CN102143354B - Method for recognizing and calculating galloping of transmission conductor based on video image processing - Google Patents
Method for recognizing and calculating galloping of transmission conductor based on video image processing Download PDFInfo
- Publication number
- CN102143354B CN102143354B CN 201010606740 CN201010606740A CN102143354B CN 102143354 B CN102143354 B CN 102143354B CN 201010606740 CN201010606740 CN 201010606740 CN 201010606740 A CN201010606740 A CN 201010606740A CN 102143354 B CN102143354 B CN 102143354B
- Authority
- CN
- China
- Prior art keywords
- image
- galloping
- transmission line
- transmission
- video
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a method for recognizing and calculating galloping of a transmission conductor based on video image processing, belonging to the technical field of remote digital video monitoring and image identification. The method comprises the steps of real-time transmitting a digital video signal collected by a camera to a monitoring centre through a transmission channel in the form of video stream; carrying out remote video monitoring on the field at the monitoring centre; intercepting a monitoring target image from the video stream; and calculating to obtain amplitude information of galloping through a series of image processing and identifications like image gray processing, image segmentation, transmission conductor extraction based on a chain code, Hough transformation and galloping amplitude calculation. Therefore, the galloping amplitude of the transmission conductor can be accurately, intuitively and effectively calculated; meanwhile, long-time galloping data can be saved in a database so as to provide the remote video monitoring and the image identification with original data of the galloping state of the transmission conductor, and thereby, the galloping state of the transmission conductor is convenient to analyze; and the omitted and false reports of accidents are reduced to assure the safe operation of the transmission line.
Description
Technical field
The invention belongs to remote digital video monitoring and image recognition technology field, be mainly used in the identification of Galloping of Overhead Transmission Line is calculated, relate to image gray processing, image segmentation, the transmission pressure extraction based on chain code, Hough transformation, wave that a series of images such as amplitude calculating is processed and the identification computing technique, can identify automatically calculating to waving of transmission pressure, be a kind of Galloping of Overhead Transmission Line identification computational methods based on monitoring remote video and image recognition technology.
Background technology
At present in the on-line monitoring system of transmission line, mainly be to utilize various transducers to the monitoring of Galloping of Overhead Transmission Line, the mass data that such as displacement transducer, acceleration transducer, air velocity transducer etc. obtains comes COMPREHENSIVE CALCULATING to wave size.But this method need to install transducer additional at transmission line diverse location place, particularly in order to guarantee precision, then needs at every circuit a large amount of transducers to be installed, and has increased the weight of the burden of transmission line.Simultaneously, in view of the requirement of safety in production, installation quantity and the position of transmission line upper sensor are controlled, cause the precision of the method lower.The transmission line of 220KV above many loop lines and the bundle conductors of adopting, when the bundle conductor radical is more, wire is liftoff when higher, more easily causes and waves more.Therefore when Galloping of Overhead Transmission Line surpasses the limit that circuit can bear, can cause just that gold utensil damages, disconnected thigh of wire, phase fault, shaft tower inclination even the serious accident such as tower, make electric power enterprise and people's productive life suffer great loss.Adopt remote digital video monitoring and image recognition technology that Galloping of Overhead Transmission Line is monitored, more more directly perceived, effective than existing monitoring means, and need to not install too much equipment additional at transmission pressure, avoided the impact on transmission line.
Remote digital video monitoring and image identification system are exactly with remote digital video monitoring and image is processed and recognition technology combines, at first by being installed in the equipment such as camera on the steel tower and transmission of video, the digital video signal that collects is sent back Surveillance center by transmission channel in real time in the mode of video flowing, in Surveillance center remote video monitoring is carried out at the scene, intercepting monitors target image from video flowing, by corresponding Preprocessing Technique calculating is analyzed, processes and identified to digital video image.The manifestation mode of Galloping of Overhead Transmission Line is generally level and waves, and the probability of waving on the generation vertical direction is minimum.Thereby, the present invention adopts remote digital video monitoring and image identification system, the automatic identification of waving on the realization transmission pressure horizontal direction is calculated, for guaranteeing that electric power enterprise production safety and failure diagnosis provide a kind of new intuitively means accurately, the Monitoring Data of its long-term accumulation also can provide foundation for Transmission Line Design.For helping electric power enterprise to improve the automatic monitoring level very important realistic meaning is arranged.
Summary of the invention
The purpose of this invention is to provide a kind of Galloping of Overhead Transmission Line identification computational methods of processing based on video image, it is characterized in that, the step of Galloping of Overhead Transmission Line identification computational methods is as follows,
1) utilizes the digital video signal of the camera collection transmission line of installing on the high-pressure tower;
2) digital video signal of collection transmission line sends back Surveillance center in the mode of video flowing in real time by transmission channel;
3) Surveillance center carries out remote video monitoring to the scene, and intercepting monitors Target Photo from video flowing, monitors that wherein Target Photo is divided into original image and surveillance map picture;
4) to original image carry out that gray processing, two dimensional image are cut apart, chain code searching and Hough (Hough) conversion carry out that image is processed and identification, and transmission pressure is extracted from the image background of complexity, calculate its center of gravity, center, and store;
5) the surveillance map picture is also carried out identical processing, identification and calculating;
6) use rule of three to calculate the throw amplitude value of power transmission line reality at last according to the center of gravity of transmission pressure and the equivalent line that the center calculates respectively original image and surveillance map picture transmission pressure separately in the original image that obtains and the surveillance map picture, and according to the ultimate range between these two equivalent lines.
The transmission pressure image that the original nothing of described original image indication is waved; Described surveillance map similarly is the transmission pressure image that intercepts in the observation process.
The described Two-dimensional Maximum Ostu method that is based on simulated annealing and particle swarm optimization that adopts when the transmission pressure road image behind the gray processing is carried out binary segmentation, it combines the advantage of simulated annealing, particle swarm optimization and Two-dimensional Maximum Ostu method, and can effectively overcome separately shortcoming, compare with general image partition method and to have better segmentation effect and robustness, be more suitable for cutting apart the lower complicated image of contrast.
It is described that what the transmission pressure image behind the binary segmentation was carried out adopting when the transmission pressure image extracts is improved Freeman chain code searching algorithm, this algorithm is on the basis of Freeman chain code representation, in image, generally be the vertically characteristics of linear vertical distribution according to transmission pressure in the actual environment, once only need lower, bottom right, the right side, the left side by counterclockwise sequential search image slices vegetarian refreshments, the pixel of 5 the field directions in lower-left, and unlike traditional chain code searching method each 8 pixels all searching for neighborhood, thereby greatly reduced amount of calculation.
The invention has the beneficial effects as follows that the present invention is take truncated picture in the transmission line live video streams as research object, process and identifying by a series of image, draw the amplitude information of waving, can accurately, intuitively, effectively calculate the amplitude of waving of transmission pressure, simultaneously can in database, preserve the long-term data of waving, the initial data of the state of waving of transmission pressure is provided for monitoring remote video and image recognition, is convenient to the state analysis of waving to transmission pressure; The generation that the minimizing accident is failed to report and misrepresented deliberately is to guarantee the safe operation of transmission line.
Description of drawings
Fig. 1 is transmission pressure original image and monitoring objective image, wherein (a) original image (b) monitoring objective image.
Fig. 2 is the gray-scale map of transmission pressure image preprocessing process, wherein (a) original image (b) monitoring objective image.
Fig. 3 is the binary segmentation result figure of the transmission pressure image preprocessing process of Fig. 1, Fig. 2, wherein (a) original image, (b) monitoring objective image.
Fig. 4 is the as a result figure of transmission pressure image behind the chain code searching, and wherein (a) is to the result of original image behind the chain code searching, (b) to the result of monitoring objective image behind the chain code searching.
Fig. 5 is the as a result figure of transmission pressure image behind the segmentation Hough transformation, and wherein (a) is to the result of original image behind the segmentation Hough transformation, (b) to the result of monitoring objective image behind the segmentation Hough transformation.
Fig. 6 is the program flow diagram of Galloping of Overhead Transmission Line identification.
Embodiment
The invention provides a kind of Galloping of Overhead Transmission Line identification computational methods of processing based on video image, be explained below in conjunction with accompanying drawing, according to the program flow diagram of Galloping of Overhead Transmission Line identification shown in Figure 6.Process description to Galloping of Overhead Transmission Line identification computational methods is as follows:
One, coloured image gray processing
Fig. 1 is original image and the monitoring objective image of transmission pressure.Because the data volume of coloured image is large, and the monitoring objective Characteristic of Image is complicated, is unfavorable for the needed characteristic quantity of rapid extraction.And gray-scale map is the simplicial graph picture that only contains monochrome information, and the gray scale denotation of graph is quantized into 0 to 255 totally 256 ranks to brightness value usually, 0 the darkest (complete black), 255 the brightest (complete white).Therefore, need to carry out gray processing to image processes.
Fig. 1 is carried out the gray processing operation, give up to fall complicated colouring information, obtain gray-scale map result such as Fig. 2.
Two, employing is based on the Two-dimensional Maximum Ostu method split image of simulated annealing and particle swarm optimization.
Threshold segmentation method commonly used has maximum variance between clusters (OTSU) method, maximum entropy method (MEM) etc.The half-tone information that these methods all are based on image comes calculated threshold.But the signal to noise ratio of working as image is low, and gray difference is not obvious, and when target area was less, these methods just were difficult to obtain preferably segmentation effect.And Two-dimensional Maximum inter-class variance algorithm (2D-OTSU) is based on two-dimensional histogram, and it has not only considered the half-tone information of image, has also considered the relevant information of neighborhood space.Therefore, Two-dimensional Maximum inter-class variance algorithm has higher segmentation precision and robustness with comparing based on the histogrammic partitioning algorithm of one dimension, is more suitable for the lower complicated image of contrast.Although ask for the effect of image optimum segmentation threshold with the Two-dimensional Maximum Ostu method fine, computing time is long, poor practicability.Genetic algorithm and particle swarm optimization (PSO) can search out globally optimal solution rapidly, and particle swarm optimization is compared with genetic algorithm, and desired parameters is few and amount of calculation is little, and convergence rate is faster.But they all can't guarantee to converge on optimal solution, easily are absorbed in local optimum.Simulated annealing (SA) be in theory a kind of convergence with probability 1 in the optimization method of globally optimal solution, have the ability of breaking away from locally optimal solution, but need to spend the more time in order to search out globally optimal solution.For these problems, the present invention is on Two-dimensional Maximum Ostu method basis, a kind of Two-dimensional Maximum Ostu method based on simulated annealing and particle swarm optimization has been proposed, the method has preferably convergence and computational speed in the search globally optimal solution, the threshold value that calculates makes transmission pressure class and the maximization of the variance between background classes in the image.Wire in the transmission pressure image presents linear vertical distribution, and be between the background of irregular distribution the significant difference of existence is arranged, therefore the transmission pressure image can partly filter out the complex background in the image after the Two-dimensional Maximum Ostu method is cut apart, and effectively highlights the transmission pressure target in the image.
When utilizing any vector (s, t) in the two-dimensional histogram to Image Segmentation Using, the variance between image object and background classes is as follows:
B(s,t)=ω
0(s,t)[(u
0i-u
zi)
2+(u
0j-u
zj)
2]+ω
1(s,t)[(u
1i-u
zi)
2+(u
1j-u
zj)
2 (1)
In the following formula: 0≤s, t≤L-1, ω
0(s, t) and ω
1(s, t) represents respectively the ratio of background classes and the shared entire image of target class, u
0iAnd u
1iRepresent respectively the gray average that background classes and target class are corresponding, u
0jAnd u
1jThe average that represents respectively the neighboring mean value j that background classes and target class pixel are corresponding, u
ZiThe gray average of expression entire image, u
ZjThe average of expression entire image neighborhood of pixel points average j.
Optimal segmenting threshold (s
0, t
0) when being taken at B (s, t) for maximum:
What particle swarm optimization was described is that each individuality is with certain speed flight, and this flying speed is dynamically adjusted according to the flying experience of individuality and the flying experience of colony in a n dimension search volume.The evolution iterative equation of PSO is as follows:
v
ij(t+1)=wv
ij(t)+c
1r
1j(t)(p
ij(t)-x
ij(t))+c
2r
2j(t)(p
gj(t)-x
ij(t)) (3)
x
ij(t+1)=x
ij(t)+v
ij(t+1) (4)
In the following formula: X
i=(x
I1, x
I2..., x
In) current location of expression particulate i, V
i=(v
I1, v
I2..., v
In) present speed of expression particulate i, P
i=(p
I1, p
I2..., p
In) the individual desired positions of expression particulate i, the position with best adaptive value that namely particulate i experiences.P
g=(p
G1, p
G2..., p
Gn) be the desired positions that all particulates live through in the population, i.e. overall desired positions.Inertia weight w generally regulates from big to small gradually in the evolution, and span is 0.2~1.2; c
1, c
2Be acceleration constant, usually 0~2 value; r
1, r
2For obeying two mutually independent random variables of U (0,1).
Simulated annealing produces the solution of combinatorial optimization problem with the Metropolis algorithm, and by transition probability P corresponding to Metropolis criterion
iDetermine whether to accept the transfer from current solution i to new explanation j.
In the following formula: t
kThe expression temperature control parameter, f is fitness function.Allow t when simulated annealing is initial
kGet larger value, after carrying out a certain amount of transfer, slowly reduce t
kValue so repeats, until algorithm stops.
Concrete steps based on the Two-dimensional Maximum Ostu method of simulated annealing and particle swarm optimization are as follows:
1) parameter of simulated annealing and particle swarm optimization is set.For example establish initial temperature t
0=100, temperature damping's function t
K+1=α t
k(0<α<1), population scale m=10, w
Max=1.2, w
Min=0.8, c1=c2=1.6.
2) initialization population.The position initialization of particulate is certain gray value of 0 to 255, and the dimension n of particulate=2 set initial velocity.Calculate the desired positions P of each particulate according to fitness function formula (1)
i, get wherein maximum value as current desired positions P
g
3) establish the initial value Y=P of simulated annealing
g, carry out step search according to criterion (5) and obtain a new explanation Y.
4) according to iterative formula (3) and (4) speed and the position of each particulate are upgraded, obtained current desired positions P
g
5) make formula (1) be fitness function f, if f (Y)=f (P
g), and simulated annealing and particle swarm optimization all restrain, and then algorithm finishes; If f (Y)>f (P
g), then in m particulate, choose at random a particulate i, make its current location X
iWith current desired positions P
iBe Y, return step 3; If f (Y)<f (P
g), then make Y=P
g, return step 3.
Employing based on the Two-dimensional Maximum Ostu method of simulated annealing and particle swarm optimization to the result of transmission pressure Image Segmentation Using as shown in Figure 3.
Three, tentatively extract the transmission pressure image based on improved chain code search method
The background more complicated of the transmission line image that collects in actual environment, the background that wherein exists various shapes when image meeting discovery after cutting apart is remaining, disturbs the identification to target, as shown in Figure 3.The image that is partitioned into like this can't be used for identifying calculating.But can find that from Fig. 3 transmission pressure has the geometric properties that significantly is linear vertical distribution, and the pixel of transmission pressure part also there is stronger connectedness in the image.Therefore the present invention at first adopts the preliminary transmission pressure image that extracts the image of improved chain code search method after cutting apart according to above characteristics.
The present invention adopts improved Freeman chain code representation, with lower, the bottom right of pixel, the right side, a left side, lower left to being encoded to respectively 0,1,2,3,4, and coding is divided into two parts, right half part forms by 0,1,2, the direction of search is taken as+and 1, left-half forms by 3,4, and the direction of search is taken as-1.When in the right half part search procedure, not finding pixel non-vanishing and that be not labeled, then change the direction of search into 1, proceed the left-half search, otherwise include this point in chained list, continue in its neighborhood, to keep original direction of search search next node; If do not find pixel non-vanishing and that be not labeled in the first half search equally, then the direction of search is changed into-1, proceed the latter half search, if 5 pixels all do not satisfy condition then the chain end of list (EOL).For the chained list after having searched for set, also to reject length less than the chained list of certain threshold value, through lot of experiment validation, threshold value is generally got 0.25 times of picture altitude and can be met the demands.The chain code search method at most only needs 5 field directions of search after improving, and needn't all search for nearly 8 field pixel at every turn, thereby has increased search efficiency, has significantly cut down amount of calculation, has improved arithmetic speed.Result behind the chain code searching as shown in Figure 4.
Four, adopt the segmentation Hough transform method to identify the transmission pressure line segment
In order to identify exactly the position of transmission pressure, the present invention also need adopt Hough transform method that transmission pressure is further extracted after the chain code search method tentatively extracts target.But, because common Hough transformation generally all is used for extracting the regular figures such as straight line and circle, and the transmission pressure of waving in the real video often has irregular radian, so the present invention adopts the conversion of segmentation Hough transform method to go out N section short lines, and simulate the transmission pressure with radian with these short lines, the value of N and the residing actual environment of image and accuracy of identification require relevant, and N gets 5 and can satisfy required precision generally speaking.Through lot of experiment validation, segmentation Hough method of changing can identify the transmission pressure line segment comparatively exactly on the basis of chain code searching, and recognition result as shown in Figure 5.
Five, calculate center line and the center of gravity line of transmission pressure image, and ask for final equivalent line according to center line and center of gravity line.
Every phase bundle conductor is locked by fixture, and the integral body of single-phase transmission line is waved with the amplitude of waving of single wire basically identical.Therefore to use at first respectively center method and gravity model appoach be a transmission pressure with single-phase transmission pressure equivalence in the present invention, and the center abscissa that transmission pressure i is capable and the computing formula of center of gravity abscissa are as follows:
In the following formula: x
ImidExpression transmission pressure central point abscissa, x
IleftExpression transmission pressure left hand edge point abscissa, x
IrightExpression transmission pressure right hand edge point abscissa, above abscissa all be single-phase transmission pressure i capable in the abscissa of respective point.
In the following formula: x
IgravityThe every row focus point of expression transmission pressure abscissa, x
I1, x
I2X
InThe abscissa of the point on every wire in the presentation video delegation successively from left to right, n is the division number of single-phase transmission line, above abscissa all be single-phase transporting circuit i capable in the abscissa of respective point.
Behind the abscissa of the central point that calculates the every row of single-phase transmission pressure and focus point, its center line and center of gravity line just can be determined.
In theory, the Galloping of Overhead Transmission Line amplitude of single monitoring is a definite value, so the result of calculation of gravity model appoach and center method should be close, still, because inevitably there is difference between the two in the error that image is processed.Therefore, the present invention is by asking for by center line and the center of gravity line final equivalent line of center line conduct between the two, to reduce error, that is:
In the following formula: x
iRepresent the capable abscissa of i on the final equivalent line.
Six, calculate the Galloping of Overhead Transmission Line amplitude
Behind the final equivalent line of the target image that obtains respectively original image and monitoring, just can by the capable horizontal displacement absolute difference of asking between the two, use at last rule of three to calculate the Galloping of Overhead Transmission Line amplitude.
If the distance between the left and right edges wire of known single-phase transmission pressure is D, single pixel represents distance value d, and then wire i is capable waves range value w
iFor:
In the following formula: x
IoriginalThe capable abscissa of i on the final equivalent line of expression original image, x
ImonitorThe capable abscissa of i on the final equivalent line of expression monitoring objective image.
Actual Galloping of Overhead Transmission Line amplitude W gets all w
iIn maximum namely:
W=max(w
1,w
2,…w
height) (10)
In the following formula: height represents the height of transmission pressure image, i.e. the sum of all pixels of image vertical direction.
According to the capable center abscissa x of transmission pressure i in the original image that obtains and the monitoring objective image
ImidWith center of gravity abscissa x
Igravity, behind the equivalent line that calculates respectively original image and surveillance map picture transmission pressure separately, just can use rule of three to calculate the throw amplitude value W of transmission pressure reality according to the ultimate range between these two equivalent lines.
At last result of calculation is deposited in the database, can be with its important evidence of whether waving as transmission pressure.Simultaneously, in database long-term preserve wave data, can analyze for waving of transmission pressure initial data is provided, and provide useful help for Transmission Line Design and safe operation.
Claims (3)
1. Galloping of Overhead Transmission Line identification computational methods of processing based on video image is characterized in that, the step of Galloping of Overhead Transmission Line identification computational methods is as follows;
1) utilizes the digital video signal of the camera collection transmission line of installing on the high-pressure tower;
2) digital video signal of collection transmission line sends back Surveillance center in the mode of video flowing in real time by transmission channel;
3) Surveillance center carries out remote video monitoring to the scene, and intercepting monitors Target Photo from video flowing, monitors that wherein Target Photo is divided into original image and surveillance map picture;
4) to original image carry out that gray processing, two dimensional image are cut apart, chain code searching and Hough (Hough) conversion carry out that image is processed and identification, and transmission pressure is extracted from the image background of complexity, calculate its center of gravity, center, and store;
5) the surveillance map picture is also carried out identical processing, identification and calculating;
6) use rule of three to calculate the throw amplitude value of power transmission line reality at last according to the center of gravity of transmission pressure and the equivalent line that the center calculates respectively original image and surveillance map picture transmission pressure separately in the original image that obtains and the surveillance map picture, and according to the ultimate range between these two equivalent lines.
2. the described Galloping of Overhead Transmission Line of processing based on video image is identified computational methods according to claim 1, it is characterized in that, the described Two-dimensional Maximum Ostu method that is based on simulated annealing and particle swarm optimization that original image is carried out gray processing, adopts when two dimensional image is cut apart, it combines simulated annealing, particle swarm optimization and Two-dimensional Maximum Ostu method can overcome separately shortcoming effectively, compare with traditional image partition method and to have better segmentation effect and robustness, be more suitable for cutting apart the lower complicated image of contrast.
3. the described Galloping of Overhead Transmission Line of processing based on video image is identified computational methods according to claim 1, it is characterized in that, described original image is carried out gray processing, what the transmission pressure image after two dimensional image is cut apart carried out adopting when the transmission pressure image extracts is improved Freeman chain code searching algorithm, this algorithm is on the basis of Freeman chain code representation, in image, be the vertically characteristics of linear vertical distribution according to transmission pressure in the actual environment, once only need to press the lower of counterclockwise sequential search image slices vegetarian refreshments, the bottom right, right, left, the pixel of 5 the neighborhood directions in lower-left, and unlike traditional chain code searching method each 8 pixels all searching for neighborhood, thereby greatly reduced amount of calculation.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201010606740 CN102143354B (en) | 2010-12-16 | 2010-12-16 | Method for recognizing and calculating galloping of transmission conductor based on video image processing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201010606740 CN102143354B (en) | 2010-12-16 | 2010-12-16 | Method for recognizing and calculating galloping of transmission conductor based on video image processing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102143354A CN102143354A (en) | 2011-08-03 |
CN102143354B true CN102143354B (en) | 2013-03-06 |
Family
ID=44410542
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 201010606740 Expired - Fee Related CN102143354B (en) | 2010-12-16 | 2010-12-16 | Method for recognizing and calculating galloping of transmission conductor based on video image processing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102143354B (en) |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102155933B (en) * | 2011-03-08 | 2013-04-24 | 西安工程大学 | Method for measuring galloping of transmission conductor on the basis of video difference analysis |
CN102494760A (en) * | 2011-11-28 | 2012-06-13 | 西安雷迪维护***设备有限公司 | Video monitoring system for swinging of lead wire of high-voltage power transmission line |
CN103247060B (en) * | 2012-02-06 | 2016-03-02 | 华北电力科学研究院有限责任公司 | A kind of method for drafting of gallop distribution map and device thereof |
CN103093192B (en) * | 2012-12-28 | 2016-12-28 | 昆山市工业技术研究院有限责任公司 | The recognition methods that high voltage transmission line is waved |
CN103442209B (en) * | 2013-08-20 | 2017-02-22 | 北京工业大学 | Video monitoring method of electric transmission line |
CN103530620A (en) * | 2013-10-30 | 2014-01-22 | 华北电力大学 | Method for identifying bird nest on electric transmission line tower |
CN108734704B (en) * | 2018-05-07 | 2021-11-09 | 西安工程大学 | Transmission conductor strand breakage detection method based on gray variance normalization |
CN110276787B (en) * | 2019-06-27 | 2021-02-26 | 合肥工业大学智能制造技术研究院 | Conductor galloping monitoring method based on marker image detection |
CN110349172B (en) * | 2019-06-28 | 2022-12-16 | 华南理工大学 | Power transmission line external damage prevention early warning method based on image processing and binocular stereo ranging |
CN114219696A (en) * | 2021-11-09 | 2022-03-22 | 国网江苏省电力有限公司盐城供电分公司 | Line galloping real-time video monitoring system based on FPGA and DSP |
CN116046076A (en) * | 2023-03-09 | 2023-05-02 | 合肥工业大学 | Online detection system for power transmission line galloping based on machine vision technology |
CN116993743B (en) * | 2023-09-28 | 2024-03-19 | 南方电网数字电网研究院有限公司 | Method, device, equipment and storage medium for detecting galloping amplitude of power transmission line |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101430195A (en) * | 2008-12-19 | 2009-05-13 | 华北电力大学 | Method for computing electric power line ice-covering thickness by using video image processing technology |
-
2010
- 2010-12-16 CN CN 201010606740 patent/CN102143354B/en not_active Expired - Fee Related
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101430195A (en) * | 2008-12-19 | 2009-05-13 | 华北电力大学 | Method for computing electric power line ice-covering thickness by using video image processing technology |
Non-Patent Citations (1)
Title |
---|
*** 等.大型电力变压器漏磁场及附加损耗的研究.《华北电力大学学报》.1997,第24卷(第1期),17-22. * |
Also Published As
Publication number | Publication date |
---|---|
CN102143354A (en) | 2011-08-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102143354B (en) | Method for recognizing and calculating galloping of transmission conductor based on video image processing | |
CN111353413B (en) | Low-missing-report-rate defect identification method for power transmission equipment | |
CN103839065B (en) | Extraction method for dynamic crowd gathering characteristics | |
CN109670446A (en) | Anomaly detection method based on linear dynamic system and depth network | |
CN105426820B (en) | More people's anomaly detection methods based on safety monitoring video data | |
CN109738970A (en) | The method, apparatus and storage medium for realizing Lightning Warning are excavated based on lightning data | |
CN112560745B (en) | Method for discriminating personnel on electric power operation site and related device | |
CN112183643B (en) | Hard rock tension-shear fracture identification method and device based on acoustic emission | |
CN114998830A (en) | Wearing detection method and system for safety helmet of transformer substation personnel | |
CN108256447A (en) | A kind of unmanned plane video analysis method based on deep neural network | |
CN109241814A (en) | Pedestrian detection method based on YOLO neural network | |
CN108898098A (en) | Early stage video smoke detection method based on monitor supervision platform | |
CN116012762A (en) | Traffic intersection video image analysis method and system for power equipment | |
CN113298817A (en) | High-accuracy semantic segmentation method for remote sensing image | |
CN113569801B (en) | Distribution construction site live equipment and live area identification method and device thereof | |
CN110378892A (en) | A kind of method of quick detection electric transmission line channel hidden danger | |
CN112232235B (en) | Intelligent factory remote monitoring method and system based on 5G sum algorithm | |
CN106022977A (en) | Transformer substation job site safety analysis method | |
CN113536944A (en) | Distribution line inspection data identification and analysis method based on image identification | |
CN108549882A (en) | A kind of real-time railway mouth multi-cam pedestrian detection method based on deep learning | |
CN113328986A (en) | Network flow abnormity detection method based on combination of convolutional neural network and LSTM | |
Yang et al. | Abnormal Object Detection with an Improved YOLOv8 in the Transmission Lines | |
CN110211323A (en) | Forest fire recognition methods based on cascade sort | |
CN116152699B (en) | Real-time moving target detection method for hydropower plant video monitoring system | |
Weifu et al. | Research on the application of target detection based on deep learning technology in power grid operation inspection |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130306 Termination date: 20171216 |
|
CF01 | Termination of patent right due to non-payment of annual fee |