CN109145905A - A kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness - Google Patents
A kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness Download PDFInfo
- Publication number
- CN109145905A CN109145905A CN201811000906.8A CN201811000906A CN109145905A CN 109145905 A CN109145905 A CN 109145905A CN 201811000906 A CN201811000906 A CN 201811000906A CN 109145905 A CN109145905 A CN 109145905A
- Authority
- CN
- China
- Prior art keywords
- region
- pixel
- straightway
- transmission line
- area
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- 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
- G06V10/255—Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/176—Urban or other man-made structures
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of transmission line of electricity accessory detection methods of view-based access control model conspicuousness, video image information acquisition is carried out to transmission line of electricity to be detected using the unmanned plane for carrying binocular vision picture pick-up device, unmanned plane during flying is controlled by the staff of profession, and unmanned plane during flying is oriented parallel to electric force lines distribution direction.The video image information that will acquire carries out taking frame, entire video is considered as a frame sequence, and image preprocessing, region of interesting extraction, accessory defects detection, result output and feedback successively are carried out to all picture frames, complete the record and feedback operation of the defects of transmission line of electricity accessory detection process accessory information.The transmission line of electricity accessory detection method of a kind of view-based access control model conspicuousness provided by the invention, with real-time is good, accuracy is high, the advantage of strong antijamming capability.
Description
Technical field
The present invention relates to a kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness, belong to computer vision with it is defeated
Electric line inspection technical field.
Background technique
Currently, the continuous development with national overall national strength is promoted, science and technology has obtained development at full speed, city process
Change is also gradually being accelerated, also more and more for the demand of resource, and social development is so that the construction needs of transmission line of electricity are higher
Requirement.Transmission line of electricity is indispensable pith in power network.Its main function is conveying, distribution and exchange electricity
Energy.Meanwhile it is exactly to connect several independent power grids there are also another prior effect, formed interconnected network or
Integrated power system, to improve the advisability of power system security power supply.
Transmission line of electricity can be divided into two class of overhead transmission line and cable run from structure.High-voltage power line, that is, overhead transmission line is used
Wire erection is power network and electric system in the power line on shaft tower by the transmission lines of electricity such as insulator and electric armour clamp accessory
Important component is highly prone to external influence and damage.It is run since transmission line of electricity is in for a long time under outdoor, so that line
Road accessory will not only bear normal mechanical, electric power load, and it is various severe to also suffer wind and frost sleet, thunder and lightning and atmosphere pollution etc.
Natural conditions influence, these influences can all jeopardize the safe operation of transmission line of electricity.For guarantee transmission line of electricity safe operation,
It is necessary to the electric power accessories such as insulator, stockbridge damper to transmission line of electricity to detect.
Traditional transmission line of electricity accessory detection is usually artificial inspection in place, and not only human resources consumption is big for this mode,
And risk is high.Under the big area coverage of power line and the demand of diversified environment, manually power line fittings is detected in place
Low efficiency, real-time is poor, tends not to the covering surface for meeting electric power line inspection and instantaneity requirement.In addition, traditional is artificial defeated
Electric line accessory detection method judges often by eye-observation according to state of the experience of staff to accessory,
Experience and state to staff excessively rely on, and can not detect automatically to transmission line of electricity accessory defect and early warning, easily occur
Erroneous detection and detection leakage phenomenon are not able to satisfy the accuracy requirement of electric power line inspection.
Summary of the invention
Purpose: in order to overcome the technology vacancy in transmission line of electricity accessory detection field in the prior art, the present invention provides one
The transmission line of electricity accessory detection method of kind view-based access control model conspicuousness, it is intended to improve the mode of transmission line of electricity accessory detection, raising is patrolled
The real-time and accuracy rate of inspection.
Technical solution: in order to solve the above technical problems, the technical solution adopted by the present invention are as follows:
A kind of transmission line of electricity accessory detection system of view-based access control model conspicuousness, comprising: UAV Video image information collecting
Module, image pre-processing module, region of interesting extraction module, accessory defects detection module, result export feedback module；
The UAV Video image information collecting module is obtained defeated using the unmanned plane for carrying binocular vision picture pick-up device
Electric line distributed intelligence；
Described image preprocessing module locates frame image for carrying out taking frame to the video image that unmanned plane acquires in advance
Science and engineering is made, including image denoising and image gradient extract；
The region of interesting extraction module is based on straightway growth algorithm and algorithm of convex hull, by defeated in detection frame image
Electric line obtains area-of-interest；
The accessory defects detection module is emerging to sense using quaternary number phase spectral analysis and the significant confidence calculations of super-pixel
Interesting region is handled to obtain notable figure, obtains the salient region of binaryzation；
The result output feedback module is used for feeding back to transmission line of electricity defect accessory information, using k-means
Algorithm is to salient region area given threshold, when salient region area is greater than threshold value, feeds back at computer software interface
The specifying information of accessory, recording means detection time, geographical location are significant to this when salient region area is less than threshold value
Property region is not processed.
Preferably, the binocular vision picture pick-up device includes that identical two video image acquisitions of specifications parameter are set
It is standby, work is arranged in the form of the purpose of left and right respectively, video image information is acquired simultaneously with fixed viewpoint；Unmanned plane carries out power line
When inspection, straight line is carried out above power line with the route parallel with power line and is flied at a constant speed, acquired in sequence of video images
Electric force lines distribution direction is parallel with unmanned plane during flying direction.
A kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness, includes the following steps:
Step 1: transmission line of electricity video data being backed up in data field, and inputs shooting time and place；By video data
Video is carried out according to certain time interval to take frame, forms a frame sequence；
Step 2: frame image being pre-processed, including image denoising and image gradient extract；Use Sobel operator extraction
Gradient calculates gradient direction vector；
Step 3: based on straightway growth algorithm, passing through pixel growth method, line segment region-growing method and algorithm of convex hull
For frame image zooming-out area-of-interest；
Step 4: area-of-interest being handled using quaternary number phase spectral analysis and super-pixel significant confidence calculations
Notable figure is obtained, the salient region of binaryzation is calculated；
Step 5: using k-means algorithm to salient region area given threshold, when salient region area is greater than threshold
When value, in the specifying information of computer interface feedback defect accessory, defect accessory detection time, geographical location are recorded, when significant
Property region area be less than threshold value when, which is not processed.
Preferably, the video data carries out transmission line of electricity using UAV flight's binocular vision picture pick-up device
It takes photo by plane and acquires relevant information, collected video is stored in the storage equipment carried to unmanned plane, and utilize mobile radio network
Network is transmitted to backstage.
Preferably, the binocular vision picture pick-up device pixel be not less than 500W, unmanned plane fly at a constant speed direction with
Transmission line of electricity distribution arrangement is parallel, and unmanned plane during flying process is at the uniform velocity stable.
Preferably, the step 3 includes:
3.1: choosing a pixel in a certain region of frame image as seed point, region expansion is carried out using pixel growth method
Open detection；The difference given threshold at direction vector angle and current region direction vector angle to pixel, when the direction of pixel
When vectorial angle and current region direction vector angular difference value are less than threshold value, pixel is included into current region, obtains straightway region；
Wherein, the direction vector of pixel is gradient direction vector, the calculation formula of region direction vectorial angle are as follows:
θ is region direction vectorial angle, angle in formulaiFor the direction vector angle of ith pixel point in region；
Straightway region direction angle is region direction vectorial angle, and straightway regional center position is the center of gravity in the region,
Position of centre of gravity (nx,ny) are as follows:
In formula, P (x, y) is pixel coordinate in straightway region, and (x, y) is pixel direction vector, and S is straightway area
Domain；Based on center of gravity (nx,ny) and region direction vectorial angle θ, diagonal line is square with straightway region direction vector, center of gravity is
Diagonal line midpoint determines a square covering whole region, and square diagonal line is straightway；The image is repeated above-mentioned
Step obtains other straightways, to obtain a plurality of Discrete line segments；
3.2: straight line section is chosen in a plurality of Discrete line segments of acquisition as seed straightway, using line segment region
Growth method carries out region clustering division to Discrete line segments, judges whether Discrete line segments belong to a branch of straight line；It collects all
The location information of Discrete line segments two-end-point calculates the angular deviation S of seed straightway and other straightwaysθWith position deviation Sd,
It and is SθAnd SdThreshold value is set, S is worked asθAnd SdWhen less than threshold value, the straight line that current straightway is included into where seed straightway；
Wherein, SθAnd SdCalculation formula it is as follows:
Sθ=| θl1-θl2|
Sd=min (P11-P21|,|P11-P22|,|P12-P21|,|P12-P22|)
In formula, θl1、θl2Represent straight line l1、l2With the angle of horizontal direction, i.e. rectilinear direction；P11、P12And P21、P22Respectively
For l1、l2Extreme coordinates；Line segment region growing is carried out to the image, obtains belonging to collinear straightway aggregation zone；
3.3: similarly with 3.1, direction vector angle and the center of gravity of straightway aggregation zone are calculated separately, with straightway accumulation regions
The direction vector in domain is square diagonal line, and center of gravity is that diagonal line midpoint determines a square covering whole region, square
Diagonal line is final straightway, i.e., transmission line of electricity is distributed；
3.4: establishing two-dimensional coordinate system in frame image by x-axis of final straightway, there is other discrete straight lines in coordinate system
Section takes Discrete line segments endpoint abscissa minimum and maximum two points and the point farthest from x-axis according to algorithm of convex hull, meter respectively
It calculates convex polygon and covers maximum region, as area-of-interest.
