution line | |
| CN109446964A (en) | Face detection analysis method and device based on end-to-end single-stage multiple scale detecting device | |
| CN113888462A (en) | Crack identification method, system, readable medium and storage medium | |
| CN106650735B (en) | A kind of LED character automatic positioning recognition methods | |
| CN111524121A (en) | Road and bridge fault automatic detection method based on machine vision technology | |
| CN111339927A (en) | Intelligent work state identification system for personnel in electric power business hall | |
| CN117197700B (en) | Intelligent unmanned inspection contact net defect identification system | |
| CN114757962A (en) | Overhead line forest fire identification method | |
| CN110516565A (en) | A kind of sea ship recognition methods based on Zernike and SVM | |
| CN104376322B (en) | A kind of intelligent evaluating method of container number pretreatment quality | |
| CN110363749A (en) | Stockbridge damper corrosion degree evaluation method based on image procossing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Zheng Jiachun Inventor after: Huang Lianghao Inventor after: Tang Kai Inventor after: You Shumin Inventor after: Zhuang Wei Inventor after: Chen Weimin Inventor after: Zhao Bing Inventor after: Li Jie Inventor after: Lu Linhua Inventor after: Liang Zhongwei Inventor before: Tang Kai Inventor before: Zheng Jiachun Inventor before: You Shumin Inventor before: Liang Zhongwei Inventor before: Zhuang Wei Inventor before: Chen Weimin Inventor before: Zhao Bing |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |