CN115601565A - Large-span steel structure fixed feature extraction method based on minimum valley distance - Google Patents

Large-span steel structure fixed feature extraction method based on minimum valley distance Download PDF

Info

Publication number
CN115601565A
CN115601565A CN202211609628.2A CN202211609628A CN115601565A CN 115601565 A CN115601565 A CN 115601565A CN 202211609628 A CN202211609628 A CN 202211609628A CN 115601565 A CN115601565 A CN 115601565A
Authority
CN
China
Prior art keywords
hypergraph
vertex
point cloud
steel structure
span steel
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202211609628.2A
Other languages
Chinese (zh)
Other versions
CN115601565B (en
Inventor
孙长银
李建
吴巧云
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.)
Anhui University
Original Assignee
Anhui University
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 Anhui University filed Critical Anhui University
Priority to CN202211609628.2A priority Critical patent/CN115601565B/en
Publication of CN115601565A publication Critical patent/CN115601565A/en
Application granted granted Critical
Publication of CN115601565B publication Critical patent/CN115601565B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/30—Noise filtering
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

The invention discloses a method for extracting fixed features of a large-span steel structure based on a minimum valley distance, which comprises the following steps: acquiring point cloud data of a large-span steel structure; filtering the point cloud data; dividing the point cloud after filtering by using an octree, and performing parallel processing on octree leaf nodes by using a multithreading technology; constructing a hypergraph, and determining the vertex and the edge of the hypergraph; calculating the residual error of each vertex to each super edge thereof, wherein the residual error is measured by the Simpson distance; evaluating vertices of the hypergraph; simplifying the hypergraph based on the evaluation result of the vertex and the entropy of the prior probability of the vertex, and deleting the vertex which does not meet the condition; calculating the minimum valley distance of the hypergraph vertex based on the preference of the hypergraph vertex on the hyperedge; selecting a hypergraph vertex based on the minimum valley distance, wherein the parameter in the hypergraph vertex corresponds to the shape parameter of the point cloud, and determining the point cloud fixing characteristic. The extraction of the fixed characteristics of the large-span steel structure can be rapidly and accurately realized.

Description

Large-span steel structure fixed feature extraction method based on minimum valley distance
Technical Field
The invention belongs to the technical field of building structure monitoring, and particularly relates to a method for extracting fixed features of a large-span steel structure based on a minimum valley distance.
Background
With the rapid development of national economy, the demand of people for large public buildings is higher and higher. The steel structure has high strength, light weight, good plasticity, low cost, short construction period, high mechanization degree and convenient installation, so the steel structure is widely applied to large buildings such as airports, gymnasiums, bridges and the like. However, the construction procedure and the construction process of the large-span steel structure building are heavy, and in addition, the general space span of the building using the steel structure is large, so the steel structure is easy to deform after long-time bearing, and therefore, the large-span steel structure building needs to be quickly and effectively detected.
The fixed characteristics of the large-span steel structure refer to the axial characteristics of the pipeline and the spherical center characteristics of the welding balls, and the characteristics can comprehensively and accurately reflect the overall deformation condition of the large-span steel structure and are important carriers for deformation detection. However, some difficulties are faced in the extraction of fixed features. First, scanning a large-span steel structure results in a very large amount of point cloud data, which can be very time consuming to process. Secondly, the point cloud data obtained by scanning has a certain amount of noise and abnormal points due to various reasons, and the noise and the abnormal points have certain influence on the accuracy of fixed feature extraction.
Aiming at the problem that the fixed characteristic extraction of a large-span steel structure is difficult to realize quickly and accurately in the prior art, an effective solution is not provided at present.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a method for extracting the fixed features of the large-span steel structure based on the minimum valley cost distance, so that the fixed features of the large-span steel structure can be quickly and accurately extracted, the field work time is greatly reduced, and the shape features of the large-span steel structure can be well reflected.
In order to achieve the technical purpose, the technical scheme adopted by the invention is as follows:
a method for extracting large-span steel structure fixed features based on a minimum valley distance, wherein the fixed features comprise axis features of a steel structure pipeline and center features of a steel structure welding ball, and the method comprises the following steps:
scanning the large-span steel structure by using a three-dimensional laser scanner to obtain point cloud data of the large-span steel structure;
filtering the acquired large-scale point cloud data;
dividing the point cloud after filtering by using an octree, and performing parallel processing on leaf nodes of the octree by using a multithreading technology, namely each thread corresponds to one leaf node of the octree;
constructing a hypergraph for leaf nodes by using the point cloud data in each thread, and determining vertexes and edges of the hypergraph;
calculating residual errors of each vertex to each superedge thereof, measured by the Simpson distance, based on the constructed hypergraph;
evaluating the vertex of the hypergraph based on the residual error of the Simpson distance measurement and a kernel function;
simplifying the hypergraph based on the evaluation result of the vertex and the entropy of the prior probability of the vertex, and deleting the vertex which does not meet the condition;
calculating the minimum valley distance of the hypergraph vertex based on the preference of the hypergraph vertex on the hyperedge;
selecting a hypergraph vertex based on the minimum valley distance, wherein the parameter in the hypergraph vertex corresponds to the shape parameter of the point cloud, and determining the point cloud fixing characteristic.
In order to optimize the technical scheme, the specific measures adopted further comprise:
foretell use three-dimensional laser scanner scans the large-span steel construction, acquires the point cloud data of large-span steel construction, includes: the method comprises the steps of scanning a steel structure by using a three-dimensional laser scanner at a plurality of stations around a large-span steel structure to obtain different point cloud data of a plurality of positions, splicing the different point cloud data of the plurality of positions, and obtaining the integral point cloud data of the large-span steel structure.
The filtering process for the acquired large-scale point cloud data includes: for the data of the ground and the wall surface, detecting the curvature, and removing the part of the curvature which is lower than a threshold value; and removing the system noise by adopting a statistical filtering algorithm.
The above-mentioned point cloud after using the octree to divide and filter, and use multithread technology to process the octree leaf node in parallel, that is, each thread corresponds to a leaf node of the octree, including: the method comprises the steps of wrapping point cloud by using a minimum cube, dividing the cube into 8 equally divided sub-squares, equally dividing each non-empty sub-square again, stopping equally dividing a certain node when the point cloud number of the node is equal to the point cloud number of a parent node of the node, stopping the whole dividing process when the preset minimum edge length of the cube is reached, and simultaneously processing a plurality of leaf nodes in all the subsequent steps by using a multithreading technology.
The above-mentioned building a hypergraph by using point cloud data for leaf nodes inside each thread, and determining the vertexes and edges of the hypergraph, includes: all points in the point cloud are used as the super edges of the hypergraph; selecting a point set which can determine the minimum number of point cloud shapes from the point clouds, and determining parameters of the point cloud shapes by adopting the point set; the determined parameter is used as a vertex of the hypergraph; the vertex represents a space shape, euclidean distances from all points in the point cloud to the surface of the space shape table are calculated, when the distances are smaller than a certain threshold value, the corresponding points are inner points, and otherwise, the corresponding points are outer points; all the inner points are used as the corresponding super edges of the vertexes and are connected with each vertex and the super edge thereof in the super graph; repeating the steps for multiple times to establish a hypergraph of the point cloud.
Calculating the residual error of each vertex to each superedge thereof measured by the Simpson distance based on the constructed hypergraph, which comprises the following steps: first, for each vertex in the hypergraph
Figure 846255DEST_PATH_IMAGE001
And a super edge thereof
Figure 287601DEST_PATH_IMAGE002
To obtain its residual terms
Figure 681673DEST_PATH_IMAGE003
Wherein, in the step (A),
Figure 250189DEST_PATH_IMAGE004
is an expression of the shape represented by the point cloud; the residual term is then measured by the Simpson distance, i.e.
Figure 46107DEST_PATH_IMAGE005
Wherein the content of the first and second substances,
Figure 923933DEST_PATH_IMAGE006
。
the above-mentioned vertex of the hypergraph is evaluated based on the residual error of the simpson distance measure and the kernel function, and the vertex comprises: for each vertex in the hypergraph
Figure 805301DEST_PATH_IMAGE007
Using the formula:
Figure 911928DEST_PATH_IMAGE008
the vertices are evaluated, wherein,
Figure 562353DEST_PATH_IMAGE009
is a vertex
Figure 611080DEST_PATH_IMAGE007
Degree of (i.e. vertex)
Figure 979744DEST_PATH_IMAGE001
The number of corresponding interior points;
Figure 811434DEST_PATH_IMAGE010
has a value of
Figure 191731DEST_PATH_IMAGE011
Is that
Figure 286726DEST_PATH_IMAGE001
At an inner point of 1, in
Figure 267320DEST_PATH_IMAGE011
Is that
Figure 840384DEST_PATH_IMAGE007
The outer point of (a) is 0;
Figure 343697DEST_PATH_IMAGE012
is I Pan Nieqie familyA Function;
Figure 609593DEST_PATH_IMAGE013
is the residual measured in simpson distance;
Figure 811904DEST_PATH_IMAGE014
is a vertex
Figure 454238DEST_PATH_IMAGE007
The corresponding inner points account for the proportion of the point cloud in the total point cloud;
Figure 809127DEST_PATH_IMAGE015
whereinnIs the number of the excess edges.
The above entropy based on the evaluation result of the vertex and the prior probability of the vertex simplifies the hypergraph, and deletes the vertex which does not satisfy the condition, including: the set of vertices of the hypergraph is
Figure 245925DEST_PATH_IMAGE016
The evaluation result obtained by evaluating each vertex is:
Figure 935532DEST_PATH_IMAGE017
，
Figure 850398DEST_PATH_IMAGE018
the number of vertexes in the hypergraph; computing vertices
Figure 325373DEST_PATH_IMAGE019
And the difference between the maximum evaluation result and the evaluation result of (c):
Figure 933072DEST_PATH_IMAGE020
(ii) a Computing vertices
Figure 109975DEST_PATH_IMAGE019
Prior probability of (2)
Figure 828533DEST_PATH_IMAGE021
: when in use
Figure 158014DEST_PATH_IMAGE022
Is greater than 0 and the content of the active ingredient,
Figure 936614DEST_PATH_IMAGE023
otherwise, otherwise
Figure 600814DEST_PATH_IMAGE021
Is a positive constant; the entropy of all vertices is then calculated:
Figure 857483DEST_PATH_IMAGE024
(ii) a And then, screening the hypergraph vertexes by using the entropy values of the vertexes, wherein the reserved vertexes are as follows:
Figure 304120DEST_PATH_IMAGE025
。
the calculating the minimum valley distance of the hypergraph vertex based on the preference of the hypergraph vertex to the hyperedge includes:
for a vertex in the hypergraph
Figure 253621DEST_PATH_IMAGE019
And a super edge
Figure 139538DEST_PATH_IMAGE011
And calculating the preference of the vertex to the excess edge: when in use
Figure 199898DEST_PATH_IMAGE026
When the temperature of the water is higher than the set temperature,
Figure 238392DEST_PATH_IMAGE027
otherwise, the value is 0, wherein,
Figure 358795DEST_PATH_IMAGE028
is a constant; computing vertices
Figure 263166DEST_PATH_IMAGE029
The preference of all the super edges connected with the super edges is sorted according to the order from big to small, and all the sorted preference values are used as preference vectors
Figure 861637DEST_PATH_IMAGE030
Represents; calculating out
Figure 20217DEST_PATH_IMAGE019
Minimum valley distance of (c):
Figure 577101DEST_PATH_IMAGE031
wherein:
Figure 640872DEST_PATH_IMAGE032
represents and vertices
Figure 839772DEST_PATH_IMAGE019
A set of neighboring vertices with more common superedges;
Figure 852858DEST_PATH_IMAGE033
to represent
Figure 580643DEST_PATH_IMAGE034
And with
Figure 662868DEST_PATH_IMAGE035
The proportion of the number of the edges with the public excess edges to the total number of the edges;
Figure 134301DEST_PATH_IMAGE036
is shown by
Figure 993105DEST_PATH_IMAGE037
Middle evaluation value greater than
Figure 626212DEST_PATH_IMAGE038
A set of vertices of (1);
Figure 461312DEST_PATH_IMAGE039
is shown by
Figure 470857DEST_PATH_IMAGE040
And
Figure 192956DEST_PATH_IMAGE041
the valley distance therebetween.
Selecting a hypergraph vertex based on the minimum valley distance and determining point cloud fixing characteristics simultaneously comprises the following steps: sorting the minimum valley distances of all the vertexes in the hypergraph from large to small; comparing the distance between two adjacent minimum valleys to find the minimum valley distance with the maximum difference
Figure 262543DEST_PATH_IMAGE042
And
Figure 319361DEST_PATH_IMAGE043
(ii) a Discarding all smallest valleys less than
Figure 132596DEST_PATH_IMAGE044
And the rest vertexes represent one or more groups of shape parameters, an expression of the point cloud shape is determined according to the shape parameters, and the fixed characteristics of the steel structure are extracted by using the point cloud shape.
The invention has the following beneficial effects:
the method realizes the rapid high-precision extraction of the fixed features of the large-span steel structure based on the minimum valley distance of the top point of the hypergraph, has high data processing precision and efficiency, and has higher robustness aiming at the serious noise in the point cloud data.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of three-dimensional point cloud data of a roof steel structure of a factory building;
FIG. 3 is a diagram illustrating a division result of a certain layer of octree division of three-dimensional point cloud data of a building roof steel framework of a certain factory;
FIG. 4 is a diagram illustrating processing results of a thread according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Although the steps in the present invention are arranged by using reference numbers, the order of the steps is not limited, and the relative order of the steps can be adjusted unless the order of the steps is explicitly stated or other steps are required for the execution of a certain step. It is to be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
As shown in fig. 1, the invention relates to a method for extracting a fixed feature of a large-span steel structure based on a minimum valley distance, wherein the fixed feature comprises an axial feature of a steel structure pipeline and a spherical center feature of a steel structure welding ball, and the method comprises the following steps:
scanning the large-span steel structure by using a three-dimensional laser scanner to obtain point cloud data of the large-span steel structure;
filtering the acquired large-scale point cloud data;
dividing the point cloud after filtering by using an octree, and performing parallel processing on leaf nodes of the octree by using a multithreading technology, namely each thread corresponds to one leaf node of the octree;
in each thread, constructing a hypergraph by using point cloud data for leaf nodes, and determining vertexes and edges of the hypergraph;
calculating residual errors of each vertex to each superedge thereof, measured by the Simpson distance, based on the constructed hypergraph;
evaluating the vertex of the hypergraph based on the residual error of the Simpson distance measurement and a kernel function;
simplifying the hypergraph based on the evaluation result of the vertex and the entropy of the prior probability of the vertex, and deleting the vertex which does not meet the condition;
calculating the minimum valley distance of the hypergraph vertex based on the preference of the hypergraph vertex on the hyperedge;
selecting a hypergraph vertex based on the minimum valley distance, wherein the parameter in the hypergraph vertex corresponds to the shape parameter of the point cloud, and determining the point cloud fixing characteristic.
Further, foretell use three-dimensional laser scanner to scan the large-span steel construction, acquire the point cloud data of large-span steel construction, include: the method comprises the steps of using a three-dimensional laser scanner, scanning a steel structure by a plurality of stations around a large-span steel structure to obtain different point cloud data of a plurality of positions, splicing the different point cloud data of the plurality of positions, and acquiring the integral point cloud data of the large-span steel structure. Here, a multi-site scan is performed on a roof steel structure of a factory building, and then multi-site point clouds are registered, and the overall point cloud is shown in fig. 2.
Further, the filtering process performed on the acquired large-scale point cloud data includes: for the data of the ground and the wall, curvature detection can be carried out, and the parts (ground and wall point clouds) with the curvature lower than a threshold value are removed; and removing the system noise by adopting a statistical filtering algorithm. Based on the above, the smooth large-span steel structure measurement point cloud data is obtained.
Further, the above-mentioned point cloud after being processed by using octree partition filtering and performing parallel processing on octree leaf nodes by using multithread technology, that is, each thread corresponds to a leaf node of the octree, includes: firstly, wrapping point cloud by using a minimum cube, dividing the cube into 8 equally-divided sub-squares, equally dividing each non-empty sub-square again, stopping equally dividing a certain node when the point cloud number of the certain node is the same as that of a parent node of the certain node, and stopping the whole dividing process when the preset minimum edge length of the cube is reached, wherein fig. 3 shows that all nodes (one cube represents one node) of a certain layer of an octree are obtained after the point cloud data of a rigid architecture of a building roof of a certain factory is subjected to octree division. All subsequent steps are processed simultaneously for multiple leaf nodes using multi-threading.
Further, the above-mentioned constructing a hypergraph by using the point cloud data for the leaf nodes inside each thread, and determining the vertices and edges of the hypergraph, includes: all points in the point cloud are used as the super edges of the hypergraph; selecting a point set (if a plane needs to select three points, and a spherical surface needs to select four points) capable of determining the minimum number of point cloud shapes from the point clouds, and determining parameters of the point cloud shapes by using the point set (such as a spherical equation of a large-span steel structure welding ball:
Figure 974781DEST_PATH_IMAGE045
in (1)
Figure 949691DEST_PATH_IMAGE046
(ii) a The determined parameter is used as a vertex of the hypergraph; the vertex represents a space shape, euclidean distances from all points in the point cloud to the surface of the space shape table are calculated, when the distances are smaller than a certain threshold value, the corresponding points are inner points, and otherwise, the corresponding points are outer points; all the inner points are used as the corresponding super edges of the vertexes and are connected with each vertex and the super edge thereof in the super graph; repeating the above steps for a plurality of times (for example 10000 times or enumerating all combinations capable of determining the shape), the hypergraph of the point cloud can be established.
Further, the above-mentioned calculating the residual error of each vertex to each superedge thereof measured by simpson distance based on the constructed supergraph includes: first, for each vertex in the hypergraph
Figure 493805DEST_PATH_IMAGE001
And a super edge thereof
Figure 110731DEST_PATH_IMAGE002
To obtain its residual terms
Figure 541843DEST_PATH_IMAGE003
Wherein, in the process,
Figure 687654DEST_PATH_IMAGE004
is an expression of the shape represented by the point cloud; the residual term is then measured by the Simpson distance, i.e.
Figure 719064DEST_PATH_IMAGE047
Wherein, in the step (A),
Figure 874101DEST_PATH_IMAGE006
。
further, the above-mentioned vertex evaluation of the hypergraph based on the residual error of the simpson distance metric and the kernel function includes: for each vertex in the hypergraph
Figure 422370DEST_PATH_IMAGE001
Using the formula:
Figure 4661DEST_PATH_IMAGE048
the vertices are evaluated, wherein,
Figure 523367DEST_PATH_IMAGE009
is a vertex
Figure 216516DEST_PATH_IMAGE001
Degree of (i.e. vertex)
Figure 887800DEST_PATH_IMAGE001
The number of corresponding interior points;
Figure 640993DEST_PATH_IMAGE010
has a value of
Figure 381416DEST_PATH_IMAGE011
Is that
Figure 878256DEST_PATH_IMAGE007
At the inner point of (2) is 1
Figure 325418DEST_PATH_IMAGE011
Is that
Figure 124878DEST_PATH_IMAGE001
The outer point of (a) is 0;
Figure 493542DEST_PATH_IMAGE012
is the i Pan Nieqie Kefu kernel;
Figure 653128DEST_PATH_IMAGE013
is the residual measured in simpson distance;
Figure 158059DEST_PATH_IMAGE014
is a vertex
Figure 128420DEST_PATH_IMAGE001
The corresponding inner points account for the proportion of the point cloud in the total point cloud;
Figure 718801DEST_PATH_IMAGE015
whereinnIs the number of the excess edges.
Further, the above entropy based on the evaluation result of the vertex and the prior probability of the vertex simplifies the hypergraph, and the deleting of the vertex which does not satisfy the condition includes: constructed hypergraphs with a set of vertices
Figure 682078DEST_PATH_IMAGE016
The evaluation result obtained by evaluating each vertex is as follows:
Figure 388653DEST_PATH_IMAGE017
，
Figure 654549DEST_PATH_IMAGE018
is the number of vertices in the hypergraph; computing vertices
Figure 856860DEST_PATH_IMAGE019
And the difference between the maximum evaluation result and the evaluation result of (c):
Figure 233615DEST_PATH_IMAGE020
(ii) a Computing vertices
Figure 588504DEST_PATH_IMAGE019
Prior probability of (2)
Figure 25302DEST_PATH_IMAGE021
: when in use
Figure 714909DEST_PATH_IMAGE022
Is greater than 0, and is greater than the total weight of the rubber,
Figure 895355DEST_PATH_IMAGE023
otherwise
Figure 370329DEST_PATH_IMAGE049
Is a positive constant; and then calculate all verticesEntropy:
Figure 978028DEST_PATH_IMAGE024
(ii) a And then, screening the hypergraph vertexes by using the entropy values of the vertexes, wherein the reserved vertexes are as follows:
Figure 561456DEST_PATH_IMAGE025
。
further, the calculating the minimum valley distance of the hypergraph vertex based on the hypergraph vertex-to-hypergraph edge preference includes:
for a vertex in the hypergraph
Figure 670227DEST_PATH_IMAGE019
And a super edge
Figure 202970DEST_PATH_IMAGE011
And calculating the preference of the vertex to the excess edge: when in use
Figure 778308DEST_PATH_IMAGE050
When the temperature of the water is higher than the set temperature,
Figure 52295DEST_PATH_IMAGE051
otherwise, the value is 0, wherein,
Figure 699177DEST_PATH_IMAGE028
is a constant; computing vertices
Figure 7798DEST_PATH_IMAGE019
The preference of all the superedges connected with the preference vector is sorted according to the descending order of the preference values, and all the sorted preference values are used as the preference vector
Figure 829736DEST_PATH_IMAGE030
Representing; computing
Figure 856598DEST_PATH_IMAGE019
Minimum valley distance of (c):
Figure 41592DEST_PATH_IMAGE052
wherein:
Figure 470299DEST_PATH_IMAGE053
represents and vertices
Figure 200489DEST_PATH_IMAGE019
A set of neighboring vertices with more common superedges;
Figure 449067DEST_PATH_IMAGE033
to represent
Figure 703331DEST_PATH_IMAGE034
And
Figure 720966DEST_PATH_IMAGE041
the proportion of the number of the edges with the public excess edges to the total number of the edges;
Figure 887636DEST_PATH_IMAGE054
is shown by
Figure 623511DEST_PATH_IMAGE037
Middle evaluation value greater than
Figure 415886DEST_PATH_IMAGE019
A set of vertices of (1);
Figure 553607DEST_PATH_IMAGE055
is shown by
Figure 891178DEST_PATH_IMAGE034
And
Figure 114349DEST_PATH_IMAGE041
the valley distance therebetween.
Further, the selecting the hypergraph vertex based on the minimum valley distance and determining the point cloud fixing characteristics includes: sorting the minimum valley distances of all the vertexes in the hypergraph from large to small; comparing the distance between two adjacent minimum valleys to find the minimum valley distance with the maximum difference
Figure 710416DEST_PATH_IMAGE042
And
Figure 702642DEST_PATH_IMAGE043
(ii) a Discarding all smallest valleys less than
Figure 132487DEST_PATH_IMAGE042
The remaining vertices represent one or more sets of shape parameters (e.g., defining a sphere)
Figure 545468DEST_PATH_IMAGE046
) The expression of the point cloud shape (such as the spherical equation of the welding ball and the cylindrical equation of the pipeline) can be determined according to the shape parameters, and the fixed characteristics (such as the spherical center) of the steel structure can be extracted by using the point cloud shape
Figure 555013DEST_PATH_IMAGE056
). The processing results obtained by a thread according to all the procedures of the present invention are shown in fig. 4.
In the embodiment, a comprehensive, accurate and efficient automatic processing algorithm is provided for the problem of extraction of the fixed features of the large-span steel structure. The method can greatly save the working time of field operation and effectively reduce the difficulty of extracting the fixed features of the large-span steel structure.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.

Claims (10)

1. A method for extracting fixed features of a large-span steel structure based on a minimum valley distance, wherein the fixed features comprise axial features of a steel structure pipeline and spherical center features of a steel structure welding ball, and the method comprises the following steps:
scanning the large-span steel structure by using a three-dimensional laser scanner to obtain point cloud data of the large-span steel structure;
filtering the acquired large-scale point cloud data;
dividing the point cloud after filtering by using an octree, and performing parallel processing on leaf nodes of the octree by using a multithreading technology, namely each thread corresponds to one leaf node of the octree;
constructing a hypergraph for leaf nodes by using the point cloud data in each thread, and determining vertexes and edges of the hypergraph;
calculating residual errors of each vertex to each superedge thereof, measured by the Simpson distance, based on the constructed hypergraph;
evaluating the vertex of the hypergraph based on the residual error of the Simpson distance measurement and a kernel function;
simplifying the hypergraph based on the evaluation result of the vertex and the entropy of the prior probability of the vertex, and deleting the vertex which does not meet the condition;
calculating the minimum valley distance of the hypergraph vertex based on the preference of the hypergraph vertex on the hyperedge;
selecting a hypergraph vertex based on the minimum valley distance, wherein the parameter in the hypergraph vertex corresponds to the shape parameter of the point cloud, and determining the point cloud fixing characteristic.
2. The method for extracting the fixed features of the large-span steel structure based on the minimum valley distance as claimed in claim 1, wherein the scanning of the large-span steel structure by using the three-dimensional laser scanner to obtain the point cloud data of the large-span steel structure comprises: the method comprises the steps of scanning a steel structure by using a three-dimensional laser scanner at a plurality of stations around a large-span steel structure to obtain different point cloud data of a plurality of positions, splicing the different point cloud data of the plurality of positions, and obtaining the integral point cloud data of the large-span steel structure.
3. The method for extracting the fixed features of the large-span steel structure based on the minimum valley distance as claimed in claim 1, wherein the filtering process of the acquired large-scale point cloud data comprises: for the data of the ground and the wall surface, detecting the curvature, and removing the part of the curvature which is lower than a threshold value; and removing the system noise by adopting a statistical filtering algorithm.
4. The method for extracting the fixed features of the large-span steel structure based on the minimum valley distance as claimed in claim 1, wherein the point cloud after the filtering processing is divided by using the octree, and the leaf nodes of the octree are processed in parallel by using a multithread technology, that is, each thread corresponds to one leaf node of the octree, the method comprises the following steps: the method comprises the steps of firstly wrapping point clouds by a minimum cube, dividing the cube into 8 equally-divided sub-squares, equally dividing each non-empty sub-square again, stopping equally dividing a certain node when the point cloud number of the node is the same as the point cloud number of a father node of the certain node, stopping the whole division process when the preset minimum edge length of the cube is reached, and simultaneously processing a plurality of leaf nodes in all the subsequent steps by using a multithreading technology.
5. The method for extracting the fixed features of the large-span steel structure based on the minimum valley distance as claimed in claim 1, wherein the step of constructing the hypergraph by using the point cloud data for the leaf nodes in each thread and determining the vertexes and the edges of the hypergraph comprises the following steps: all points in the point cloud are used as the super edges of the hypergraph; selecting a point set which can determine the minimum number of point cloud shapes from the point clouds, and determining parameters of the point cloud shapes by adopting the point set; the determined parameter is used as a vertex of the hypergraph; the vertex represents a space shape, euclidean distances from all points in the point cloud to the surface of the space shape table are calculated, when the distances are smaller than a certain threshold value, the corresponding points are inner points, and otherwise, the corresponding points are outer points; all the inner points are used as the corresponding super edges of the vertexes and are connected with each vertex and the super edge thereof in the super graph; repeating the steps for multiple times to establish a hypergraph of the point cloud.
6. The method for extracting the fixed features of the large-span steel structure based on the minimum valley distance as claimed in claim 1, wherein the step of calculating the residual error of each vertex to each super edge thereof measured by the simpson distance based on the constructed super map comprises the following steps: first, for each vertex in the hypergraph
Figure 237472DEST_PATH_IMAGE001
And a super edge thereof
Figure 757446DEST_PATH_IMAGE002
To obtain its residual terms
Figure 213835DEST_PATH_IMAGE003
Wherein, in the step (A),
Figure 359514DEST_PATH_IMAGE004
is an expression of the shape represented by the point cloud; the residual term is then measured in terms of Simpson distance, i.e.
Figure 686590DEST_PATH_IMAGE005
Wherein, in the process,
Figure 377466DEST_PATH_IMAGE006
。
7. the method for extracting the fixed features of the large-span steel structure based on the minimum valley distance as claimed in claim 1, wherein the residual features based on the Simpson distance measureThe difference and kernel function evaluates vertices of the hypergraph, including: for each vertex in the hypergraph
Figure 321151DEST_PATH_IMAGE007
Using the formula:
Figure 996153DEST_PATH_IMAGE008
the vertices are evaluated, wherein,
Figure 708894DEST_PATH_IMAGE009
is a vertex
Figure 305092DEST_PATH_IMAGE010
Degree of (i.e. vertex)
Figure 470494DEST_PATH_IMAGE011
The number of corresponding interior points;
Figure 567763DEST_PATH_IMAGE012
has a value of
Figure 525223DEST_PATH_IMAGE013
Is that
Figure 151377DEST_PATH_IMAGE014
At an inner point of 1, in
Figure 210600DEST_PATH_IMAGE013
Is that
Figure 111560DEST_PATH_IMAGE001
The outer point of (a) is 0;
Figure 657947DEST_PATH_IMAGE015
is the i Pan Nieqie Keff Kelvin function;
Figure 720581DEST_PATH_IMAGE016
is residue measured by Simpson distanceDifference;
Figure 735942DEST_PATH_IMAGE017
is a vertex
Figure 440593DEST_PATH_IMAGE018
The corresponding inner points account for the proportion of the point cloud in the total point cloud;
Figure 107066DEST_PATH_IMAGE019
whereinnIs the number of the excess edges.
8. The method for extracting the fixed features of the large-span steel structure based on the minimum valley distance as claimed in claim 1, wherein the step of simplifying the hypergraph based on the entropy of the evaluation result of the vertexes and the prior probability of the vertexes and deleting the vertexes which do not satisfy the condition comprises the steps of: the set of vertices of the hypergraph is
Figure 606181DEST_PATH_IMAGE020
The evaluation result obtained by evaluating each vertex is as follows:
Figure 967892DEST_PATH_IMAGE021
，
Figure 86021DEST_PATH_IMAGE022
is the number of vertices in the hypergraph; computing vertices
Figure 482367DEST_PATH_IMAGE023
And the maximum evaluation result:
Figure 279946DEST_PATH_IMAGE024
(ii) a Computing vertices
Figure 394533DEST_PATH_IMAGE023
Prior probability of (2)
Figure 785194DEST_PATH_IMAGE025
: when the temperature is higher than the set temperature
Figure 301626DEST_PATH_IMAGE026
Is greater than 0 and the content of the active ingredient,
Figure 267177DEST_PATH_IMAGE027
otherwise
Figure 869059DEST_PATH_IMAGE028
Is a positive constant; the entropy of all vertices is then calculated:
Figure 328991DEST_PATH_IMAGE029
(ii) a And then, screening the hypergraph vertexes by using the entropy values of the vertexes, wherein the reserved vertexes are as follows:
Figure 699929DEST_PATH_IMAGE030
。
9. the method for extracting the fixed features of the large-span steel structure based on the minimum valley distance as claimed in claim 1, wherein the step of calculating the minimum valley distance of the hypergraph vertex based on the preference of the hypergraph vertex to the hyper-edge comprises the following steps:
for a vertex in the hypergraph
Figure 836381DEST_PATH_IMAGE031
And a super edge
Figure 394402DEST_PATH_IMAGE032
And calculating the preference of the vertex to the excess edge: when in use
Figure 782658DEST_PATH_IMAGE033
When the temperature of the water is higher than the set temperature,
Figure 883469DEST_PATH_IMAGE034
if the value is not 0, otherwise,wherein, the first and the second end of the pipe are connected with each other,
Figure 66188DEST_PATH_IMAGE035
is a constant; computing vertices
Figure 767297DEST_PATH_IMAGE036
The preference of all the super edges connected with the super edges is sorted according to the order from big to small, and all the sorted preference values are used as preference vectors
Figure 693665DEST_PATH_IMAGE037
Represents; computing
Figure 383403DEST_PATH_IMAGE031
Minimum valley distance of (c):
Figure 2603DEST_PATH_IMAGE038
，
wherein:
Figure 928359DEST_PATH_IMAGE039
represents and vertices
Figure 658417DEST_PATH_IMAGE031
A set of neighboring vertices with more common superedges;
Figure 468241DEST_PATH_IMAGE040
represent
Figure 992764DEST_PATH_IMAGE041
And
Figure 278251DEST_PATH_IMAGE042
the proportion of the number of the edges with the public excess edges to the total number of the edges;
Figure 671056DEST_PATH_IMAGE043
is shown by
Figure 725599DEST_PATH_IMAGE044
Middle evaluation value greater than
Figure 561968DEST_PATH_IMAGE045
A set of vertices of (1);
Figure 69173DEST_PATH_IMAGE046
is shown by
Figure 265668DEST_PATH_IMAGE047
And
Figure 440297DEST_PATH_IMAGE048
the valley distance therebetween.
10. The method for extracting the fixed features of the large-span steel structure based on the minimum valley distance as claimed in claim 1, wherein the step of selecting the hypergraph vertex based on the minimum valley distance and determining the point cloud fixed features comprises the following steps: sorting the minimum valley distances of all the vertexes in the hypergraph from large to small; comparing the distance between two adjacent minimum valleys to find the minimum valley distance with the maximum difference
Figure 916409DEST_PATH_IMAGE049
And
Figure 442068DEST_PATH_IMAGE050
(ii) a Discarding all smallest valleys less than
Figure 911096DEST_PATH_IMAGE051
And the rest vertexes represent one group or a plurality of groups of shape parameters, the expression of the point cloud shape is determined according to the shape parameters, and the fixed characteristics of the steel structure are extracted by using the point cloud shape.
CN202211609628.2A 2022-12-15 2022-12-15 Large-span steel structure fixed feature extraction method based on minimum valley distance Active CN115601565B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211609628.2A CN115601565B (en) 2022-12-15 2022-12-15 Large-span steel structure fixed feature extraction method based on minimum valley distance

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211609628.2A CN115601565B (en) 2022-12-15 2022-12-15 Large-span steel structure fixed feature extraction method based on minimum valley distance

Publications (2)

Publication Number Publication Date
CN115601565A true CN115601565A (en) 2023-01-13
CN115601565B CN115601565B (en) 2023-03-31

Family

ID=84853919

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211609628.2A Active CN115601565B (en) 2022-12-15 2022-12-15 Large-span steel structure fixed feature extraction method based on minimum valley distance

Country Status (1)

Country Link
CN (1) CN115601565B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015169029A1 (en) * 2014-05-05 2015-11-12 中兴通讯股份有限公司 Graph data partitioning method and device
CN106373199A (en) * 2016-08-31 2017-02-01 中测新图(北京)遥感技术有限责任公司 Rapid oblique photography building model extraction method
CN108876885A (en) * 2018-06-29 2018-11-23 山东鲁能智能技术有限公司 The Processing Method of Point-clouds and device of power equipment
CN110222564A (en) * 2018-10-30 2019-09-10 上海市服装研究所有限公司 A method of sex character is identified based on three-dimensional data
CN111222516A (en) * 2020-01-06 2020-06-02 西南电子技术研究所（中国电子科技集团公司第十研究所） Method for extracting key outline characteristics of point cloud of printed circuit board
WO2022235637A1 (en) * 2021-05-04 2022-11-10 Trax Technology Solutions Pte Ltd. Methods and systems for retail environments

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015169029A1 (en) * 2014-05-05 2015-11-12 中兴通讯股份有限公司 Graph data partitioning method and device
CN106373199A (en) * 2016-08-31 2017-02-01 中测新图(北京)遥感技术有限责任公司 Rapid oblique photography building model extraction method
CN108876885A (en) * 2018-06-29 2018-11-23 山东鲁能智能技术有限公司 The Processing Method of Point-clouds and device of power equipment
CN110222564A (en) * 2018-10-30 2019-09-10 上海市服装研究所有限公司 A method of sex character is identified based on three-dimensional data
CN111222516A (en) * 2020-01-06 2020-06-02 西南电子技术研究所（中国电子科技集团公司第十研究所） Method for extracting key outline characteristics of point cloud of printed circuit board
WO2022235637A1 (en) * 2021-05-04 2022-11-10 Trax Technology Solutions Pte Ltd. Methods and systems for retail environments

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
刘雅艳等: "BIM与三维GIS在古建筑信息模型中的应用研究", 《科学技术创新》 *

Also Published As

Publication number Publication date
CN115601565B (en) 2023-03-31

Similar Documents

Publication Publication Date Title
CN107948930B (en) Indoor positioning optimization method based on position fingerprint algorithm
Raumonen et al. Massive-scale tree modelling from TLS data
CN106023298B (en) Point cloud Rigid Registration method based on local Poisson curve reestablishing
CN105488770A (en) Object-oriented airborne laser radar point cloud filtering method
CN110060338B (en) Prefabricated part point cloud identification method based on BIM model
CN106091923A (en) The central point rapid assay methods of industrial bolt circular hole based on three-dimensional laser scanning technique
CN110544233A (en) Depth image quality evaluation method based on face recognition application
CN111369606B (en) Cultural relic object high-precision micro-deformation monitoring method based on uncontrolled scanning point cloud
CN110660056B (en) Building crack width measuring method based on image processing
CN114898118A (en) Automatic statistical method and system for power transmission line house removal amount based on multi-source point cloud
CN116008671A (en) Lightning positioning method based on time difference and clustering
CN115601565B (en) Large-span steel structure fixed feature extraction method based on minimum valley distance
CN108765446B (en) Power line point cloud segmentation method and system based on random field and random forest
CN112131752B (en) Super-collapse pollution rate tolerance estimation algorithm based on quasi-calibration
CN117292181A (en) Sheet metal part hole group classification and full-size measurement method based on 3D point cloud processing
CN116664823A (en) Small sample SAR target detection and recognition method based on meta learning and metric learning
CN111060922A (en) Tree point cloud extraction method based on airborne laser radar point cloud spatial distribution characteristics
CN115953421A (en) Harris honeycomb vertex extraction method for detecting regularity of honeycomb structure
CN114092545A (en) Self-adaptive grid searching method suitable for spherical target fitting
CN111583243B (en) Method for reconstructing adjacent point cell for detecting cellular regularity
CN112986948B (en) Building deformation monitoring method and device based on InSAR technology
CN113518308B (en) Optimal AP screening method in indoor positioning
CN112284287B (en) Stereoscopic vision three-dimensional displacement measurement method based on structural surface gray scale characteristics
CN114092395A (en) Method for evaluating concrete peeling disease condition of prefabricated part based on three-dimensional laser point cloud
CN111369610A (en) Point cloud data gross error positioning and eliminating method based on credibility information

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant