[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

CN104732588B - A kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud - Google Patents

A kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud Download PDF

Info

Publication number
CN104732588B
CN104732588B CN201510145070.0A CN201510145070A CN104732588B CN 104732588 B CN104732588 B CN 104732588B CN 201510145070 A CN201510145070 A CN 201510145070A CN 104732588 B CN104732588 B CN 104732588B
Authority
CN
China
Prior art keywords
power line
laser radar
radar point
line
point cloud
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN201510145070.0A
Other languages
Chinese (zh)
Other versions
CN104732588A (en
Inventor
林祥国
张继贤
段敏燕
臧艺
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chinese Academy of Surveying and Mapping
Original Assignee
Chinese Academy of Surveying and Mapping
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Chinese Academy of Surveying and Mapping filed Critical Chinese Academy of Surveying and Mapping
Priority to CN201510145070.0A priority Critical patent/CN104732588B/en
Publication of CN104732588A publication Critical patent/CN104732588A/en
Application granted granted Critical
Publication of CN104732588B publication Critical patent/CN104732588B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Length Measuring Devices By Optical Means (AREA)

Abstract

A kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud, the data of described airborne laser radar point cloud are correctly classified, described method comprises the following steps: step one, load overhead transmission line correctly classify after airborne laser radar point cloud data and initial tracks data; Step 2, extracts the precise information of electricity tower position, quantity and transmitting line track, and determines total gear number of transmitting line; Step 3, it is determined that each grade of two-dimensional space scope of transmitting line and power line laser radar point; Step 4, carries out cluster to each grade of power line laser radar point determined; Step 5, carries out power line three-dimensional reconstruction. Parameter required for the present invention is few, level of automation height, robustness and universality better, there is the radical to power line, power line type, power line space matching structure, thick not good enough, the point insensitive advantage of factor such as the irregular fracture of cloud and line length, and reconstruction model has higher reconstruction precision.

Description

A kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud
Technical field
The present invention relates to one utilizes airborne laser radar point cloud to carry out power line three-dimensional rebuilding method, specifically based on a long distance overhead transmission line power line three-dimensional rebuilding method for airborne laser radar point cloud data, belong to laser radar point cloud data information extraction technology field.
Background technology
Overhead transmission line is an important national basis facility. In order to prevent and stop the generation of power grid security accident, power-line patrolling is the important process that power grid operation maintenance management department needs to carry out. Manually patrol that line labor intensive is big, working conditions is arduous, efficiency is low, patrol cycle length again, to patrol and examine data accuracy not high.
In recent years, airborne laser radar measuring technology obtains increasingly extensive application in helicopter power line inspection. At present, the correlative study of airborne laser radar point cloud power-line patrolling mainly concentrates on six aspects such as the laser point cloud classification of power line corridor, the detection of power line point, potential safety hazard monitoring (dangerous point detection), electricity tower three-dimensional reconstruction, power line corridor three-dimensional visualization, power line three-dimensional reconstruction. Wherein, power line three-dimensional reconstruction is the basis of the important application such as dangerous point detection, the alternate discrepancy in elevation measurement of wire, separation measurement, the measurement of wire distance to the ground, three-dimensional visualization, conducting wire sag analysis, wire icing analysis, wire windage yaw analysis, become the emphasis that industry is paid close attention to, also it is difficult point.
At present, most power line three-dimensional rebuilding methods based on airborne laser radar point cloud is relatively applicable to landform flat site, the simple power line three-dimensional reconstruction of electric network composition, but the power line three-dimensional reconstruction of long distance extra-high voltage, ultra-high voltage, high pressure overhead power line is also existed five problems:
The first, power line three-D space structure and layout are considered not enough. Can there is the arrangement mode of the multiple configurations such as triangle, level, vertical, mixing arrangement in the power line in high-tension line one grade, and the Hough transform method of classics cannot detect out many power lines arranged vertically; Further, for many power lines of vertical structure, the separation that cannot be carried out different root power line laser radar point by height value feature can be caused due to sag phenomenon.
2nd, the type understanding of power line is not enough. With patrol and examine closely related power line and can be divided into wire and lightning conducter, the former can be subdivided into again S.C., division wire (also known as wire bundle, can be further subdivided into two, four, six, eight divisions etc.). In laser radar point cloud, the morphological specificity of S.C., lightning conducter, the arbitrary division of division wire (bundle) is extremely similar; And on the whole, division wire is different from the morphological specificity of S.C., lightning conducter. Meanwhile, in power-line patrolling, division wire is looked as a whole, namely without the need to taking the division number of wire into account, rebuild whole division wire in process of reconstruction. But most existing method for reconstructing only considers the three-dimensional reconstruction of S.C., lightning conducter, a branch of division wire, and do not take the three-dimensional reconstruction of division wire into account. For division wire, owing to conductor spacer can connect each division, and differentiation conductor spacer laser radar point that cannot be detailed in airborne laser radar point cloud, power line laser radar point, this can cause segment space cluster algorithm to gather for cluster by each division, can cause the mistake of subsequent reconstruction.
3rd, the Information Pull of tower bar is not enough. Tower bar information may be used for determining the shelves at power line laser spots place; The point of tangent determining each power line can also be helped. At present, power department has rough tower bar positional information; Airborne laser radar point cloud can also be distinguished tower bar laser radar point. But existing method ignores the potential value of tower bar information in power line is rebuild.
4th, slightly poor robustness is not enough. It is slightly poor to exist in airborne laser radar point cloud, and power line laser radar point cloud is no exception. These are slightly poor has remarkably influenced to the precision based on classical least square fitting model, and most existing power line extracts and method for reconstructing ignores slightly poor negative impact so that the robustness of method is not high.
5th, the power line three-dimensional model complexity of employing is various, and part model exists theoretical mistake. Such as, the mistake of model occurs when cover strong for the para-curve formula of two-dimensional space is incorporated into three-dimensional space by part method, can cause there will be reconstruction errors when the power line matching being applied to true three-dimensional space.
The problems referred to above cause existing power line three-dimensional rebuilding method exist the more parameter of needs, level of automation low, need the problems such as a large amount of manual intervention, robustness and universality are poor, reconstruction precision is low, intelligent grid cannot be met to demand quick, that high-precision electric patrols line.
Summary of the invention
For the deficiencies in the prior art, the present invention proposes a kind of long distance extra-high voltage, ultra-high voltage, high pressure overhead power line power line three-dimensional rebuilding method based on airborne laser radar point cloud data, it can make full use of tower bar information, there is the radical to power line, power line type, power line space matching structure, slightly almost, put the insensitive advantages of factor such as the irregular fracture of cloud, line length, and reconstruction model has higher reconstruction precision.
The present invention solves the technical scheme that its technical problem takes: a kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud, is characterized in that, the data of described airborne laser radar point cloud are correctly classified, and described method comprises the following steps:
Step one, load overhead transmission line correctly classify after airborne laser radar point cloud data and initial tracks data;
Step 2, extracts the precise information of electricity tower position, quantity and transmitting line track, and determines total gear number of transmitting line;
Step 3, it is determined that each grade of two-dimensional space scope of transmitting line and power line laser radar point;
Step 4, carries out cluster to each grade of power line laser radar point determined;
Step 5, carries out power line three-dimensional reconstruction.
Preferably, in step 2, carry out based on electricity tower laser radar point cloud and initial tracks data extracting accurate electricity tower positional information, electricity tower quantity information and tracks information, and determine total gear number that transmitting line comprises.
Preferably, in step 3, determine the two-dimensional space scope of each grade of transmitting line in XOY plane and corresponding power line laser radar point based on accurate electricity tower position and tracks data.
Preferably, in step 4, carry out single grade of power line laser radar point cloud cluster based on segmentation projection and k-means cluster, to isolate the laser radar point cloud of every root power line.
Preferably, in step 5, carry out the three-dimensional reconstruction of each power line in conjunction with straight line model and parabola model.
Further, described step 2 specifically comprises the following steps: (1) prepares the data of input; (2) irregular triangulation network TIN is constructed; (3) the long limit in TIN is deleted; (4) connect component analysis based on TIN is carried out; (5) minimum enclosed rectangle and the geometric center point thereof of each Connected component is asked for; (6) electric power circuit trace information renewal is carried out.
Further, described step 3 specifically comprises the following steps: (1) obtains two adjacent electric tower information in the overhead transmission line that each shelves relate to; (2) ask for each geometric center point and with corresponding geometric center point belonging to the parallel straight line of the longest edge of minimum enclosed rectangle; (3) the extension straight-line segment of each electric tower is determined; (4) spatial dimension of each grade of power line is determined; (5) the power line laser radar point of each grade is determined.
Further, described step 4 specifically comprises the following steps: (1) carries out the linear fit of whole shelves power line; (2) length of this grade of power line and the scale factor of each laser radar point is asked for; (3) laser radar point of this grade of power line is carried out segmentation; (4) each section of this grade of power line is carried out segmentation projection, segmentation cluster and classification transmission process between section and section.
Further, described step 5 specifically comprises the following steps: (1) sets up power line three-dimensional model; (2) straight line model is asked for; (3) parabola model is asked for; (4) combination is asked for straight line model and parabola model carry out power line three-dimensional reconstruction.
The useful effect of the present invention is as follows:
(1) the present invention proposes a kind of equation of line and single grade of single root power line three-dimensional model that One-place 2-th Order polynomial expression combines, polynomial expression uses scale factor as parameter, it is to increase the precision of model.
(2) the present invention proposes a kind of based on segmentation projection and single grade of power line laser radar point cloud cluster process method of k-means cluster. In cluster process, first on XOY plane, this grade of whole power line laser radar point has been carried out overall linear fit, then by power line laser radar point segmentation, then by the section (this plane orthogonal in fitting a straight line) of each section of laser radar point reprojection to power line, and use k-means clustering method to carry out the cluster of each section of subpoint, transmission and the regularization of classification is carried out between last section and section by projected centre point.
(3) existing laser radar point cloud power line three-dimensional rebuilding method is compared, parameter required for the present invention is few, level of automation height, robustness and universality better, there is the radical to power line, power line type, power line space matching structure, thick not good enough, the point insensitive advantage of factor such as the irregular fracture of cloud and line length, and reconstruction model has higher reconstruction precision.
Accompanying drawing explanation
Fig. 1 is the method flow diagram of the present invention;
Fig. 2 (a) is the schematic diagram (experimental datas one in two scape data) of the airborne laser radar point cloud classified;
Fig. 2 (b) is the laser radar point of 5 power lines and the schematic diagram of the laser radar point of 3 electric towers;
Fig. 2 (c) is the TIN and partial enlargement figure thereof of electricity tower laser radar point;
Minimum enclosed rectangle that Fig. 2 (d) is Connected component and partial enlargement figure thereof;
Fig. 2 (e) is the schematic diagram of initial electric tower position point;
Fig. 3 is the schematic diagram (having two grades in figure, the limit of each grade of four corresponding limit Polygonss is endowed a kind of color) that each grade of power line two-dimensional space scope is determined;
Fig. 4 (a) is single grade of power line laser radar point schematic diagram and partial side view thereof and cross-sectional perspective view;
Fig. 4 (b) is the schematic diagram of fitting of a straight line and two scale factor extreme points;
Fig. 4 (c) is the schematic diagram of local power line laser radar points and the k-means cluster result of subpoint thereof;
The cluster result schematic diagram (every root power line is endowed a kind of random coloration) that Fig. 4 (d) is certain grade of power line laser radar point;
The schematic diagram (every root power line is endowed a kind of random coloration) that Fig. 4 (e) is the three-dimensional reconstruction model of 5 power lines in Fig. 4 (a);
Fig. 5 is the power line three-dimensional model schematic diagram that straight line model and parabola model combine;
Fig. 6 is the overall numerical evaluation schema of the present invention;
Fig. 7 is the schematic diagram that in Fig. 2 (a), power line three-dimensional reconstruction model cover is incorporated into classification point cloud;
Fig. 8 (a) is the schematic diagram of the experimental data two in two scape data;
Fig. 8 (b) is the result schematic diagram of Fig. 8 (a) power line three-dimensional reconstruction.
Embodiment
For clearly demonstrating the technology feature of this scheme, below by embodiment, and in conjunction with its accompanying drawing, the present invention will be described in detail. Disclosing hereafter provides many different embodiments or example is used for realizing the different structure of the present invention. In order to simplify disclosing of the present invention, hereinafter parts and setting to specific examples are described. In addition, the present invention can in different example repeat reference numerals and/or letter. This kind repeats to be to simplify and clearly object, itself does not indicate the relation between discussed various embodiment and/or setting. It should be noted that parts illustrated in the accompanying drawings are not necessarily drawn in proportion. Present invention omits the description to known assemblies and treatment technology and process to avoid unnecessarily limiting the present invention.
For the needs of the long distance overhead transmission line power line three-dimensional reconstruction based on airborne laser radar point cloud, the present invention provides a kind of long distance extra-high voltage based on airborne laser radar point cloud data, ultra-high voltage, high pressure overhead power line power line three-dimensional rebuilding method, especially a kind of equation of line is proposed and single grade of single root power line three-dimensional model that One-place 2-th Order polynomial expression combines, a kind of based on single grade of power line laser radar point cloud clustering method of segmentation projection and k-means cluster, to improve the precision of airborne laser radar point cloud power line three-dimensional reconstruction, level of automation, robustness and universality, can be applicable to electric power line inspection, electric power circuit corridor three-dimensional visualization, electric power circuit is planned, the technical fields such as intelligent grid. in addition, it is necessary to it is emphasized that, the present invention stresses airborne laser radar point cloud power line three-dimensional reconstruction, and does not stress the target detect of the classification of power line corridor airborne laser radar point cloud data and power line point, electricity tower point, the airborne laser radar point cloud data used in the present invention is correctly classified.
As shown in Figure 1, a kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud of the present invention, is characterized in that, the data of described airborne laser radar point cloud are correctly classified, and described method comprises the following steps:
Step one, load overhead transmission line correctly classify after airborne laser radar point cloud data and initial tracks data;
Step 2, extracts the precise information of electricity tower position, quantity and transmitting line track, and determines total gear number of transmitting line;
Step 3, it is determined that each grade of two-dimensional space scope of transmitting line and power line laser radar point;
Step 4, carries out cluster to each grade of power line laser radar point determined;
Step 5, carries out power line three-dimensional reconstruction.
Further, described step 2 specifically comprises the following steps:
(1) data of input are prepared
The laser radar point isolating separately electricity tower and power line two classifications from the airborne laser radar point cloud data (as Suo Shi Fig. 2 (a)) classified is as input data. 3 the electric towers isolated from Fig. 2 (a) and the laser radar point of 5 power lines, two classifications are as shown in Fig. 2 (b).
(2) irregular triangulation network TIN is constructed
The horizontal coordinate of electricity tower laser radar point is utilized to construct irregular triangulation network TIN. By the TIN of electricity tower laser radar point structure in Fig. 2 (b) as shown in Fig. 2 (c).
(3) the long limit in TIN is deleted
The reason deleting the long limit in TIN has two aspects. On the one hand, the shelves of overhead transmission line are longer apart from (horizontal throws between two adjacent shaft tower wire point of tangent), are at least some hundred meters. On the other hand, the density of overhead transmission line corridor airborne laser radar point cloud data is higher, has at least some points in every square metre, and namely the horizontal dot spacing of laser radar point is less. Therefore, if a certain bar limit in TIN is very long, can judge that this limit is connected to the laser radar point of two adjacent electricity towers instead of the laser radar point of same electricity tower. This patent selects 10m as length threshold, deletes length in TIN and is greater than the limit of 10m.
(4) connect component analysis based on TIN is carried out
Delete the TIN behind long limit describe electricity tower laser radar point space topological and be connected relation, after carrying out connect component analysis, the corresponding electric tower of each Connected component. This can determine the electricity overall number N of tower and total gear number N-1 of this circuit. The connect component analysis result of electricity tower laser radar point is as shown in Fig. 2 (c) in Fig. 2 (b), totally three Connected component in result, the corresponding electric tower of each Connected component.
(5) minimum enclosed rectangle and the geometric center point thereof of each Connected component is asked for
In XOY plane, seek minimum enclosed rectangle and the geometric center point thereof of each Connected component, Fig. 2 (c) illustrates the minimum enclosed rectangle of each Connected component simultaneously.
(6) electric power circuit trace information renewal is carried out
Utilizing geometric center point to upgrade the existing electric power circuit trace information of electric inspection process department (comprising electricity tower position, electricity tower neighborhood information), Fig. 2 (e) illustrates the initial electric tower position point that Fig. 2 (a) data relate to. Seek electric tower point nearest in the electric tower point that each geometric center point comprises to existing tracks, and utilize this geometric center point to substitute this electricity tower point. Electric power circuit trace information after renewal can assist the range information determining electric power circuit each grade.
Further, described step 3 specifically comprises the following steps:
(1) two adjacent electric tower information in the overhead transmission line that each shelves relate to are obtained
As previously mentioned, each shelves relate to two electric towers, every corresponding minimum enclosed rectangle of electric tower and geometric center point thereof, and the information of rectangle and geometric center point is known.
(2) that ask for each geometric center point and parallel with the minimum enclosed rectangle longest edge belonging to geometric center point straight line
In XOY plane, the profile of electricity tower is rectangular, and the short limit of rectangle is moved towards limit basically identical, long and power line and moved towards substantially vertical with power line. In addition, the point of tangent of power line is generally positioned near rectangular centre line. Therefore, the straight line that this step is asked for is the optimum boundary divided by each electric tower left and right sides power line laser radar point.
(3) the extension straight-line segment of each electric tower is determined
The straight line direction asked for along previous step and extend length b (unit: m) respectively to both sides centered by the minimum enclosed rectangle geometric center point of correspondence. Then each the corresponding straight-line segment of electric tower, this straight-line segment and power line move towards near normal, and its length is 2b.
(4) spatial dimension of each grade of power line is determined
Connect the straight-line segment end points that each shelves relate to two electric towers along clockwise direction in turn, a tetragon can be formed, this tetragon is the two-dimensional space scope of this grade, and Fig. 3 illustrates each grade of two-dimensional space scope of the electric power circuit that laser radar point cloud in Fig. 2 (a) relates to.
(5) the power line laser radar point of each grade is determined
After each grade of corresponding tetragon is determined, it is possible to ask for the power line laser radar point falling into each tetragon, it is designated as this grade of corresponding power line laser radar point.
Further, described step 4 specifically comprises the following steps:
(1) linear fit of whole shelves power line is carried out
In XOY plane, utilizing the horizontal coordinate of whole power line laser radar points of a certain shelves to carry out overall least square linear fit, straight-line equation have employed the normal formula shown in formula (1):
D=x*cos ��+y* �� (1)
In formula, the implication of d and �� is respectively: crossing the vertical line section that initial point does to straight line, the pitch angle of this vertical line section place straight line is ��, d is the length of this line segment; And set the intersection point of this vertical line section and fitting a straight line as P (xfootprint,yfootprint). Fig. 4 (a) illustrates the laser radar point cloud of in Fig. 2 (a) certain grade, and Fig. 4 (b) illustrates the fitting a straight line of Fig. 4 (a).
(2) length of this grade of power line and the scale factor of each laser radar point is asked for
If the horizontal coordinate of arbitrary laser radar point is Q (x0,y0), its vertical projection to the subpoint coordinate of above-mentioned fitting a straight line be Q ' (x '0,y��0), the scale factor s of this vertical projection point is calculated by formula (2) or formula (3).
As fabs (sin ��) >=fabs (cos ��),
s = x 0 ′ - x footprint - sin α - - - ( 2 )
As fabs (sin ��) < fabs (cos ��),
s = y 0 &prime; - y footprint cos &alpha;&alpha; - - - ( 3 )
Visible, the corresponding scale factor of each laser radar point. And then, seek the scale factor of this grade of all power line laser radar point, and obtain maximum ratio factor smaximum, minimum proportion factor sminimumVertical projection point M (x corresponding respectivelys_maximum,ys_maximum)��N(xs_minimum,ys_minimum). Horizontal throw between point M and N is designated as the length l of this grade of power linespan. Fig. 4 (b) illustrates two extreme value end points of Fig. 4 (a) mid point cloud subpoint in fitting a straight line simultaneously, the straight line segment length 437.12m between end points.
(3) laser radar point of this grade of power line is carried out segmentation
The power line laser radar point of this grade is carried out staging treating. The calculation formula of total section of number m of segmentation is such as formula shown in (4):
m = INT ( l span l sec tion ) - - - ( 4 )
In formula, lsectionIt is local length value, it is an empirical value. Then, proportionally power line point is sorted by factor size, and proportionally big young pathbreaker's power line point of the factor is divided into m section uniformly. Then i-th (i=0,1,2 ..., m-1) in section the scope of the scaled factor of power line laser radar point such as formula shown in (5):
s min imum + s max imun - s min imum m * i < < s < < s min imum + s max imum - s min imum m * ( i + 1 ) - - - ( 5 )
Power line laser spots in Fig. 4 (a) is with lsection=5.0m carries out segmentation, divide into 87 sections altogether, and wherein the some cloud of the 53rd section is as shown in Fig. 4 (c) upper part.
(4) segmentation projection, segmentation cluster and classification transmission between section and section is carried out
In operating process, first m-1 section power line laser spots before process; Then last section of line of force laser spots is processed.
For the some cloud of front m-1 section, respectively carry out following process by little to big order according to segment number.
Assume to process i-th (i=0,1,2 ..., m-2) section point cloud, the maximum scale factor of this section of power line laser radar point in theory value be. Meanwhile, the point remembered in the fitting a straight line that this maximum scale factor is corresponding is W (xi,yi), xi��yiCalculation formula respectively such as formula shown in (6) and formula (7):
x i = x footprint - sin &alpha; * s max inum - s min imum m * ( i + 1 ) - - - ( 6 )
y i = y footprint + cos &alpha; * s max imum - s min imum m * ( i + 1 ) - - - ( 7 )
Point W (xi,yi) corresponding space three-dimensional point is designated as W ' (xi,yi, 0). In three dimensions, cross three-dimensional point W ' and it is perpendicular to fitting a straight line and draw a plane Pplane_i��
Two sub cases are divided to process below:
As i=0. Then, by the power line point vertical projection in this section to plane Pplane_i, then the corresponding vertical projection point of each power line point. Then, utilizing the three-dimensional coordinate of vertical projection point to carry out k-means cluster, in cluster process, classification number is power line radical n, and asks for the central point of each cluster; And give corresponding laser spots the label of each subpoint cluster, this achieve the cluster to this section of power line laser spots, and the number of cluster is n, the corresponding cluster centre point of each cluster. It needs to be noted, a little, and central point is positioned at plane P in the institute that each cluster centre point can represent suchplane_iOn. Therefore, preserve n cluster centre point and cluster label thereof.
As i > 0. Then, by vertical projection while of the power line point in this section and upper one section of n cluster centre point produced to plane Pplane_i, then each power line point corresponding vertical projection point, each cluster centre point also corresponding vertical projection point. Then, utilizing the three-dimensional coordinate of vertical projection point to carry out k-means cluster, in cluster process, classification number is still power line radical n, and asks for the central point of each cluster. Process below, start the difference occurred with i=0 processes, search the label of one section of each cluster centre point in this section of k-means cluster, and subpoint identical for label in this section of k-means cluster is labeled as the label of corresponding upper one section of this cluster centre point, achieve classification transmission and stdn like this. Finally, corresponding laser spots is given the label of each subpoint cluster. In Fig. 4 (a), the k-means cluster effect of the 53rd section of some cloud subpoint is as shown in lower part of Fig. 4 (c), wherein has 5 clusters, and the laser spots of the corresponding power line of each cluster; Near the central point of the projected centre point of this section of laser point cloud 5 clusters 5 circles in figure (c).
2. for the some cloud of m section, base conditioning is similar to other section, but difference is: only need to consider the laser spots of this section and upper one section of cluster centre point, it is not necessary to consider the cluster centre point problem of k-means cluster.
Above-mentioned 4 steps achieve the cluster to whole shelves power line laser radar point, the cluster result of Fig. 4 (a) laser point cloud is as shown in Fig. 4 (d), the corresponding power line laser spots of each color, totally 5 clusters, and cluster result are entirely true. Need it is specifically intended that uniquely need the parameter specified to be locally length lsection, its value must be reasonable, should embody the Parallel Signature of same section of each power line, embody the continuity between adjacent segment again, and the length being also greater than fracture is effectively to overcome the negative impact of power line laser spots fracture. In this patent, lsectionValue 2.0m to 10.0m can satisfy the demands.
Further, described step 5 specifically comprises the following steps:
(1) three-dimensional model is determined
As shown in Figure 5, single root power line three-dimensional model that this patent adopts comprises two portions, and first part is the fitting a straight line model of power line laser radar point at XOY plane inner projection point; Second section is parabola model.
(2) straight line model is asked for
The same formula of straight line model in XOY plane (1). After straight line model is determined, ask for the scale factor of each power line laser radar point, shown in (2) or formula (3).
(3) parabola model is asked for
Parabola model have employed formula (8):
Z=a0*s2+a1*s+a2(8)
In formula, a0-a3Being two order polynomial coefficients, s is scale factor. It needs to be noted, the three-dimensional model that this patent proposes is suitable for the three-dimensional modeling of any type power line (comprising: S.C., lightning conducter, the arbitrary division of division wire (bundle), division wire entirety).
(4) combination is asked for straight line model and parabola model carry out power line three-dimensional reconstruction
After the three-dimensional model of power line is determined, the three-dimensional reconstruction of single root power line can be carried out one by one, thus complete. The modeling result of shown in Fig. 4 (a) grade power line, as shown in Fig. 4 (e), has rebuild 5 power lines altogether.
The present invention creatively proposes a kind of straight line and One-place 2-th Order polynomial expression combines single grade of single root power line three-dimensional model, a kind of based on single grade of power line laser radar point cloud clustering method of segmentation projection and k-means cluster, the precision that utilize the new power line three-dimensional model simplifying complexity of Model Reconstruction, improve Model Reconstruction, utilizes new power line laser radar point cloud clustering method to improve the level of automation of single root power line laser radar point identification, robustness, universality and reconstructing three-dimensional model precision. Also it is integrated with that the classical irregular triangulation network generates algorithm, connect component analysis algorithm, minimum enclosed rectangle ask for algorithm, k-means cluster algorithm etc. simultaneously, defines the techniqueflow of the long distance overhead transmission line three-dimensional reconstruction based on airborne laser radar point cloud of complete set.
As shown in Figure 6, to carry out the numerical evaluation flow process of overhead transmission line power line three-dimensional reconstruction as follows in the present invention:
(I) airborne laser radar point cloud and initial electricity two kinds of data sources such as tower position and tracks that overhead transmission line is correctly classified are loaded;
(II) extract accurate electricity tower positional information, electricity tower quantity information, tracks information based on electricity tower laser radar point cloud and initial tracks data, and determine total gear number that transmitting line comprises;
(III) the two-dimensional space scope of a certain shelves in XOY plane and corresponding power line laser radar point is determined based on the accurately electricity data such as tower position and initial tracks;
(IV) single grade of power line laser radar point cloud cluster is realized based on segmentation projection and k-means cluster, to isolate the laser radar point cloud of every root power line;
(V) a certain power line three-dimensional reconstruction is carried out based on the model that linearly combines with two order polynomials.
In order to verify the validity of the power line three-dimensional rebuilding method based on airborne laser radar point cloud, it may also be useful to after the classification in two scape overhead transmission line corridors, airborne laser radar point cloud data carries out power line three-dimensional reconstruction experiment. In two scape data, experimental data one and experimental data two are respectively as shown in Fig. 2 (a) and Fig. 8 (a), and the airborne laser radar system CP-560 by Riegl company obtains. The first order laser device that this system use safety the highest grade, the height of flight contains low latitude to high and medium (30 meters to 3200 meters); Maximum laser transmitting frequency is up to 240,000 hertz. The average dot spacing of two scape experimental datas is 10cm, and the power line of two scape data is the mixed structure of horizontal and vertical permutation and combination.
As shown in Fig. 2 (a), this grade of power line in experimental data one is a part for 500 kilovolts of imperial main tracks, and overall length is about 1096.50m, adds up to 58213 points, comprises 4 power lines, wherein 2 lightning conducters, 2 division wires (4 division). Two sections of discrepancy in elevation of this grade of power line are about 169.41m, but each power line presents parallel state on the whole, and obvious fracture does not occur in power line. As shown in Fig. 8 (a), comprising 24 grades of power lines in experimental data two, overall length is about 10000m, is a part for 500 kilovolts of imperial main tracks; Each shelves comprise 4 power lines, wherein 2 lightning conducters, 2 division wires (4 division). Experimentation, the power line radical n of two data is respectively 6 and 4, and power line width 2b is 100m, and section length a is respectively 200m and 500m.
Experiment is under identical experiment platform, and the configuration of experiment porch: ThinkPadW520 notebook, CPU is Intel Duo i7-2760QM2.4GHz, RAM2.98GB, assembling WindowsXP system.
Experimentation, the power line radical n of two data is respectively 6,4, and power line width 2b is 100.0m, section length lsectionIt is respectively 5.0m, 10.0m. Fig. 7 illustrates the effect that power line three-dimensional reconstruction model cover in Fig. 2 (a) is incorporated in classification point cloud; Fig. 8 (b) illustrates the three-dimensional reconstruction result of power line in Fig. 8 (a).
In the evaluation of three-dimensional reconstruction automatization, robustness, universality, have employed accuracy, complete rate two indices. Accuracy, the complete rate of two scape experiment power line three-dimensional reconstructions are 100%.
Reconstruction accuracy evaluate in, have employed raw power line laser radar points to reconstruction power line subpoint three-dimensional distance as judgement criteria, calculate the average d of both three-dimensional distances respectivelymean, maximum value dmax, minimum value dminThree indexs. The precision statistics of two scape experiment power line three-dimensional reconstructions the results are shown in Table 1.
Table 1:
The average d of two scape experiment power line three-dimensional reconstructionsmeanIt is respectively 2.8749cm, 15.1752cm, all less. Meanwhile, maximum value dmax, minimum value dminAlso similar rule is shown. In addition, the precision of the precision difference of two scape experimental data reconstructed results and two scape airborne laser radar point clouds itself is closely related, little with the method for reconstructing relation of the present invention.
Visible, the overhead transmission line corridor airborne laser radar point cloud power line three-dimensional rebuilding method of patent of the present invention achieves good effect, the three-dimensional reconstruction of S.C., the division broad variety power line such as wire, lightning conducter can be realized, have without the need to manual intervention, rebuild precision height, robustness height, advantage that adaptivity is strong. The method obtains above-mentioned advantage and benefits from following factor:
(1) make use of power department existing electricity tower position and electric power circuit move towards information, avoid the randomness of data organization;
(2) utilize the accurate power line position of electricity tower laser radar point cloud reflection and move towards information, determine the spatial dimension of electric power circuit each grade and corresponding power line laser spots accurately;
(3) most overhead transmission line power line Local approximation line electricity line of force number parallel, same is taken full advantage of and priori that power line width remains unchanged;
(4) in units of the section of local, carry out tissue and the cluster of power line laser radar point, avoid the negative impact of laser radar point cloud power line phenomenon of rupture;
(6) the ripe algorithm such as TIN, connect component analysis, k-means cluster is employed, it is ensured that the robustness of the inventive method, universality.
The above is the preferred embodiment of the present invention, for those skilled in the art, under the premise without departing from the principles of the invention, it is also possible to make some improvements and modifications, and these improvements and modifications are also regarded as protection scope of the present invention.

Claims (7)

1. based on a power line three-dimensional rebuilding method for airborne laser radar point cloud, it is characterized in that, the data of described airborne laser radar point cloud are correctly classified, and described method comprises the following steps:
Step one, load overhead transmission line correctly classify after airborne laser radar point cloud data and initial tracks data;
Step 2, extracts the precise information of electricity tower position, quantity and transmitting line track, and determines total gear number of transmitting line;
Step 3, it is determined that each grade of two-dimensional space scope of transmitting line and power line laser radar point;
Step 4, carries out single grade of power line laser radar point cloud cluster based on segmentation projection and k-means cluster to each grade of power line laser radar point determined, to isolate the laser radar point cloud of every root power line;
Step 5, carries out power line three-dimensional reconstruction;
Described step 4 specifically comprises the following steps: (1) carries out the linear fit of whole shelves power line; (2) length of this grade of power line and the scale factor of each laser radar point is asked for; (3) laser radar point of this grade of power line is carried out segmentation; (4) each section of this grade of power line is carried out segmentation projection, segmentation k-means cluster and classification transmission process between section and section.
2. a kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud according to claim 1, it is characterized in that, in step 2, carry out based on electricity tower laser radar point cloud and initial tracks data extracting accurate electricity tower positional information, electricity tower quantity information and tracks information, and determine total gear number that transmitting line comprises.
3. a kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud according to claim 1, it is characterized in that, in step 3, determine the two-dimensional space scope of each grade of transmitting line in XOY plane and corresponding power line laser radar point based on accurate electricity tower position and tracks data.
4. a kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud according to claim 1, is characterized in that, in step 5, carries out the three-dimensional reconstruction of each power line in conjunction with straight line model and parabola model.
5. a kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud according to the arbitrary item of Claims 1-4, is characterized in that, described step 2 specifically comprises the following steps: (1) prepares the data of input; (2) irregular triangulation network TIN is constructed; (3) the long limit in TIN is deleted; (4) connect component analysis based on TIN is carried out; (5) minimum enclosed rectangle and the geometric center point thereof of each Connected component is asked for; (6) electric power circuit trace information renewal is carried out.
6. a kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud according to the arbitrary item of Claims 1-4, it is characterized in that, described step 3 specifically comprises the following steps: (1) obtains two adjacent electric tower information in the overhead transmission line that each shelves relate to; (2) ask for each geometric center point and with corresponding geometric center point belonging to the parallel straight line of the longest edge of minimum enclosed rectangle; (3) the extension straight-line segment of each electric tower is determined; (4) spatial dimension of each grade of power line is determined; (5) the power line laser radar point of each grade is determined.
7. a kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud according to the arbitrary item of Claims 1-4, is characterized in that, described step 5 specifically comprises the following steps: (1) sets up power line three-dimensional model; (2) straight line model is asked for; (3) parabola model is asked for; (4) combination is asked for straight line model and parabola model carry out power line three-dimensional reconstruction.
CN201510145070.0A 2015-03-30 2015-03-30 A kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud Expired - Fee Related CN104732588B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510145070.0A CN104732588B (en) 2015-03-30 2015-03-30 A kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510145070.0A CN104732588B (en) 2015-03-30 2015-03-30 A kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud

Publications (2)

Publication Number Publication Date
CN104732588A CN104732588A (en) 2015-06-24
CN104732588B true CN104732588B (en) 2016-06-01

Family

ID=53456456

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510145070.0A Expired - Fee Related CN104732588B (en) 2015-03-30 2015-03-30 A kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud

Country Status (1)

Country Link
CN (1) CN104732588B (en)

Families Citing this family (27)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105488847B (en) * 2015-12-11 2018-03-02 中国测绘科学研究院 The single grade of single power wire three-dimensional rebuilding method based on airborne laser radar point cloud
CN108780154B (en) * 2016-03-14 2023-06-09 亿目朗欧洲股份有限公司 3D point cloud processing method
CN106157361A (en) * 2016-05-31 2016-11-23 中国科学院遥感与数字地球研究所 A kind of multiple fission conductor full-automatic three-dimensional method for reconstructing based on LiDAR point cloud
CN106772424B (en) * 2016-11-22 2019-04-19 中国南方电网有限责任公司超高压输电公司检修试验中心 A kind of friendship cross-line determines method and apparatus
CN106772425B (en) * 2016-11-25 2019-08-20 北京拓维思科技有限公司 Data processing method and device
CN107329129B (en) * 2017-05-18 2019-12-31 陕西长岭电子科技有限责任公司 Method for fitting high-voltage line by using high-voltage line echo characteristic points
CN107817504B (en) * 2017-10-27 2021-05-28 广东电网有限责任公司机巡作业中心 Airborne laser radar point cloud data processing method
US10962650B2 (en) 2017-10-31 2021-03-30 United States Of America As Represented By The Administrator Of Nasa Polyhedral geofences
CN107958486A (en) * 2017-11-21 2018-04-24 北京煜邦电力技术股份有限公司 A kind of generation method and device of conducting wire vector model
CN108061901A (en) * 2017-12-01 2018-05-22 国家电网公司 The method that 3D electric power line models are rebuild based on airborne laser radar point cloud data
CN108230336B (en) * 2017-12-29 2021-09-24 国网通用航空有限公司 Point cloud tower extraction method and device
CN108734022B (en) * 2018-04-03 2021-07-02 安徽师范大学 Privacy protection track data publishing method based on three-dimensional grid division
CN108680925B (en) * 2018-08-20 2019-09-27 贵州电网有限责任公司 A kind of alternate measurement method of transmission line wire based on laser point cloud data
CN109272560A (en) * 2018-09-07 2019-01-25 北京数字绿土科技有限公司 A kind of power line vectorization method and device
CN109446640A (en) * 2018-10-25 2019-03-08 国网河南省电力公司濮阳供电公司 A kind of transmission line of electricity power line modeling extracting method based on laser point cloud
CN109341667B (en) * 2018-11-29 2020-12-18 武汉汇卓航科技有限公司 Three-dimensional power line measuring method based on region judgment selection
CN110335282B (en) * 2018-12-25 2023-04-18 广州启明星机器人有限公司 Contour line segment feature extraction method based on grids
CN109816029B (en) * 2019-01-30 2023-12-19 重庆邮电大学 High-order clustering division algorithm based on military operation chain
CN109886877B (en) * 2019-02-27 2023-01-13 同济汽车设计研究院有限公司 Method for fitting waypoints and splicing segmented routes
CN110060289B (en) * 2019-04-26 2021-08-17 深圳市镭神智能系统有限公司 Power line extraction method, laser radar system, and storage medium
CN110163882B (en) * 2019-05-15 2019-12-03 贵州电网有限责任公司 A kind of power line hitch point space coordinate extracting method based on laser point cloud data
CN110084819B (en) * 2019-05-15 2023-03-28 贵州电网有限责任公司 Laser point cloud power transmission line segmentation method based on spatial relationship
CN111490489A (en) * 2020-04-16 2020-08-04 北京天下宏图空间信息技术有限公司 Unmanned aerial vehicle trajectory planning method and device for power transmission line inspection
CN111443730B (en) * 2020-05-12 2021-02-12 江苏方天电力技术有限公司 Unmanned aerial vehicle track automatic generation method and device for power transmission line inspection
CN112034481B (en) * 2020-09-02 2024-08-02 亿嘉和科技股份有限公司 Automatic cable identification method based on reflective sticker and laser radar
CN114578374B (en) * 2022-01-26 2024-10-22 国网江苏省电力有限公司南京供电分公司 Real-time monitoring method for power transmission line laser radar point cloud acquisition and external damage prevention
CN115685222B (en) * 2022-11-14 2023-06-23 国网湖北省电力有限公司超高压公司 Automatic power line tower detection method based on laser point cloud data

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103413133A (en) * 2013-06-28 2013-11-27 广东电网公司电力科学研究院 Automatically-extracting power line method in random laser point cloud data
CN104020475A (en) * 2014-06-20 2014-09-03 西安煤航信息产业有限公司 Power line extracting and modeling method based on airborne LiDAR data

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101404655B1 (en) * 2014-04-18 2014-06-09 국방과학연구소 Power line extraction using eigenvalues ratio of 3d raw data of laser radar

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103413133A (en) * 2013-06-28 2013-11-27 广东电网公司电力科学研究院 Automatically-extracting power line method in random laser point cloud data
CN104020475A (en) * 2014-06-20 2014-09-03 西安煤航信息产业有限公司 Power line extracting and modeling method based on airborne LiDAR data

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
余洁等.机载LiDAR点云数据中电力线的提取方法研究.《武汉大学学报&#8226 *
信息科学版》.2011,第36卷(第11期),全文. *

Also Published As

Publication number Publication date
CN104732588A (en) 2015-06-24

Similar Documents

Publication Publication Date Title
CN104732588B (en) A kind of power line three-dimensional rebuilding method based on airborne laser radar point cloud
CN104751481B (en) A kind of single grade of power line LiDAR point cloud clustering method based on stratified random smapling
CN109829199B (en) Power line rapid layered extraction method based on LiDAR point cloud
CA2839165C (en) Method and system for processing image data
CN105488847B (en) The single grade of single power wire three-dimensional rebuilding method based on airborne laser radar point cloud
CN103456192B (en) Terminal area prevailing traffic flow recognizing method based on track spectral clusters
CN109100742A (en) The method for carrying out power-line patrolling based on airborne laser radar
CN109085604A (en) A kind of system and method for power-line patrolling
CN109146836A (en) A kind of system and method carrying out power transmission tower and wireline inspection based on point cloud data
CN111047542B (en) Power line point supplementing method
CN104599023A (en) Typhoon weather transmission line time-variant reliability calculation method and risk evaluation system
CN111929698A (en) Method for identifying hidden danger of tree obstacle in corridor area of power transmission line
CN108663669A (en) Airborne laser radar system for being applied in power-line patrolling
CN102946069A (en) Three-dimensional visualized intelligent monitoring system for extra-high voltage transmission line
CN114972631A (en) Online reconstruction method and system for laser point cloud model of power transmission line
CN109147044A (en) A kind of system that power transmission tower and wireline inspection are carried out based on point cloud data
CN111146866B (en) Power transmission channel control method
CN111062949B (en) Power line point cloud extraction method based on airborne laser radar
CN115641406A (en) Partial overlapping segmentation and clustering method for power line laser point cloud data
CN110969798A (en) Power transmission line forest fire early warning method and system
Guo et al. Research on point cloud power line segmentation and fitting algorithm
Zhou et al. Automatic extraction of power lines from UAV lidar point clouds using a novel spatial feature
CN112630792A (en) Power grid transmission line working condition simulation and dangerous point detection method and detection system
CN117437631A (en) Overhead line point cloud detection method based on catenary face hough transformation
Chi et al. Research on power line segmentation and tree barrier analysis

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20160601

Termination date: 20170330