Preferably, the step 4 includes:
4.1: the region of interest area image of triple channel is transformed to single pass quaternionic matrix IQ, calculate quaternionic matrix
Quaternary number DCT parameter；
4.2: initial notable figure S is obtained based on quaternary number dct transformDCT(IQ)；
4.3: for the super-pixel p in region of interest area image, defining the Euclidean distance between super-pixel is dEc(p,pi),
piFor any super-pixel of super-pixel p affiliated area；
4.4: being based on piBelong to the same area, calculate the area of super-pixel p affiliated area:
In formula, N is super-pixel number, the tolerance of σ Euclidean distance between super-pixel；
4.5: for each super-pixel p in area-of-interest, its significant confidence level is equal to:
In formula, TsFor the sum of all pixels point conspicuousness in super-pixel p, Areasize is the face of super-pixel p affiliated area
Product, to significant confidence level Wslc(p) it is normalized to obtain normalized significant confidence level
4.6: initial notable figure is optimized according to obtained significant confidence level, obtains the salient region of binaryzation,
Accessory picture after optimizing, majorized function are as follows:
Wherein, λ is the coefficient of balance of foreground and background,For the background confidence level W of super-pixelBgd(p)=1-Wslc
(p), Wi(p-pi) smooth item between super-pixel.
Preferably, the σ takes 10.
The utility model has the advantages that a kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness provided by the invention, utilizes view
Feel that conspicuousness technology detects transmission line of electricity accessory, the main unmanned plane for carrying binocular vision picture pick-up device by analysis exists
Then the video taken during inspection, the computation vision notable figure in the transmission line of electricity area-of-interest detected pass through
K-means algorithm is determined and is fed back to transmission line of electricity defect accessory.This method can in time carry out transmission line of electricity accessory
Automatic detection improves working efficiency to reduce working strength.View-based access control model conspicuousness technology transmission line of electricity accessory detection with
Traditional artificial detection is compared, with real-time is good, accuracy is high, the advantage of strong antijamming capability.
Detailed description of the invention
Fig. 1 is detection system function structure chart of the present invention；
Fig. 2 is that present system runs topology diagram；
Fig. 3 is detection method flow diagram；
Fig. 4 is area growth process figure.
Specific embodiment
The present invention will be further explained with reference to the accompanying drawing.
A kind of transmission line of electricity accessory detection system of view-based access control model conspicuousness, as shown in Figure 1, comprising: UAV Video figure
As information acquisition module, image pre-processing module, region of interesting extraction module, accessory defects detection module, result output are anti-
Present module.
The UAV Video image information collecting module is obtained defeated using the unmanned plane for carrying binocular vision picture pick-up device
Electric line distributed intelligence, wherein binocular vision picture pick-up device refers to the identical two video image acquisition equipment of specifications parameter, point
Work is arranged not in the form of the purpose of left and right, and video image information is acquired simultaneously with fixed viewpoint.Unmanned plane carries out electric power line inspection
When, wireless remote control is carried out to unmanned plane by professional staff, is carried out above power line with the route parallel with power line straight
Line flies at a constant speed, and the electric force lines distribution direction acquired in sequence of video images is parallel with unmanned plane during flying direction.
Described image preprocessing module carries out current frame image for carrying out taking frame to the video image that unmanned plane acquires
Pretreatment work, including image denoising and image gradient extract.Wherein, video sequence is taken into frame according to certain time interval,
A frame image sequence is obtained, each image is pre-processed.
The region of interesting extraction module is based on straightway growth algorithm and algorithm of convex hull, by defeated in detection frame image
Electric line obtains area-of-interest.
The accessory defects detection module is emerging to sense using quaternary number phase spectral analysis and the significant confidence calculations of super-pixel
Interesting region is handled to obtain notable figure, obtains the salient region of binaryzation.
The result output feedback module is used for feeding back to transmission line of electricity defect accessory information, using k-means
Algorithm is to salient region area given threshold, when salient region area is greater than threshold value, feeds back at computer software interface
The specifying information of accessory, recording means detection time, geographical location are significant to this when salient region area is less than threshold value
Property region is not processed.The above processing successively is carried out to the frame image of all acquirements, completes the defects of inspection process accessory letter
The record of breath.
Present system topological structure, as shown in Fig. 2, using the unmanned plane of binocular vision picture pick-up device is carried to be detected
Transmission line of electricity carries out video image information acquisition, and unmanned plane during flying is controlled by the staff of profession, and unmanned plane during flying direction is flat
Row is in electric force lines distribution direction.The video image information that will acquire carries out taking frame, and entire video is considered as a frame sequence, and according to
It is secondary that image preprocessing, region of interesting extraction, accessory defects detection, result output and feedback are carried out to all picture frames, it completes
The record and feedback operation of the defects of transmission line of electricity accessory detection process accessory information.
A kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness, as shown in figure 3, including the following steps:
Step 1: when the transmission line of electricity accessory in a certain area is detected, being taken the photograph using UAV flight's binocular vision
As equipment take photo by plane to transmission line of electricity acquiring relevant information, wherein binocular vision picture pick-up device pixel is not less than 500W.Nobody
Machine is manipulated by the staff of profession, and the unmanned plane direction that flies at a constant speed is parallel with transmission line of electricity distribution arrangement, unmanned plane during flying mistake
Journey is at the uniform velocity stable, and binocular vision picture pick-up device need to be mounted on one stabilised platform of unmanned plane, and collected video is stored
In the storage equipment carried to unmanned plane, and backstage is transmitted to using mobile wireless network.
Step 2: the video data that will acquire from the background is backed up in data field, and inputs shooting time and ground by staff
Point；Video data is sent to image pre-processing module again and handled by backstage, is carried out according to certain time interval to video
Frame is taken, entire video is considered as a frame sequence, it is continuous between frame image, but do not include identical content.
Step 3: frame image being pre-processed, including image denoising and image gradient extract.Wherein, image denoising reduces
The noise of frame image promotes picture quality；Sobel operator extraction gradient is used for frame image, calculates gradient direction vector.
Step 4: based on straightway growth algorithm, passing through pixel growth method, line segment region-growing method and algorithm of convex hull
For frame image zooming-out area-of-interest, concrete operations are as follows:
4.1: choosing a pixel in a certain region of frame image as seed point, region expansion is carried out using pixel growth method
Open detection；Concrete operations are to work as picture to the direction vector angle of pixel and the difference given threshold at current region direction vector angle
When the direction vector angle of vegetarian refreshments and current region direction vector angular difference value are less than threshold value, pixel is included into current region, is obtained
Straightway region, area growth process are as shown in Figure 4, wherein the direction vector of pixel is gradient direction vector, region direction
The calculation formula of vectorial angle are as follows:
θ is region direction vectorial angle, angle in formulaiFor the direction vector angle of ith pixel point in region；
Straightway region direction angle is region direction vectorial angle, and straightway regional center position is the center of gravity in the region,
Position of centre of gravity (nx,ny) are as follows:
In formula, P (x, y) is pixel coordinate in straightway region, and (x, y) is pixel direction vector, and S is straightway area
Domain；Based on center of gravity (nx,ny) and region direction vectorial angle θ, diagonal line is square with straightway region direction vector, center of gravity is
Diagonal line midpoint determines a square covering whole region, and square diagonal line is straightway；The image is repeated above-mentioned
Step obtains other straightways, to obtain a plurality of Discrete line segments；
4.2: straight line section is chosen in a plurality of Discrete line segments of acquisition as seed straightway, using line segment region
Growth method carries out region clustering division to Discrete line segments, judges whether Discrete line segments belong to a branch of straight line.Concrete operations
For the location information for collecting all Discrete line segments two-end-points, the angular deviation S of seed straightway and other straightways is calculatedθWith
Position deviation Sd, and be SθAnd SdThreshold value is set, S is worked asθAnd SdWhen less than threshold value, current straightway is included into where seed straightway
Straight line.Wherein, SθAnd SdCalculation formula it is as follows:
Sθ=| θl1-θl2|
Sd=min (| P11-P21|,|P11-P22|,|P12-P21|,|P12-P22|)
In formula, θl1、θl2Represent straight line l1、l2With the angle of horizontal direction, i.e. rectilinear direction；P11、P12And P21、P22Respectively
For l1、l2Extreme coordinates；Line segment region growing is carried out to the image, obtains belonging to collinear straightway aggregation zone.
4.3: similarly with 4.1, direction vector angle and the center of gravity of straightway aggregation zone are calculated separately, with straightway accumulation regions
The direction vector in domain is square diagonal line, and center of gravity is that diagonal line midpoint determines a square covering whole region, square
Diagonal line is final straightway, i.e., transmission line of electricity is distributed.
4.4: establishing two-dimensional coordinate system in frame image by x-axis of final straightway, there is other discrete straight lines in coordinate system
Section takes Discrete line segments endpoint abscissa minimum and maximum two points and the point farthest from x-axis according to algorithm of convex hull, meter respectively
It calculates convex polygon and covers maximum region, as area-of-interest, also include accessory side since the region not only includes accessory
Power line, the environment of edge, therefore step 5 processing need to be carried out.
Step 5: area-of-interest being handled using quaternary number phase spectral analysis and super-pixel significant confidence calculations
Notable figure is obtained, the salient region of binaryzation is obtained.
5.1: the region of interest area image of triple channel is transformed to single pass quaternionic matrix IQ, calculate quaternionic matrix
Quaternary number DCT parameter；
5.2: initial notable figure S is obtained based on quaternary number dct transformDCT(IQ)；
5.3: for the super-pixel p in region of interest area image, defining the Euclidean distance between super-pixel is dEc(p,pi),
piFor any super-pixel of super-pixel p affiliated area；
5.4: being based on piBelong to the same area, calculate the area of super-pixel p affiliated area:
In formula, N is super-pixel number, and the tolerance of σ Euclidean distance between super-pixel is taken as 10；
5.5: for each super-pixel p in area-of-interest, its significant confidence level is equal to:
In formula, TsFor the sum of all pixels point conspicuousness in super-pixel p, Areasize is the face of super-pixel p affiliated area
Product, to significant confidence level Wslc(p) it is normalized to obtain normalized significant confidence level
5.6: initial notable figure is optimized according to obtained significant confidence level, obtains the salient region of binaryzation,
Accessory picture after optimizing, majorized function are as follows:
Wherein, λ is the coefficient of balance of foreground and background,For the background confidence level W of super-pixelBgd(p)=1-Wslc
(p), Wi(p-pi) smooth item between super-pixel, for smooth adjacent super-pixel.
Step 6: using k-means algorithm to salient region area given threshold, when salient region area is greater than threshold
When value, in the specifying information of computer interface feedback defect accessory, defect accessory detection time, geographical location are recorded, when significant
Property region area be less than threshold value when, which is not processed.The above processing successively is carried out to the frame image of all acquirements, is completed
The record of the defects of inspection process accessory information.
The above is only a preferred embodiment of the present invention, it should be pointed out that: for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (8)
1. a kind of transmission line of electricity accessory detection system of view-based access control model conspicuousness, it is characterised in that: include: UAV Video image
Information acquisition module, image pre-processing module, region of interesting extraction module, accessory defects detection module, result output feedback
Module；
The UAV Video image information collecting module obtains power transmission line using the unmanned plane for carrying binocular vision picture pick-up device
Road distributed intelligence；
Described image preprocessing module carries out pretreatment work to frame image for carrying out taking frame to the video image that unmanned plane acquires
Make, including image denoising and image gradient extract；
The region of interesting extraction module is based on straightway growth algorithm and algorithm of convex hull, passes through power transmission line in detection frame image
Road obtains area-of-interest；
The accessory defects detection module is using quaternary number phase spectral analysis and the significant confidence calculations of super-pixel to region of interest
Domain is handled to obtain notable figure, obtains the salient region of binaryzation；
The result output feedback module is used for feeding back to transmission line of electricity defect accessory information, using k-means algorithm
To salient region area given threshold, when salient region area is greater than threshold value, accessory is fed back at computer software interface
Specifying information, recording means detection time, geographical location, when salient region area be less than threshold value when, to the conspicuousness area
Domain is not processed.
2. a kind of transmission line of electricity accessory detection system of view-based access control model conspicuousness according to claim 1, it is characterised in that:
The binocular vision picture pick-up device includes the identical two video image acquisition equipment of specifications parameter, respectively in the form of the purpose of left and right
Work is arranged, video image information is acquired simultaneously with fixed viewpoint；When unmanned plane carries out electric power line inspection, with parallel with power line
Route carry out straight line above power line and fly at a constant speed, acquire the electric force lines distribution direction in sequence of video images and unmanned plane
Heading is parallel.
3. a kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness, characterized by the following steps:
Step 1: transmission line of electricity video data being backed up in data field, and inputs shooting time and place；By video data according to
Certain time interval carries out video to take frame, forms a frame sequence；
Step 2: frame image being pre-processed, including image denoising and image gradient extract；Use Sobel operator extraction ladder
Degree calculates gradient direction vector；
Step 3: being frame by pixel growth method, line segment region-growing method and algorithm of convex hull based on straightway growth algorithm
Image zooming-out area-of-interest；
Step 4: area-of-interest being handled to obtain using quaternary number phase spectral analysis and super-pixel significant confidence calculations
The salient region of binaryzation is calculated in notable figure；
Step 5: using k-means algorithm to salient region area given threshold, when salient region area is greater than threshold value,
In the specifying information of computer interface feedback defect accessory, defect accessory detection time, geographical location are recorded, salient region is worked as
When area is less than threshold value, which is not processed.
4. a kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness according to claim 3, it is characterised in that:
The video data take photo by plane to transmission line of electricity acquiring relevant information using UAV flight's binocular vision picture pick-up device, will adopt
The video collected stores in the storage equipment carried to unmanned plane, and is transmitted to backstage using mobile wireless network.
5. a kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness according to claim 4, it is characterised in that:
The binocular vision picture pick-up device pixel is not less than 500W, and the unmanned plane direction that flies at a constant speed is parallel with transmission line of electricity distribution arrangement,
Unmanned plane during flying process is at the uniform velocity stable.
6. a kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness according to claim 3, it is characterised in that:
The step 3 includes:
3.1: choosing a pixel in a certain region of frame image as seed point, zone broadening inspection is carried out using pixel growth method
It surveys；The difference given threshold at direction vector angle and current region direction vector angle to pixel, when the direction vector of pixel
When angle and current region direction vector angular difference value are less than threshold value, pixel is included into current region, obtains straightway region；
Wherein, the direction vector of pixel is gradient direction vector, the calculation formula of region direction vectorial angle are as follows:
θ is region direction vectorial angle, angle in formulaiFor the direction vector angle of ith pixel point in region；
Straightway region direction angle is region direction vectorial angle, and straightway regional center position is the center of gravity in the region, center of gravity
Position (nx,ny) are as follows:
In formula, P (x, y) is pixel coordinate in straightway region, and (x, y) is pixel direction vector, and S is straightway region；
Based on center of gravity (nx,ny) and region direction vectorial angle θ, diagonal line is square with straightway region direction vector, center of gravity is diagonal
Line midpoint determines a square covering whole region, and square diagonal line is straightway；It repeats the above steps to the image
Other straightways are obtained, to obtain a plurality of Discrete line segments；
3.2: straight line section is chosen in a plurality of Discrete line segments of acquisition as seed straightway, using line segment region growing
Method carries out region clustering division to Discrete line segments, judges whether Discrete line segments belong to a branch of straight line；It collects all discrete
The location information of straightway two-end-point calculates the angular deviation S of seed straightway and other straightwaysθWith position deviation Sd, and be
SθAnd SdThreshold value is set, S is worked asθAnd SdWhen less than threshold value, the straight line that current straightway is included into where seed straightway；
Wherein, SθAnd SdCalculation formula it is as follows:
Sθ=| θl1-θl2|
Sd=min (| P11-P21|,|P11-P22|,|P12-P21|,|P12-P22|)
In formula, θl1、θl2Represent straight line l1、l2With the angle of horizontal direction, i.e. rectilinear direction；P11、P12And P21、P22Respectively l1、
l2Extreme coordinates；Line segment region growing is carried out to the image, obtains belonging to collinear straightway aggregation zone；
3.3: similarly with 3.1, direction vector angle and the center of gravity of straightway aggregation zone are calculated separately, with straightway aggregation zone
Direction vector is square diagonal line, and center of gravity is that diagonal line midpoint determines a square covering whole region, and square is diagonal
Line is final straightway, i.e., transmission line of electricity is distributed；
3.4: two-dimensional coordinate system is established in frame image by x-axis of final straightway, there are other Discrete line segments in coordinate system, point
It does not take Discrete line segments endpoint abscissa minimum and maximum two points and the point farthest from x-axis according to algorithm of convex hull, calculates
Convex polygon covers maximum region, as area-of-interest.
7. a kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness according to claim 3, it is characterised in that:
The step 4 includes:
4.1: the region of interest area image of triple channel is transformed to single pass quaternionic matrix IQ, calculate the four of quaternionic matrix
First number DCT parameter；
4.2: initial notable figure S is obtained based on quaternary number dct transformDCT(IQ)；
4.3: for the super-pixel p in region of interest area image, defining the Euclidean distance between super-pixel is dEc(p,pi), piFor
Any super-pixel of super-pixel p affiliated area；
4.4: being based on piBelong to the same area, calculate the area of super-pixel p affiliated area:
In formula, N is super-pixel number, the tolerance of σ Euclidean distance between super-pixel；
4.5: for each super-pixel p in area-of-interest, its significant confidence level is equal to:
In formula, TsFor the sum of all pixels point conspicuousness in super-pixel p, Areasize is the area of super-pixel p affiliated area, right
Significant confidence level Wslc(p) it is normalized to obtain normalized significant confidence level Wi Fg(p)；
4.6: initial notable figure is optimized according to obtained significant confidence level, obtains the salient region of binaryzation, i.e., it is excellent
Accessory picture after change, majorized function are as follows:
Wherein, λ is the coefficient of balance of foreground and background, Wi BgdFor the background confidence level W of super-pixelBgd(p)=1-Wslc(p), Wi
(p-pi) smooth item between super-pixel.
8. a kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness according to claim 7, it is characterised in that:
The σ takes 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811000906.8A CN109145905A (en) | 2018-08-29 | 2018-08-29 | A kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811000906.8A CN109145905A (en) | 2018-08-29 | 2018-08-29 | A kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109145905A true CN109145905A (en) | 2019-01-04 |
Family
ID=64829235
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811000906.8A Pending CN109145905A (en) | 2018-08-29 | 2018-08-29 | A kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109145905A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110245701A (en) * | 2019-06-11 | 2019-09-17 | 云南电网有限责任公司曲靖供电局 | A kind of electric power line detecting method based on unmanned plane image |
| CN110413003A (en) * | 2019-07-31 | 2019-11-05 | 广东电网有限责任公司 | Inspection method, device, equipment and the computer readable storage medium of transmission line of electricity |
| CN112985263A (en) * | 2021-02-09 | 2021-06-18 | 中国科学院上海微系统与信息技术研究所 | Method, device and equipment for detecting geometrical parameters of bow net |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105957077A (en) * | 2015-04-29 | 2016-09-21 | 国网河南省电力公司电力科学研究院 | Detection method for foreign body in transmission lines based on visual saliency analysis |
| CN106356757A (en) * | 2016-08-11 | 2017-01-25 | 河海大学常州校区 | Method for inspecting electric power lines by aid of unmanned aerial vehicle on basis of human vision characteristics |
| CN108317953A (en) * | 2018-01-19 | 2018-07-24 | 东北电力大学 | A kind of binocular vision target surface 3D detection methods and system based on unmanned plane |
-
2018
- 2018-08-29 CN CN201811000906.8A patent/CN109145905A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105957077A (en) * | 2015-04-29 | 2016-09-21 | 国网河南省电力公司电力科学研究院 | Detection method for foreign body in transmission lines based on visual saliency analysis |
| CN106356757A (en) * | 2016-08-11 | 2017-01-25 | 河海大学常州校区 | Method for inspecting electric power lines by aid of unmanned aerial vehicle on basis of human vision characteristics |
| CN108317953A (en) * | 2018-01-19 | 2018-07-24 | 东北电力大学 | A kind of binocular vision target surface 3D detection methods and system based on unmanned plane |
Non-Patent Citations (1)
| Title |
|---|
| 万迪明等: "一种基于视觉显著性分析的输电线路异物检测方法", 《监管与检测》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110245701A (en) * | 2019-06-11 | 2019-09-17 | 云南电网有限责任公司曲靖供电局 | A kind of electric power line detecting method based on unmanned plane image |
| CN110413003A (en) * | 2019-07-31 | 2019-11-05 | 广东电网有限责任公司 | Inspection method, device, equipment and the computer readable storage medium of transmission line of electricity |
| CN112985263A (en) * | 2021-02-09 | 2021-06-18 | 中国科学院上海微系统与信息技术研究所 | Method, device and equipment for detecting geometrical parameters of bow net |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110554704B (en) | Unmanned aerial vehicle-based fan blade autonomous inspection method | |
| CN110297498B (en) | Track inspection method and system based on wireless charging unmanned aerial vehicle | |
| CN110879601B (en) | Unmanned aerial vehicle inspection method for unknown fan structure | |
| CN108109437B (en) | Unmanned aerial vehicle autonomous route extraction and generation method based on map features | |
| CN106203265B (en) | A kind of Construction Fugitive Dust Pollution source monitors automatically and coverage forecasting system and method | |
| CN104865971B (en) | The control method and unmanned plane of a kind of polling transmission line unmanned plane | |
| WO2021115124A1 (en) | Edge-cloud coordinated three-dimensional reconstruction method for farmland site | |
| CN106547814A (en) | A kind of power transmission line unmanned machine patrols and examines the structuring automatic archiving method of image | |
| WO2020221284A1 (en) | Unmanned aerial vehicle monitoring method and system for basin-wide flood scene | |
| CN105023014B (en) | A kind of shaft tower target extraction method in unmanned plane inspection transmission line of electricity image | |
| CN106708073B (en) | A kind of quadrotor system of independent navigation power-line patrolling fault detection | |
| CN109739254B (en) | Unmanned aerial vehicle adopting visual image positioning in power inspection and positioning method thereof | |
| CN110989658B (en) | High-voltage transmission line crossing inclined photographic point cloud acquisition method | |
| Bian et al. | A novel monocular-based navigation approach for UAV autonomous transmission-line inspection | |
| CN109145905A (en) | A kind of transmission line of electricity accessory detection method of view-based access control model conspicuousness | |
| CN109002048B (en) | Multi-rotor unmanned aerial vehicle large-scale centralized photovoltaic power station image data acquisition method | |
| US11021246B2 (en) | Method and system for capturing images of asset using unmanned aerial vehicles | |
| CN111244822B (en) | Fixed-wing unmanned aerial vehicle line patrol method, system and device in complex geographic environment | |
| CN109389056B (en) | Space-based multi-view-angle collaborative track surrounding environment detection method | |
| CN115275870B (en) | Inspection system based on high-altitude line maintenance | |
| CN107221006A (en) | A kind of communication single pipe tower slant detection method based on unmanned plane imaging platform | |
| CN114038193B (en) | Intelligent traffic flow data statistics method and system based on unmanned aerial vehicle and multi-target tracking | |
| CN112947526B (en) | Unmanned aerial vehicle autonomous landing method and system | |
| Xi et al. | A vision-based inspection strategy for large-scale photovoltaic farms using an autonomous UAV | |
| CN107741233A (en) | A kind of construction method of the outdoor map of three-dimensional |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190104 |
|
| RJ01 | Rejection of invention patent application after publication |