A deformation detection method applied to an aerial work vehicle boom

Abstract

The application relates to the technical field of deformation detection, in particular to a deformation detection method applied to an aerial work vehicle boom, which comprises the following steps: acquiring point cloud data of the aerial work vehicle boom; constructing a local form matrix of the point cloud data according to local point cloud data of the point cloud data, and acquiring a local anisotropy degree of the point cloud data; acquiring a discrete reference weight of the point cloud data and a reflection gradient of the point cloud data according to the spatial position and the reflection intensity of each local point cloud data of the point cloud data, and then acquiring a local dispersion degree of the point cloud data; combining the local anisotropy degree of the point cloud data to acquire an adaptive adjustment coefficient of the point cloud data, so as to judge whether the point cloud data is a flying point, and then to perform deformation detection on the aerial work vehicle boom. The application can analyze the characteristics of the flying point in the point cloud data, accurately remove the flying point in the point cloud data, and improve the precision of detecting the deformation of the aerial work vehicle boom.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of deformation detection, and in particular to a deformation detection method applied to a high-altitude operation vehicle jib. BACKGROUND

[0002] During use of the high-altitude operation vehicle, deformation of the jib structure is directly related to safety and stability of operation, in order to ensure safety and stability of operation of the high-altitude operation vehicle, deformation detection of the jib of the high-altitude operation vehicle needs to be performed regularly, and when deformation detection of the jib of the high-altitude operation vehicle is performed by using a traditional three-dimensional laser scanner, flying points are prone to be generated in high-reflection areas such as metal patches and reflective markers in the jib of the high-altitude operation vehicle, thereby interfering with deformation detection of the jib of the high-altitude operation vehicle, and in order to accurately detect whether deformation of the jib of the high-altitude operation vehicle occurs, the flying points need to be removed.

[0003] When a traditional statistical filtering method with a fixed threshold value is used to remove the flying points in the high-reflection areas, if a global threshold value set is too large, the flying points in the high-reflection areas cannot be removed, and if the threshold value set is too small, real point cloud data near the flying points is likely to be determined as the flying points, and when the statistical filtering algorithm with the fixed global threshold value is used to remove the flying points in the high-reflection areas, point cloud data with unevenness in a rust area is likely to be identified as the flying points, but the point cloud data with unevenness in the rust area is also real point cloud data on the surface of the jib; that is, the traditional statistical filtering method with the fixed threshold value cannot remove the flying points in the point cloud data collected by the three-dimensional laser scanner, and thus cannot accurately detect whether deformation of the jib of the high-altitude operation vehicle occurs. SUMMARY

[0004] The application provides a deformation detection method applied to a jib of a high-altitude operation vehicle, to solve the existing problem that a traditional statistical filtering method with a fixed threshold value cannot remove flying points in point cloud data collected by a three-dimensional laser scanner, and thus cannot accurately detect whether deformation of the jib of the high-altitude operation vehicle occurs.

[0005] The deformation detection method applied to the jib of the high-altitude operation vehicle adopts the following technical scheme:

[0006] The method comprises the following steps:

[0007] Obtaining point cloud data of the jib of the high-altitude operation vehicle;

[0008] Obtaining local point cloud data of the point cloud data; constructing a local morphology matrix of the point cloud data according to the local point cloud data of the point cloud data; and obtaining a local anisotropy degree of the point cloud data according to the local morphology matrix of the point cloud data.

[0009] Based on the spatial location of each local point cloud data, obtain the discrete reference weight of each local point cloud data to the point cloud data; based on the reflection intensity of the point cloud data and its local point cloud data, obtain the reflection gradient of the point cloud data; and combine the discrete reference weight of each local point cloud data to the point cloud data to obtain the local dispersion of the point cloud data.

[0010] Based on the local dispersion and local anisotropy of the point cloud data, the adaptive adjustment coefficient of the point cloud data is obtained; based on the adaptive adjustment coefficient of the point cloud data, it is determined whether the point cloud data is a flying point, and flying points in the point cloud data of the aerial work vehicle boom are removed, thereby performing deformation detection on the aerial work vehicle boom.

[0011] Preferably, the specific method for acquiring the local point cloud data of the point cloud data is as follows:

[0012] Preset a local radius The average distance between all point cloud data in the boom point cloud data is used as the baseline distance. For the first... The point cloud data, with the first Construct a radius of [number] points based on the point cloud data. A sphere at twice the reference distance is obtained. The local space of the point cloud data will be the first Point cloud data within a local space of the first point cloud data, as the first point cloud data... Local point cloud data of a point cloud dataset.

[0013] Preferably, the specific method for constructing a local morphology matrix of the point cloud data based on local point cloud data is as follows:

[0014] For the The point cloud data is obtained to acquire the first... Each local point cloud data point points to the first... The vector of the local spatial centroid of the i-th point cloud data, denoted as the i-th The deviation vector of each local point cloud data in the point cloud data; according to the first... The deviation vector of each local point cloud data in the nth point cloud data is used to construct the nth point cloud data. The local spatial covariance matrix of the i-th point cloud data, denoted as the j-th... The local morphological matrix of a point cloud data.

[0015] Preferably, the specific method for obtaining the local anisotropy of the point cloud data based on the local morphology matrix of the point cloud data is as follows:

[0016] For the The local morphological matrix of the i-th point cloud data, for the i-th perform eigenvalue decomposition on the local shape matrix of the point cloud data to obtain a plurality of eigenvalues of the local shape matrix of the point cloud data

[0017] obtain the ratio of the minimum eigenvalue to the maximum eigenvalue of the local shape matrix of the point cloud data, and obtain the difference between 1 and the ratio as the local anisotropy of the point cloud data

[0018] Preferably, the method for obtaining the discrete reference weight of each local point cloud data of the point cloud data according to the spatial position of each local point cloud data of the point cloud data comprises the following specific steps:

[0019] obtain the normal vector of the point cloud data according to the local point cloud data of the point cloud data by using the point cloud normal vector estimation method

[0020] obtain the normal vector of each local point cloud data of the point cloud data according to the respective local point cloud data of each local point cloud data of the point cloud data

[0021] for the point cloud data and the local point cloud data thereof, if the angle between the normal vector of the point cloud data and the normal vector of the local point cloud data thereof is less than is a preset coplanar angle threshold;

[0022] obtain the discrete reference weight of the local point cloud data of the point cloud data according to the distance between the point cloud data and the local point cloud data thereof, and the coplanarity of the point cloud data and the local point cloud data thereof ​​​​​​​​​​​​​​​​​​​​​​​​​​

[0023] Preferably, the acquisition of the first The first point cloud data The local point cloud data for the first The discrete reference weights for each point cloud data point include the following specific methods:

[0024]

[0025] In the formula, Indicates the first The first point cloud data The local point cloud data for the first Discrete reference weights for each point cloud data; Indicates the first The point cloud data and its first The distance between local point cloud data; Indicates the reference distance; Indicates the first The point cloud data and its first Coplanarity of point cloud data; This represents an exponential function with the natural constant as its base.

[0026] Preferably, the specific method for obtaining the reflection gradient of the point cloud data based on the reflection intensity of the point cloud data and its local point cloud data is as follows:

[0027] For the The point cloud data, based on the first point cloud data, The normal vector of the i-th point cloud data is used to construct the i-th point cloud data. The tangent plane of the point cloud data will be the first... Projecting all local point cloud data of the first point cloud data onto the second point cloud data... In the tangent plane of the point cloud data, the first... Scatter plot of tangent plane of point cloud data;

[0028] For the first The reflection intensity of the point cloud data corresponding to all scatter points in the tangent plane scatter plot of the point cloud data is fitted to a surface to obtain the first... Local point cloud data reflection intensity surface of point cloud data; according to the first point cloud data; The reflection intensity surface of the local point cloud data of the nth point cloud data is obtained by combining the basic principles of calculus. The reflection intensity gradient of a point cloud data point in its local point cloud reflection intensity surface is denoted as the i-th point cloud data point. The reflection gradient of a point cloud data.

[0029] Preferably, the specific method for obtaining the local dispersion of the point cloud data is as follows:

[0030] According to the Each local point cloud data point of the first point cloud data point is paired with the first... Discrete reference weights of point cloud data, the first The reflection intensity of each local point cloud data in the point cloud data, combined with the first point cloud data, The reflection gradient of the i-th point cloud data is obtained. The local dispersion of a point cloud dataset is calculated using the following formula:

[0031]

[0032] In the formula, Indicates the first Local dispersion of point cloud data; Indicates the first The number of local point cloud data in a point cloud dataset; Indicates the first The first point cloud data The local point cloud data for the first Discrete reference weights for each point cloud data; Indicates the first The first point cloud data The reflection intensity of a local point cloud data; Indicates the first The mean reflection intensity of all local point cloud data of a point cloud dataset; Indicates the first The reflection gradient of a point cloud data; This indicates the preset enhancement coefficient.

[0033] Preferably, the method for obtaining the adaptive adjustment coefficient of the point cloud data based on the local dispersion and local anisotropy of the point cloud data includes:

[0034] For the Point cloud data, with a preset local anisotropy adjustment coefficient. With local dispersion adjustment coefficient , obtain the The local anisotropy of a point cloud data and The product of 1 and 1, plus the sum of the product, is denoted as the first . The local anisotropy factor of the nth point cloud data; obtain the nth The local dispersion of point cloud data and The product of and , minus 1 to get the difference, is taken as the first . Local discrete factors of point cloud data;

[0035] The first The local anisotropy factor of the point cloud data and the first point cloud data. a product of a local discrete factor of the point cloud data as the adaptive adjustment coefficient of the point cloud data. a product of a local discrete factor of the point cloud data as the adaptive adjustment coefficient of the point cloud data.

[0036] Preferably, the adaptive adjustment coefficient of the point cloud data is used to determine whether the point cloud data is a flying point, and the specific method comprises the following steps:

[0037] For the first point cloud data, a statistical filtering algorithm is used to determine whether the first point cloud data is a flying point according to the adaptive adjustment coefficient of the point cloud data and each local point cloud data of the first point cloud data. For the first point cloud data, a statistical filtering algorithm is used to determine whether the first point cloud data is a flying point according to the adaptive adjustment coefficient of the point cloud data and each local point cloud data of the first point cloud data. For the first point cloud data, a statistical filtering algorithm is used to determine whether the first point cloud data is a flying point according to the adaptive adjustment coefficient of the point cloud data and each local point cloud data of the first point cloud data.

[0038] The beneficial effects of the technical scheme of the present application are as follows: the local morphology matrix of the point cloud data is constructed according to the local point cloud data of the point cloud data, the local anisotropy degree of the point cloud data is obtained, the local morphology matrix of the point cloud data is specifically used to quantitatively evaluate the distribution characteristics of the point cloud data in the three-dimensional space, can reflect the geometric morphology of the local position surface of the point cloud data, is used to identify whether the point cloud data is located in a high reflection area or a rust area, and the greater the local anisotropy degree of the point cloud data is, the more likely the local position of the point cloud is in the high reflection area, which provides a basis for subsequently adaptively assigning a best filtering threshold to each point cloud data.

[0039] According to the spatial position and the reflection intensity of each local point cloud data of the point cloud data, the discrete reference weight of each local point cloud data of the point cloud data to the point cloud data and the reflection gradient of the point cloud data are obtained, the flying points in the point cloud data of the aerial work vehicle boom are not uniformly distributed but are concentrated, and the reflection intensity is suddenly changed or sharply transitions in space, so that the gradient of the reflection intensity of the local point cloud data of the point cloud data is used to distinguish the flying points in the point cloud data of the aerial work vehicle boom from the real smooth area, the adaptive adjustment coefficient of the point cloud data is obtained by combining the local anisotropy degree of the point cloud data and the local dispersion degree of the point cloud data, whether the point cloud data is a flying point is determined, all flying points in the point cloud data of the aerial work vehicle boom are obtained and removed, the high-precision point cloud data is obtained by three-dimensional laser scanning in the non-deformation state of the boom, the reference digital model is constructed after denoising processing, and the reference standard is stored to detect whether the aerial work vehicle boom is deformed. BRIEF DESCRIPTION OF DRAWINGS

[0040] ​​In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative effort.

[0041] Figure 1 A step flow chart of a deformation detection method applied to an aerial work vehicle boom. DETAILED DESCRIPTION

[0042] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined object, the following will combine the drawings and preferred embodiments to specifically describe the specific implementation, structure, features and effects of the deformation detection method applied to an aerial work vehicle boom according to the present application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0044] The following will specifically describe the specific scheme of the deformation detection method applied to an aerial work vehicle boom according to the present application.

[0045] Please refer to Figure 1 which shows a step flow chart of the deformation detection method applied to an aerial work vehicle boom according to one embodiment of the present application. The method comprises the following steps:

[0046] Step S001: Obtain point cloud data of the aerial work vehicle boom.

[0047] It should be noted that the existing detection of the deformation of the aerial work vehicle boom mainly relies on manual inspection, strain gauge sensing or ordinary three-dimensional laser scanning. Manual inspection is low in efficiency and strong in subjectivity, and strain gauge sensing is susceptible to environmental interference and has limited scope of application. Therefore, the present embodiment proposes a deformation detection method applied to an aerial work vehicle boom, which specifically analyzes the point cloud data of the aerial work vehicle boom to detect the deformation of the aerial work vehicle boom. Therefore, the point cloud data of the aerial work vehicle boom needs to be collected first.

[0048] Specifically, the point cloud data of the aerial work vehicle boom is collected by a three-dimensional laser scanner.

[0049] Step S002: Obtain local point cloud data; construct a local morphology matrix of the point cloud data based on the local point cloud data; obtain the local anisotropy of the point cloud data based on the local morphology matrix of the point cloud data.

[0050] It should be noted that the boom of the aerial work vehicle contains highly reflective areas such as metal patches and reflective markings, as well as rusted areas composed of point cloud data with uneven surfaces, large local curvature changes, and high geometric complexity due to corrosion. Due to the strong reflective characteristics of high-reflectivity areas, point cloud data will contain densely distributed anomalous points (flying points). To avoid interference from flying points in high-reflectivity areas on boom deformation detection, it is necessary to remove flying points in these areas. However, traditional statistical filtering algorithms using a fixed global threshold to remove flying points in high-reflectivity areas are problematic. If the global threshold is set too high, it will fail to remove flying points; if the threshold is set too low, it will easily identify real point cloud data near the flying points as flying points. Furthermore, traditional statistical filtering algorithms using a fixed global threshold to remove flying points in high-reflectivity areas can easily identify uneven point cloud data in rusted areas as flying points. However, uneven point cloud data in rusted areas is also real point cloud data on the boom surface. To prevent real point cloud data in rusted areas from being identified as flying points and removed, and to accurately identify flying points in high-reflectivity areas while avoiding identifying real point cloud data near flying points as flying points, this embodiment adaptively assigns an optimal filtering threshold to each point cloud data point to accurately remove flying points from the point cloud data, thereby accurately detecting the deformation of the aerial work platform boom.

[0051] Preferably, in a specific embodiment of the present invention, a local radius is preset. The The specific value can be set according to the actual situation. This embodiment does not make a hard requirement. In this embodiment, it is used as... Taking the example of the boom point cloud data, the average distance between all point cloud data points is used as the baseline distance. For the first... The point cloud data, with the first Construct a radius of [number] points based on the point cloud data. A sphere at twice the reference distance is obtained. The local space of the point cloud data will be the first Point cloud data within a local space of the first point cloud data, as the first point cloud data... Local point cloud data of a point cloud dataset;

[0052] Furthermore, to obtain the first Each local point cloud data point points to the first... The vector of the local spatial centroid of the i-th point cloud data, denoted as the i-th The deviation vector of each local point cloud data in the point cloud data; according to the first... The deviation vector of each local point cloud data in the nth point cloud data is used to construct the nth point cloud data. The local spatial covariance matrix of the i-th point cloud data, denoted as the j-th... The local morphological matrix of the point cloud data; since the process of obtaining the covariance matrix is ​​a well-known prior art, it will not be described in detail in this embodiment.

[0053] It should be noted that the local morphology matrix of point cloud data is specifically used to quantitatively evaluate the distribution characteristics of point cloud data in three-dimensional space. It can reflect the geometric shape of the surface at the local location of the point cloud data, and is used to identify whether the point cloud data is located in a high-reflection area or a rusted area, providing a basis for subsequently adaptively assigning an optimal filtering threshold to each point cloud data.

[0054] Preferably, in a specific embodiment of the present invention, for the first The local morphological matrix of the i-th point cloud data, for the i-th Eigenvalue decomposition is performed on the local morphological matrix of the point cloud data (since the ... The local morphological matrix of the point cloud data is the covariance matrix. The process of eigenvalue decomposition of the covariance matrix is ​​a well-known existing technique, and therefore will not be elaborated upon in this embodiment. This yields the... Several eigenvalues ​​of the local morphological matrix of a point cloud data (since point cloud data is three-dimensional data, the number of eigenvalues ​​of the local morphological matrix of point cloud data is three).

[0055] Furthermore, to obtain the first The ratio of the minimum eigenvalue to the maximum eigenvalue of the local morphological matrix of a point cloud dataset, and the difference obtained by subtracting the ratio from 1, is taken as the value of the first eigenvalue. Local anisotropy of point cloud data.

[0056] It should be noted that the largest eigenvalue of the local morphology matrix of point cloud data represents the intensity of change in the main extension direction of the surface at the local location of the point cloud. The larger the value, the more concentrated the distribution of the point cloud in a certain main direction, the more obvious the extension, and the closer the surface is to a planar or linear structure, exhibiting strong directional consistency. The smallest eigenvalue of the local morphology matrix of point cloud data represents the flatness of the surface in the normal direction at the local location of the point cloud. The smaller the value, the smaller the undulation of the point cloud in the normal direction, the smoother and flatter the surface, exhibiting highly stable planar characteristics. Therefore, the greater the local anisotropy of the point cloud data, the more likely the local location of the point cloud is in a high-reflectivity area. This allows for the subsequent adaptive assignment of an optimal filtering threshold to the point cloud data to accurately remove flying points and accurately detect deformation of the boom of aerial work vehicles.

[0057] At this point, the local anisotropy degree of the point cloud data is obtained.

[0058] Step S003: According to the spatial position of each local point cloud data of the point cloud data, the discrete reference weight of each local point cloud data of the point cloud data to the point cloud data is obtained; according to the reflection intensity of the point cloud data and the local point cloud data thereof, the reflection gradient of the point cloud data is obtained, and the local discrete degree of the point cloud data is obtained in combination with the discrete reference weight of each local point cloud data of the point cloud data to the point cloud data.

[0059] It should be noted that the reflection intensity information obtained by three-dimensional laser scanning not only reflects the reflectivity difference of the material surface, but also is jointly affected by the surface geometry and the incident angle. In the actual arm support scene, the reflection intensity of the painted area is the highest, the reflection intensity of the metal patch area is the second, and the reflection intensity of the rust area is the lowest; and in the curved surface area, due to the angle effect, the greater the incident angle, the weaker the reflection intensity. Based on these rules, the high reflection area often shows a local sharp increase in reflection intensity accompanied by outliers (flying points), and the rust area shows a low overall reflection intensity but a large fluctuation. Therefore, the embodiment does not stop at the description of the reflection intensity difference, but further introduces the reflection intensity information into the subsequent operation: by calculating the weighted variance of the reflection intensity in the local range, combining the normal consistency constraint and the distance decay factor to form a spatial weighting mechanism; at the same time, a reflection intensity gradient enhancement factor is introduced to amplify the response of the edge or abnormal mutation area. In this way, the reflection intensity logic and the subsequent discrete degree calculation operation correspond closely, which can accurately identify flying points in high reflection areas while avoiding false deletion of real concave-convex points in rust areas, providing a quantifiable feature basis for the construction of an adaptive filtering threshold.

[0060] Preferably, in a specific embodiment of the present application, for the first point cloud data, according to the local point cloud data of the first point cloud data, the normal vector of the first point cloud data is obtained by a point cloud normal vector estimation method, since the point cloud normal vector estimation method is a known prior art, which will not be described in detail in this embodiment. Similarly, according to each local point cloud data of the first point cloud data, the normal vector of each local point cloud data of the first point cloud data is obtained.

[0061] Similarly, according to each local point cloud data of the first point cloud data, the normal vector of each local point cloud data of the first point cloud data is obtained. For the first point cloud data and the first local point cloud data thereof, if the normal vector of the first point cloud data and the first local point cloud data thereof are not consistent, the first local point cloud data of the first point cloud data is deleted.

[0062] For the first point cloud data and the first local point cloud data thereof, if the normal vector of the first point cloud data and the first local point cloud data thereof are not consistent, the first local point cloud data of the first point cloud data is deleted. For the first point cloud data and the first local point cloud data thereof, if the normal vector of the first point cloud data and the first local point cloud data thereof are not consistent, the first local point cloud data of the first point cloud data is deleted. ​​​​​An angle between normal vectors of the first local point cloud data and the second local point cloud data is less than The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. An angle between normal vectors of the first local point cloud data and the second local point cloud data is greater than or equal to The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0.

[0063] Further, according to the distance between the first point cloud data and the second local point cloud data, and the coplanarity of the first point cloud data and the second point cloud data, the first local point cloud data of the first point cloud data is obtained. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0.

[0064] The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1.

[0065] The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1. The coplanarity of the first point cloud data and the second point cloud data is recorded as 0. The coplanarity of the first point cloud data and the second point cloud data is recorded as 1.

[0066] It should be noted that the closer the distance between the point cloud data and its local point cloud data, the more the local point cloud data of the point cloud data should be taken as a reference to evaluate the dispersion of the point cloud data, and the smaller the angle between the normal vector of the point cloud data and the normal vector of its local point cloud data, the more likely the point cloud data and its local point cloud data are located in the same plane, and when the angle between the normal vector of the point cloud data and the normal vector of its local point cloud data is less than a preset coplanar angle threshold , the normal vector of the point cloud data and its local point cloud data are located in the same plane, thereby giving the local point cloud data of the point cloud data a dispersion reference weight for the point cloud data.

[0067] It should be further noted that the flying points in the point cloud data of the aerial work vehicle boom are often not uniformly distributed, but are concentrated, and their reflection intensities show sudden changes or sharp transitions in space. Therefore, the gradient of the reflection intensity of the local point cloud data of the point cloud data is used to distinguish the flying points in the point cloud data of the aerial work vehicle boom from the real smooth area.

[0068] Preferably, in a specific embodiment of the present application, for the first point cloud data, a tangent plane of the first point cloud data is constructed according to the normal vector of the first point cloud data, all local point cloud data of the first point cloud data is projected into the tangent plane of the first point cloud data, and a tangent plane scatter plot of the first point cloud data is obtained.

[0069] Further, the reflection intensity of the point cloud data corresponding to all scatter points in the tangent plane scatter plot of the first point cloud data is fitted by the least square method to obtain a local point cloud data reflection intensity surface of the first point cloud data, and the reflection intensity gradient of the first point cloud data in the local point cloud data reflection intensity surface thereof is obtained according to the local point cloud data reflection intensity surface of the first point cloud data and the basic principles of calculus, and is recorded as the reflection gradient of the first point cloud data. Since the least square method and the basic principles of calculus are well-known prior art, they will not be described in detail in this embodiment.

[0070] It should be noted that the reflection gradient of the point cloud data is large, indicating that the local of the point cloud data has a strong reflection intensity mutation, and the strong reflection intensity mutation is often located at the edge of the high reflection area or the area composed of flying points. The reflection gradient of the point cloud data is small, indicating that the local reflection intensity distribution of the point cloud data is relatively flat, that is, the high reflection area and the corrosion area are different in reflection intensity distribution. Therefore, according to the reflection gradient of the point cloud data, in combination with the discrete reference weight of the local point cloud data of the point cloud data and the reflection intensity of each local point cloud data of the point cloud data, the local discrete degree of the point cloud data is obtained, which provides a basis for subsequent identification of flying points.

[0071] Preferably, in a specific embodiment of the present application, for the first point cloud data, according to the discrete reference weight of each local point cloud data of the first point cloud data to the first point cloud data, the reflection intensity of each local point cloud data of the first point cloud data, in combination with the reflection gradient of the first point cloud data, the local discrete degree of the first point cloud data is obtained, and the specific calculation formula is:

[0072]

[0073] The specific value of a can be set by itself in combination with the actual situation, and the present embodiment does not make a hard requirement. In the present embodiment, a is taken as an example for description. ​​​​​​​​​​​​​​​​​​​​​​​​

[0074] It should be noted that the local dispersion of the point cloud data is obtained by comprehensively reflecting the fluctuation of the reflection intensity and the spatial structure characteristics in the local range, the dispersion degree of the reflection intensity in the local range is represented by the weighted variance, and different neighborhood points are given different weights by combining the spatial distance attenuation and the normal consistency, so that the point cloud data close to and coplanar contributes more, and the point far away or not coplanar contributes less; meanwhile, the reflection gradient factor is introduced to amplify the local reflection intensity mutation, so that the fly point aggregation can be more sensitively identified in the high reflection area, and the dispersion degree can comprehensively and accurately represent the abnormal characteristics of the local surface in the rust area.

[0075] At this point, the local dispersion of the point cloud data is obtained.

[0076] Step S004: According to the local dispersion of the point cloud data and the local anisotropy of the point cloud data, an adaptive adjustment coefficient of the point cloud data is obtained; according to the adaptive adjustment coefficient of the point cloud data, it is judged whether the point cloud data is a fly point, and the fly point in the point cloud data of the aerial work vehicle boom is removed, so as to detect the deformation of the aerial work vehicle boom.

[0077] It should be noted that the local anisotropy of the point cloud data and the local dispersion of the point cloud data are obtained by steps S002 and S003 respectively, the local anisotropy reflects the geometric shape characteristics of the local surface of the point cloud, when the anisotropy is larger, it indicates that the area is closer to the plane or the regular surface, at this time the noise recognition threshold should be appropriately increased, in order to enhance the removal ability of the fly point in the high reflection area; secondly, the local dispersion reflects the reflection intensity fluctuation of the local point cloud, when the dispersion is larger, it indicates that there is strong reflection intensity mutation or fly point aggregation in the area, at this time the noise recognition threshold should be reduced to accurately identify the fly point. Further, by constructing the local anisotropy factor and the local dispersion factor respectively, and combining the two to form the adaptive adjustment coefficient, the differential and adaptive filtering threshold adjustment in the high reflection area and the rust area is realized, so as to accurately remove the fly point of the point cloud data of the aerial work vehicle boom, and improve the accuracy of detecting the deformation of the aerial work vehicle boom.

[0078] Preferably, in one specific embodiment of the present application, for the first point cloud data, a local anisotropy adjustment coefficient and a local dispersion adjustment coefficient are preset, the specific values of and can be set by oneself according to the actual situation, and the present embodiment does not make hard requirements, in the present embodiment, the first , are taken as examples to describe the local anisotropy of the first point cloud data and the product of the local anisotropy factor of the first point cloud data and the local dispersion factor of the first point cloud data, subtracting 1 from the difference obtained by the product, as the adaptive adjustment coefficient of the first point cloud data. the product of the local anisotropy factor of the first point cloud data and the local dispersion factor of the first point cloud data, subtracting 1 from the difference obtained by the product, as the adaptive adjustment coefficient of the first point cloud data. the product of the local anisotropy factor of the first point cloud data and the local dispersion factor of the first point cloud data, subtracting 1 from the difference obtained by the product, as the adaptive adjustment coefficient of the first point cloud data.

[0079] Further, the product of the local anisotropy factor of the first point cloud data and the local dispersion factor of the first point cloud data, subtracting 1 from the difference obtained by the product, as the adaptive adjustment coefficient of the first point cloud data. the product of the local anisotropy factor of the first point cloud data and the local dispersion factor of the first point cloud data, subtracting 1 from the difference obtained by the product, as the adaptive adjustment coefficient of the first point cloud data.

[0080] It should be noted that after obtaining the adaptive adjustment coefficient of each point cloud data of the aerial work vehicle boom, the adaptive adjustment coefficient can be used to identify and remove all flying points of the point cloud data of the aerial work vehicle boom, and whether the aerial work vehicle boom is deformed can be detected according to the point cloud data of the aerial work vehicle boom after removing the flying points.

[0081] Preferably, in a specific embodiment of the present application, for the first point cloud data, the statistical filtering algorithm is used to determine whether the first point cloud data is a flying point according to the adaptive adjustment coefficient of the first point cloud data and each local point cloud data of the first point cloud data. Since the statistical filtering algorithm is a known prior art, it will not be described in detail in this embodiment.

[0082] Further, all flying points of the point cloud data of the aerial work vehicle boom are obtained and removed. In the non-deformed state of the boom, high-precision point cloud data is obtained by three-dimensional laser scanning, a reference model is constructed after denoising processing, and the reference model is stored as a reference standard. The point cloud data collected each time after removing the flying points is spatially aligned with the reference model through a feature matching algorithm (such as bolt holes, welds, or structural inflection points in the key area), and the deformation amount can be calculated based on distance deviation detection, for example, the perpendicular distance of each point in the real-time point cloud to the corresponding surface of the reference model is calculated, and the area exceeding the preset threshold (such as ±2 mm) is marked as a deformation area.

[0083] Thus, the embodiment is completed.

[0084] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.​​​​​​​​

Claims

1. A method for detecting deformation applied to a jib of an aerial work vehicle, characterized by, The method comprises the following steps: Obtain point cloud data of the aerial work vehicle boom; Obtain local point cloud data of the point cloud data; construct a local morphology matrix of the point cloud data according to the local point cloud data of the point cloud data; and obtain a local anisotropy degree of the point cloud data according to the local morphology matrix of the point cloud data; According to the spatial position of each local point cloud data of the point cloud data, obtain the discrete reference weight of each local point cloud data of the point cloud data on the point cloud data; The specific method for obtaining the discrete reference weight of each local point cloud data of the point cloud data on the point cloud data according to the spatial position of each local point cloud data of the point cloud data comprises: According to the first local point cloud data of the first point cloud data, the normal vector of the first point cloud data is obtained by a point cloud normal vector estimation method. According to the first local point cloud data of each local point cloud data of the first normal vector of each local point cloud data of the first For the The point cloud data and its first If the local point cloud data, the first The normal vector of the i-th point cloud data and its i-th point cloud data The angle between the normal vectors of the local point cloud data is less than 1. Then the first The point cloud data and its first The coplanarity of the nth point cloud data is denoted as 1. If the nth point cloud data is coplanar, then... The normal vector of the i-th point cloud data and its i-th point cloud data The angle between the normal vectors of a local point cloud data is greater than or equal to the angle between the normal vectors of the local point cloud data. Then the first The point cloud data and its first The coplanarity of the point cloud data is denoted as 0; This is a preset coplanar angle threshold; According to the The point cloud data and its first The distance between local point cloud data points, combined with the first The point cloud data and its first The coplanarity of the nth point cloud data is obtained to determine the nth point cloud data. The first point cloud data The local point cloud data for the first Discrete reference weights for each point cloud data; According to the reflection intensity of the point cloud data and the local point cloud data thereof, obtain a reflection gradient of the point cloud data, combine the discrete reference weight of each local point cloud data of the point cloud data on the point cloud data, and obtain a local discrete degree of the point cloud data; The specific method for obtaining the local discrete degree of the point cloud data comprises: According to the first local point cloud data of each point cloud data, the discrete reference weight of the first point cloud data, the reflection intensity of each local point cloud data of the first point cloud data, and the reflection gradient of the first point cloud data, a local discreteness of the first point cloud data is obtained, and a specific calculation formula is as follows: wherein, denotes a local dispersion of the i-th point cloud data; denotes a number of local point cloud data of the i-th point cloud data; denotes a i-th local point cloud data of the i-th point cloud data; denotes a dispersion reference weight of the i-th local point cloud data of the i-th point cloud data to the j-th point cloud data; denotes a reflection intensity of the i-th local point cloud data of the i-th point cloud data; denotes a reflection intensity of the i-th local point cloud data of the i-th point cloud data; denotes a reflection intensity mean of all local point cloud data of the i-th point cloud data; denotes a reflection gradient of the i-th point cloud data; denotes a preset enhancement coefficient;​​​​​​​ According to the local discrete degree of the point cloud data and the local anisotropy degree of the point cloud data, obtain an adaptive adjustment coefficient of the point cloud data; according to the adaptive adjustment coefficient of the point cloud data, judge whether the point cloud data is a flying point; remove the flying point in the point cloud data of the aerial work vehicle boom, so as to perform deformation detection on the aerial work vehicle boom.

2. The method for deformation detection of a boom for an aerial work vehicle according to claim 1, characterized in that, The specific method for obtaining the local point cloud data of the point cloud data comprises: A local radius is preset The average distance between all point cloud data in the arm support point cloud data is taken as a reference distance, and for the i th point cloud data, a sphere with a radius of i times the reference distance is constructed with the i th point cloud data as the center to obtain a local space of the i th point cloud data, and the point cloud data in the local space of the i th point cloud data is taken as local point cloud data of the i th point cloud data.

3. The method for deformation detection of a boom for an aerial work vehicle according to claim 1, characterized in that, The specific method for constructing the local morphology matrix of the point cloud data according to the local point cloud data of the point cloud data comprises: For the The point cloud data is obtained to acquire the first... Each local point cloud data point points to the first... The vector of the local spatial centroid of the i-th point cloud data, denoted as the i-th The deviation vector of each local point cloud data in the point cloud data; according to the first... The deviation vector of each local point cloud data in the nth point cloud data is used to construct the nth point cloud data. The local spatial covariance matrix of the i-th point cloud data, denoted as the j-th... The local morphological matrix of a point cloud data.

4. The method for deformation detection of an aerial work vehicle boom according to claim 1, wherein The specific method for obtaining the local anisotropy degree of the point cloud data according to the local morphology matrix of the point cloud data comprises: For the local shape matrix of the first point cloud data , the local shape matrix of the first point cloud data is subjected to eigenvalue decomposition to obtain a plurality of eigenvalues of the local shape matrix of the first point cloud data . The ratio of the minimum eigenvalue and the maximum eigenvalue of the local shape matrix of the first point cloud data is obtained, and a difference obtained by subtracting the ratio from 1 is taken as the local anisotropy degree of the first point cloud data. The ratio of the minimum eigenvalue and the maximum eigenvalue of the local shape matrix of the first point cloud data is obtained, and a difference obtained by subtracting the ratio from 1 is taken as the local anisotropy degree of the first point cloud data. The ratio of the minimum eigenvalue and the maximum eigenvalue of the local shape 5. The method for deformation detection of an aerial device vehicle jib according to claim 1, wherein The acquisition of the first local point cloud data of the first discrete reference weight of the first point cloud data includes specific methods: In the formula, Indicates the first The first point cloud data The local point cloud data for the first Discrete reference weights for each point cloud data; Indicates the first The point cloud data and its first The distance between local point cloud data; Indicates the reference distance; Indicates the first The point cloud data and its first Coplanarity of point cloud data; This represents an exponential function with the natural constant as its base.

6. The method for deformation detection of an aerial device vehicle jib according to claim 1, wherein The specific method for obtaining the reflection gradient of the point cloud data according to the reflection intensity of the point cloud data and the local point cloud data thereof comprises: For the first point cloud data, according to the normal vector of the first point cloud data, a tangent plane of the first point cloud data is constructed, all local point cloud data of the first point cloud data is projected into the tangent plane of the first point cloud data, and a tangent plane scatter plot of the first point cloud data is obtained. The point cloud data of the first The point cloud data of the first The point cloud data of the first The point cloud data of the first The point cloud data of the first The point cloud data of the first 7. The method for deformation detection of an aerial device vehicle jib according to claim 1, wherein The specific method for obtaining the adaptive adjustment coefficient of the point cloud data according to the local discrete degree of the point cloud data and the local anisotropy degree of the point cloud data comprises: For the Point cloud data, with a preset local anisotropy adjustment coefficient. With local dispersion adjustment coefficient , obtain the The local anisotropy of a point cloud data and The product of 1 and 1, plus the sum of the product, is denoted as the first . The local anisotropy factor of the nth point cloud data; obtain the nth The local dispersion of point cloud data and The product of and , minus 1 to get the difference, is taken as the first . Local discrete factors of point cloud data; The first The local anisotropy factor of the point cloud data and the first point cloud data. The product of the local discrete factors of the nth point cloud data is used as the product of the nth point cloud data. Adaptive adjustment coefficients for point cloud data.

8. The method for deformation detection of an aerial device vehicle jib according to claim 1, wherein, The specific method for judging whether the point cloud data is a flying point according to the adaptive adjustment coefficient of the point cloud data comprises: For the first point cloud data, a statistical filtering algorithm is used to determine whether the first point cloud data is a flying point based on an adaptive adjustment coefficient of the first point cloud data and each local point cloud data of the first point cloud data.

Images