CN114119606A - Intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring - Google Patents
Intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring Download PDFInfo
- Publication number
- CN114119606A CN114119606A CN202210065172.1A CN202210065172A CN114119606A CN 114119606 A CN114119606 A CN 114119606A CN 202210065172 A CN202210065172 A CN 202210065172A CN 114119606 A CN114119606 A CN 114119606A
- Authority
- CN
- China
- Prior art keywords
- power line
- point
- tree
- image
- hidden danger
- 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.)
- Granted
Links
Images
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C1/00—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people
- G07C1/20—Checking timed patrols, e.g. of watchman
-
- 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
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring, which is characterized in that matching is carried out according to the longitude and latitude of a line tower and the longitude and latitude of an inspection photo, and the inspection photo is automatically classified to the nearest tower; the method has the advantages that the method can improve data processing efficiency and reduce the technical and economic threshold for preventing the hidden danger of the tree obstacle.
Description
Technical Field
The invention belongs to the technical field of power grid inspection, and particularly relates to an intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring.
Background
Overhead transmission line is in the mountain area mostly, and the area is big, trees are many, and dangerous hidden danger has also been many, and the personnel of patrolling the line often meet the tree obstacle problem. The tree barriers have an influence on the safe operation of the line, and the tree barriers are tripped if the tree barriers are not safe, and lead wires are broken or towers collapse if the tree barriers are serious, so that power failure of a large area is caused, and the safety of a power grid is seriously influenced. Traditional tree obstacle search means mainly relies on artifical naked eye to judge or the point cloud data analysis that unmanned aerial vehicle gathered, the naked eye is judged to have very big possibility to cause the judgement mistake, laser point cloud data acquisition equipment is expensive, it is with high costs to patrol, and the collection of visible light point cloud is with low costs, but data analysis is complicated, consequently, it is high to need one set of degree of automation, and easy operation handles convenient tree obstacle hidden danger analytic system, intelligent analysis goes out tree obstacle hidden danger on the basis of the visible light photo that unmanned aerial vehicle patrolled the line and shoots.
CN113537180A discloses a method and an apparatus for identifying tree barriers, a computer device and a storage medium. The method comprises the following steps: acquiring a shot image of a power transmission and distribution corridor; performing pixel-level dense point cloud data recovery on the shot image to obtain target dense three-dimensional point cloud data; the target dense three-dimensional point cloud data comprises chroma characteristic information; performing three-dimensional semantic segmentation on the target dense three-dimensional point cloud data to obtain a three-dimensional semantic segmentation result; the three-dimensional semantic segmentation result comprises a point cloud data set corresponding to each semantic; and determining potential fault points of the power transmission and distribution line tree barriers according to the three-dimensional semantic segmentation result.
CN111929698A discloses a method for identifying hidden danger of tree obstacle in a corridor area of a power transmission line, which comprises the following steps: calculating sag of the power transmission line between adjacent base towers by using a catenary equation and a sag equation for suspended points in the point cloud data of the corridor area of the power transmission line so as to fit each power transmission line between the adjacent base towers; then, classifying the point cloud data to obtain power transmission line point cloud data, base tower point cloud data and vegetation point cloud data; constructing a columnar exploration space with the adjacent base towers, the power transmission lines between the adjacent base towers and the ground as boundaries on the basis of the point cloud data obtained after classification; and finally, obtaining the distance between the vegetation in the columnar exploration space and the power transmission line, and identifying the trees of which the distance between the vegetation and the power transmission line is smaller than a set value. The method can find the hidden danger of the tree obstacle in the power transmission line in time, prevent the trees from threatening the power transmission line, and has higher identification accuracy compared with the traditional manual measurement elimination.
In the method for identifying the tree barriers in the prior art, after the visible light image of the line channel acquired by the unmanned aerial vehicle generates the visible light three-dimensional point cloud, the conducting wire can be extracted only by complex operation, the sag modeling is required, wherein, a great amount of manual operation is needed to be carried out on the pictures for fitting the wires, for the pictures of the towers only the wires can not be seen in the tower interval, the wires at the lowest part of the pictures can be distinguished from the pictures by repeated comparison and analysis, high requirement on the professional of personnel, low intelligent degree of data processing and low analysis efficiency of hidden troubles of tree obstacles because of manual classification of the ground point clouds under the line, and the dense three-dimensional point cloud data is used for three-dimensional semantic segmentation to determine the hidden danger points of the power transmission and distribution line, the accuracy of the result of the three-dimensional semantic segmentation cannot be guaranteed, a large error exists, the accuracy of the analysis result cannot be guaranteed, and the practicability is not high.
Disclosure of Invention
In order to overcome the defects that operation is complex and analysis efficiency of hidden danger of tree obstacles is low in the process of tree obstacles identification after visible light three-dimensional point cloud is generated in the prior art, the invention provides an intelligent hidden danger analysis method based on visible light photo power line coloring, which is used for automatically classifying polling pictures, rapidly identifying and coloring power lines in the polling pictures by using an intelligent identification technology, then generating three-dimensional point cloud, automatically extracting power lines according to power line identification, generating a tower section, automatically classifying point clouds below the power lines according to the generated power lines, rapidly and automatically calculating and analyzing the hidden danger of the power lines, and improving data processing efficiency. The technical and economic threshold for preventing the hidden danger of the tree obstacle is reduced.
In order to solve the technical problem, the invention provides an intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring, which comprises the following steps:
s1, automatic photo classification, namely automatically classifying the collected inspection photos, performing distance matching according to the longitude and latitude of the line tower and the longitude and latitude of the inspection photos, and automatically classifying and binding the collected inspection photos to the nearest line tower;
step S2, coloring power line identification, namely opening inspection photos corresponding to towers, identifying a lowest power line hanging point for each tower, wherein the lowest power line is a power line needing to analyze the hidden danger of the tree barrier, automatically identifying the power line needing to analyze the hidden danger of the tree barrier in each inspection photo according to the identified power line by using an intelligent identification technology, and coloring the power line;
step S3, point cloud generation: converting the routing inspection photo with the marked and colored power line into a corresponding visible light dense three-dimensional point cloud;
step S4, generating a tower section and a power line: generating tower sections by every two adjacent towers according to the sequence of the line towers, generating three-dimensional point clouds after the air-space-three calculation, wherein power lines in the three-dimensional point clouds are colored, dividing three-dimensional space areas through tower areas, wherein the colored point clouds with the height in the areas ranging from tower height/3 to tower height are power line point clouds, and classifying the power lines into the corresponding tower sections;
step S5, point cloud classification: automatically classifying point clouds below the power lines, and identifying all point clouds in an area which is long when the power lines needing to analyze the hidden danger of the tree barrier vertically downwards and has unlimited width on two sides as ground point clouds so as to analyze the hidden danger of the tree barrier in the next step;
step S6, analyzing hidden danger of tree obstacle;
and step S7, one-key export report.
Further preferably, the process of coloring the power line identifier in step S2 is as follows:
s21, preprocessing the inspection photo, converting the color image into gray scale, and binarizing the image to obtain a binary image area Z;
s22, image enhancement: filtering the binary gray level image by using a boundary operator;
s23, marking a lead wire: searching the connected region in the image obtained by filtering in the step S22 to obtain the boundary of the connected region, and initializing the coordinate point p (x) of the marked power line hanging point0，y0) Then, a connected region S surrounding the point is screened out, and all points in the connected region S and coordinate points p (x) are selected0，y0) Finding the vector direction to obtain a set pi (theta) of unit vectors, wherein the unit vectors theta are defined as follows:
in the formula, x and y are the abscissa and the ordinate of the pixel point of the communicated region S;
counting a set pi (theta) of the unit vectors to obtain a mean value point theta ' of the density peak value, wherein theta ' is the direction of the lowermost power line, and pixel points of the unit vectors close to theta ' are all pixel points on the lowermost power line and are represented by a coordinate point p (x)0，y0) Taking theta ' as an initial direction, calculating a current straight line L, taking a pixel point p ' closest to the tail end on the straight line, recording the pixel point of a binary image region Z in the range of 8 adjacent regions of the pixel point p ', and marking the pixel point as m; if m is not empty, the gravity center f of the pixel point in m is solved, and p (x) is calculated0，y0) Replacing the straight line L with the straight line to the gravity center f, if m is empty, calculating the next m, and constructing the previous gravity center f until m is not emptyn-1To the current center of gravity fnN is the number of the gravity center, the pixel points on the line segment are marked as M ', the process is repeated until 5M continuous calculation values are null values, or the process exits when the image edge is reached, the image breakpoints at the dark light positions of the wires are avoided in the binarization process, and M ' is led into a mark set M each time when M ' is obtained;
s24, coloring the conducting wire: and taking all the marks M' in the mark set M, performing morphological dilation operation once to obtain a final mark image, and finally, marking the mark image on the original image to finish the coloring of the image.
Further preferably, in step S21, in the process of converting the color image into the Gray scale Gray, the R, G, B three color channels are split, and then the following psychology formula is used for conversion:
further preferably, in step S21, the image is binarized, and binarization is performed by using a specified threshold value, and the formula of the binarization process is as follows:
wherein maxval is the set final threshold, src (x, y) is the pixel in the original inspection photo, thresh is the set threshold, dst (x, y) is the pixel in the target image, otherwise represents the condition that is less than or equal to thresh.
Further preferably, step S3 includes the following steps
S31: performing matching turning on the inspection photos, wherein the matching turning is to extract and match characteristic points of adjacent inspection photos in the air belt and between the air belts so as to obtain enough homonymous points, and the homonymous points are imaging points of the same point on the ground on different inspection photos;
s32: performing adjustment calculation on the inspection photos, wherein the adjustment calculation is to obtain coordinates of homonymous points according to input rough values of internal parameters and external parameters of the inspection photos and matching transfer points, the internal parameters comprise camera focal length, pixel size and distortion, the external parameters comprise spatial positions and postures of phase image shooting moments, and the optimal internal and external parameters of each inspection photo and a large number of corresponding ground point coordinates of homonymous points are calculated by repeated iterative calculation by utilizing a collinear equation;
s33: processing the optimal internal and external parameters and the coordinates of the homonymous points of the inspection photo, obtaining visible light dense three-dimensional point cloud of the line tower by utilizing a pixel-by-pixel dense matching algorithm of computer vision, integrating the obtained three-dimensional point cloud data to obtain an engineering file, and loading the three-dimensional point cloud to realize visual presentation.
Further preferably, the tree obstacle hidden danger analysis process is as follows: and calculating the power line point cloud generated in the step S4 and the ground point cloud automatically classified in the step S5 to obtain distances from each ground point to the power line, including horizontal distances, vertical distances and clearance distances, setting a line tree safety distance standard according to the power tree obstacle hidden danger classification standard, obtaining tree obstacle defect points with different severity levels, positions of the tree obstacle defect points and distances from towers, previewing positions of the tree obstacle defect points on aerial photographs according to spatial position matching, and displaying the tree obstacle defect points by using circle identification.
Further preferably, the step S7 automatically integrates the information of the tree fault defect point into the report after confirming that the information of the tree fault defect point is correct, and the implementation process includes making a report template, designing and defining statistical variables, commands and inspection pictures according to requirements by using the data source related to the control in the report template and the information of the tree fault defect point analyzed in the step S6, transmitting related data to the control in the template through the data source, filling the statistical variables into the report template by using scripts according to logic operation in the report module to obtain results, and finally calling an office component to render and generate a standard tree fault office report.
Further preferably, the report content comprises tower intervals, small tower distances, defect levels, longitudes, latitudes, horizontal distances, vertical distances, clearance distances, analysts, defect pictures and section views.
The invention supports various unmanned aerial vehicles or human-computer visible light data sources, such as images shot by common visible light cameras carried by unmanned aerial vehicles such as fixed wings and multi-rotor wings, or photos shot by professional measuring cameras carried by human-computer systems such as helicopters and large-scale fixed wings. The method comprises the steps of classifying visible light images according to a pole tower, quickly identifying the power line at the lowest position in a patrol picture by using an intelligent identification technology, coloring, generating visible light point cloud after air-to-three encryption processing, quickly fitting a power line track according to the power line colored by the identification, automatically classifying the ground point cloud below the power line, and accurately calculating and measuring the accurate distance between the power line and the ground. And finally, automatically analyzing the hidden danger and outputting a real-time operating condition tree obstacle danger point analysis report.
The invention can obtain professional precision without human intervention: by automatically calculating the external orientation elements of the original image, automatically calibrating the image and automatically generating the precision report, the quality of the result can be quickly and correctly evaluated. The method provides detailed and quantitative precision of automatic air-ground separation, area network adjustment and ground control points. The whole process is fully automated and has higher accuracy. The automation degree is high, only need simply identify the power line that needs analysis tree obstacle hidden danger, the analysis can be developed automatically to the system, intelligent data analysis, and the operating efficiency is high, saves time.
Drawings
FIG. 1 is a flow chart of the present invention.
Fig. 2 is an imaging schematic diagram of a photograph of a drone.
Detailed Description
The present invention will be explained in further detail with reference to examples.
And step S1, automatically classifying the photos, namely collecting visible light photos on two sides of a route channel by the high-precision positioning unmanned aerial vehicle according to a certain flight rule to serve as inspection photos, wherein the overlapping degree of the inspection photos and the route is more than 85%, the side overlapping degree is more than 35%, and the condition that the operation covers enough power transmission line towers, ground areas and precision is guaranteed. The collected inspection photos are automatically classified, distance matching is carried out according to the longitude and latitude of the line tower and the longitude and latitude of the inspection photos, the collected inspection photos are automatically classified and bound to the nearest line tower, so that the tower section to which the power line in the inspection photos belongs can be distinguished, and the corresponding tower section can be automatically classified during later hidden danger analysis.
And step S2, coloring the power line identification, namely opening the inspection photos corresponding to the towers, identifying the lowest power line hanging point for each tower, identifying the power line needing to analyze the hidden danger of the tree barrier in the lowest power line, automatically identifying the power line needing to analyze the hidden danger of the tree barrier in each inspection photo according to the identified power line by using an intelligent identification technology, and coloring the power line.
S21, preprocessing the inspection photo, converting the color image into gray scale, and binarizing the image to obtain a binary image area Z.
In a general process of converting color image into Gray scale Gray, R, G, B three color channels are split and then converted by using a psychological formula as follows:
because the green part of the inspection photo is mostly the tree background, and the blue light wave band of the wire part is too much, therefore, optimize and improve the psychological formula specially, make the wire region more obvious, the psychological formula after this method improves is:
image binarization, namely performing binarization by adopting a specified threshold value, wherein the formula of the binarization process is as follows:
wherein maxval is the set final threshold, src (x, y) is the pixel in the original inspection photo, thresh is the set threshold, dst (x, y) is the pixel in the target image, otherwise represents the condition that is less than or equal to thresh.
And S22, enhancing the image. Filtering the binarized gray level image by using a boundary operator, wherein the operator comprises the following steps:
and S23, marking a lead. Searching the connected region in the image obtained by filtering in the step S22 to obtain the boundary of the connected region, and initializing the coordinate point p (x) of the marked power line hanging point0，y0) Then, a connected region S surrounding the point is screened out, and all points in the connected region S and coordinate points p (x) are selected0，y0) Finding the vector direction to obtain a set pi (theta) of unit vectors, wherein the unit vectors theta are defined as follows:
in the formula, x and y are the abscissa and the ordinate of the pixel point of the communicated region S;
counting a set pi (theta) of the unit vectors to obtain a mean value point theta ' of the density peak value, wherein theta ' is the direction of the lowermost power line, and pixel points of the unit vectors close to theta ' are all pixel points on the lowermost power line and are represented by a coordinate point p (x)0，y0) Taking theta ' as an initial pixel point, calculating a current straight line L, taking a pixel point p ' closest to the tail end on the straight line, recording pixel points of a binary image region Z in the region range of 8 adjacent (closest 8 pixel points) regions of the pixel point p ', and marking the pixel points as m; if m is not empty, the gravity center f of the pixel point in m is solved, and p (x) is calculated0，y0) The straight line to the center of gravity f replaces the straight line L, if m is empty, the next m is calculated (at most 5 times), and until m is not empty, the last center of gravity f is constructedn-1To the current center of gravity fnAnd (3) marking pixel points on the line segment as M ' by taking n as the serial number of the gravity center, repeating the process until 5M continuous calculated values are null values or quitting when the image edge is reached, wherein the image breakpoint at the dark light position of the wire in the binarization process is avoided, and when M ' is obtained every time, introducing M ' into the mark set M.
And S24, coloring the conducting wire. And taking all the marks M' in the mark set M, performing morphological dilation operation once to obtain a final mark image, and finally, marking the mark image on the original image to finish the coloring of the image.
And step S3, generating a point cloud. Converting a patrol photo with a colored mark on a power line into a corresponding visible light dense three-dimensional point cloud, comprising the following steps
S31: performing matching turning on the inspection photos, wherein the matching turning is to extract and match characteristic points of adjacent inspection photos in the air belt and between the air belts so as to obtain enough homonymous points, and the homonymous points are imaging points of the same point on the ground on different inspection photos;
s32: performing adjustment calculation on the inspection photos, wherein the adjustment calculation is to obtain coordinates of homonymous points according to input rough values of internal parameters and external parameters of the inspection photos and matching transfer points, the internal parameters comprise camera focal length, pixel size and distortion, the external parameters comprise spatial positions and postures of phase image shooting moments, and the optimal internal and external parameters of each inspection photo and a large number of corresponding ground point coordinates of homonymous points are calculated by repeated iterative calculation by utilizing a collinear equation;
s33: the optimal internal and external parameters and the coordinates of the homonymous points of the inspection photo are processed, the pixel-by-pixel dense matching algorithm of computer vision is utilized, the visible light dense three-dimensional point cloud of the line tower can be obtained, the colored marks of the power lines needing to analyze the hidden danger of the tree barrier in the three-dimensional point cloud are obtained, the obtained three-dimensional point cloud data are integrated to obtain an engineering file, the three-dimensional point cloud is loaded to realize visual presentation, and operations such as plane and section view viewing, plane and three-dimensional visual angle viewing, zooming, rotation and the like are provided.
Step S4, generating a tower section and a power line: according to the sequence of the line towers, generating tower sections by every two adjacent towers, generating three-dimensional point clouds after the air-to-three calculation, coloring power lines in the three-dimensional point clouds, dividing three-dimensional space areas through tower areas, wherein the colored point clouds with the height in the area within the range of [ tower height/3, tower height ] are the power line point clouds, and the power lines are classified into the corresponding tower sections.
Step S5, point cloud classification: the method comprises the steps of automatically classifying point clouds below a power line, wherein the power line needing to analyze the hidden danger of the tree obstacle is vertically downwards long, and all point cloud marks in the region with unlimited width on two sides are ground point clouds, so that the hidden danger of the power line can be analyzed on the next step.
Step S6, analyzing the hidden danger of the tree obstacle: the power line point cloud generated by S4 and the ground point cloud automatically classified by S5 are calculated together to obtain the distance from each ground point to the power line, the distance comprises a horizontal distance, a vertical distance and a clearance distance, a line tree safety distance standard (different voltage levels and corresponding safety distances can be configured) is set according to the power tree obstacle hidden danger classification standard, tree obstacle defect points with different severity levels are obtained, the positions of the tree obstacle defect points and the distance from a tower are obtained, the positions of the tree obstacle defect points on aerial photographs can be matched and previewed according to the spatial positions, and the tree obstacle defect points are displayed by using circle identification, wherein the coordinate conversion matching steps of the spatial coordinate positions of the point cloud object and the pixels of the aerial photographs are as follows:
s61, as shown in fig. 2, where W represents the aerial camera of the drone, a represents a certain point on the ground, and a represents an image of a on the aerial camera. According to the imaging principle of unmanned aerial vehicle photo, ground object passes through on light projection to the image, satisfies central projection principle. Therefore, the coordinate conversion relationship from the ground object coordinate to the image pixel can be described by the collinear equation as follows:
the variable lambda is a collinear scaling coefficient, the resolutions of the images are different, the scaling coefficient needs to be added, i and j are collinear linear quantities and are coordinates on image pixels, X, Y and Z are lower positions of an actual ground space, P is a projection matrix, and the P matrix is composed of internal parameters and external parameters of the images. The intrinsic parameters refer to the focal length, the pixel size and the distortion of the camera; the external parameters refer to the spatial position and the attitude of the phase image at the time of shooting.
And S62, after coordinate conversion, matching the coordinate position of the pixel of the corresponding inspection photo based on the space position of the fault point of the tree barrier, displaying the corresponding aerial inspection photo after matching, and displaying the position of the fault point of the matched tree barrier on the photo by using a circle mark.
And step S7, one-key export report. And automatically integrating the information of the defect points into the report after the information of the defect points is confirmed to be correct, wherein the realization process comprises the steps of manufacturing a report template, designing and defining statistical variables, commands, routing inspection pictures and the like according to requirements by a data source related to a control in the report template, transmitting related data to the control in the template through the data source, filling the obtained result into the report template by using a script according to the logic operation in the report module according to the statistics in the report, and finally calling an office component to render and generate a standard tree obstacle office report. The report content is graphically and contented, the defect number is counted according to pole tower intervals, the critical, serious, general and attention point defect numbers are counted according to the defect levels, and tree obstacle defect information is described in detail item by item, wherein the tree obstacle defect information comprises information such as pole tower intervals, tower distances from small towers, defect levels, longitudes, latitudes, horizontal distances, vertical distances, clearance distances, analysts, defect pictures, profile diagrams and the like, defect records can be sorted from small to large according to line pole tower intervals, and the same pole tower intervals are sorted according to the defect levels.
Claims (8)
1. An intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring is characterized by comprising the following steps:
s1, automatic photo classification, namely automatically classifying the collected inspection photos, performing distance matching according to the longitude and latitude of the line tower and the longitude and latitude of the inspection photos, and automatically classifying and binding the collected inspection photos to the nearest line tower;
step S2, coloring power line identification, namely opening inspection photos corresponding to towers, identifying a lowest power line hanging point for each tower, wherein the lowest power line is a power line needing to analyze the hidden danger of the tree barrier, automatically identifying the power line needing to analyze the hidden danger of the tree barrier in each inspection photo according to the identified power line by using an intelligent identification technology, and coloring the power line;
step S3, point cloud generation: converting the routing inspection photo with the marked and colored power line into a corresponding visible light dense three-dimensional point cloud;
step S4, generating a tower section and a power line: generating tower sections by every two adjacent towers according to the sequence of the line towers, generating three-dimensional point clouds after the air-space-three calculation, wherein power lines in the three-dimensional point clouds are colored, dividing three-dimensional space areas through tower areas, wherein the colored point clouds with the height in the areas ranging from tower height/3 to tower height are power line point clouds, and classifying the power lines into the corresponding tower sections;
step S5, point cloud classification: automatically classifying point clouds below the power lines, and identifying all point clouds in an area which is long when the power lines needing to analyze the hidden danger of the tree barrier vertically downwards and has unlimited width on two sides as ground point clouds so as to analyze the hidden danger of the tree barrier in the next step;
step S6, analyzing hidden danger of tree obstacle;
and step S7, one-key export report.
2. The intelligent obstacle-tree potential analysis method based on visible light photo power line coloring as claimed in claim 1, wherein the power line identifier coloring process in step S2 is as follows:
s21, preprocessing the inspection photo, converting the color image into gray scale, and binarizing the image to obtain a binary image area Z;
s22, image enhancement: filtering the binary gray level image by using a boundary operator;
s23, marking a lead wire: searching the connected region in the image obtained by filtering in the step S22 to obtain the boundary of the connected region, and initializing the coordinate point p (x) of the marked power line hanging point0，y0) Then, a connected region S surrounding the point is screened out, and all points in the connected region S and coordinate points p (x) are selected0，y0) Finding the vector direction to obtain a set pi (theta) of unit vectors, wherein the unit vectors theta are defined as follows:
in the formula, x and y are the abscissa and the ordinate of the pixel point in the communicated region S;
counting a set pi (theta) of the unit vectors to obtain a mean value point theta ' of the density peak value, wherein theta ' is the direction of the lowermost power line, and pixel points of the unit vectors close to theta ' are all pixel points on the lowermost power line and are represented by a coordinate point p (x)0，y0) Taking theta ' as an initial direction, calculating a current straight line L, taking a pixel point p ' closest to the tail end on the straight line, recording the pixel point of a binary image region Z in the range of 8 adjacent regions of the pixel point p ', and marking the pixel point as m; if m is not empty, the gravity center f of the pixel point in m is solved, and p (x) is calculated0，y0) Replacing the straight line L with the straight line to the gravity center f, if m is empty, calculating the next m, and constructing the previous gravity center f until m is not emptyn-1To the current center of gravity fnN is the number of the center of gravity, and images on the line segmentsMarking the prime point as M ', repeating the process until 5M continuous null values appear in the calculation or quitting when the image edge is reached, wherein the image breakpoint appears at the dark light position of the conducting wire in the binarization process is avoided, and when M ' is obtained each time, introducing M ' into a mark set M;
s24, coloring the conducting wire: and taking all the marks M' in the mark set M, performing morphological dilation operation once to obtain a final mark image, and finally, marking the mark image on the original image to finish the coloring of the image.
3. The intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring as claimed in claim 2, wherein in step S21, in the process of converting the color image into Gray level Gray, the R, G, B three color channels are split, and then the following psychological formula is used for conversion:
4. the intelligent tree-obstacle hidden danger analysis method based on visible light photo power line coloring as claimed in claim 2, wherein in step S21, the image is binarized, a specified threshold value is used for binarization, and a formula of a binarization process is as follows:
wherein maxval is the set final threshold, src (x, y) is the pixel in the original inspection photo, thresh is the set threshold, dst (x, y) is the pixel in the target image, otherwise represents the condition that is less than or equal to thresh.
5. The intelligent obstacle-tree potential analysis method based on visible light photo power line coloring as claimed in claim 1, wherein the step S3 includes the following steps
S31: performing matching turning on the inspection photos, wherein the matching turning is to extract and match characteristic points of adjacent inspection photos in the air belt and between the air belts so as to obtain enough homonymous points, and the homonymous points are imaging points of the same point on the ground on different inspection photos;
s32: performing adjustment calculation on the inspection photos, wherein the adjustment calculation is to obtain coordinates of homonymous points according to input rough values of internal parameters and external parameters of the inspection photos and matching transfer points, the internal parameters comprise camera focal length, pixel size and distortion, the external parameters comprise spatial positions and postures of phase image shooting moments, and the optimal internal and external parameters of each inspection photo and a large number of corresponding ground point coordinates of homonymous points are calculated by repeated iterative calculation by utilizing a collinear equation;
s33: processing the optimal internal and external parameters and the coordinates of the homonymous points of the inspection photo, obtaining visible light dense three-dimensional point cloud of the line tower by utilizing a pixel-by-pixel dense matching algorithm of computer vision, integrating the obtained three-dimensional point cloud data to obtain an engineering file, and loading the three-dimensional point cloud to realize visual presentation.
6. The intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring as claimed in claim 1, wherein the tree obstacle hidden danger analysis process is as follows: and calculating the power line point cloud generated in the step S4 and the ground point cloud automatically classified in the step S5 to obtain distances from each ground point to the power line, including horizontal distances, vertical distances and clearance distances, setting a line tree safety distance standard according to the power tree obstacle hidden danger classification standard, obtaining tree obstacle defect points with different severity levels, positions of the tree obstacle defect points and distances from towers, previewing positions of the tree obstacle defect points on aerial photographs according to spatial position matching, and displaying the tree obstacle defect points by using circle identification.
7. The intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring as claimed in claim 1, wherein the step S7 is implemented by automatically integrating the information of the tree obstacle defect point into the report after confirming that the information is correct, and the implementation process is that a report template is made, the data source related to the control in the report template is derived from the information of the tree obstacle defect point analyzed in the step S6, statistical variables, commands and inspection pictures are designed and defined as required, the data source transmits related data to the control in the template, statistics in the report are filled in the report template by using a script according to logic operation in the report module, and finally an office component is called to render and generate a standard tree obstacle office report.
8. The intelligent obstacle-tree potential analysis method based on visible light photo power line coloring as claimed in claim 7, wherein the report content comprises tower intervals, tower distance to small size, defect level, longitude, latitude, horizontal distance, vertical distance, clearance distance, analysts, defect photo, and profile.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210065172.1A CN114119606B (en) | 2022-01-20 | 2022-01-20 | Intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210065172.1A CN114119606B (en) | 2022-01-20 | 2022-01-20 | Intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114119606A true CN114119606A (en) | 2022-03-01 |
| CN114119606B CN114119606B (en) | 2022-05-10 |
Family
ID=80360933
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210065172.1A Active CN114119606B (en) | 2022-01-20 | 2022-01-20 | Intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114119606B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115063442A (en) * | 2022-06-30 | 2022-09-16 | 山东信通电子股份有限公司 | Power transmission line hidden danger target tracking method, equipment and medium |
| CN117132915A (en) * | 2023-10-27 | 2023-11-28 | 国网江西省电力有限公司电力科学研究院 | Power transmission line tree obstacle hidden danger analysis method based on automatic classification of point cloud |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779170A (en) * | 2016-11-22 | 2017-05-31 | 国网安徽省电力公司 | A kind of method and device of automatic discrimination passage screen of trees hidden danger |
| CN107103599A (en) * | 2017-04-06 | 2017-08-29 | 云南电网有限责任公司电力科学研究院 | A kind of transmission line of electricity trees hidden danger prediction analysis method based on LiDAR |
| WO2018028103A1 (en) * | 2016-08-11 | 2018-02-15 | 河海大学常州校区 | Unmanned aerial vehicle power line inspection method based on characteristics of human vision |
| CN108761271A (en) * | 2018-03-30 | 2018-11-06 | 广州中科云图智能科技有限公司 | A kind of power grid screen of trees detection method and system |
| CN110595442A (en) * | 2019-08-13 | 2019-12-20 | 中国南方电网有限责任公司超高压输电公司昆明局 | Transmission line channel tree obstacle detection method, storage medium and computer equipment |
| CN111985496A (en) * | 2020-07-13 | 2020-11-24 | 南方电网数字电网研究院有限公司 | Tree barrier hidden danger rapid detection method based on vegetation three-dimensional convex hull and terminal |
| CN112085036A (en) * | 2020-09-17 | 2020-12-15 | 广东电网有限责任公司 | Region identification method, device, equipment and medium |
| CN112184903A (en) * | 2020-09-25 | 2021-01-05 | 广东电网有限责任公司韶关供电局 | Method, device, equipment and medium for detecting high-voltage line tree obstacle risk points |
| US20210090263A1 (en) * | 2019-09-24 | 2021-03-25 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for detecting ground point cloud points |
| CN113222914A (en) * | 2021-04-27 | 2021-08-06 | 南方电网数字电网研究院有限公司 | Registration point cloud-based tree obstacle hidden danger rapid detection method |
-
2022
- 2022-01-20 CN CN202210065172.1A patent/CN114119606B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018028103A1 (en) * | 2016-08-11 | 2018-02-15 | 河海大学常州校区 | Unmanned aerial vehicle power line inspection method based on characteristics of human vision |
| CN106779170A (en) * | 2016-11-22 | 2017-05-31 | 国网安徽省电力公司 | A kind of method and device of automatic discrimination passage screen of trees hidden danger |
| CN107103599A (en) * | 2017-04-06 | 2017-08-29 | 云南电网有限责任公司电力科学研究院 | A kind of transmission line of electricity trees hidden danger prediction analysis method based on LiDAR |
| CN108761271A (en) * | 2018-03-30 | 2018-11-06 | 广州中科云图智能科技有限公司 | A kind of power grid screen of trees detection method and system |
| CN110595442A (en) * | 2019-08-13 | 2019-12-20 | 中国南方电网有限责任公司超高压输电公司昆明局 | Transmission line channel tree obstacle detection method, storage medium and computer equipment |
| US20210090263A1 (en) * | 2019-09-24 | 2021-03-25 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for detecting ground point cloud points |
| CN111985496A (en) * | 2020-07-13 | 2020-11-24 | 南方电网数字电网研究院有限公司 | Tree barrier hidden danger rapid detection method based on vegetation three-dimensional convex hull and terminal |
| CN112085036A (en) * | 2020-09-17 | 2020-12-15 | 广东电网有限责任公司 | Region identification method, device, equipment and medium |
| CN112184903A (en) * | 2020-09-25 | 2021-01-05 | 广东电网有限责任公司韶关供电局 | Method, device, equipment and medium for detecting high-voltage line tree obstacle risk points |
| CN113222914A (en) * | 2021-04-27 | 2021-08-06 | 南方电网数字电网研究院有限公司 | Registration point cloud-based tree obstacle hidden danger rapid detection method |
Non-Patent Citations (5)
| Title |
|---|
| ZHAOWEI MA: "UAV low-altitude obstacle detection based on the fusion of LiDAR and camera", 《AUTONOMOUS INTELLIGENT SYSTEMS》 * |
| 吴锦秋等: "可见光影像与激光雷达点云融合技术在配网树障巡检中的应用", 《湖南电力》 * |
| 曾忱等: "基于无人机可见光点云的输电线路树障隐患智能分析研究", 《电网运维》 * |
| 朱从军: "无人机输电线路巡检树障隐患自动分析技术研究", 《中国优秀硕士学位论文全文数据库》 * |
| 陈驰等: "大型无人机电力巡检 LiDAR 点云安全距离诊断方法", 《电网技术》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115063442A (en) * | 2022-06-30 | 2022-09-16 | 山东信通电子股份有限公司 | Power transmission line hidden danger target tracking method, equipment and medium |
| CN115063442B (en) * | 2022-06-30 | 2024-05-31 | 山东信通电子股份有限公司 | Method, equipment and medium for tracking hidden danger targets of power transmission line |
| CN117132915A (en) * | 2023-10-27 | 2023-11-28 | 国网江西省电力有限公司电力科学研究院 | Power transmission line tree obstacle hidden danger analysis method based on automatic classification of point cloud |
| CN117132915B (en) * | 2023-10-27 | 2024-03-12 | 国网江西省电力有限公司电力科学研究院 | Power transmission line tree obstacle hidden danger analysis method based on automatic classification of point cloud |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114119606B (en) | 2022-05-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114119606B (en) | Intelligent tree obstacle hidden danger analysis method based on visible light photo power line coloring | |
| CN112633535B (en) | Intelligent inspection method and system for photovoltaic power station based on unmanned aerial vehicle image | |
| CN108037770B (en) | Unmanned aerial vehicle power transmission line inspection system and method based on artificial intelligence | |
| CN108694386B (en) | Lane line detection method based on parallel convolution neural network | |
| CN106951836B (en) | crop coverage extraction method based on prior threshold optimization convolutional neural network | |
| CN108734143A (en) | A kind of transmission line of electricity online test method based on binocular vision of crusing robot | |
| CN112801230B (en) | Intelligent acceptance method for unmanned aerial vehicle of power distribution line | |
| CN115439424A (en) | Intelligent detection method for aerial video image of unmanned aerial vehicle | |
| CN112327906A (en) | Intelligent automatic inspection system based on unmanned aerial vehicle | |
| CN115240093B (en) | Automatic power transmission channel inspection method based on visible light and laser radar point cloud fusion | |
| CN114240868A (en) | Unmanned aerial vehicle-based inspection analysis system and method | |
| CN107564111A (en) | Power line space safety analysis method based on computer vision | |
| CN111931559A (en) | Method for classifying tree species in corridor area of power transmission line | |
| CN113673444A (en) | Intersection multi-view target detection method and system based on angular point pooling | |
| CN112163639B (en) | Crop lodging grading method based on height distribution feature vector | |
| CN115240087A (en) | Tree barrier positioning analysis method and system based on binocular stereo vision and laser point cloud | |
| CN109000621A (en) | A kind of power grid screen of trees detection method | |
| CN115841633A (en) | Power tower and power line associated correction power tower and power line detection method | |
| CN113378754A (en) | Construction site bare soil monitoring method | |
| CN117406789A (en) | Automatic planning method for multi-unmanned aerial vehicle bridge support inspection route based on image analysis | |
| CN109902607A (en) | A kind of city Automatic Optimal modeling based on inclined camera | |
| CN112016518A (en) | Crowd distribution form detection method based on unmanned aerial vehicle and artificial intelligence | |
| CN115755088A (en) | Laser point cloud-based automatic measurement method for power transmission line engineering construction parameters | |
| CN112287787B (en) | Crop lodging grading method based on gradient histogram characteristics | |
| CN114924260A (en) | Multi-laser radar point cloud registration method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |