CN113848560A - Dam surface image unmanned aerial vehicle rapid and safe acquisition method and system - Google Patents

Dam surface image unmanned aerial vehicle rapid and safe acquisition method and system Download PDF

Info

Publication number
CN113848560A
CN113848560A CN202111134127.9A CN202111134127A CN113848560A CN 113848560 A CN113848560 A CN 113848560A CN 202111134127 A CN202111134127 A CN 202111134127A CN 113848560 A CN113848560 A CN 113848560A
Authority
CN
China
Prior art keywords
inspection
point
dam
points
edge
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
Application number
CN202111134127.9A
Other languages
Chinese (zh)
Inventor
王远慧
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hunan Desen Jiuchuang Technology Co ltd
Original Assignee
Hunan Desen Jiuchuang Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hunan Desen Jiuchuang Technology Co ltd filed Critical Hunan Desen Jiuchuang Technology Co ltd
Priority to CN202111134127.9A priority Critical patent/CN113848560A/en
Publication of CN113848560A publication Critical patent/CN113848560A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/86—Combinations of lidar systems with systems other than lidar, radar or sonar, e.g. with direction finders
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B64—AIRCRAFT; AVIATION; COSMONAUTICS
    • B64C—AEROPLANES; HELICOPTERS
    • B64C39/00—Aircraft not otherwise provided for
    • B64C39/02—Aircraft not otherwise provided for characterised by special use
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C11/00—Photogrammetry or videogrammetry, e.g. stereogrammetry; Photographic surveying
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • 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
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Health & Medical Sciences (AREA)
  • Pathology (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Immunology (AREA)
  • Chemical & Material Sciences (AREA)
  • Signal Processing (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Electromagnetism (AREA)
  • Quality & Reliability (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a rapid and safe acquisition method and a rapid and safe acquisition system for dam surface images by an unmanned aerial vehicle, wherein a dam three-dimensional finite element refined model is established; calculating the stress deformation of the dam finite element model under a given working condition, and carrying out partition management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result; extracting dam surface nodes and the information of the dam surface nodes and the affiliated units in the finite element model under the current deformation working condition; planning an internal inspection point of the inspection surface according to the inspection surface, the node on the surface and the external normal information; and automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface, and constructing the dam unmanned aerial vehicle automatic safety inspection route by adopting a path planning algorithm, edge point transition and a mode of preferential inspection in the inspection surface. The invention can realize automatic image acquisition and save the inspection time; the system ensures the safety and high efficiency of close-range image acquisition, and is particularly suitable for routing inspection areas with complex and variable structures.

Description

Dam surface image unmanned aerial vehicle rapid and safe acquisition method and system
Technical Field
The invention relates to the technical field of dam safety monitoring, and particularly discloses a dam surface image unmanned aerial vehicle rapid and safe acquisition method and system.
Background
The unmanned aerial vehicle is adopted to collect images on the surface of the dam, so that defects such as cracks, leakage and the like on the surface of the dam can be effectively found in time, and the unmanned aerial vehicle has important significance for guaranteeing the safety and stability of the dam. However, the dam belongs to a large building, the defects may be only in millimeter level (such as cracks), the unmanned aerial vehicle can effectively save manpower, but still needs to acquire in a short distance, which results in long time consumption, at the present stage, the unmanned aerial vehicle mostly plans routes by means of three-dimensional point cloud, oblique photography, visual navigation or manual operation, and the like, and is not combined with methods such as finite element and the like capable of effectively analyzing the dam structure safety, so that only regional acquisition or a full-coverage type low-efficiency acquisition method depending on the experience of professionals can be adopted, and meanwhile, in the near distance route planning, especially in an inspection area with a complex and variable structure, the safety problem is also the key point of inspection tour,
therefore, how to use the unmanned aerial vehicle to realize quick safe acquisition to the dam image will have very strong practical value.
Disclosure of Invention
The invention provides an unmanned aerial vehicle rapid and safe acquisition method and system for dam surface images, and aims to realize rapid and safe acquisition of dam images by using an unmanned aerial vehicle.
The invention relates to a dam surface image rapid and safe unmanned aerial vehicle acquisition method, which comprises the following steps:
an unmanned aerial vehicle is adopted to carry a three-dimensional laser radar and a high-definition camera to obtain the data of the surface and peripheral scanning point clouds of the dam and the point cloud data of oblique photography dense matching, and then a three-dimensional finite element refined model of the dam is established by combining site and design data;
calculating the stress deformation of the dam finite element model under a given working condition, and carrying out partition management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result;
extracting dam surface nodes and the information of the dam surface nodes and the affiliated units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal lines of the unit surfaces to form routing inspection surfaces, and fusing the routing inspection surface information into dam three-dimensional point cloud data;
planning internal inspection points of the inspection surface, including a spatial position and a camera position, according to the inspection surface, the node on the surface and the external normal information so as to ensure the integrity of surface image acquisition;
calculating whether the edge nodes of the inspection point and the inspection surface exist points intersected with the dam and the surrounding three-dimensional point cloud within a sphere with the safety obstacle avoidance distance as the radius, and if so, deleting the inspection point and the edge nodes of the inspection surface;
and automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface, and constructing the dam unmanned aerial vehicle automatic safety inspection route by adopting a path planning algorithm, edge point transition and a mode of preferential inspection in the inspection surface.
And further, calculating the stress deformation of the dam finite element model under a given working condition, and performing partition management on the acquisition region by taking the units of the finite element model as basic units according to the calculation result, wherein the partition management is formulated according to the finite element calculation result under the given working condition, and the region is divided into a core region and a non-core region.
Further, extracting dam surface nodes and the information of the dam surface nodes and the corresponding units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal line of the unit surfaces to form routing inspection surfaces, integrating the routing inspection surface information into dam three-dimensional point cloud data, and determining translation distance jointly according to a shooting object, camera imaging quality and a partition acquisition area; setting a smaller translation distance or increasing the zooming times of the camera to ensure the shooting quality of the inspection object in the core area; for the non-core area, a larger translation distance is set to reduce the number of captured images.
Further, the step of calculating whether the edge nodes of the inspection point and the inspection surface are intersected with the points of the dam and the surrounding three-dimensional point cloud in the sphere range taking the safe obstacle avoidance distance as the radius, and if the edge nodes of the inspection point and the inspection surface are intersected with the points of the dam and the surrounding three-dimensional point cloud, deleting the edge nodes of the inspection point and the inspection surface comprises the following steps:
constructing KD-Tree for the dam surface and surrounding point cloud data, and inquiring the points closest to the inspection points and the edge nodes in the surface point cloud data;
and calculating the distance according to the inquired points closest to the inspection point and the edge node, and discarding the inspection point and the edge node of the inspection surface when the distance is smaller than the safe obstacle avoidance distance.
Further, according to the finally determined inspection point and the edge node of the inspection surface, the unmanned aerial vehicle inspection route is automatically planned by adopting a path planning algorithm, edge point transition and a mode of preferential inspection in the inspection surface, and the step of constructing the dam unmanned aerial vehicle automatic safety inspection route comprises the following steps:
step 610, regarding the routing inspection surfaces in which the edge node connection can be reached as the same category, and routing inspection surfaces of different categories need to be separately routed:
step 620, routing inspection path planning is carried out on the routing inspection surfaces of the same category;
step 610 includes:
611, constructing a set C containing all the inspection surfaces;
step 612, selecting one routing inspection surface from the set C as an initial routing inspection surface, searching for other routing inspection surfaces in the set C, regarding routing inspection surfaces with common edge nodes with the initial routing inspection surface as the same category, and otherwise, regarding routing inspection surfaces without common edge nodes with the initial routing inspection surface as a single category, and jumping to step 614;
step 613, circularly searching other polling surfaces in the set C, and regarding the polling surfaces with the same edge nodes as the type in the step 612 as the same type until all the polling surfaces in the set C are traversed or the manually specified upper limit number of the polling surfaces of the same type is reached;
step 614, updating the set C, that is, a new routing inspection surface set established by removing the routing inspection surfaces included in the category in step 613 from the set C in step 613;
615, repeating 612-614 until all the inspection surfaces are classified;
step 620 includes:
621, forming centroids of inspection points in the inspection surfaces into a path planning point set, constructing a point set and a weighted undirected complete graph with the distance between the two points as a weight, designating a starting point and an end point, and solving a shortest Hamilton path between the two points; when the inspection surfaces are transferred, the inspection points are transferred in a mode that the edge points are intermediate transition points; for adjacent routing inspection surfaces, the sum of the shortest distances from the intersected edge points to the routing inspection points of the two routing inspection surfaces is obtained, and the edge point corresponding to the minimum value is the selected edge point; for nonadjacent routing inspection surfaces, firstly, constructing a weighted undirected graph by using all edge points and connecting paths formed on the routing inspection surfaces, then calculating the edge point with the shortest distance between the two routing inspection surfaces, and preferably calculating the shortest path between the two points by using the edge point at one end as a starting point and the edge point at the other end as a terminal point by using a Floyd algorithm;
step 622, using the inspection point with the shortest connection distance with the edge point as the end point of one inspection surface and the start point of the other inspection surface, if the start point and the end point of the same inspection surface are the same point, modifying the end point, selecting the point with the second shortest connection distance with the edge point of the end point as a new end point, and calculating the start point and the end point of all the inspection surfaces; and constructing a weighted undirected complete graph with the distance between the two points as a weight value for the routing inspection points in the same routing inspection plane, knowing a starting point and an end point, and solving the shortest Hamilton path of the two points.
Another aspect of the invention relates to an unmanned aerial vehicle rapid and safe acquisition system for dam surface images, comprising:
the building module is used for acquiring dam surface and peripheral scanning point cloud and oblique photography dense matching point cloud data by adopting an unmanned aerial vehicle carrying a three-dimensional laser radar and a high-definition camera, and then building a dam three-dimensional finite element refined model by combining site and design data;
the partitioning module is used for calculating the stress deformation of the dam finite element model under a given working condition and performing partitioning management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result;
the merging module is used for extracting dam surface nodes and the information of the dam surface nodes and the affiliated units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal lines of the unit surfaces to form routing inspection surfaces, and merging the routing inspection surface information into the dam three-dimensional point cloud data;
the planning module is used for planning internal inspection points of the inspection surface, including spatial positions and camera orientations, according to the inspection surface, the nodes on the surface and the external normal information so as to ensure the integrity of surface image acquisition;
the deleting module is used for calculating whether a point intersected with the dam and the surrounding three-dimensional point cloud exists in the sphere range with the safety obstacle avoidance distance as the radius of the edge node of the inspection point and the inspection surface, and if the point intersected with the dam and the surrounding three-dimensional point cloud exists, deleting the edge node of the inspection point and the edge node of the inspection surface;
and the building module is used for automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface where the inspection point is located by adopting a path planning algorithm, edge point transition and a mode of preferentially inspecting in the inspection surface, and building the dam unmanned aerial vehicle automatic safety inspection route.
Further, the partition module includes a partition unit,
and the partition unit is used for partition management to be formulated according to a finite element calculation result under a given working condition and dividing the area into a core area and a non-core area.
Further, the blending module comprises a translation unit,
the translation unit is used for determining translation distance jointly according to the shooting object, the camera imaging quality and the subarea acquisition area; setting a smaller translation distance or increasing the zooming times of the camera to ensure the shooting quality of the inspection object in the core area; for the non-core area, a larger translation distance is set to reduce the number of captured images.
Further, the deletion module includes:
the query unit is used for constructing KD-Tree for the dam surface and surrounding point cloud data and querying points which are nearest to the inspection point and the edge node in the surface point cloud data;
and the abandoning unit is used for calculating the distance of the point according to the inquired point closest to the inspection point and the edge node, and abandoning the inspection point and the edge node of the inspection surface when the distance is smaller than the safe obstacle avoidance distance.
Further, the building module comprises a category calculating unit and a path rule unit,
the class calculation unit is used for regarding the routing inspection surfaces which can be reached by the edge node connection as the same class, and routing inspection surfaces of different classes need to be separately routed;
the route rule unit is used for planning routing inspection routes for routing inspection surfaces of the same category;
the category calculation unit includes:
the construction subunit is used for constructing a set C containing all the routing inspection surfaces;
the selecting subunit is used for randomly selecting one routing inspection surface from the set C as an initial routing inspection surface, searching other routing inspection surfaces in the set C, and regarding the routing inspection surfaces with common edge nodes with the initial routing inspection surface as the same category, or regarding the routing inspection surfaces without common edge nodes with the initial routing inspection surface as a single category;
the circulating subunit is used for circularly searching other routing inspection surfaces in the set C, and regarding the routing inspection surfaces with the same edge nodes as the routing inspection surfaces in the initial and selected subunits as the same type until all the routing inspection surfaces in the set C are traversed or the manually specified upper limit number of the routing inspection surfaces in the same type is reached;
the updating subunit is used for updating the set C, namely a new routing inspection surface set which is established by removing the routing inspection surfaces contained in the category of the circulating subunit from the set C of the circulating subunit;
the classification subunit is used for repeating the class calculation until all the inspection surfaces are classified;
the path rule unit includes:
the first calculating subunit is used for forming a path planning point set by centroids of inspection points in all inspection surfaces, constructing a point set and a weighted undirected complete graph with the distance between two points as a weight, designating a starting point and an end point, and solving the shortest Hamilton path between the two points; when the inspection surfaces are transferred, the inspection points are transferred in a mode that the edge points are intermediate transition points; for adjacent routing inspection surfaces, the sum of the shortest distances from the intersected edge points to the routing inspection points of the two routing inspection surfaces is obtained, and the edge point corresponding to the minimum value is the selected edge point; for nonadjacent routing inspection surfaces, firstly, constructing a weighted undirected graph by using all edge points and connecting paths formed on the routing inspection surfaces, then solving the edge point with the shortest distance between the two routing inspection surfaces, and preferably solving by using a Floyd algorithm, wherein the shortest path between the two points is calculated by using the edge point at one end as a starting point and the edge point at the other end as a terminal point;
the second calculation subunit is used for taking the routing inspection point with the shortest connection distance with the edge point as the end point of one routing inspection surface and the start point of the other routing inspection surface, modifying the end point if the start point and the end point of the same routing inspection surface are the same point, selecting the point with the second shortest connection distance with the edge point of the end point as a new end point, and calculating the start point and the end point of all routing inspection surfaces; and constructing a weighted undirected complete graph with the distance between the two points as a weight value for the routing inspection points in the same routing inspection plane, knowing a starting point and an end point, and solving the shortest Hamilton path of the two points.
The beneficial effects obtained by the invention are as follows:
the invention provides a rapid and safe acquisition method and a rapid and safe acquisition system for dam surface images by an unmanned aerial vehicle, wherein the unmanned aerial vehicle is adopted to carry a three-dimensional laser radar and a high-definition camera to obtain dam surface and peripheral scanning point cloud and oblique photography dense matching point cloud data, and then a dam three-dimensional finite element refined model is established by combining site and design data; calculating the stress deformation of the dam finite element model under a given working condition, and carrying out partition management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result; extracting dam surface nodes and the information of the dam surface nodes and the affiliated units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal lines of the unit surfaces to form routing inspection surfaces, and fusing the routing inspection surface information into dam three-dimensional point cloud data; planning internal inspection points of the inspection surface, including a spatial position and a camera position, according to the inspection surface, the node on the surface and the external normal information so as to ensure the integrity of surface image acquisition; calculating whether the edge nodes of the inspection point and the inspection surface exist points intersected with the dam and the surrounding three-dimensional point cloud within a sphere with the safety obstacle avoidance distance as the radius, and if so, deleting the inspection point and the edge nodes of the inspection surface; and automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface, and constructing the dam unmanned aerial vehicle automatic safety inspection route by adopting a path planning algorithm, edge point transition and a mode of preferential inspection in the inspection surface. According to the method and the system for rapidly and safely acquiring the dam surface image by the unmanned aerial vehicle, the dam surface image is acquired by combining the three-dimensional point cloud data of the unmanned aerial vehicle and the finite element, so that automatic image acquisition can be realized, a key area can be focused according to the stress characteristics of the dam structure, and inspection time is saved; the automatic routing method for the inspection points and the inspection route is provided by combining the attributes of the finite element units and the three-dimensional point cloud information, and meanwhile, an inspection mode of unit edge point transition and preferential inspection in the inspection surface is provided, so that the unmanned aerial vehicle and the surface of the dam are always kept at a safe distance, the close-range image acquisition is ensured to be safe and efficient, and the method is particularly suitable for inspection areas with complex and variable structures.
Drawings
FIG. 1 is a schematic flow chart of an embodiment of an unmanned aerial vehicle rapid and safe acquisition method for dam surface images provided by the invention;
fig. 2 is a schematic view of a detailed flow of an embodiment of a step of calculating whether a point intersecting a dam and a surrounding three-dimensional point cloud exists in a sphere range with a safety obstacle avoidance distance as a radius of an edge node of a routing inspection point and a routing inspection surface shown in fig. 1, and if the point intersects the dam and the surrounding three-dimensional point cloud exists, deleting the routing inspection point and the edge node of the routing inspection surface;
fig. 3 is a detailed flow diagram of an embodiment of the step of automatically planning the unmanned aerial vehicle inspection route and constructing the dam unmanned aerial vehicle automatic safety inspection route according to the finally determined inspection point and the edge node of the inspection surface in fig. 1 by adopting a path planning algorithm, edge point transition and preferential inspection in the inspection surface;
FIG. 4 is a flowchart illustrating a detailed process of an embodiment of the step shown in FIG. 3, in which routing inspection surfaces reachable by edge node connections are treated as the same class, and routing inspection is separately required for routing inspection surfaces of different classes;
fig. 5 is a detailed flowchart of an embodiment of the routing inspection path planning step for the routing inspection surfaces of the same category shown in fig. 3;
FIG. 6 is a functional block diagram of an embodiment of a rapid and secure acquisition system for dam surface image unmanned aerial vehicle according to the present invention;
FIG. 7 is a functional block diagram of one embodiment of a deletion module shown in FIG. 6;
FIG. 8 is a functional block diagram of one embodiment of the building block shown in FIG. 6;
FIG. 9 is a functional block diagram of an embodiment of the category calculating unit shown in FIG. 8;
fig. 10 is a functional block diagram of an embodiment of the path rule unit shown in fig. 8.
The reference numbers illustrate:
10. establishing a module; 20. a partitioning module; 30. a merging module; 40. a planning module; 50. a deletion module; 60. building a module; 51. a query unit; 52. a discarding unit; 61. a category calculation unit; 62. a path rule unit; 611. constructing a subunit; 612. selecting a subunit; 613. a recycling subunit; 614. updating the subunit; 615. a classification subunit; 621. a first calculation subunit; 622. a second calculation subunit.
Detailed Description
In order to better understand the technical solution, the technical solution will be described in detail with reference to the drawings and the specific embodiments.
As shown in fig. 1, a first embodiment of the present invention provides an unmanned aerial vehicle rapid and safe acquisition method for dam surface images, which includes the following steps:
and S100, acquiring the data of the dam surface and peripheral scanning point clouds and oblique photography dense matching point clouds by adopting an unmanned aerial vehicle carrying a three-dimensional laser radar and a high-definition camera, and then establishing a dam three-dimensional finite element refined model by combining site and design data.
The dam three-dimensional finite element refined model mainly takes design data as a main part, the integrity of a building is checked in combination with a site, and the finite element model can be well matched with the three-dimensional space position of current point cloud data only by calculating corresponding working conditions, so that the point cloud data has a reference function when the finite element model is modeled, and has a checking function when a deformation calculation result is analyzed.
And S200, calculating the stress deformation of the dam finite element model under a given working condition, and performing partition management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result.
The partition management is not unique and is formulated according to finite element calculation results under given working conditions; the regions may be generally classified into core regions and non-core regions according to two categories. The calculation of the stress deformation field and the plastic region by finite calculations is conventional in the art and will not be described in detail herein.
S300, extracting dam surface nodes and the information of the dam surface nodes and the corresponding units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal lines of the unit surfaces to form routing inspection surfaces, and integrating the routing inspection surface information into dam three-dimensional point cloud data.
The translation distance is determined according to the shooting object, the camera imaging quality and the subarea acquisition area; for the core area, a smaller translation distance can be set or the zoom multiple of the camera can be increased to ensure the shooting quality of the inspection object, such as cracks; for the non-core area, a larger translation distance can be set to reduce the number of shot images, so that the shooting efficiency is improved; the translation distance cannot be less than 2 meters. The integration of the inspection surface information and the dam three-dimensional point cloud data mainly depends on establishing the same three-dimensional coordinate system, and corresponding coordinate transformation is required for different three-dimensional coordinate systems.
And S400, planning internal inspection points of the inspection surface, including spatial positions and camera orientations, according to the inspection surface, the nodes on the surface and the external normal information, so as to ensure the integrity of surface image acquisition.
The internal inspection point of the inspection surface unit is determined by the area of the surface unit and the shooting area of the camera, the unit surface is generally a plane for the dam, even if the arch dam is an arc-shaped surface, the surface unit of a local area can be assumed to be the plane, the external rectangle of the inspection surface is generated, the size of the shooting area of the camera is used as a sliding window to slide on the external rectangle to obtain an image, and the image can be completely obtained. The central point of the sliding window is the inspection point, and if the sliding window is not in the surface unit, the sliding window is not regarded as the inspection point. Preferably, the inspection unit surfaces which have a common node and the same normal direction and are in the same partition can be merged to reduce the number of inspection points, so that the inspection efficiency is improved.
And S500, calculating whether the edge nodes of the inspection point and the inspection surface exist points intersected with the dam and the surrounding three-dimensional point cloud within a sphere with the safety obstacle avoidance distance as the radius, and if so, deleting the inspection point and the edge nodes of the inspection surface.
Preferably, referring to fig. 2, step S500 includes:
and step S510, constructing KD-Tree for the dam surface and surrounding point cloud data, and inquiring the points closest to the inspection points and the edge nodes in the surface point cloud data.
And step S520, calculating the distance of the point according to the point which is closest to the inspection point and the edge node, and discarding the inspection point and the edge node of the inspection surface when the distance is smaller than the safe obstacle avoidance distance.
The safe obstacle avoidance distance is different according to different values of objects, and if the power transmission line needs to be divided according to voltage grades.
And the safe obstacle avoidance distance of other general buildings is not less than 2 meters, points closest to the inspection point and the edge node in the surface point cloud data are inquired by constructing KD-Tree for the surface point cloud data and the peripheral point cloud data of the dam, the Euclidean distance between the two points is calculated by utilizing the three-dimensional coordinates of the point cloud, and when the distance is less than the safe obstacle avoidance distance, the inspection point and the edge node are abandoned.
And S600, automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface, and constructing the dam unmanned aerial vehicle automatic safety inspection route by adopting a path planning algorithm, edge point transition and inspection-in-surface priority inspection modes.
Preferably, referring to fig. 3 to 5, step S600 includes:
step 610, regarding the routing inspection surfaces in which the edge node connection can be reached as the same category, and routing inspection surfaces of different categories need to be separately routed:
step 620, routing inspection path planning is carried out on the routing inspection surfaces of the same category;
referring to fig. 4, step 610 includes:
and 611, constructing a set C containing all the routing inspection surfaces.
And step 612, selecting one routing inspection surface from the set C as an initial routing inspection surface, searching for other routing inspection surfaces in the set C, regarding routing inspection surfaces with common edge nodes with the initial routing inspection surface as the same category, and otherwise, regarding routing inspection surfaces without common edge nodes with the initial routing inspection surface as a single category, and skipping to step 614.
Step 613, circularly searching other polling surfaces in the set C, and regarding the polling surfaces with the edge nodes common to the types in the step 612 as the same type until all the polling surfaces in the set C are traversed or the manually specified upper limit number of the polling surfaces of the same type is reached.
Step 614, updating the set C, that is, the set C of new routing inspection surfaces created by removing the routing inspection surfaces included in the category in step 613 from the set C in step 613.
And 615, repeating the steps 612 to 614 until all the inspection surfaces are classified into different types.
Referring to fig. 5, step 620 includes:
621, forming centroids of inspection points in the inspection surfaces into a path planning point set, constructing a point set and a weighted undirected complete graph with the distance between the two points as a weight, designating a starting point and an end point, and solving a shortest Hamilton path between the two points; when the inspection surfaces are transferred, the inspection points are transferred in a mode that the edge points are intermediate transition points; for adjacent routing inspection surfaces, the sum of the shortest distances from the intersected edge points to the routing inspection points of the two routing inspection surfaces is obtained, and the edge point corresponding to the minimum value is the selected edge point; for nonadjacent routing inspection surfaces, firstly, constructing a weighted undirected graph by using all edge points and connecting paths formed on the routing inspection surfaces, then calculating the edge point with the shortest distance between the two routing inspection surfaces, and preferably calculating the shortest path between the two points by using the edge point at one end as a starting point and the edge point at the other end as a terminal point by using a Floyd algorithm.
Step 622, using the inspection point with the shortest connection distance with the edge point as the end point of one inspection surface and the start point of the other inspection surface, if the start point and the end point of the same inspection surface are the same point, modifying the end point, selecting the point with the second shortest connection distance with the edge point of the end point as a new end point, and calculating the start point and the end point of all the inspection surfaces; and constructing a weighted undirected complete graph with the distance between the two points as a weight value for the routing inspection points in the same routing inspection plane, knowing a starting point and an end point, and solving the shortest Hamilton path of the two points.
Compared with the prior art, the unmanned aerial vehicle is adopted to carry a three-dimensional laser radar and a high-definition camera to obtain dam surface and peripheral scanning point cloud and oblique photography dense matching point cloud data, and then a dam three-dimensional finite element refined model is established by combining site and design data; calculating the stress deformation of the dam finite element model under a given working condition, and carrying out partition management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result; extracting dam surface nodes and the information of the dam surface nodes and the affiliated units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal lines of the unit surfaces to form routing inspection surfaces, and fusing the routing inspection surface information into dam three-dimensional point cloud data; planning internal inspection points of the inspection surface, including a spatial position and a camera position, according to the inspection surface, the node on the surface and the external normal information so as to ensure the integrity of surface image acquisition; calculating whether the edge nodes of the inspection point and the inspection surface exist points intersected with the dam and the surrounding three-dimensional point cloud within a sphere with the safety obstacle avoidance distance as the radius, and if so, deleting the inspection point and the edge nodes of the inspection surface; and automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface, and constructing the dam unmanned aerial vehicle automatic safety inspection route by adopting a path planning algorithm, edge point transition and a mode of preferential inspection in the inspection surface. According to the unmanned aerial vehicle rapid and safe acquisition method for the dam surface image, the dam surface image is acquired by combining three-dimensional point cloud data of the unmanned aerial vehicle and a finite element, so that automatic image acquisition can be realized, a key area can be focused according to the stress characteristics of a dam structure, and inspection time is saved; the automatic routing method for the inspection points and the inspection route is provided by combining the attributes of the finite element units and the three-dimensional point cloud information, and meanwhile, an inspection mode of unit edge point transition and preferential inspection in the inspection surface is provided, so that the unmanned aerial vehicle and the surface of the dam are always kept at a safe distance, the close-range image acquisition is ensured to be safe and efficient, and the method is particularly suitable for inspection areas with complex and variable structures.
Further, please refer to fig. 6, where fig. 6 is a functional block diagram of an embodiment of a rapid and secure acquisition system for an image of a dam surface by an unmanned aerial vehicle according to the present invention, in this embodiment, the rapid and secure acquisition system for an image of a dam surface by an unmanned aerial vehicle includes an establishing module 10, a partitioning module 20, an integrating module 30, a planning module 40, a deleting module 50, and a constructing module 60, where the establishing module 10 is configured to obtain data of a cloud of scanned points and a cloud of oblique dense matching points on the dam surface and its periphery by using the unmanned aerial vehicle carrying a three-dimensional laser radar and a high-definition camera, and then establish a three-dimensional finite element refinement model of the dam by combining with site and design data. And the partitioning module 20 is used for calculating the stress deformation of the dam finite element model under a given working condition, and performing partitioning management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result. And the blending module 30 is used for extracting dam surface nodes and the information of the dam surface nodes and the affiliated units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal lines of the unit surfaces to form routing inspection surfaces, and blending the routing inspection surface information into the dam three-dimensional point cloud data. And the planning module 40 is used for planning internal inspection points of the inspection surface, including spatial positions and camera orientations, according to the inspection surface, the nodes on the surface and the external normal information, so as to ensure the integrity of surface image acquisition. And the deleting module 50 is used for calculating whether the edge nodes of the inspection point and the inspection surface exist points intersected with the dam and the surrounding three-dimensional point cloud within a sphere with the safety obstacle avoidance distance as the radius, and if so, deleting the inspection point and the edge nodes of the inspection surface. And the construction module 60 is used for automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface where the inspection point is located by adopting a path planning algorithm, edge point transition and a mode of preferentially inspecting in the inspection surface, and constructing the dam unmanned aerial vehicle automatic safety inspection route.
Preferably, referring to fig. 6 to 10, in this embodiment, the dam three-dimensional finite element refinement model established by the establishing module 10 is based on design data, and integrity of a building is checked in combination with a site, and since the finite element model needs to calculate corresponding working conditions to be well matched with the three-dimensional space position of the current point cloud data, the point cloud data plays a reference role in modeling the finite element, and plays a check role in analyzing a deformation calculation result.
The partitioning module 20 includes a partitioning unit, wherein the partitioning unit is configured to partition a region into a core region and a non-core region according to a finite element calculation result under a given working condition. The partition management is not unique and is established according to finite element calculation results under given working conditions; the regions may be generally classified into core regions and non-core regions according to two categories.
The blending module 30 includes a translation unit, wherein the translation unit is configured to determine a translation distance jointly according to the shooting object, the camera imaging quality, and the partition acquisition area; setting a smaller translation distance or increasing the zooming times of the camera to ensure the shooting quality of the inspection object in the core area; for the non-core area, a larger translation distance is set to reduce the number of captured images. The translation distance is determined according to the shooting object, the camera imaging quality and the partitioned acquisition area; for the core area, a smaller translation distance can be set or the zoom multiple of the camera can be increased to ensure the shooting quality of the inspection object, such as cracks; for the non-core area, a larger translation distance can be set to reduce the number of shot images, so that the shooting efficiency is improved; the translation distance cannot be less than 2 meters. The integration of the inspection surface information and the dam three-dimensional point cloud data mainly depends on establishing the same three-dimensional coordinate system, and corresponding coordinate transformation is required for different three-dimensional coordinate systems.
The internal routing inspection point setting of the routing inspection surface unit in the planning module 40 is determined by the area of the surface unit and the shooting area of the camera, for a dam, the unit surface is generally a plane, even if the surface is an arc-shaped surface like an arch dam, the surface unit of a local area can be assumed to be the plane, and by generating a circumscribed rectangle of the routing inspection unit surface, the size of the shooting area of the camera is taken as a sliding window to slide the circumscribed rectangle to obtain an image, so that the image can be completely obtained. The central point of the sliding window is the inspection point, and if the sliding window is not in the surface unit, the sliding window is not regarded as the inspection point. Preferably, the inspection unit surfaces which have a common node and the same normal direction and are in the same partition can be merged to reduce the number of inspection points, so that the inspection efficiency is improved.
Referring to fig. 7, fig. 7 is a functional module schematic diagram of an embodiment of the deletion module shown in fig. 6, in the embodiment, the deletion module 50 includes a query unit 51 and a discarding unit 52, where the query unit 51 is configured to construct KD-Tree for dam surface and peripheral point cloud data, and query points closest to the inspection point and the edge node in the surface point cloud data. The discarding unit 52 is configured to calculate an euclidean distance between two points according to the searched points closest to the inspection point and the edge node and by using the three-dimensional coordinates of the point cloud, and discard the inspection point and the edge node of the inspection surface when the distance is smaller than the safety obstacle avoidance distance.
In this embodiment, the safe obstacle avoidance distance is different according to different values of objects, for example, the power transmission line needs to be divided according to voltage classes.
And the safe obstacle avoidance distance of other general buildings is not less than 2 meters, points closest to the inspection point and the edge node in the surface point cloud data are inquired by constructing KD-Tree for the surface point cloud data and the peripheral point cloud data of the dam, the Euclidean distance between the two points is calculated by utilizing the three-dimensional coordinates of the point cloud, and when the distance is less than the safe obstacle avoidance distance, the inspection point and the edge node are abandoned.
Referring to fig. 8 and fig. 8 are schematic functional module diagrams of an embodiment of the building module shown in fig. 6, in this embodiment, the building module 60 includes a category calculating unit 61 and a path rule unit 62, where the category calculating unit 61 is configured to regard routing inspection surfaces where edge node connections are reachable as a same category, and routing inspection surfaces of different categories need to be separately routed. And the path rule unit 62 is used for planning the routing inspection paths for the routing inspection surfaces of the same category.
Referring to fig. 9, fig. 9 is a functional module schematic diagram of an embodiment of the category calculating unit shown in fig. 8, in which the category calculating unit 61 includes a constructing subunit 611, a selecting subunit 612, a circulating subunit 613, an updating subunit 614, and a classifying subunit 615,
a construction subunit 611, configured to construct a set C including all inspection surfaces; a loop subunit 613, an update subunit 614 and
a selecting subunit 612, configured to select one inspection surface from the set C as an initial inspection surface, search for other inspection surfaces in the set C, regard inspection surfaces having edge nodes common to the initial inspection surface as a same category, and otherwise, regard inspection surfaces having no edge nodes common to the initial inspection surface as a single category;
a circulation subunit 613, configured to circularly search for other inspection surfaces in the set C, and regard the inspection surfaces having the same edge node as the category of the selection subunit as the same category until all the inspection surfaces in the set C are traversed or the manually specified upper limit number of the inspection surfaces in the same category is reached;
the updating subunit 614 is configured to update the set C, that is, a new routing inspection surface set created by removing the routing inspection surfaces included in the category of the cyclic subunit from the set C of the cyclic subunit;
and the classification subunit 615 is used for repeating the class calculation until all the routing inspection surfaces are classified into classes.
Referring to fig. 10, the path rule unit 62 includes:
the first calculating subunit 621 is configured to form a path planning point set by using the centroids of the inspection points in the inspection surfaces, construct a weighted undirected complete graph with the point set and the distance between the two points as a weight, specify a starting point and an end point, and solve the shortest Hamilton path between the two points; when the inspection surfaces are transferred, the inspection points are transferred in a mode that the edge points are intermediate transition points; for adjacent routing inspection surfaces, the sum of the shortest distances from the intersected edge points to the routing inspection points of the two routing inspection surfaces is obtained, and the edge point corresponding to the minimum value is the selected edge point; for nonadjacent routing inspection surfaces, firstly, constructing a weighted undirected graph by using all edge points and connecting paths formed on the routing inspection surfaces, then solving the edge point with the shortest distance between the two routing inspection surfaces, and preferably solving by using a Floyd algorithm, wherein the shortest path between the two points is calculated by using the edge point at one end as a starting point and the edge point at the other end as a terminal point;
a second calculating subunit 622, configured to use the routing inspection point with the shortest connection distance to the edge point as an end point of one routing inspection surface and a start point of another routing inspection surface, modify the end point if the start point and the end point of the same routing inspection surface are the same point, select a point with the next shortest connection distance to the edge point of the end point as a new end point, and obtain the start point and the end point of all routing inspection surfaces; and constructing a weighted undirected complete graph with the distance between the two points as a weight value for the routing inspection points in the same routing inspection plane, knowing a starting point and an end point, and solving the shortest Hamilton path of the two points.
Compared with the prior art, the unmanned aerial vehicle is adopted to carry a three-dimensional laser radar and a high-definition camera to obtain dam surface and peripheral scanning point cloud and oblique photography dense matching point cloud data, and then a dam three-dimensional finite element refined model is established by combining site and design data; calculating the stress deformation of the dam finite element model under a given working condition, and carrying out partition management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result; extracting dam surface nodes and the information of the dam surface nodes and the affiliated units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal lines of the unit surfaces to form routing inspection surfaces, and fusing the routing inspection surface information into dam three-dimensional point cloud data; planning internal inspection points of the inspection surface, including a spatial position and a camera position, according to the inspection surface, the node on the surface and the external normal information so as to ensure the integrity of surface image acquisition; calculating whether the edge nodes of the inspection point and the inspection surface exist points intersected with the dam and the surrounding three-dimensional point cloud within a sphere with the safety obstacle avoidance distance as the radius, and if so, deleting the inspection point and the edge nodes of the inspection surface; and automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface, and constructing the dam unmanned aerial vehicle automatic safety inspection route by adopting a path planning algorithm, edge point transition and a mode of preferential inspection in the inspection surface. According to the unmanned aerial vehicle rapid and safe acquisition system for the dam surface image, the dam surface image is acquired by combining three-dimensional point cloud data of the unmanned aerial vehicle and a finite element, so that automatic image acquisition can be realized, a key area can be focused according to the stress characteristics of a dam structure, and inspection time is saved; the automatic routing method for the inspection points and the inspection route is provided by combining the attributes of the finite element units and the three-dimensional point cloud information, and meanwhile, an inspection mode of unit edge point transition and preferential inspection in the inspection surface is provided, so that the unmanned aerial vehicle and the surface of the dam are always kept at a safe distance, the close-range image acquisition is ensured to be safe and efficient, and the method is particularly suitable for inspection areas with complex and variable structures.
The embodiment provides a dam surface image unmanned aerial vehicle rapid and safe acquisition method and system, which can realize rapid and safe acquisition of dam surface images and are particularly suitable for routing inspection areas with complex and variable structures. In the aspect of time consumption, according to the proportion of 0.5 to 0.5 between a core area and a non-core area, according to the hydraulic concrete design specification SL191, the core area is used for detecting 0.1 mm-width cracks at minimum, the non-core area is used for detecting 0.3 mm-width cracks of 2 types of reinforced concrete according to the environment category as required, the time consumption of the flight process of each inspection point is not calculated, namely the time consumption is assumed to be the same, the number of the inspection points of the non-core area and the time consumption are known to be 1/3 times of the inspection points of the core area, and the time consumption of the acquisition method is 0.5 x 1+0.5/3 times of the 0.1 mm-width cracks detected at minimum in the whole area as 0.67 times. Since the dam is an statically indeterminate structure, the core stress or damage range is generally small, and if the core area and the non-core area are in a ratio of 0.1 to 0.9, the time consumption is 0.1 x 1+0.9/3 times that of the full-area minimum detection of 0.1 mm-level cracks, namely 0.4 times. If daily inspection is performed only on the core area, and the proportion is calculated according to 0.1, the time consumption is 0.1 × 1+0/3 times of the minimum detection 0.1mm level crack in the whole area.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention. It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.

Claims (10)

1. An unmanned aerial vehicle rapid and safe acquisition method for dam surface images is characterized by comprising the following steps:
an unmanned aerial vehicle is adopted to carry a three-dimensional laser radar and a high-definition camera to obtain the data of the surface and peripheral scanning point clouds of the dam and the point cloud data of oblique photography dense matching, and then a three-dimensional finite element refined model of the dam is established by combining site and design data;
calculating the stress deformation of the dam finite element model under a given working condition, and carrying out partition management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result;
extracting dam surface nodes and the information of the dam surface nodes and the affiliated units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal lines of the unit surfaces to form routing inspection surfaces, and fusing the routing inspection surface information into dam three-dimensional point cloud data;
planning internal inspection points of the inspection surface, including a spatial position and a camera position, according to the inspection surface, the node on the surface and the external normal information so as to ensure the integrity of surface image acquisition;
calculating whether the edge nodes of the inspection point and the inspection surface exist points intersected with the dam and the surrounding three-dimensional point cloud within a sphere with the safety obstacle avoidance distance as the radius, and if so, deleting the inspection point and the edge nodes of the inspection surface;
and automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface, and constructing the dam unmanned aerial vehicle automatic safety inspection route by adopting a path planning algorithm, edge point transition and a mode of preferential inspection in the inspection surface.
2. The unmanned aerial vehicle rapid and safe acquisition method of dam surface images as claimed in claim 1, wherein the step of calculating the stress deformation of the dam finite element model under the given working condition, and performing partition management on the acquisition region by using the elements of the finite element model as basic units according to the calculation result, wherein the partition management is formulated according to the finite element calculation result under the given working condition, and the region is divided into a core region and a non-core region.
3. The unmanned aerial vehicle rapid and safe acquisition method of dam surface images according to claim 1, characterized in that in the step of extracting dam surface nodes and the information of the dam surface nodes and the corresponding units in the finite element model under the current deformation condition, constructing surface units, translating the unit surfaces according to the direction of the external normal line to form routing inspection surfaces, and integrating the routing inspection surface information into dam three-dimensional point cloud data, the translation distance is determined jointly according to the shooting object, the camera imaging quality and the partition acquisition area; setting a smaller translation distance or increasing the zooming times of the camera to ensure the shooting quality of the inspection object in the core area; for the non-core area, a larger translation distance is set to reduce the number of captured images.
4. The unmanned aerial vehicle rapid and safe collection method for the dam surface image according to claim 1, wherein the step of calculating whether the edge nodes of the inspection point and the inspection surface have points intersecting with the dam and the surrounding three-dimensional point cloud within a sphere range taking the safe obstacle avoidance distance as the radius, and if so, deleting the inspection point and the edge nodes of the inspection surface comprises the following steps:
constructing KD-Tree for the dam surface and surrounding point cloud data, and inquiring the points closest to the inspection points and the edge nodes in the surface point cloud data;
and calculating the distance according to the inquired points closest to the inspection point and the edge node, and discarding the inspection point and the edge node of the inspection surface when the distance is smaller than the safe obstacle avoidance distance.
5. The unmanned aerial vehicle rapid and safe collection method for dam surface images according to claim 1, wherein the step of automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface where the inspection point is located by adopting a path planning algorithm, edge point transition and preferential inspection in the inspection surface, and constructing the dam unmanned aerial vehicle automatic and safe inspection route comprises the following steps:
step 610, regarding the routing inspection surfaces in which the edge node connection can be reached as the same category, and routing inspection surfaces of different categories need to be separately routed:
step 620, routing inspection path planning is carried out on the routing inspection surfaces of the same category;
the step 610 includes:
611, constructing a set C containing all the inspection surfaces;
step 612, selecting one routing inspection surface from the set C as an initial routing inspection surface, searching for other routing inspection surfaces in the set C, regarding routing inspection surfaces with common edge nodes with the initial routing inspection surface as the same category, and otherwise, regarding routing inspection surfaces without common edge nodes with the initial routing inspection surface as a single category, and jumping to step 614;
step 613, circularly searching other polling surfaces in the set C, and regarding the polling surfaces with the same edge nodes as the type in the step 612 as the same type until all the polling surfaces in the set C are traversed or the manually specified upper limit number of the polling surfaces of the same type is reached;
step 614, updating the set C, that is, a new routing inspection surface set established by removing the routing inspection surfaces included in the category in step 613 from the set C in step 613;
615, repeating 612-614 until all the inspection surfaces are classified;
the step 620 includes:
621, forming centroids of inspection points in the inspection surfaces into a path planning point set, constructing a point set and a weighted undirected complete graph with the distance between the two points as a weight, designating a starting point and an end point, and solving a shortest Hamilton path between the two points; when the inspection surfaces are transferred, the inspection points are transferred in a mode that the edge points are intermediate transition points; for adjacent routing inspection surfaces, the sum of the shortest distances from the intersected edge points to the routing inspection points of the two routing inspection surfaces is obtained, and the edge point corresponding to the minimum value is the selected edge point; for nonadjacent routing inspection surfaces, firstly, constructing a weighted undirected graph by using all edge points and connecting paths formed on the routing inspection surfaces, then calculating the edge point with the shortest distance between the two routing inspection surfaces, and preferably calculating the shortest path between the two points by using the edge point at one end as a starting point and the edge point at the other end as a terminal point by using a Floyd algorithm;
step 622, using the inspection point with the shortest connection distance with the edge point as the end point of one inspection surface and the start point of the other inspection surface, if the start point and the end point of the same inspection surface are the same point, modifying the end point, selecting the point with the second shortest connection distance with the edge point of the end point as a new end point, and calculating the start point and the end point of all the inspection surfaces; and constructing a weighted undirected complete graph with the distance between the two points as a weight value for the routing inspection points in the same routing inspection plane, knowing a starting point and an end point, and solving the shortest Hamilton path of the two points.
6. The utility model provides a quick safe collection system of dam surface image unmanned aerial vehicle which characterized in that includes:
the system comprises an establishing module (10) and a control module, wherein the establishing module is used for acquiring dam surface and peripheral scanning point cloud and oblique photography dense matching point cloud data by adopting an unmanned aerial vehicle carrying a three-dimensional laser radar and a high-definition camera, and then establishing a dam three-dimensional finite element refined model by combining site and design data;
the partitioning module (20) is used for calculating the stress deformation of the dam finite element model under a given working condition and performing partitioning management on the acquisition region by taking the elements of the finite element model as basic units according to the calculation result;
the merging module (30) is used for extracting dam surface nodes and the information of the dam surface nodes and the affiliated units in the finite element model under the current deformation working condition, constructing surface units, translating the unit surfaces according to the direction of the external normal lines of the unit surfaces to form routing inspection surfaces, and merging the routing inspection surface information into the dam three-dimensional point cloud data;
the planning module (40) is used for planning internal inspection points of the inspection surface, including spatial positions and camera orientations, according to the inspection surface, the nodes on the surface and the external normal information so as to ensure the integrity of surface image acquisition;
the deleting module (50) is used for calculating whether the edge nodes of the inspection point and the inspection surface exist points intersected with the dam and the surrounding three-dimensional point cloud within a sphere taking the safety obstacle avoidance distance as the radius, and if the edge nodes exist points, deleting the inspection point and the edge nodes of the inspection surface;
and the construction module (60) is used for automatically planning the unmanned aerial vehicle inspection route according to the finally determined inspection point and the edge node of the inspection surface where the inspection point is located by adopting a path planning algorithm, edge point transition and a mode of preferentially inspecting in the inspection surface, and constructing the dam unmanned aerial vehicle automatic safety inspection route.
7. The unmanned aerial vehicle rapid and safe dam surface image acquisition system according to claim 6, wherein the partition module (20) comprises a partition unit,
and the partition unit is used for partition management to be formulated according to a finite element calculation result under a given working condition and dividing the area into a core area and a non-core area.
8. The unmanned aerial vehicle rapid and safe dam surface image acquisition system according to claim 6, wherein the blending module (30) comprises a translation unit,
the translation unit is used for determining translation distance according to the shooting object, the camera imaging quality and the subarea acquisition area; setting a smaller translation distance or increasing the zooming times of the camera to ensure the shooting quality of the inspection object in the core area; for the non-core area, a larger translation distance is set to reduce the number of captured images.
9. The unmanned aerial vehicle rapid and safe dam surface image acquisition system according to claim 6, wherein the deletion module (50) comprises:
the query unit (51) is used for constructing KD-Tree for the dam surface and surrounding point cloud data and querying the points which are nearest to the inspection points and the edge nodes in the surface point cloud data;
and the abandon unit (52) is used for calculating the distance of the point closest to the inspection point and the edge node according to the inquired point, and abandoning the inspection point and the edge node of the inspection surface when the distance is smaller than the safe obstacle avoidance distance.
10. Dam surface image unmanned aerial vehicle fast and safe acquisition system according to claim 6, characterized in that the construction module (60) comprises a category calculation unit (61) and a path rule unit (62),
the class calculation unit (61) is used for regarding routing inspection surfaces in which edge node connection can be reached as the same class, and routing inspection surfaces of different classes need to be separately routed;
the path rule unit (62) is used for planning routing inspection paths for routing inspection surfaces of the same type;
the category calculation unit (61) includes:
a construction subunit (611) for constructing a set C containing all the inspection surfaces;
a selecting subunit (612) for selecting one routing inspection surface from the set C as an initial routing inspection surface, searching other routing inspection surfaces in the set C, and regarding the routing inspection surfaces with common edge nodes with the initial routing inspection surface as a same category, or regarding the routing inspection surfaces without common edge nodes with the initial routing inspection surface as a single category;
the circulating subunit (613) is used for circularly searching other routing inspection surfaces in the set C, and regarding the routing inspection surfaces with the same edge nodes as the selected subunit in the category as the same category until all the routing inspection surfaces in the set C are traversed or the manually specified upper limit number of the routing inspection surfaces in the same category is reached;
the updating subunit (614) is used for updating the set C, namely a new routing inspection surface set which is established by removing the routing inspection surfaces contained in the category of the circulating subunit from the set C of the circulating subunit;
a classification subunit (615) for repeating the class calculation until all the inspection surfaces are classified;
the path rule unit (62) comprises:
the first calculating subunit (621) is used for forming a path planning point set by centroids of the inspection points in all the inspection surfaces, constructing a point set and a weighted undirected complete graph taking the distance between the two points as a weight, designating a starting point and an end point, and solving the shortest Hamilton path between the two points; when the inspection surfaces are transferred, the inspection points are transferred in a mode that the edge points are intermediate transition points; for adjacent routing inspection surfaces, the sum of the shortest distances from the intersected edge points to the routing inspection points of the two routing inspection surfaces is obtained, and the edge point corresponding to the minimum value is the selected edge point; for nonadjacent routing inspection surfaces, firstly, constructing a weighted undirected graph by using all edge points and connecting paths formed on the routing inspection surfaces, then solving the edge point with the shortest distance between the two routing inspection surfaces, and preferably solving by using a Floyd algorithm, wherein the shortest path between the two points is calculated by using the edge point at one end as a starting point and the edge point at the other end as a terminal point;
a second calculating subunit (622) for taking the patrol inspection point with the shortest connection distance with the edge point as the end point of one patrol inspection surface and the start point of the other patrol inspection surface, if the start point and the end point of the same patrol inspection surface are the same point, modifying the end point, selecting the point with the second shortest connection distance with the edge point of the end point as a new end point, and calculating the start point and the end point of all patrol inspection surfaces; and constructing a weighted undirected complete graph with the distance between the two points as a weight value for the routing inspection points in the same routing inspection plane, knowing a starting point and an end point, and solving the shortest Hamilton path of the two points.
CN202111134127.9A 2021-09-27 2021-09-27 Dam surface image unmanned aerial vehicle rapid and safe acquisition method and system Pending CN113848560A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111134127.9A CN113848560A (en) 2021-09-27 2021-09-27 Dam surface image unmanned aerial vehicle rapid and safe acquisition method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111134127.9A CN113848560A (en) 2021-09-27 2021-09-27 Dam surface image unmanned aerial vehicle rapid and safe acquisition method and system

Publications (1)

Publication Number Publication Date
CN113848560A true CN113848560A (en) 2021-12-28

Family

ID=78980572

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111134127.9A Pending CN113848560A (en) 2021-09-27 2021-09-27 Dam surface image unmanned aerial vehicle rapid and safe acquisition method and system

Country Status (1)

Country Link
CN (1) CN113848560A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115297303A (en) * 2022-09-29 2022-11-04 国网浙江省电力有限公司 Image data acquisition and processing method and device suitable for power grid power transmission and transformation equipment
CN117218743A (en) * 2023-11-07 2023-12-12 诺比侃人工智能科技(成都)股份有限公司 Intelligent inspection control method and system based on machine vision

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115297303A (en) * 2022-09-29 2022-11-04 国网浙江省电力有限公司 Image data acquisition and processing method and device suitable for power grid power transmission and transformation equipment
CN115297303B (en) * 2022-09-29 2022-12-27 国网浙江省电力有限公司 Image data acquisition and processing method and device suitable for power grid power transmission and transformation equipment
CN117218743A (en) * 2023-11-07 2023-12-12 诺比侃人工智能科技(成都)股份有限公司 Intelligent inspection control method and system based on machine vision
CN117218743B (en) * 2023-11-07 2024-02-09 诺比侃人工智能科技(成都)股份有限公司 Intelligent inspection control method and system based on machine vision

Similar Documents

Publication Publication Date Title
US20220028163A1 (en) Computer Vision Systems and Methods for Detecting and Modeling Features of Structures in Images
CN110148196B (en) Image processing method and device and related equipment
CN109614935B (en) Vehicle damage assessment method and device, storage medium and electronic equipment
CN111968229B (en) High-precision map making method and device
Zhang et al. Rapid data quality oriented laser scan planning for dynamic construction environments
JP2022514974A (en) Object detection methods, devices, electronic devices, and computer programs
Forlani et al. C omplete classification of raw LIDAR data and 3D reconstruction of buildings
Hichri et al. Review of the “as-buit BIM” approaches
Wang et al. A bayesian approach to building footprint extraction from aerial lidar data
CN113848560A (en) Dam surface image unmanned aerial vehicle rapid and safe acquisition method and system
CN111652929A (en) Visual feature identification and positioning method and system
CN107679458B (en) Method for extracting road marking lines in road color laser point cloud based on K-Means
CN114842139A (en) Building three-dimensional digital model construction method based on spatial analysis
CN106338736A (en) Full-3D occupation volume element landform modeling method based on laser radar
KR102414307B1 (en) 3D map change area update system and method
CN112634340A (en) Method, device, equipment and medium for determining BIM (building information modeling) model based on point cloud data
CN108803659B (en) Multi-window heuristic three-dimensional space path planning method based on magic cube model
Gordon et al. Automating building element detection for deconstruction planning and material reuse: A case study
CN113804100A (en) Method, device, equipment and storage medium for determining space coordinates of target object
CN114556445A (en) Object recognition method, device, movable platform and storage medium
Nevatia et al. Automatic and interactive modeling of buildings in urban environments from aerial images
Valero et al. An integrated scan-to-bim approach for buildings energy performance evaluation and retrofitting
CN115578525B (en) Engineering line selection optimization system and method for complex environment
CN113514053B (en) Method and device for generating sample image pair and method for updating high-precision map
CN115187736A (en) Target map generation method and device, and AR map generation method and device

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
CB02 Change of applicant information
CB02 Change of applicant information

Address after: 410000 Area 217, 2nd Floor, Building A3, Maker Center, Tongguan Street, Wangcheng District, Changsha City, Hunan Province

Applicant after: Hunan Desen jiuchuang Technology Co.,Ltd.

Address before: 410203 3-812 shengtongcheng, Wangcheng District, Changsha City, Hunan Province

Applicant before: Hunan Desen jiuchuang Technology Co.,Ltd.