CN106705866B - A kind of powerline ice-covering detection method based on visible images - Google Patents
A kind of powerline ice-covering detection method based on visible images Download PDFInfo
- Publication number
- CN106705866B CN106705866B CN201611151446.XA CN201611151446A CN106705866B CN 106705866 B CN106705866 B CN 106705866B CN 201611151446 A CN201611151446 A CN 201611151446A CN 106705866 B CN106705866 B CN 106705866B
- Authority
- CN
- China
- Prior art keywords
- line segment
- conducting wire
- icing
- line
- image
- 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
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/02—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
- G01B11/06—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness for measuring thickness ; e.g. of sheet material
- G01B11/0616—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness for measuring thickness ; e.g. of sheet material of coating
- G01B11/0683—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness for measuring thickness ; e.g. of sheet material of coating measurement during deposition or removal of the layer
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
Abstract
The powerline ice-covering detection method based on visible images that the present invention relates to a kind of, including edge detection is carried out to original image by binaryzation and obtains bianry image；Edge detection is carried out to bianry image using edge detection operator, obtains its binary edge map；Piecemeal processing is carried out to binary edge map, obtains the shaft tower piecemeal for having shaft tower image；Line segment is carried out to shaft tower piecemeal according to approximation, continuity and the synteny between line segment and merges acquisition conducting wire；According to the difference between the lower edges pixel of conducting wire in binary edge map, ice covering thickness is calculated.The present invention is directed to the visible images of transmission line of electricity, the method for proposing to be carried out in five steps ice covering thickness detection, after the visible images for obtaining transmission line of electricity, just wire icing of transmission line thickness can be quickly obtained according to formula, for the status for identifying conducting wire difficulty in visible images, the present invention proposes first to identify the shaft tower in visible images, then the method according to shaft tower position identification conducting wire.
Description
Technical field
The present invention relates to transmission line faultlocating technical fields more particularly to a kind of transmission line of electricity based on visible images to cover
Ice detection method.
Background technique
In transmission line faultlocating technical field, it is commonly encountered powerline ice-covering phenomenon, so-called powerline ice-covering phenomenon is
Refer to after bad weather generation, icy water drop condenses on electric wire in cloud or in mist, wire icing is formed, if a model
Institute's live wire in enclosing all is encased by ice, and here it is transmission line icings.Transmission line icing makes thin electric wire become thick ice lolly, for
For the high-tension bus-bar of long distance powedr transmission, the steel tower for supporting high-voltage line is made to increase weight bearing, serious transmission line icing make steel tower without
Power supports these electric wires and collapses.It can only operate a switch if having icing on the insulator chain of steel tower, stop transmission line of electricity defeated
Electricity then causes the power breakdown of large area.China is as one of powerline ice-covering Chong Zai state in the world, powerline ice-covering
Accident frequently occurs, to avoid powerline ice-covering disaster that power grid is caused to be paralysed, when needing to reache a certain level for ice cover
Effective early warning is obtained, and corresponding ice-melt measure is taken in time to powerline ice-covering phenomenon, so, scene is carried out to icing phenomenon
Exploration and information collection, especially carrying out detection to electric power line ice-covering thickness is an extremely important problem.
There are mainly two types of electric power line ice-covering thickness detection methods common at present: mechanical analysis and video monitoring image point
Analysis.Wherein, the pulling force that mechanical analyzing method mainly passes through insulator reflects that the weight of icing and icing add to conducting wire
Tension sensor or stress section sensor are specifically sealed in the electricity that insulator chain or insulator chain are vertically connected with by load
In power fitting, the information such as detection wire icing Gravity changer, windage yaw angle, insulator tilt angle and wind speed, wind direction are established
Mathematical model calculates the ice covering thickness of current line.Video monitoring image analysis method is adopted by installing image on shaft tower
Collect terminal, by being handled to obtain ice covering thickness to the image that terminal is passed back.
However, by mechanical analyzing method carry out ice covering thickness detection when, since there are many factor of icing, various factors it
Between correlation it is complicated, the relationship between insulator tension and transmission line icing distribution, icing pattern etc. is failed at present accurate true
It is fixed, and tension sensor used belongs to nonstandard electric armour clamp, it is inaccurate so as to cause mechanical analysis result.Pass through video
When monitoring image analysis method carries out video monitoring, under the severe situation of weather condition, video monitoring camera is also possible to
There is icing, cause institute's acquired image not clear enough, to influence the accuracy of ice covering thickness calculating.Therefore, it is badly in need of one
The method that kind can accurately calculate electric power line ice-covering thickness.
Summary of the invention
To overcome the problems in correlation technique, the present invention provides a kind of powerline ice-covering based on visible images
Detection method.
The embodiment of the present invention provides a kind of powerline ice-covering detection method based on visible images, comprising:
Edge detection is carried out to original image by binaryzation and obtains bianry image；
Edge detection is carried out to the bianry image from 8 different directions respectively using edge detection operator, extracts institute
The edge feature for stating bianry image obtains its binary edge map；
Piecemeal processing is carried out to the binary edge map, obtains the shaft tower piecemeal for having shaft tower image；
Line segment merging is carried out to the shaft tower piecemeal according to approximation, continuity and the synteny between line segment, is obtained
Conducting wire；
According to the difference between the lower edges pixel of conducting wire in the binary edge map, icing thickness is calculated
Degree.
Preferably, it according to the difference between the lower edges pixel of conducting wire in the binary edge map, is calculated
After ice covering thickness, further includes: according to the inclination angle of the conducting wire before the inclination angle of conducting wire described after icing and icing, to the icing
Thickness correction obtains accurate ice covering thickness.
It is preferably, described that edge detection acquisition bianry image is carried out to original image by binaryzation, comprising:
Gaussian smoothing is carried out to the original image and obtains Gaussian kernel, the original image is carried out using the Gaussian kernel
Convolution, the image after obtaining Gaussian smoothing；
It is differentiated using gradient operator to the image after the Gaussian smoothing, obtains gradient image；
Non-maxima suppression is carried out to the gradient value of the gradient image, obtains pixel local maximum and pixel office
The non-maximum in portion；
By the pixel locally corresponding gray value zero setting of non-maximum, the bianry image is obtained.
Preferably, the gradient operator is the Prewitt operator or Sobel Operator in 8 directions.
Preferably, piecemeal processing is carried out to the binary edge map and obtains the shaft tower piecemeal with shaft tower image, comprising:
Original line segments in the binary edge map are carried out with the Blob analysis of connectivity structure, treated for acquisition finally
Line segment, and the final line segment quantity in each direction is less than or equal to 128；
Piecemeal processing is carried out to the binary edge map, obtains block image, and horizontal piecemeal in the block image
Number is 16, and vertical block count is 4；
Count respectively horizontal direction in the block image, vertical direction, oblique direction and obliquely downward direction it is described most
Finish line segment number；
If the final line segment quantity of horizontal direction is greater than the final line segment of 3, vertical direction in the piecemeal
Quantity is greater than 3 and the final line segment quantity of oblique direction is more than or equal to 3；Or in the piecemeal horizontal direction institute
It states final line segment quantity and is greater than 3 and the obliquely downward final line segment number in direction greater than the final line segment quantity of 3, vertical direction
Amount is more than or equal to 3, then determines the block memory in shaft tower；
The horizontal direction is -10 °~10 °；The vertical direction is greater than 85 ° or less than -85 °；The oblique upper
To being 10 °~85 °；The direction obliquely downward is -85 °~-10 °.
Preferably, the approximation according between line segment, continuity and synteny carry out line to the shaft tower piecemeal
Section merges, and obtains conducting wire, comprising:
According to the direction of the final line segment and angle, divides six direction to be grouped on the final line segment, obtain tight condition
Under line segment group, it is described it is tight under the conditions of line segment group include: horizontal line section group, upper inclined horizontal line section group, upper hading straightway group,
Inclined horizontal line section group, lower hading straightway group, vertical segment group down；
Each line segment group is scanned for using the method for two-stage iterative processing, obtains best merging line segment；
The method of the two-stage iterative processing includes:
The final line segment for traversing all non-merging treatments in each line segment group reads a line segment A, the line
The length of section A is L1；
It traverses all final line segments in addition to line segment A in each line segment group, reads another line segment S, it is described
The length of line segment S is L2；
Obtain the distance between the line segment A and line segment S L；
If meeting L≤2 × min (L1,L2), then determine that the line segment A and line segment S have approximation；
Obtain the angle theta 1 of the line segment A and horizontal direction, the angle theta 2 and angle threshold of the line segment S and horizontal direction
θthreshold；
If met | θ1-θ2|≤θthreshold, then determine that the line segment A and line segment S have continuity；
Obtain the width line1_width of the line segment A, the width line2_width and threshold value k of the line segment S；
If met distance dist≤k-max (line1_width, line2_width), then the line segment A and line are determined
Section S has synteny；
If the line segment A and line segment S meet approximation, continuity and synteny simultaneously, the distance L is put into level-one
It measures in array；
All final line segments in addition to line segment A, line segment S in each line segment group are traversed, another line segment is read
X；
Judge approximation, continuity and the synteny of the line segment X Yu the line segment A；
If the line segment X and the line segment A meet approximation, continuity and synteny, by the line segment X with it is described
The distance of line segment A is put into the level metric array；
Traversed all final line segments in addition to line segment A, line segment S, line segment X in each line segment group, obtain with
Line segment A meets the line segment of approximation, continuity and synteny, and is put into level metric array；
In the level metric array, according to the line segment of the shortest distance, obtains level-one and most preferably merge line segment；
Merge the line segment A and most preferably merge line segment with level-one, obtains new line segment C；
The final line segment all in each line segment group is traversed, a line segment D is read；
Judge approximation, continuity and the synteny of the line segment D Yu the line segment C；
If the line segment D and the line segment D meet approximation, continuity and synteny, by the line segment D with it is described
The distance of line segment C is put into second level measurement array；
All final line segments in addition to line segment D in each line segment group are traversed；
In second level measurement array, according to the line segment of the shortest distance, best merging line segment E is obtained；
Merge the line segment C with most preferably merges line segment E, obtain it is described tightly under the conditions of merging line segment；
Traverse remaining interior non-merging treatment of each line segment group does not merge final line segment, at the two-stage iteration
The method of reason obtains the best merging line segment for not merging final line segment；
Do not merge final line segment and the best merging line segment for not merging final line segment described in merging, obtains the tight item
Merging line segment under part；
According to it is described it is tight under the conditions of merging line segment, obtain conducting wire.
Preferably, the approximation according between line segment, continuity and synteny carry out line to the shaft tower piecemeal
Section merges, and obtains conducting wire, further includes:
By it is described it is tight under the conditions of the grouping of merging line segment, obtain the line segment group under the conditions of pine, the line segment under the conditions of the pine
Group includes that loose condition lower line segment group and line segment angle of the line segment angle between -75 °~75 ° are being less than -75 ° or or are being greater than
Loose condition lower line segment group between 75 °；
The line segment for traversing all non-merging treatments in the line segment group of the loose condition, using the side of the two-stage iterative processing
Method obtains the merging line segment under the conditions of the pine.
Preferably, the difference between the lower edges pixel according to conducting wire in binary edge map, is calculated
Ice covering thickness, comprising:
Obtain the vertical coordinate p1 (i) of single-row top edge pixel in the binary edge map；
Obtain the vertical coordinate p2 (i) of single-row lower edge pixel described in the binary edge map；
According to the vertical coordinate p1 and vertical coordinate p2, using p (i)=| p1 (i)-p2 (i) |, the list is calculated
The pixel difference of column border pixel；
According to the pixel difference, utilizeThe lower edges average thickness d1 of the conducting wire before acquisition icing
And after icing the conducting wire lower edges average thickness d2；
According to the lower edges average thickness d2 of conducting wire after the lower edges average thickness d1 of conducting wire, icing before the icing
And wire diameter D, it utilizesThe thickness DM between conducting wire lower edges is calculated；
According to the thickness between the conducting wire lower edges, utilizeThe ice covering thickness of conducting wire is calculated.
Preferably, the inclination angle according to the conducting wire before the inclination angle of conducting wire described after icing and icing, to the icing
Thickness correction obtains accurate ice covering thickness, comprising:
According to the measurement thickness e g' of lower edges after icing, the actual (real) thickness eg of lower edges, utilizes eg=after icing
The angle of inclination beta of conducting wire is calculated in eg' × cos β；
According to conducting wire angle of inclination beta after icing1, conducting wire angle of inclination beta before icing2, thickness DM between the conducting wire lower edges and lead
Line outer diameter D utilizesWire icing thickness is calculated.
Preferably, the angle threshold θthresholdIt is 2~12, the threshold value k value is 1~3.
The technical solution that the embodiment of the present invention provides can include the following benefits:
The present invention is directed to the visible images of transmission line of electricity, and the method for proposing to be carried out in five steps ice covering thickness detection obtains
It, just can be according to formula after the visible images of transmission line of electricityQuickly show that the icing of transmission line wire is thick
Degree, it is different in conjunction with the inclination angle before and after wire icing in practice, utilize formulaTo transmission of electricity
The ice covering thickness of route is corrected, and wire icing of transmission line THICKNESS CALCULATION can be made more accurate, to melt to be subsequent
Ice measure provides reliable foundation.For the status that conducting wire identification in visible images is difficult, present invention proposition first identifies visible
Shaft tower in light image, then according to the method for shaft tower position identification conducting wire, can effectively solve the problem that and identify and lead in visible images
The problem of line, to guarantee that this method can accurately calculate the ice covering thickness of transmission line of electricity.
It should be understood that above general description and following detailed description be only it is exemplary and explanatory, not
It can the limitation present invention.
Detailed description of the invention
The drawings herein are incorporated into the specification and forms part of this specification, and shows and meets implementation of the invention
Example, and be used to explain the principle of the present invention together with specification.
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, for those of ordinary skill in the art
Speech, without any creative labor, is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of process of the powerline ice-covering detection method based on visible images provided in an embodiment of the present invention
Schematic diagram；
Fig. 2 is the signal for obtaining the shaft tower piecemeal with shaft tower image in the embodiment of the present invention according to binary edge map
Figure；
Fig. 3 is two-stage iterative processing method flow diagram in the embodiment of the present invention；
Fig. 4 is the approximation schematic diagram in the embodiment of the present invention between two line segments；
Fig. 5 is the synteny schematic diagram in the embodiment of the present invention between two line segments；
Fig. 6 is that ice covering thickness corrects schematic diagram in the embodiment of the present invention.
Symbol indicates:
The distance between the length of L1- line segment A, the length of L2- line segment S, L- line segment A and line segment S, θ 1- line segment A and level
The angle in direction, the angle of θ 2- line segment S and horizontal direction, the upper edge of AD- ice coating wire, the lower edge of BC- ice coating wire, β-are led
After the inclination angle of line, eg'- icing after the measurement thickness of lower edges, eg- icing lower edges actual (real) thickness.
Specific embodiment
Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to
When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment
Described in embodiment do not represent all embodiments consistented with the present invention.On the contrary, they be only with it is such as appended
The example of device and method being described in detail in claims, some aspects of the invention are consistent.
The binaryzation of image is to set the gray value of the pixel on image to 0 or 255, that is, by whole image
It shows and significantly there was only black and white visual effect, by carrying out binary conversion treatment to original image, original graph can be obtained
The bianry image of picture can obtain two-value after carrying out edge detection and edge extracting using edge detection operator to bianry image
The binary edge map of image.
The embodiment of the present invention is directed to the powerline ice-covering visible images obtained by unmanned plane, using based on visible light
The powerline ice-covering detection method of image, realization accurately calculate ice covering thickness.Below in conjunction with attached drawing, to the present embodiment
It is further discussed in detail, Fig. 1 is a kind of powerline ice-covering based on visible images provided in an embodiment of the present invention
The flow diagram of detection method.Specific step is as follows:
S110: edge detection is carried out to original image by binaryzation and obtains bianry image.
Specifically, edge detection can be carried out to original image in the following way:
Firstly, carrying out Gaussian smoothing to original image obtains Gaussian kernel, convolution is carried out to original image using Gaussian kernel, is gone
Image except significant noise, after obtaining Gaussian smoothing.
Secondly, differentiating using gradient operator to the image after Gaussian smoothing, gradient image is obtained.Here gradient operator can
To select the Prewitt operator or Sobel Operator in 8 directions.Prewitt operator namely average calculus of finite differences, since the operator is first
Averaging seeks difference again to acquire gradient, and the average calculating operation in algorithm can be played the role of reducing or eliminating noise, and
And 8 the Prewitt operator in direction detect the image border in 8 directions, can guarantee gradient than more comprehensively acquiring image information
Image is more clear.Sobel Operator is edge detection operator typically based on first derivative, due to including class in the operator
Like the operation of local average, there is smoothing effect to noise, be able to suppress noise, is conducive to obtain more accurately bianry image,
To be conducive to ice covering thickness calculating.
After obtaining gradient image, non-maxima suppression is carried out to the gradient value of gradient image, i.e., searching pixel is local most
The local non-maximum of big value and pixel.
Finally, the pixel locally corresponding gray value zero setting of non-maximum is obtained bianry image.
S120: edge detection is carried out to bianry image from 8 different directions respectively using edge detection operator, extracts two
It is worth the edge feature of image, obtains its binary edge map.
Specifically, edge detection is carried out to bianry image from 8 different directions respectively using Prewitt operator, rejected
The point of non-edge, extracts the edge feature of bianry image, obtains its binary edge map.
S130: piecemeal processing is carried out to binary edge map, obtains the shaft tower piecemeal for having shaft tower image.
After the binary edge map for obtaining transmission line of electricity, in order to determine the position of conducting wire in image, first to the area of shaft tower
Domain is positioned, and is conducive to reduce search range in this way, is able to solve the problem that conducting wire identification is difficult in visible images.Due to
Significant characteristics of the transmission line structure in natural scene are its tower structures, although electric power line pole tower have different purposes,
Its structure is also different, it is contemplated that the difference of inspection shooting angle, electric power line pole tower has common nearly symmetrical friendship in the picture
Structure composition is pitched, so defaulting the part with common nearly symmetrical chiasma structure in binary edge map in the present invention is bar
Tower piecemeal.
It is to obtain the shaft tower piecemeal for having shaft tower image in the embodiment of the present invention according to binary edge map with reference to Fig. 2, Fig. 2
Schematic diagram.Specifically, the step of obtaining the shaft tower piecemeal for having shaft tower image according to binary edge map is as follows:
S1301: the Blob for carrying out connectivity structure to the original line segments in binary edge map is analyzed, and treated for acquisition most
Finish line section, and the final line segment quantity in each direction is less than or equal to 128.
Blot analysis is carried out to the connected region of same pixel in binary edge map, the final line segment that obtains that treated,
Since edge detection operator uses Prewitt operator, final line segment is divided into 8 directions, to be processed final in each direction
Line segment is up to 128.
S1302: carrying out piecemeal processing to binary edge map, obtains block image, and horizontal block count in block image
It is 16, vertical block count is 4.
Since the electric power line pole tower image of unmanned plane inspection shooting runs through the upper and lower of image mostly, and shaft tower presence is engraved
Empty region, horizontal piecemeal is larger, and vertical piecemeal is smaller, so carrying out typical piecemeal to binary edge map in the present embodiment
Mode: horizontal block count is 16, and vertical block count is 4.After obtaining block image, judge in each piecemeal with the presence or absence of line segment
Symmetrical chiasma characteristic judges the shaft tower image in each piecemeal.
S1303: count respectively: horizontal direction in block image, vertical direction, oblique direction and direction obliquely downward it is final
Line segment quantity.
Be -10 °~10 ° according to horizontal direction, vertical direction be greater than 85 ° or less than -85 °, oblique direction be 10 °~
85 °, obliquely downward direction be -85 °~-10 °, count the final line segment quantity of four direction in each block image respectively.
S1304: if the final line segment quantity of horizontal direction is big greater than 3, the final line segment quantity of vertical direction in piecemeal
The final line segment quantity that final line segment quantity in 3 and oblique direction is more than or equal to horizontal direction in 3 or piecemeal is greater than
3, the final line segment quantity of vertical direction is greater than 3 and the final line segment quantity in direction is more than or equal to 3 obliquely downward, then determines piecemeal
Inside there is shaft tower.
According to symmetrical chiasma characteristic, it is determined to have shaft tower piecemeal there are two types of in the case of:
A, in piecemeal the final line segment quantity of horizontal direction be greater than 3, the final line segment quantity of vertical direction be greater than 3 and tiltedly on
The final line segment quantity in direction is more than or equal to 3；
B, the final line segment quantity of horizontal direction is greater than 3, the final line segment quantity of vertical direction and is greater than 3 and obliquely downward in piecemeal
The final line segment quantity in direction is more than or equal to 3.
S140: line segment merging is carried out to shaft tower piecemeal according to approximation, continuity and the synteny between line segment, is obtained
Conducting wire.
The positioning of conducting wire can be regarded as the merging of sub-line section on the same direction, the difference being distributed in the picture according to line segment
Situation is divided to tight condition and loose two kinds of situations of condition to calculate the best search for merging line segment.The merging of tight condition lower line segment be by
It is influenced interrupted straight-line segment by background texture to merge, it is that the line segment that nearly straight line or curve are interrupted closes that loose condition lower line segment, which merges,
And.This mode that final line segment is handled according to different situations, can reduce the interference of the small line segment of background texture, also can reduce
Influence to transmission line of electricity line segment arranged side by side, be conducive to improve image selection accuracy, thus be conducive to it is subsequent accurately calculate it is defeated
The ice covering thickness of electric line.
Specifically, line segment merging is carried out to shaft tower piecemeal according to approximation, continuity and the synteny between line segment, obtained
The step of taking conducting wire is as follows:
S1401: according to the direction of final line segment and angle, dividing six direction to be grouped on final line segment, under the conditions of obtaining tightly
Line segment group, it is tight under the conditions of line segment group include: horizontal line section group, upper inclined horizontal line section group, upper hading straightway group, lower inclined water
Horizontal line section group, lower hading straightway group, vertical segment group.
S1402: scanning for each line segment group using the method for two-stage iterative processing, obtains best merging line segment.
It is two-stage iterative processing method flow diagram in the embodiment of the present invention, the method for two-stage iterative processing with reference to Fig. 3, Fig. 3
Include the following steps:
1) the final line segment for traversing all non-merging treatments in each line segment group, reads a line segment A, the length of line segment A
For L1；
2) all final line segments in addition to line segment A in each line segment group are traversed, another line segment S, the length of line segment S are read
For L2；
3) Fig. 4 is referred to, Fig. 4 is the approximation schematic diagram in the embodiment of the present invention between two line segments.Obtain line segment A and line segment
The distance between S L, distance L is the distance between endpoint nearest on line segment A and line segment S herein, rather than between two lines section
Vertical range；
If 4) meet L≤2 × min (L1,L2), then determine that line segment A and line segment S have approximation；
5) Fig. 5 is referred to, Fig. 5 is the synteny schematic diagram in the embodiment of the present invention between two line segments.Obtain line segment A and level side
To angle theta 1, the angle theta 2 and angle threshold θ of line segment S and horizontal directionthreshold, angle threshold θ hereinthresholdTake 2~
12；
If 6) met | θ1-θ2|≤θthreshold, then determine that line segment A and line segment S have continuity；
7) the width line1_width of line segment A, the width line2_width and threshold value k of line segment S are obtained, herein threshold value
K takes 2~12；
If 8) meet dist≤k-max (line1_width, line2_width), then determine that line segment A and line segment S have altogether
Linearly；
9) if line segment A and line segment S meet approximation, continuity and synteny simultaneously, distance L is put into level metric number
In group；
10) the 2) step is returned, all final line segments in addition to line segment A, line segment S in each line segment group is traversed, reads another
Line segment X；
11) judge the approximation, continuity and synteny of line segment X and line segment A；
12) if line segment X and line segment A meet approximation, continuity and synteny, line segment X is put at a distance from line segment A
Enter in level metric array；
13) all final line segments in addition to line segment A, line segment S, line segment X in each line segment group have been traversed, have been obtained and line segment A
Meet the line segment of approximation, continuity and synteny, and is put into level metric array；
14) it in level metric array, according to the line segment of the shortest distance, obtains level-one and most preferably merges line segment；
15) merge line segment A and most preferably merge line segment with level-one, obtain new line segment C, recalculate the attribute of new line segment C,
16) the 2) step is returned, is iterated that search is best to merge line segment, it may be assumed that is traversed all final in each line segment group
Line segment reads a line segment D；
Judge the approximation, continuity and synteny of line segment D and line segment C；
If line segment D and line segment D meet approximation, continuity and synteny, line segment D is put at a distance from line segment C
Second level is measured in array；
All final line segments in addition to line segment D in each line segment group are traversed；
In second level measurement array, according to the line segment of the shortest distance, best merging line segment E is obtained；
Merge line segment C and merges line segment E with best, a merging line segment under the conditions of obtaining tightly；
Return to the 1) step, traverse remaining untreated final line segment, until terminating, obtain it is tight in each line segment group under the conditions of
All line segments merging line segment.
The method that the best selection for merging line segment uses two-stage iterative processing, it is significant for capable of making the conducting wire extracted processing
, complete edge contour, line segment is arranged according to line segment length, retains longer line segment, so as to more fully
Retain the profile information of transmission line of electricity.Angle threshold θ in the present inventionthreshold2~12 are taken, threshold value k takes 2~12, in tight condition
Under, θthresholdIt is smaller with k value, under the conditions of pine, θthresholdIt is larger with k value.
S1403: according to the merging line segment under the conditions of tight, conducting wire is obtained.
Merge the method that the selection of line segment also uses two-stage iterative processing under the conditions of pine, with it is tight under the conditions of merge the choosing of line segment
Take the difference is that, the merging of loose condition lower line segment be by it is above-mentioned it is tight under the conditions of merge after line segment be divided into two groups: including
Line segment group and line segment angle of the line segment angle between -75 °~75 ° are less than -75 ° or or greater than the line segment group between 75 °.
The choosing method of specific line segment with tightly under the conditions of choosing method it is identical, which is not described herein again.
S150: according to the difference between the lower edges pixel of conducting wire in binary edge map, icing thickness is calculated
Degree.
Specifically, steps are as follows for ice covering thickness calculating:
01) the vertical coordinate p1 (i) of single-row top edge pixel in binary edge map is obtained；
02) the vertical coordinate p2 (i) of single-row lower edge pixel described in binary edge map is obtained；
03) according to vertical coordinate p1 and vertical coordinate p2, using p (i)=| p1 (i)-p2 (i) |, single-row side is calculated
The pixel difference of edge pixel；
04) it according to pixel difference, utilizesObtain icing before conducting wire lower edges average thickness d1 and
The lower edges average thickness d2 of conducting wire after icing；
05) according to the lower edges average thickness d2 of conducting wire after the lower edges average thickness d1 of conducting wire, icing before icing
And wire diameter D, it utilizesThe thickness DM between conducting wire lower edges is calculated；
06) it according to the thickness between conducting wire lower edges, utilizesThe ice covering thickness of conducting wire is calculated.
The calculating of ice covering thickness is mainly the difference of the pixel of the lower edges according to icing, the i.e. lower edges of icing
Pixel vertical coordinate difference, the difference of the pixel is the thickness after icing.Due in the thickness after icing herein
Include the thickness of overhead stranded conductor, so needing to remove twisted wire rear and then obtaining ice covering thickness using certain calculation method.
Correspondingly, it needs to be known in advance in the case where non-icing when calculating ice covering thickness, in binary edge map on overhead stranded conductor
Pixel difference shared by lower edge.In binary edge map, each column can all have up contour point, and (i.e. there is edge in the first row
Point) and down contour point (i.e. the point at edge occurs in last line), the pixel vertical coordinate difference of lower edges is thickness, this
Thickness refers to the thickness on binary edge map, it is also necessary to by Parameter Switch be actual (real) thickness.
Indicate that the pixel number of overhead transmission line thickness, p1 (i) and p2 (i) respectively indicate the upper of overhead transmission line with p (i)
The vertical coordinate of pixel corresponding to lower edge, then using p (i)=| p1 (i)-p2 (i) | the thickness on single-row is calculated
Pixel p (i) utilizesThe average thickness d of the lower edges of overhead transmission line is calculated, it is assumed that do not covered
The mean pixel of ice overhead transmission line with a thickness of d1, make somebody a mere figurehead with a thickness of d2 by the mean pixel of the overhead transmission line after obtaining icing
Wire diameter is D on power transmission line, utilizes formulaThe thickness of the overhead transmission line lower edges of icing is calculated
Spend DM.From the point of view of current a large amount of environment icing investigation situations, icing form overview gets up and can be divided into two classes: cross sectional shape has rule
It is then one kind, cross sectional shape is random for one kind.Usually in the ice covering thickness on computing electric power line by icing section
It is approximately circle, so utilizing formulaIce covering thickness d is calculated.
Since overhead transmission line is in that catenary curve is sagging under gravity, in addition to unmanned plane shooting angle is from upper vertical
Downwards or other than lower online vertical two kinds of situations, overhead transmission line is generally not straight line in practice, and therefore, it is necessary to meter
Obtained ice covering thickness is corrected, i.e., is corrected to pixel shared by ice covering thickness in binary edge map.
Specifically, correcting the step of obtaining accurate ice covering thickness to ice covering thickness includes:
According to the measurement thickness e g' of lower edges after icing, the actual (real) thickness eg of lower edges, utilizes eg=after icing
The angle of inclination beta of conducting wire is calculated in eg' × cos β；
According to conducting wire angle of inclination beta after icing1, conducting wire angle of inclination beta before icing2, thickness DM between the conducting wire lower edges and lead
Line outer diameter D utilizesWire icing thickness is calculated.
It is that ice covering thickness corrects schematic diagram in the embodiment of the present invention with reference to Fig. 6, Fig. 6.AD is icing (or overhead power transmission
Line) on edge, BC is icing (or overhead transmission line) lower edge, and straight line L is obtained dummy line after AD and BC processing, and eg' is to cover
The measurement thickness of lower edges after ice, and the actual (real) thickness of lower edges is eg=eg' × cos β after icing.The solution of straight line L
According to compromise principle, binary edge map is with top edge and lower edge, each being aligned of lower edges and in actual state
It is disjoint, so the single-row point on straight line L is the intermediate point of lower edges point, i.e.,It asks in this way
The point on straight line L that solution comes out is possible to not point-blank, so its slope can not be determined, also can not just be calculated accurately
Therefore eg value does first order curve fitting to L (i), so utilizing formulaCalculate correction
Ice covering thickness d afterwards, wherein β1And β2Respectively indicate after icing inclining for overhead transmission line before the inclination angle of overhead transmission line and icing
Angle.Wire icing thickness is corrected, more accurate ice covering thickness can be obtained, is conducive to power supply enterprise for ice cover pair
Powerline ice-covering phenomenon takes corresponding ice-melt measure in time.
Those skilled in the art will readily occur to of the invention its after considering specification and the disclosure invented here of practice
Its embodiment.This application is intended to cover any variations, uses, or adaptations of the invention, these modifications, purposes or
Person's adaptive change follows general principle of the invention and including the undocumented common knowledge in the art of the present invention
Or conventional techniques.The description and examples are only to be considered as illustrative, and true scope and spirit of the invention are by following
Claim is pointed out.
It should be understood that the present invention is not limited to the precise structure already described above and shown in the accompanying drawings, and
And various modifications and changes may be made without departing from the scope thereof.The scope of the present invention is limited only by the attached claims.
Claims (7)
1. a kind of powerline ice-covering detection method based on visible images characterized by comprising
Edge detection is carried out to original image by binaryzation and obtains bianry image；
Edge detection is carried out to the bianry image from 8 different directions respectively using edge detection operator, extracts described two
It is worth the edge feature of image, obtains its binary edge map；
Piecemeal processing is carried out to the binary edge map, obtains the shaft tower piecemeal for having shaft tower image；
Line segment merging is carried out to the shaft tower piecemeal according to approximation, continuity and the synteny between line segment, obtains conducting wire；
According to the difference between the lower edges pixel of conducting wire in the binary edge map, ice covering thickness is calculated；
According to the inclination angle of the conducting wire before the inclination angle of conducting wire described after icing and icing, the ice covering thickness is corrected to obtain accurate
Ice covering thickness；
Wherein, icing thickness is calculated in the difference between the lower edges pixel according to conducting wire in binary edge map
Degree, comprising:
Obtain the vertical coordinate p1 (i) of single-row top edge pixel in the binary edge map；
Obtain the vertical coordinate p2 (i) of single-row lower edge pixel described in the binary edge map；
According to the vertical coordinate p1 and vertical coordinate p2, using p (i)=| p1 (i)-p2 (i) |, the single-row side is calculated
The pixel difference of edge pixel；
According to the pixel difference, utilizeObtain icing before the conducting wire lower edges average thickness d1 and
The lower edges average thickness d2 of the conducting wire after icing；
According to the lower edges average thickness d2 of conducting wire after the lower edges average thickness d1 of conducting wire, icing before the icing and
Wire diameter D is utilizedThe thickness DM between conducting wire lower edges is calculated；
According to the thickness between the conducting wire lower edges, utilizeThe ice covering thickness of conducting wire is calculated；
Wherein, the inclination angle according to the conducting wire before the inclination angle of conducting wire described after icing and icing, to the ice covering thickness school
Just obtaining accurate ice covering thickness, comprising:
According toThe single-row point L (i) on straight line L is calculated, first order curve fitting is done to L (i), according to
Slope of a curve after fitting obtains conducting wire angle of inclination beta after icing1With conducting wire angle of inclination beta before icing2, wherein straight line L be icing (or
Person's overhead transmission line) on along and the processing of icing (or overhead transmission line) lower edge after obtained dummy line；
According to conducting wire angle of inclination beta after icing1, conducting wire angle of inclination beta before icing2, outside thickness DM and conducting wire between the conducting wire lower edges
Diameter D is utilizedWire icing thickness is calculated.
2. the powerline ice-covering detection method according to claim 1 based on visible images, which is characterized in that described
Edge detection is carried out to original image by binaryzation and obtains bianry image, comprising:
Gaussian smoothing is carried out to the original image and obtains Gaussian kernel, the original image is rolled up using the Gaussian kernel
Product, the image after obtaining Gaussian smoothing；
It is differentiated using gradient operator to the image after the Gaussian smoothing, obtains gradient image；
Non-maxima suppression is carried out to the gradient value of the gradient image, pixel local maximum is obtained and pixel is locally non-
Maximum；
By the pixel locally corresponding gray value zero setting of non-maximum, the bianry image is obtained.
3. the powerline ice-covering detection method according to claim 2 based on visible images, which is characterized in that described
Gradient operator is the Prewitt operator or Sobel Operator in 8 directions.
4. the powerline ice-covering detection method according to claim 1 based on visible images, which is characterized in that institute
It states binary edge map and carries out the shaft tower piecemeal that piecemeal processing acquisition has shaft tower image, comprising:
The Blob analysis that original line segments in the binary edge map are carried out with connectivity structure, obtains treated most finish line
Section, and the final line segment quantity in each direction is less than or equal to 128；
Piecemeal processing is carried out to the binary edge map, obtains block image, and horizontal block count is in the block image
16, vertical block count is 4；
Horizontal direction in the block image, vertical direction, oblique direction and the most finish line in direction obliquely downward are counted respectively
Segment number；
If the final line segment quantity of horizontal direction is greater than the final line segment quantity of 3, vertical direction in the piecemeal
Greater than 3, and the final line segment quantity of oblique direction horizontal direction more than or equal to 3 or in the piecemeal is described
The final line segment quantity that final line segment quantity is greater than 3, vertical direction is greater than 3 and the obliquely downward final line segment quantity in direction
More than or equal to 3, then determine the block memory in shaft tower；
The horizontal direction is -10 °~10 °；The vertical direction is greater than 85 ° or less than -85 °；The oblique direction is
10 °~85 °；The direction obliquely downward is -85 °~-10 °.
5. the powerline ice-covering detection method according to claim 4 based on visible images, which is characterized in that described
Line segment merging is carried out to the shaft tower piecemeal according to approximation, continuity and the synteny between line segment, obtains conducting wire, packet
It includes:
According to the direction of the final line segment and angle, six direction is divided to be grouped on the final line segment, under the conditions of obtaining tightly
Line segment group, it is described it is tight under the conditions of line segment group include: horizontal line section group, upper inclined horizontal line section group, upper hading straightway group, it is lower partially
Horizontal line section group, lower hading straightway group, vertical segment group；
Each line segment group is scanned for using the method for two-stage iterative processing, obtains best merging line segment；
The method of the two-stage iterative processing includes:
The final line segment for traversing all non-merging treatments in each line segment group reads line segment an A, the line segment A
Length be L1；
All final line segments in addition to line segment A in each line segment group are traversed, another line segment S, the line segment S are read
Length be L2；
Obtain the distance between the line segment A and line segment S L；
If meeting L≤2 × min (L1,L2), then determine that the line segment A and line segment S have approximation；
Obtain the angle theta 1 of the line segment A and horizontal direction, the angle theta 2 and angle threshold of the line segment S and horizontal direction
θthreshold；
If met | θ1-θ2|≤θthreshold, then determine that the line segment A and line segment S have continuity；
Obtain the width line1_width of the line segment A, the width line2_width and threshold value k of the line segment S；
If met distance dist≤k-max (line1_width, line2_width), then determine that the line segment A and line segment S has
Synteny；
If the line segment A and line segment S meet approximation, continuity and synteny simultaneously, the distance L is put into level metric
In array；
All final line segments in addition to line segment A, line segment S in each line segment group are traversed, another line segment X is read；
Judge approximation, continuity and the synteny of the line segment X Yu the line segment A；
If the line segment X and the line segment A meet approximation, continuity and synteny, by the line segment X and the line segment
The distance of A is put into the level metric array；
All final line segments in addition to line segment A, line segment S, line segment X in each line segment group have been traversed, have been obtained and line segment A
Meet the line segment of approximation, continuity and synteny, and is put into level metric array；
In the level metric array, according to the line segment of the shortest distance, obtains level-one and most preferably merge line segment；
Merge the line segment A and most preferably merge line segment with level-one, obtains new line segment C；
The final line segment all in each line segment group is traversed, a line segment D is read；
Judge approximation, continuity and the synteny of the line segment D Yu the line segment C；
If the line segment D and the line segment D meet approximation, continuity and synteny, by the line segment D and the line segment
The distance of C is put into second level measurement array；
All final line segments in addition to line segment D in each line segment group are traversed；
In second level measurement array, according to the line segment of the shortest distance, best merging line segment E is obtained；
Merge the line segment C with most preferably merges line segment E, obtain it is described tightly under the conditions of merging line segment；
Traverse remaining interior non-merging treatment of each line segment group does not merge final line segment, using the two-stage iterative processing
Method obtains the best merging line segment for not merging final line segment；
Merge it is described does not merge final line segment and the best merging line segment for not merging final line segment, obtain it is described it is tight under the conditions of
Merging line segment；
According to it is described it is tight under the conditions of merging line segment, obtain conducting wire.
6. the powerline ice-covering detection method according to claim 5 based on visible images, which is characterized in that described
Line segment merging is carried out to the shaft tower piecemeal according to approximation, continuity and the synteny between line segment, conducting wire is obtained, also wraps
It includes:
By it is described it is tight under the conditions of the grouping of merging line segment, obtain the line segment group under the conditions of pine, the line segment group packet under the conditions of the pine
Include loose condition lower line segment group of the line segment angle between -75 °~75 ° and line segment angle be less than -75 ° or or be greater than 75 ° it
Between loose condition lower line segment group；
The line segment for traversing all non-merging treatments in the line segment group of the loose condition, using the method for the two-stage iterative processing,
Obtain the merging line segment under the conditions of the pine.
7. the powerline ice-covering detection method according to claim 5 based on visible images, which is characterized in that described
Angle threshold θthresholdIt is 2~12, the threshold value k value is 1~3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611151446.XA CN106705866B (en) | 2016-12-14 | 2016-12-14 | A kind of powerline ice-covering detection method based on visible images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611151446.XA CN106705866B (en) | 2016-12-14 | 2016-12-14 | A kind of powerline ice-covering detection method based on visible images |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106705866A CN106705866A (en) | 2017-05-24 |
| CN106705866B true CN106705866B (en) | 2019-11-08 |
Family
ID=58935727
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611151446.XA Active CN106705866B (en) | 2016-12-14 | 2016-12-14 | A kind of powerline ice-covering detection method based on visible images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106705866B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107514976B (en) * | 2017-08-24 | 2020-06-23 | 国家电网公司 | Power line icing monitoring method and terminal equipment |
| CN107742285B (en) * | 2017-09-26 | 2020-11-10 | 内江金鸿曲轴有限公司 | Method for detecting thickness of grinding layer of grinding wheel on machine |
| CN107680095A (en) * | 2017-10-25 | 2018-02-09 | 哈尔滨理工大学 | The electric line foreign matter detection of unmanned plane image based on template matches and optical flow method |
| CN108334113A (en) * | 2018-04-19 | 2018-07-27 | 北京邮电大学 | Unmanned plane deicing system and method |
| CN110070530B (en) * | 2019-04-19 | 2020-04-10 | 山东大学 | Transmission line icing detection method based on deep neural network |
| CN110686608B (en) * | 2019-11-11 | 2021-08-31 | 国网湖南省电力有限公司 | Portable icing thickness measuring device for power transmission line and measuring method thereof |
| CN110954000A (en) * | 2019-12-18 | 2020-04-03 | 上海工程技术大学 | Method for measuring height of frost layer on cold surface |
| CN113034524A (en) * | 2019-12-25 | 2021-06-25 | 深圳怡化电脑股份有限公司 | Image edge detection method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103047938A (en) * | 2013-01-05 | 2013-04-17 | 山西省电力公司大同供电分公司 | Method and device for detecting icing thickness of power transmission line |
| CN103528534A (en) * | 2013-10-25 | 2014-01-22 | 江西省电力公司检修分公司 | Image monitoring based method for detecting thickness of icing on power transmission line |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104835152A (en) * | 2015-04-27 | 2015-08-12 | 国家电网公司 | Processing method and system of power transmission line inspection images |
-
2016
- 2016-12-14 CN CN201611151446.XA patent/CN106705866B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103047938A (en) * | 2013-01-05 | 2013-04-17 | 山西省电力公司大同供电分公司 | Method and device for detecting icing thickness of power transmission line |
| CN103528534A (en) * | 2013-10-25 | 2014-01-22 | 江西省电力公司检修分公司 | Image monitoring based method for detecting thickness of icing on power transmission line |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106705866A (en) | 2017-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106705866B (en) | A kind of powerline ice-covering detection method based on visible images | |
| CN108734689B (en) | Method for detecting scattered strands of conducting wires based on region growth | |
| CN109215020B (en) | High-voltage transmission line fault identification method based on computer vision | |
| CN103487729B (en) | Based on the power equipments defect detection method that ultraviolet video and infrared video merge | |
| CN105023014B (en) | A kind of shaft tower target extraction method in unmanned plane inspection transmission line of electricity image | |
| CN107179479B (en) | Visible light image-based transmission conductor strand scattering defect detection method | |
| CN110197176A (en) | Inspection intelligent data analysis system and analysis method based on image recognition technology | |
| CN111814686A (en) | Vision-based power transmission line identification and foreign matter invasion online detection method | |
| CN111402215B (en) | Contact net insulator state detection method based on robust principal component analysis method | |
| CN107742307A (en) | Based on the transmission line galloping feature extraction and parameters analysis method for improving frame difference method | |
| CN105957077B (en) | The electric line foreign matter detection method of view-based access control model significance analysis | |
| CN107843818B (en) | High-voltage insulation fault diagnosis method based on heterogeneous image temperature rise and partial discharge characteristics | |
| CN103413150A (en) | Power line defect diagnosis method based on visible light image | |
| CN106680285A (en) | Infrared image-assisted method of recognizing contamination condition of insulator by visible light image | |
| CN110288571B (en) | High-speed rail contact net insulator abnormity detection method based on image processing | |
| CN102831393A (en) | Rapid image recognizing method of power tower pole outline | |
| CN110222683A (en) | A kind of quick-fried defect recognition localization method of electric transmission line isolator component based on depth convolutional neural networks | |
| CN106525859A (en) | Method for detecting strand breakage defect of high-voltage electric transmission wire by unmanned aerial vehicle in real time | |
| CN102519846A (en) | Hyperspectrum-based composite insulator hydrophobicity detection method | |
| CN115330646B (en) | Metal component inspection method for fitness equipment paint spraying process | |
| CN109300118A (en) | A kind of high-voltage electric power circuit unmanned plane method for inspecting based on RGB image | |
| CN115512252B (en) | Unmanned aerial vehicle-based power grid inspection automation method and system | |
| CN109523543A (en) | A kind of wire strand breakage detection method based on Edge Distance | |
| CN108229371B (en) | Ice-shape-modeling-based method for identifying icing shape of cross section of power transmission conductor | |
| CN110276787B (en) | Conductor galloping monitoring method based on marker image detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |