CN115562348A - Unmanned aerial vehicle image technology method based on transformer substation - Google Patents
Unmanned aerial vehicle image technology method based on transformer substation Download PDFInfo
- Publication number
- CN115562348A CN115562348A CN202211371013.0A CN202211371013A CN115562348A CN 115562348 A CN115562348 A CN 115562348A CN 202211371013 A CN202211371013 A CN 202211371013A CN 115562348 A CN115562348 A CN 115562348A
- Authority
- CN
- China
- Prior art keywords
- aerial vehicle
- unmanned aerial
- image
- transformer substation
- researching
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000005516 engineering process Methods 0.000 title claims abstract description 49
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 31
- 238000007689 inspection Methods 0.000 claims abstract description 11
- 238000011160 research Methods 0.000 claims abstract description 8
- RZVHIXYEVGDQDX-UHFFFAOYSA-N 9,10-anthraquinone Chemical compound C1=CC=C2C(=O)C3=CC=CC=C3C(=O)C2=C1 RZVHIXYEVGDQDX-UHFFFAOYSA-N 0.000 claims abstract description 4
- 238000001514 detection method Methods 0.000 claims description 41
- JEIPFZHSYJVQDO-UHFFFAOYSA-N iron(III) oxide Inorganic materials O=[Fe]O[Fe]=O JEIPFZHSYJVQDO-UHFFFAOYSA-N 0.000 claims description 25
- 239000012212 insulator Substances 0.000 claims description 23
- 230000007547 defect Effects 0.000 claims description 19
- 235000005770 birds nest Nutrition 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 15
- 235000005765 wild carrot Nutrition 0.000 claims description 15
- 238000013135 deep learning Methods 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 9
- 238000010586 diagram Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 8
- 230000005540 biological transmission Effects 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 6
- 238000013136 deep learning model Methods 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000002372 labelling Methods 0.000 claims description 5
- 238000013527 convolutional neural network Methods 0.000 claims description 4
- 238000011161 development Methods 0.000 claims description 4
- 230000018109 developmental process Effects 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 238000005259 measurement Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 244000000626 Daucus carota Species 0.000 claims description 3
- 206010047571 Visual impairment Diseases 0.000 claims description 3
- 230000003321 amplification Effects 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 2
- 230000007797 corrosion Effects 0.000 description 7
- 238000005260 corrosion Methods 0.000 description 7
- 230000004888 barrier function Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000003709 image segmentation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- DMBHHRLKUKUOEG-UHFFFAOYSA-N diphenylamine Chemical compound C=1C=CC=CC=1NC1=CC=CC=C1 DMBHHRLKUKUOEG-UHFFFAOYSA-N 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
- G05D1/106—Change initiated in response to external conditions, e.g. avoidance of elevated terrain or of no-fly zones
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/12—Details of acquisition arrangements; Constructional details thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/17—Terrestrial scenes taken from planes or by drones
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/176—Urban or other man-made structures
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Remote Sensing (AREA)
- Water Supply & Treatment (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Public Health (AREA)
- Automation & Control Theory (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an unmanned aerial vehicle image technical method based on a transformer substation, which comprises the following research steps: the method comprises the steps of S1, comprehensively knowing an unmanned aerial vehicle positioning applicable technology, an applicable model, a route planning and anti-interference measure in the transformer substation, and S2, researching unmanned aerial vehicle flight control and a cloud deck adjustment algorithm to achieve effective collection of transformer substation inspection images.
Description
Technical Field
The invention relates to the technical field of transformer substation inspection, in particular to an unmanned aerial vehicle image technical method based on a transformer substation.
Background
The transformer substation is a place for converting voltage and current, receiving electric energy and distributing electric energy in an electric power system, the transformer substation in the power plant is a boosting transformer substation, the boosting transformer substation is used for boosting the electric energy generated by a generator and feeding the electric energy into a high-voltage power grid, in order to master the real-time condition of the transformer substation, an unmanned aerial vehicle is usually used for routing inspection, the identification of the transformer substation on obstacles, hidden dangers and the like is greatly improved due to the routing inspection of the unmanned aerial vehicle, and great guarantee is provided for power transmission;
but when unmanned aerial vehicle image acquisition of present transformer substation, can not accurate avoiding to some less obstacles, hidden danger image recognition is limited, the hidden danger of discernment transformer substation that can not be more accurate.
Disclosure of Invention
The invention provides an unmanned aerial vehicle image technical method based on a transformer substation, which can effectively solve the problems that when the unmanned aerial vehicle image of the transformer substation is acquired in the prior art, some smaller obstacles cannot be accurately avoided, the hidden danger image identification is limited, and the hidden danger of the transformer substation cannot be accurately identified.
In order to achieve the purpose, the invention provides the following technical scheme: the unmanned aerial vehicle image technology method based on the transformer substation comprises the following research steps:
s1, comprehensively knowing an unmanned aerial vehicle positioning application technology, an application model, a route planning and anti-interference measures in a transformer substation;
s2, researching a flight control and holder adjustment algorithm of the unmanned aerial vehicle, realizing effective acquisition of a transformer substation inspection image, and optimizing the image proportion and the image position distribution of a target to be detected;
s3, researching visible light data sets of bird' S nest, rust, insulator chipping and small hardware defect hidden dangers of the transformer substation, and researching standardized labeling, preprocessing and data amplification of data;
s4, researching visible light and infrared image defect identification and detection technologies to realize hidden danger target identification and detection based on a deep learning algorithm;
the S2 specifically comprises the following steps:
a1, researching an obstacle avoidance technology of the unmanned aerial vehicle, comprehensively adopting RTK positioning, UWB and binocular vision obstacle avoidance technologies to complete obstacle avoidance in the inspection process of the unmanned aerial vehicle, and ensuring reliable flight of the unmanned aerial vehicle;
a2, researching an automatic image focusing technology of an aerial camera, and effectively acquiring key equipment of a transformer substation by carrying out secondary development on an aerial cloud deck to optimize the distribution of targets in an image domain;
the UWB technology in the A1 utilizes nanosecond-picosecond-level non-sine wave narrow pulse to transmit data, a UWB positioning algorithm mainly comprises two parts, firstly, a TOF ranging algorithm is adopted to obtain the distance between a label and a fixed base station, and then, a label coordinate is obtained according to a trilateral positioning algorithm;
assuming three base stations with physical coordinates of (x _1, y _1), (x _2, y _2), and (x _3, y _3), and tag-to-base station distances of d _1, d _2, d _3, then:
(x-x_1)^2+(y-y_1)^2＝〖d_1〗^2
(x-x_2)^2+(y-y_2)^2＝〖d_2〗^2
(x-x_3)^2+(y-y_3)^2＝〖d_3〗^2
so as to solve the real-time coordinate;
the method comprises the steps that two cameras are symmetrically fixed on an unmanned aerial vehicle, the positions of objects are determined by using a binocular obstacle avoidance technology, then the angles of the two shooting positions are respectively determined, the distance measurement is carried out through the two shooting positions and the position coordinates of the objects, after images are shot and obtained, the cameras calibrate parameters, the two images are corrected, the three-dimensional information is obtained through characteristic point extraction and three-dimensional matching, the parallax is calculated, and the parallax is converted into the depth, so that the flight angle is adjusted and obstacles are avoided when the unmanned aerial vehicle flies;
the position connecting line of two shooting angles A and B is used as an X axis, a perpendicular line passing through an obstacle P is used as a Y axis to establish a rectangular coordinate system, OP is the original flight direction of the unmanned aerial vehicle, L and R points are position points of two cameras on the unmanned aerial vehicle, and two end points of the obstacle P are respectively P 1 And P 2 Represents alpha;
wherein, the coordinate of the point P is (0, y 3);
P 1 the coordinates of the point are (x 31, y 3);
P 2 the coordinates of the point are (x 32, y 3);
the coordinates of the point L are (x 1, y 1);
the coordinates of the R point are (x 2, y 2);
when the obstacle is avoided, the calculation formula of the adjusted angle is as follows:
wherein h is 1 The distance between the point P1 and the point P2 is calculated according to the formula:
h 1 ＝x 32 -x 31 (2)；
in the formula, h2 is the width of the unmanned aerial vehicle;
z is depth, and the calculation formula is as follows:
wherein f is the focal length;
d is parallax, and the calculation formula is as follows:
d＝x 2 -x 1 (4)；
and (3) calculating to adjust the flight direction according to the deviation angle when the unmanned aerial vehicle avoids the obstacle by combining the formulas (1), (2), (3) and (4).
According to the above technical solution, in S3, the research of the data set specifically includes the following steps:
b1, establishing an aerial photography data set of the unmanned aerial vehicle in a complex scene of the transformer substation, researching and capturing a network public data set from the Internet by using a script, screening images according to defect types and picture arrangement quality, communicating with provincial and municipal electric academy departments and power supply company departments to obtain aerial photography images, and simultaneously automatically shooting a common data set of the transformer substation;
b2, researching manual standardized frame selection labeling of image defects, and performing manual standardized labeling on the established data set by adopting labelImg software;
and B3, researching a data sample enhancement method, and comprehensively adopting image rotation, saturation brightness contrast adjustment and white noise adding technologies to realize expansion and feature enhancement of a small sample data set.
According to the technical scheme, in the S4, the hidden danger targets comprise rust, a nest, an insulator falling piece and a small hardware fitting;
the specific steps of identification and detection are as follows:
c1, researching deep learning target identification and positioning based on a convolutional neural network, and optimizing by adopting a YOLOv5 network model in combination with model identification and positioning effects, hardware computing power and model running speed;
c2, researching a training model built by adopting a Pythrch deep learning frame, and carrying out target identification detection aiming at the potential hazards of the transformer substation;
and C3, building an edge computing ground station, importing aerial images and infrared images acquired by autonomous flight of the unmanned aerial vehicle into a background system, and performing real-time defect detection.
According to the technical scheme, in the step C2, the transformer substation equipment corrosion detection algorithm mainly comprises power equipment target detection, foreground and background segmentation, gamma conversion based on background brightness and an improved super-red algorithm, and corrosion area detection and calculation are carried out;
the foreground image is responsible for extracting a target area, the background image is responsible for calculating average brightness, the background image and the target area are combined to carry out Gamma conversion normalization on the target image, rust area detection is completed, a rust rate is obtained according to a rust area, operation and maintenance personnel are guided to carry out hardware fitting replacement, and the rust area and the rust degree are visually displayed through a rust thermodynamic diagram;
during the identification of the bird nest, a method based on a RetinaNet deep learning model is adopted, resNet-50 is utilized to extract early-stage features, early-stage standard features are reinforced through an FPN network, a feature pyramid image is constructed to meet the detection of bird nest targets with different sizes, and then a classification subnet and a regression subnet are constructed on the basis of the feature pyramid and are respectively used for identifying the specific positions of the bird nest and the regression bird nest;
when the insulator string chipping detection is carried out, an image is input firstly, a group of feature maps S containing insulator string key point position information is obtained after neural network training, wherein S = (S1, S2, \8230;, sN), each feature map corresponds to a fixed type of key point position information, and finally, N key point feature maps output by a model are used for matching insulator string key points by using a greedy algorithm to output an insulator string framework, so that the insulator string chipping detection is realized.
According to the technical scheme, in the S1, the positioning applicable technology of the unmanned aerial vehicle is a positioning technology combining Beidou positioning and a real-time kinematic differential method (RTK) carrier phase difference division technology;
the applicable model is an octree model, and a complete substation laser point cloud three-dimensional model is established by researching the octree model;
the interference existing in the transformer substation comprises a power frequency magnetic field, pulse electromagnetism and second harmonic existing in the transformer substation;
the route planning firstly researches a rapid expansion random tree RRT algorithm to solve the problem of trajectory planning, and determines a path through collision detection, so that the reasonable planning of the transformer substation inspection trajectory under the complex environment is adapted, and an optimal route is established;
the anti-electromagnetic interference technology mainly aims at positioning, data transmission and image transmission devices of the unmanned aerial vehicle;
the anti-interference measures comprise the steps of adopting channel coding, spread spectrum anti-interference, self-adaptive anti-interference, additionally arranging envelope shielding and reasonably planning a route to avoid a strong magnetic region.
Compared with the prior art, the invention has the following beneficial effects:
1. through before patrolling and examining the transformer substation, through to big dipper location and real-time dynamic difference method, carrier phase difference technique, transformer substation's laser point cloud three-dimensional model, extend random tree RRT algorithm and carry out the orbit planning, interference is resisted with the mode that the frequency hopping fuses mutually in the direct expansion, before unmanned aerial vehicle image acquisition, makes sufficient planning to can avoid unmanned aerial vehicle proruption trouble when gathering the image, promote unmanned aerial vehicle image acquisition's efficiency.
2. The unmanned aerial vehicle monitoring system has the advantages that the obstacle avoidance technology is researched when the unmanned aerial vehicle flies to acquire images, RTK positioning is comprehensively adopted, UWB and binocular vision obstacle avoidance technologies are adopted to finish obstacle avoidance in the unmanned aerial vehicle polling process, deviation angles are calculated, the flying direction is adjusted, so that obstacles are avoided, reliable flying of the unmanned aerial vehicle is guaranteed, the automatic image focusing technology of an aerial camera is researched, secondary development is carried out on an aerial cloud deck, effective acquisition of key equipment of a transformer substation is achieved, distribution of targets in an image domain is optimized, and the unmanned aerial vehicle is helped to acquire required images safely and efficiently.
3. By establishing an aerial photography data set of the unmanned aerial vehicle in a complex scene of a transformer substation, screening images according to defect types and picture arrangement quality, carrying out manual standardized frame selection marking on image defects, researching a data sample enhancement method, comprehensively adopting image rotation, saturation brightness contrast adjustment and white noise adding technologies, realizing expansion and feature enhancement of a small sample data set, enabling the sample set to be richer, reducing overfitting of a deep learning model and improving the robustness of the model.
4. The deep learning target identification and positioning based on the convolutional neural network is researched, the model identification and positioning effect, the hardware computing force and the model running speed are considered, optimization is performed, a training model is built, target identification and detection are performed on transformer substation corrosion, bird nests, insulator falling pieces and small hardware hidden dangers, an edge computing ground station is built, aerial images acquired by autonomous flight of the unmanned aerial vehicle and infrared images are guided into a background system, real-time detection of defects is performed, obstacle avoidance efficiency is improved, and the defects are reduced.
To sum up, through the applicable technique to transformer substation unmanned aerial vehicle image acquisition, suitable model is studied, plan the airline, can fix a position unmanned aerial vehicle, through binocular range finding, calculate the parallax, the degree of depth and keep away barrier deviation angle, avoid the obstacle through adjustment flight angle, unmanned aerial vehicle flight is safer, and through discerning and resisting the interference, reduce the interference, the image that needs can be accurate is gathered, and discernment hidden danger, the real-time situation and the hidden danger of understanding the different positions of transformer substation that the image through unmanned aerial vehicle collection can be quick, the unmanned aerial vehicle who promotes the transformer substation patrols and examines efficiency.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
FIG. 1 is a flow chart of steps of the substation unmanned aerial vehicle imaging technique of the present invention;
FIG. 2 is a transformer substation cloud data generation diagram a of the present invention;
FIG. 3 is a transformer substation cloud data generation diagram b of the present invention;
FIG. 4 is an expanded view of the Xinit of the present invention;
fig. 5 is a flow chart of the binocular vision obstacle avoidance system of the present invention.
FIG. 6 is a schematic view of binocular range finding of the present invention;
FIG. 7 is a schematic view of the coordinate setting for binocular range finding of the present invention;
FIG. 8 is a schematic diagram of an obstacle avoidance deviation angle of the present invention;
FIG. 9 is an operational display diagram of the labelme of the present invention;
FIG. 10 is a flow chart of corrosion detection according to the present invention;
FIG. 11 is a diagram of a deep learning model for bird nest identification according to the present invention;
FIG. 12 is a schematic view of the insulator string drop detection of the present invention;
FIG. 13 is a comparison of aerial images and infrared images in accordance with the present invention;
FIG. 14 is a comparison of an aerial image and an infrared image of the present invention.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it should be understood that they are presented herein only to illustrate and explain the present invention and not to limit the present invention.
The embodiment is as follows: as shown in fig. 1, the invention provides a technical scheme, and an unmanned aerial vehicle image technology method based on a transformer substation comprises the following research steps:
s1, comprehensively knowing an unmanned aerial vehicle positioning application technology, an application model, a route planning and anti-interference measures in a transformer substation;
s2, researching a flight control and holder adjustment algorithm of the unmanned aerial vehicle, effectively acquiring patrol images of the transformer substation, and optimizing the image proportion and the image position distribution of the target to be detected;
s3, researching visible light data sets of bird' S nest, rust, insulator chipping and small hardware defect hidden dangers of the transformer substation, and researching standardized labeling, preprocessing and data amplification of data;
and S4, researching visible light and infrared image defect identification and detection technologies, and realizing hidden danger target identification and detection based on a deep learning algorithm.
As shown in fig. 2-4, in S1:
the positioning technology combining the Beidou positioning suitable for the unmanned aerial vehicle in the transformer substation and the real-time dynamic differential method (RTK) carrier phase differential technology is researched, and centimeter-level positioning of the unmanned aerial vehicle in the transformer substation is realized:
the RTK system is composed of 1 base station, an unmanned aerial vehicle mobile station and a radio communication system, wherein the base station comprises a Beidou receiver, a radio communication transmitting system and a base station controller part, the unmanned aerial vehicle mobile station comprises the Beidou receiver and a radio communication receiving system part, in an RTK operation mode, 1 receiver is arranged on a known high-level point (the base station) in a transformer substation as a reference station to continuously observe a Beidou satellite and send observation data and station measurement information to the unmanned aerial vehicle mobile station in real time through radio transmission equipment, the unmanned aerial vehicle mobile station receiver receives a data chain from the base station through radio receiving equipment while receiving Beidou satellite signals and collecting satellite data, and carries out carrier phase difference processing on 2 groups of data collected and received in the system to calculate three-dimensional coordinates and precision of the unmanned aerial vehicle mobile station in real time (namely, coordinate difference X, Y and delta H of the base station and the mobile station), WGS84 coordinates of each point obtained by adding the base coordinate, plane coordinates, Y and sea height H of the RTK technology are obtained according to key phase differences of the coordinate X, Y and delta coordinates of the observation data of the base station and delta coordinates, most of observation errors of the unmanned aerial vehicle mobile station are removed through a centimeter-level space difference mode, and the carrier phase difference of observation data of each base station is realized by the unmanned aerial vehicle mobile station, thereby, and the high centimeter-level difference of the high-level difference of the unmanned aerial vehicle mobile station is realized by the high-level difference of the unmanned aerial vehicle positioning errors in the unmanned aerial vehicle positioning mode.
And (3) establishing a complete three-dimensional model of the laser point cloud of the transformer substation for researching the applicable octree model:
establishing a laser point cloud three-dimensional model of all equipment in the transformer substation by using an octree model;
in the point cloud data processing by using the octree model, the octree needs to be segmented, the subdivision process can be represented by a tree with 8 degrees, a specified three-dimensional space area is divided into 8 trigrams, and 8 data elements are stored at each non-leaf node on the tree;
storing read data points in a two-dimensional array point [ N ] [3], wherein N is the number of the data points, 3 represents three-dimensional data, information stored by each volume element comprises pointers pointing to 8 child nodes, the starting point of the volume element, side lengths in three coordinate directions and indexes of points contained by the volume element in the point, the subdivision process is from top to bottom, namely, the data points are recursively divided from top to bottom, and finally, the data points are divided into each leaf node, the root node and the middle node do not store point information, namely, after the current node is divided, the indexes of the points contained by the current node in pint are deleted;
during interpolation, the normal direction of a data point needs to be estimated, a common method for estimating the normal vector is to use a least square method to approximate a set of neighboring points of discrete points to obtain a tangent plane, a single normal vector of the tangent plane is taken as the normal vector of the point, and the non-convergence feature vector can be approximately calculated by adopting the following formula:
the method solves the problem of trajectory planning by rapidly expanding a random tree RRT algorithm, and determines a path by collision detection, thereby adapting to reasonable planning of the routing inspection trajectory of the transformer substation under the complex environment and establishing an optimal route:
the RRT algorithm searches an unknown environment space by constructing a random search tree, sets a root node as an initial point, ensures probability completeness of target point search due to the randomness of search points, is beneficial to path search in a multi-dimensional space, can solve the problem of path planning with incomplete constraint, searches in a set area by using a search strategy to find a next leaf node to be expanded, then realizes random tree expansion growth by using a step strategy as a judgment condition, and finally obtains a planned path of an unmanned aerial vehicle initial point to the target point;
the RRT path planning algorithm starts from an initial point in a state space, the initial point is used as a root node, leaf nodes are added in a random sampling expansion mode to generate a random expansion tree, when the leaf nodes in the random tree comprise target points and enter a target area, at least one piece of path information from the initial point to the target points can be found in the random expansion tree, the initial point Xinit is used as the root node of the tree, when a new node is expanded, a point Xrand is randomly selected from a planning area, then a child node xnear closest to X is found in the current RRT tree, xnew is calculated according to a set expansion step length L, obstacles and threats are met in the process of expanding the new node Xnew, and the Xrand needs to be selected again for iterative calculation; and if the new node Xnew meets the requirement, adding the new node Xnew into the random tree and establishing the connection relation between the nodes, namely the Xnew is a child node of the Xnew.
In the unmanned aerial vehicle communication link, in order to exert its anti-jamming advantage, adopt the mode that directly expands and frequency hopping fuses mutually, in certain space-time range, can improve unmanned aerial vehicle's anti-jamming ability of totality effectively.
In S2:
a1, researching an obstacle avoidance technology of an unmanned aerial vehicle, comprehensively adopting RTK positioning, UWB and binocular vision obstacle avoidance technologies to complete obstacle avoidance in the inspection process of the unmanned aerial vehicle, and ensuring reliable flight of the unmanned aerial vehicle;
the UWB technology utilizes nanosecond-picosecond-level non-sine wave narrow pulse to transmit data, a UWB positioning algorithm mainly comprises two parts, firstly, a TOF ranging algorithm is adopted to obtain the distance between a label and a fixed base station, and then, the label coordinate is obtained according to a trilateral positioning algorithm:
as shown in FIGS. 5-8, assuming three base station physical coordinates of (x _1, y _1), (x _2, y _2), and (x _3, y _3), and tag-to-base station distances of d _1, d _2, d _3, there are:
(x-x_1)^2+(y-y_1)^2＝〖d_1〗^2
(x-x_2)^2+(y-y_2)^2＝〖d_2〗^2
(x-x_3)^2+(y-y_3)^2＝〖d_3〗^2
so as to solve the real-time coordinate;
the method comprises the steps that two cameras are symmetrically fixed on an unmanned aerial vehicle, the positions of an object are determined by using a binocular obstacle avoidance technology, then the angles of the two shooting positions are determined respectively, distance measurement is carried out through the two shooting positions and the position coordinates of the object, after images are shot and obtained, the cameras calibrate parameters, the two images are corrected, feature point extraction is carried out, three-dimensional information is obtained through three-dimensional matching, parallax is calculated, and the parallax is converted into depth, so that the flight angle of the unmanned aerial vehicle is adjusted when the unmanned aerial vehicle flies, and obstacles are avoided;
the position connecting line of the two shooting angles A and B is used as an X axis, and the perpendicular line passing through the barrier P is used as a Y axis to establish a rectangular coordinateThe system comprises OP as the original flying direction of the unmanned aerial vehicle, L and R points as the position points of two cameras on the unmanned aerial vehicle, and two end points of the obstacle P respectively using P 1 And P 2 Represents alpha;
wherein the coordinate of the point P is (0, y 3);
P 1 the coordinates of the point are (x 31, y 3);
P 2 the coordinates of the point are (x 32, y 3);
the coordinates of the L point are (x 1, y 1);
the coordinates of the R point are (x 2, y 2);
when the obstacle is avoided, the calculation formula of the adjusted angle is as follows:
wherein h is 1 The distance between the point P1 and the point P2 is calculated according to the formula:
h 1 ＝x 32 -x 31 (2)；
in the formula, h2 is the width of the unmanned aerial vehicle;
z is depth, and the calculation formula is as follows:
wherein f is the focal length;
d is parallax, and the calculation formula is as follows:
d＝x 2 -x 1 (4)；
and (4) calculating to adjust the flight direction according to the deviation angle when the unmanned aerial vehicle avoids the obstacle by combining the formulas (1), (2), (3) and (4).
And A2, researching an automatic image focusing technology of the aerial camera, and effectively acquiring key equipment of the transformer substation and optimizing the distribution of targets in an image domain by carrying out secondary development on the aerial cloud deck.
In S3:
b1, establishing an aerial photography data set of the unmanned aerial vehicle in a complex scene of the transformer substation, researching and utilizing a script to capture a network public data set from the Internet, screening images according to defect types and picture arrangement quality, obtaining aerial photography images through communication with departments of provinces, municipal departments and power supply companies, and simultaneously automatically shooting a common data set of the transformer substation.
B2, researching manual standardized frame selection and marking of image defects, and performing manual standardized marking on the established data set by adopting labelImg software;
as shown in fig. 9, labelimg tool is mainly used for object detection, while labelme tool creates its own data set in image segmentation field, and we can use labelimg to create its own data set training images YOLOv3 and YOLOv4 object detection models, and can create its own data set required by image segmentation training images MaskRCNN model.
And B3, researching a data sample enhancement method, comprehensively adopting image rotation, saturation brightness contrast adjustment and white noise adding technologies, realizing expansion and feature enhancement of a small sample data set, enabling the sample set to be richer, reducing overfitting of a deep learning model, and improving the robustness of the model.
In S4:
and C1, researching deep learning target identification and positioning based on the convolutional neural network, considering model identification and positioning effects, hardware calculation force and model running speed, and optimizing by adopting a YOLOv5 network model.
C2, a Pythrch deep learning frame is adopted for researching to build a training model, and target recognition detection is carried out on transformer substation corrosion, bird nests, insulator falling pieces and small hardware hidden dangers.
As shown in fig. 10, the substation equipment corrosion detection algorithm mainly includes power equipment target detection, foreground-background segmentation, gamma transformation based on background brightness, and an improved super-red algorithm, and performs corrosion area detection and calculation. The foreground image is responsible for extracting a target area, the background image is responsible for calculating average brightness, the background image and the target area are combined to carry out Gamma conversion normalization on the target image, rust area detection is completed, a rust rate is obtained according to a rust area, operation and maintenance personnel are guided to carry out hardware replacement, and the rust area and the rust degree are visually displayed through a rust thermodynamic diagram;
as shown in fig. 11, for the identification of the nests, a method based on a retinet deep learning model is adopted, the model utilizes ResNet-50 to perform early stage feature extraction, early stage standard features are reinforced through an FPN network, a feature pyramid image is constructed to meet the detection of the nest targets with different sizes, and then a classification subnet and a regression subnet are constructed on the basis of the feature pyramid to respectively identify the specific positions of the nests and the regression nests;
as shown in fig. 12, the overall process of detecting the insulator string chipping is to input an image, obtain a set of feature maps S containing information about the key point positions of the insulator string after neural network training, where S = (S1, S2, \8230;, sN), where each feature map corresponds to a fixed type of key point position information, and finally, match the key points of the insulator string by using a greedy algorithm using N key point feature maps output by a model to output a skeleton of the insulator string, thereby realizing the detection of the insulator string chipping.
As shown in fig. 13-14, C3, building an edge computing ground station, importing aerial images and infrared images acquired by autonomous flight of the unmanned aerial vehicle into a background system, and performing real-time defect detection.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments described above, or equivalents may be substituted for elements thereof. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (5)
1. Unmanned aerial vehicle image technology method based on transformer substation, its characterized in that: the method comprises the following research steps:
s1, comprehensively knowing an unmanned aerial vehicle positioning applicable technology, an applicable model, a route planning and anti-interference measures in a transformer substation;
s2, researching a flight control and holder adjustment algorithm of the unmanned aerial vehicle, effectively acquiring patrol images of the transformer substation, and optimizing the image proportion and the image position distribution of the target to be detected;
s3, researching visible light data sets of bird' S nest, rust, insulator chipping and small hardware defect hidden dangers of the transformer substation, and researching standardized labeling, preprocessing and data amplification of data;
s4, researching visible light and infrared image defect identification and detection technologies to realize hidden danger target identification and detection based on a deep learning algorithm;
the S2 specifically comprises the following steps:
a1, researching an obstacle avoidance technology of the unmanned aerial vehicle, comprehensively adopting RTK positioning, UWB and binocular vision obstacle avoidance technologies to complete obstacle avoidance in the inspection process of the unmanned aerial vehicle, and ensuring reliable flight of the unmanned aerial vehicle;
a2, researching an automatic image focusing technology of an aerial camera, and performing secondary development on an aerial cloud deck to effectively acquire key equipment of a transformer substation and optimize the distribution of targets in an image domain;
the UWB technology in the A1 utilizes nanosecond-picosecond-level non-sine wave narrow pulse to transmit data, a UWB positioning algorithm mainly comprises two parts, firstly, a TOF ranging algorithm is adopted to obtain the distance between a label and a fixed base station, and then, a label coordinate is obtained according to a trilateral positioning algorithm;
assuming three base stations with physical coordinates of (x _1, y _1), (x _2, y _2), and (x _3, y _3), and tag-to-base station distances of d _1, d _2, d _3, then:
(x-x_1)^2+(y-y_1)^2＝〖d_1〗^2
(x-x_2)^2+(y-y_2)^2＝〖d_2〗^2
(x-x_3)^2+(y-y_3)^2＝〖d_3〗^2
so as to solve the real-time coordinate;
the method comprises the steps that two cameras are symmetrically fixed on an unmanned aerial vehicle, the positions of objects are determined by using a binocular obstacle avoidance technology, then the angles of the two shooting positions are respectively determined, the distance measurement is carried out through the two shooting positions and the position coordinates of the objects, after images are shot and obtained, the cameras calibrate parameters, the two images are corrected, the three-dimensional information is obtained through characteristic point extraction and three-dimensional matching, the parallax is calculated, and the parallax is converted into the depth, so that the flight angle is adjusted and obstacles are avoided when the unmanned aerial vehicle flies;
the position connecting line of two shooting angles A and B is used as an X axis, a perpendicular line passing through an obstacle P is used as a Y axis to establish a rectangular coordinate system, OP is the original flight direction of the unmanned aerial vehicle, L and R points are position points of two cameras on the unmanned aerial vehicle, and two end points of the obstacle P are respectively used as P 1 And P 2 Represents alpha;
wherein, the coordinate of the point P is (0, y 3);
P 1 the coordinates of the point are (x 31, y 3);
P 2 the coordinates of the point are (x 32, y 3);
the coordinates of the L point are (x 1, y 1);
the coordinates of the R point are (x 2, y 2);
when the obstacle is avoided, the calculation formula of the adjusted angle is as follows:
wherein h is 1 The calculation formula is the distance between the point P1 and the point P2:
h 1 ＝x 32 -x 31 (2)；
in the formula, h2 is the width of the unmanned aerial vehicle;
z is depth, and the calculation formula is as follows:
wherein f is the focal length;
d is parallax, and the calculation formula is as follows:
d＝x 2 -x 1 (4)；
and (4) calculating to adjust the flight direction according to the deviation angle when the unmanned aerial vehicle avoids the obstacle by combining the formulas (1), (2), (3) and (4).
2. The substation-based unmanned aerial vehicle image technology method according to claim 1, wherein in S3, the research on the data set specifically comprises the following steps:
b1, establishing an aerial photography data set of the unmanned aerial vehicle in a complex scene of the transformer substation, researching and grabbing a network public data set from the Internet by using a script, screening images according to defect types and image arrangement quality, obtaining aerial photography images through communication with departments of provinces, municipal departments and power supply companies, and automatically shooting a common data set of the transformer substation;
b2, researching manual standardized frame selection and marking of image defects, and performing manual standardized marking on the established data set by adopting labelImg software;
and B3, researching a data sample enhancement method, and comprehensively adopting image rotation, saturation brightness contrast adjustment and white noise adding technologies to realize expansion and feature enhancement of a small sample data set.
3. The substation-based unmanned aerial vehicle image technology method of claim 1, wherein in S4, the hidden danger targets comprise rust, bird nests, insulator chipping and small hardware fittings;
the specific steps of identification and detection are as follows:
c1, researching deep learning target identification and positioning based on a convolutional neural network, and optimizing by adopting a YOLOv5 network model in combination with a model identification and positioning effect, hardware computing power and model operation speed;
c2, researching a training model built by a Pythrch deep learning frame, and carrying out target identification detection on the hidden danger of the transformer substation;
and C3, building an edge computing ground station, importing aerial images and infrared images acquired by autonomous flight of the unmanned aerial vehicle into a background system, and detecting defects in real time.
4. The transformer substation-based unmanned aerial vehicle image technology method according to claim 1, wherein in the step C2, a transformer substation equipment rust detection algorithm mainly comprises power equipment target detection, foreground-background segmentation, background brightness-based Gamma conversion and an improved hyper-red algorithm, and rust area detection and calculation are carried out;
the foreground image is responsible for extracting a target area, the background image is responsible for calculating average brightness, the background image and the target area are combined to carry out Gamma conversion normalization on the target image, rust area detection is completed, a rust rate is obtained according to a rust area, operation and maintenance personnel are guided to carry out hardware fitting replacement, and the rust area and the rust degree are visually displayed through a rust thermodynamic diagram;
during the identification of the bird nest, a method based on a RetinaNet deep learning model is adopted, resNet-50 is utilized to extract early-stage features, early-stage standard features are reinforced through an FPN network, a feature pyramid image is constructed to meet the detection of bird nest targets with different sizes, and then a classification subnet and a regression subnet are constructed on the basis of the feature pyramid and are respectively used for identifying the specific positions of the bird nest and the regression bird nest;
when the insulator string chipping detection is carried out, an image is input firstly, a group of feature maps S containing the position information of key points of the insulator string is obtained after neural network training, wherein S = (S1, S2, \8230;, sN), each feature map corresponds to the position information of the key points of a fixed type, and finally, N key point feature maps output by a model are used for matching the key points of the insulator string by using a greedy algorithm, and an insulator string framework is output, so that the insulator string chipping detection is realized.
5. The substation-based unmanned aerial vehicle image technology method according to claim 1, wherein in S1, the positioning applicable technology of the unmanned aerial vehicle is a positioning technology combining Beidou positioning and a real-time kinematic differential method (RTK) carrier phase difference technology;
the applicable model is an octree model, and a complete substation laser point cloud three-dimensional model is established by researching the octree model;
the interference existing in the transformer substation comprises a power frequency magnetic field, pulse electromagnetism and second harmonic existing in the transformer substation;
the route planning firstly researches a rapid expansion random tree RRT algorithm to solve the problem of trajectory planning, and determines a path through collision detection, so that the reasonable planning of the transformer substation inspection trajectory under the complex environment is adapted, and an optimal route is established;
the anti-electromagnetic interference technology mainly aims at an unmanned aerial vehicle positioning, data transmission and image transmission device;
the anti-interference measures comprise the steps of adopting channel coding, spread spectrum anti-interference, self-adaptive anti-interference, additionally arranging envelope shielding and reasonably planning a route to avoid a strong magnetic region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211371013.0A CN115562348A (en) | 2022-11-03 | 2022-11-03 | Unmanned aerial vehicle image technology method based on transformer substation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211371013.0A CN115562348A (en) | 2022-11-03 | 2022-11-03 | Unmanned aerial vehicle image technology method based on transformer substation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115562348A true CN115562348A (en) | 2023-01-03 |
Family
ID=84769638
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211371013.0A Pending CN115562348A (en) | 2022-11-03 | 2022-11-03 | Unmanned aerial vehicle image technology method based on transformer substation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115562348A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116188998A (en) * | 2023-04-26 | 2023-05-30 | 南方电网数字电网研究院有限公司 | Method, device, equipment and storage medium for identifying defects of overhead transmission line |
| CN116188470A (en) * | 2023-04-28 | 2023-05-30 | 成都航空职业技术学院 | Unmanned aerial vehicle aerial photographing identification-based fault positioning method and system |
| CN116185079A (en) * | 2023-04-28 | 2023-05-30 | 西安迈远科技有限公司 | Unmanned aerial vehicle construction inspection route planning method based on self-adaptive cruising |
| CN116482602A (en) * | 2023-06-21 | 2023-07-25 | 中国民航管理干部学院 | Unmanned aerial vehicle positioning method and positioning system based on radio waves |
| CN118220551A (en) * | 2024-05-23 | 2024-06-21 | 中飞赛维智能科技股份有限公司 | Unmanned aerial vehicle and substation patrol route planning obstacle avoidance system |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106933243A (en) * | 2015-12-30 | 2017-07-07 | 湖南基石信息技术有限公司 | A kind of unmanned plane Real Time Obstacle Avoiding system and method based on binocular vision |
| CN110231832A (en) * | 2018-03-05 | 2019-09-13 | 北京京东尚科信息技术有限公司 | Barrier-avoiding method and obstacle avoidance apparatus for unmanned plane |
| CN111413970A (en) * | 2020-03-18 | 2020-07-14 | 天津大学 | Ultra-wideband and vision integrated indoor robot positioning and autonomous navigation method |
| CN111736622A (en) * | 2019-03-25 | 2020-10-02 | 海鹰航空通用装备有限责任公司 | Unmanned aerial vehicle obstacle avoidance method and system based on combination of binocular vision and IMU |
| CN111988524A (en) * | 2020-08-21 | 2020-11-24 | 广东电网有限责任公司清远供电局 | Unmanned aerial vehicle and camera collaborative obstacle avoidance method, server and storage medium |
| CN112050814A (en) * | 2020-08-28 | 2020-12-08 | 国网智能科技股份有限公司 | Unmanned aerial vehicle visual navigation system and method for indoor transformer substation |
| CN112130579A (en) * | 2020-09-24 | 2020-12-25 | 国网智能科技股份有限公司 | Tunnel unmanned aerial vehicle inspection method and system |
| CN112508865A (en) * | 2020-11-23 | 2021-03-16 | 深圳供电局有限公司 | Unmanned aerial vehicle inspection obstacle avoidance method and device, computer equipment and storage medium |
| CN112684806A (en) * | 2019-10-18 | 2021-04-20 | 天津工业大学 | Electric power inspection unmanned aerial vehicle system based on dual obstacle avoidance and intelligent identification |
| CN112816939A (en) * | 2020-12-31 | 2021-05-18 | 广东电网有限责任公司 | Substation unmanned aerial vehicle positioning method based on Internet of things |
| US20210224512A1 (en) * | 2020-01-17 | 2021-07-22 | Wuyi University | Danet-based drone patrol and inspection system for coastline floating garbage |
| CN114022798A (en) * | 2021-10-29 | 2022-02-08 | 国网河南省电力公司洛阳供电公司 | Transformer substation inspection robot obstacle avoidance method based on digital twinning technology |
| CN114092537A (en) * | 2021-09-23 | 2022-02-25 | 国电南瑞科技股份有限公司 | Automatic inspection method and device for electric unmanned aerial vehicle of transformer substation |
| CN115047278A (en) * | 2022-07-04 | 2022-09-13 | 广东电网有限责任公司 | Transformer substation inspection method, device, equipment and storage medium |
-
2022
- 2022-11-03 CN CN202211371013.0A patent/CN115562348A/en active Pending
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106933243A (en) * | 2015-12-30 | 2017-07-07 | 湖南基石信息技术有限公司 | A kind of unmanned plane Real Time Obstacle Avoiding system and method based on binocular vision |
| CN110231832A (en) * | 2018-03-05 | 2019-09-13 | 北京京东尚科信息技术有限公司 | Barrier-avoiding method and obstacle avoidance apparatus for unmanned plane |
| CN111736622A (en) * | 2019-03-25 | 2020-10-02 | 海鹰航空通用装备有限责任公司 | Unmanned aerial vehicle obstacle avoidance method and system based on combination of binocular vision and IMU |
| CN112684806A (en) * | 2019-10-18 | 2021-04-20 | 天津工业大学 | Electric power inspection unmanned aerial vehicle system based on dual obstacle avoidance and intelligent identification |
| US20210224512A1 (en) * | 2020-01-17 | 2021-07-22 | Wuyi University | Danet-based drone patrol and inspection system for coastline floating garbage |
| CN111413970A (en) * | 2020-03-18 | 2020-07-14 | 天津大学 | Ultra-wideband and vision integrated indoor robot positioning and autonomous navigation method |
| CN111988524A (en) * | 2020-08-21 | 2020-11-24 | 广东电网有限责任公司清远供电局 | Unmanned aerial vehicle and camera collaborative obstacle avoidance method, server and storage medium |
| CN112050814A (en) * | 2020-08-28 | 2020-12-08 | 国网智能科技股份有限公司 | Unmanned aerial vehicle visual navigation system and method for indoor transformer substation |
| CN112130579A (en) * | 2020-09-24 | 2020-12-25 | 国网智能科技股份有限公司 | Tunnel unmanned aerial vehicle inspection method and system |
| CN112508865A (en) * | 2020-11-23 | 2021-03-16 | 深圳供电局有限公司 | Unmanned aerial vehicle inspection obstacle avoidance method and device, computer equipment and storage medium |
| CN112816939A (en) * | 2020-12-31 | 2021-05-18 | 广东电网有限责任公司 | Substation unmanned aerial vehicle positioning method based on Internet of things |
| CN114092537A (en) * | 2021-09-23 | 2022-02-25 | 国电南瑞科技股份有限公司 | Automatic inspection method and device for electric unmanned aerial vehicle of transformer substation |
| CN114022798A (en) * | 2021-10-29 | 2022-02-08 | 国网河南省电力公司洛阳供电公司 | Transformer substation inspection robot obstacle avoidance method based on digital twinning technology |
| CN115047278A (en) * | 2022-07-04 | 2022-09-13 | 广东电网有限责任公司 | Transformer substation inspection method, device, equipment and storage medium |
Non-Patent Citations (9)
| Title |
|---|
| 尚虎军 等: "UWB定位的智能小车避障技术研究", 单片机与嵌入式***应用, no. 1, pages 45 - 47 * |
| 张旭 等: "基于深度学习目标检测的电力设备锈蚀缺陷检测算法", 供用电, vol. 37, no. 12, pages 87 - 92 * |
| 彭浩: "基于YOLOv5的无人机巡检图像绝缘子检测技术的研究", 中国优秀硕士学位论文全文数据库工程科技Ⅱ辑, no. 04, pages 042 - 222 * |
| 李向阳 等: "基于深度学习的绝缘子串检测算法研究", 信息通信, no. 01, pages 68 - 70 * |
| 林泓宏 等: "变电站无人机自动巡检技术研究", 电器开关, no. 2, pages 39 - 42 * |
| 欧进永 等: "基于深度学习的输电线路杆塔鸟窝识别方法研究", 机电信息, no. 24, pages 22 - 23 * |
| 王焓丁: "巡检机器人目标缺陷识别模块设计与实现", 中国优秀硕士学位论文全文数据库工程科技Ⅱ辑, no. 01 * |
| 田小壮 等: "复杂环境下无人机智能巡检轨迹规划方法研究", 电子设计工程, vol. 29, no. 20, pages 77 - 81 * |
| 皇莆萍萍 等: "基于双目视觉的无人机避障***设计", 智能计算机与应用, vol. 10, no. 7, pages 129 - 131 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116188998A (en) * | 2023-04-26 | 2023-05-30 | 南方电网数字电网研究院有限公司 | Method, device, equipment and storage medium for identifying defects of overhead transmission line |
| CN116188998B (en) * | 2023-04-26 | 2023-09-15 | 南方电网数字电网研究院有限公司 | Method, device, equipment and storage medium for identifying defects of overhead transmission line |
| CN116188470A (en) * | 2023-04-28 | 2023-05-30 | 成都航空职业技术学院 | Unmanned aerial vehicle aerial photographing identification-based fault positioning method and system |
| CN116185079A (en) * | 2023-04-28 | 2023-05-30 | 西安迈远科技有限公司 | Unmanned aerial vehicle construction inspection route planning method based on self-adaptive cruising |
| CN116185079B (en) * | 2023-04-28 | 2023-08-04 | 西安迈远科技有限公司 | Unmanned aerial vehicle construction inspection route planning method based on self-adaptive cruising |
| CN116482602A (en) * | 2023-06-21 | 2023-07-25 | 中国民航管理干部学院 | Unmanned aerial vehicle positioning method and positioning system based on radio waves |
| CN116482602B (en) * | 2023-06-21 | 2023-09-12 | 中国民航管理干部学院 | Unmanned aerial vehicle positioning method and positioning system based on radio waves |
| CN118220551A (en) * | 2024-05-23 | 2024-06-21 | 中飞赛维智能科技股份有限公司 | Unmanned aerial vehicle and substation patrol route planning obstacle avoidance system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115562348A (en) | Unmanned aerial vehicle image technology method based on transformer substation | |
| CN110097553B (en) | Semantic mapping system based on instant positioning mapping and three-dimensional semantic segmentation | |
| CN109416413A (en) | Solar energy forecast | |
| CN108317953A (en) | A kind of binocular vision target surface 3D detection methods and system based on unmanned plane | |
| CN109739254B (en) | Unmanned aerial vehicle adopting visual image positioning in power inspection and positioning method thereof | |
| US20160363653A1 (en) | Above-ground building recognition method and navigation method combining laser and infrared | |
| CN107560592B (en) | Precise distance measurement method for photoelectric tracker linkage target | |
| CN103679674A (en) | Method and system for splicing images of unmanned aircrafts in real time | |
| CN112884672B (en) | Multi-frame unmanned aerial vehicle image relative radiation correction method based on contemporaneous satellite images | |
| CN109541613A (en) | Aerial high-voltage conducting wire cruising inspection system and method for inspecting based on single line laser ranging | |
| CN103093459A (en) | Assisting image matching method by means of airborne lidar point cloud data | |
| CN110889899A (en) | Method and device for generating digital earth surface model | |
| CN111914615A (en) | Fire-fighting area passability analysis system based on stereoscopic vision | |
| CN115423968B (en) | Power transmission channel optimization method based on point cloud data and live-action three-dimensional model | |
| CN113326752B (en) | Unmanned aerial vehicle-based photovoltaic power station identification method and system | |
| CN113504543B (en) | Unmanned aerial vehicle LiDAR system positioning and attitude determination system and method | |
| CN101286236A (en) | Infrared object tracking method based on multi- characteristic image and average drifting | |
| CN114581307A (en) | Multi-image stitching method, system, device and medium for target tracking identification | |
| CN111950524B (en) | Orchard local sparse mapping method and system based on binocular vision and RTK | |
| CN116338716B (en) | Multi-target association method of air-ground unmanned system based on azimuth topological structure | |
| CN116091493B (en) | Distance measurement method for hidden danger of tree obstacle of power transmission line | |
| CN109461179B (en) | Cooperative detection system for explosive-handling primary and secondary robots | |
| Wang et al. | Recognition and locating of damaged Poles in distribution network through images shot by unmanned aerial vehicle (UA V) | |
| CN112493228B (en) | Laser bird repelling method and system based on three-dimensional information estimation | |
| CN113885510B (en) | Four-foot robot obstacle avoidance and pilot following method and system |
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 |