Power line detection method and device based on point cloud processing and medium thereof

Abstract

The invention relates to the technical field of power line detection, in particular to a power line detection method and device based on point cloud processing and a medium thereof. The detection method comprises the following steps: S1, acquiring point cloud data through a radar monitoring device installed on an electric power inspection crane, and preprocessing the point cloud data; s2, performing clustering analysis on the preprocessed point cloud data based on multi-neighborhood clustering to obtain a clustering result; s3, the clustering result is input into a pre-trained machine learning model for classification, point cloud information conforming to the power line features is obtained, and the point cloud information comprises three-dimensional coordinates and attitude information of the power line. The power line detection method based on point cloud processing is high in detection precision, is not liable to be affected by the environment, and can guarantee the safety of inspection work.

Description

Technical Field

[0001] The present invention relates to the technical field of power line detection, and in particular to a power line detection method, device and medium based on point cloud processing. Background Art

[0002] During the inspection work of high-voltage substations, it involves the construction of cranes in the environment of high-voltage power lines. The crane driver cannot clearly observe the distance between the boom and the high-voltage wires during the construction process, so necessary protective measures must be taken during construction. General protective measures are mostly to set up safety observers next to the crane, always observe the environment around the crane boom, and quickly convey the message to the crane driver when danger is detected. This method will cause a waste of manpower, and manual communication may be inaccurate.

[0003] In order to solve the above problems, the prior art proposes a wire target detection method based on visual sensors, which needs to collect two-dimensional color image data of the wires and use the target detection algorithm in the field of computer vision to obtain the wire target detection results. The disadvantages of this method are: it is easily affected by the environment, has poor robustness, is prone to misidentification in harsh environments, and cannot guarantee the safety of inspection operations. Summary of the invention

[0004] The technical problem to be solved by the present invention is: in order to solve the technical problem that the existing technology is easily affected by the environment, has poor robustness, is prone to misidentification in harsh environments, and cannot ensure the safety of inspection operations. The present invention provides a power line detection method based on point cloud processing, which has high detection accuracy, is not easily affected by the environment, and can ensure the safety of inspection work.

[0005] The technical solution adopted by the present invention to solve the technical problem is: a power line detection method based on point cloud processing, the method comprising the following steps:

[0006] S1, acquiring point cloud data through a radar monitoring device installed on a power inspection crane, and preprocessing the point cloud data;

[0007] S2, clustering analysis is performed on the preprocessed point cloud data based on multi-neighborhood clustering to obtain clustering results;

[0008] S3, inputting the clustering results into a pre-trained machine learning model for classification to obtain point cloud information that meets the characteristics of the power lines, wherein the point cloud information includes: three-dimensional coordinates and posture information of the power lines.

[0009] Further, specifically, the step S2 specifically includes the following steps:

[0010] S21, selecting any point from the preprocessed point cloud data as the current point A, obtaining the adjacent point B of the current point A, and connecting the current point A and the adjacent point B to form a line segment AB;

[0011] S22, connecting the current point A and the coordinate origin O of the radar monitoring device to form a line segment AO;

[0012] S23, calculating the distance d between the current point A and the adjacent point B, and the angle between the line segment AB and the line segment AO;

[0013] If the angle is less than the set angle threshold, or the angle is greater than the set angle threshold but the distance d is less than the set distance threshold, it is determined that the current point A and the adjacent point B belong to the same class;

[0014] S24, taking the adjacent point B as the next current point, repeating steps S21 to S23 until the clustering analysis of all points in the point cloud data is completed to obtain the clustering result.

[0015] Further, specifically, the step S3 specifically includes the following steps:

[0016] The pre-trained machine learning model is used to calculate the normal direction of each point in the same category in the clustering results, and the absolute value of the angle between the normal direction of each point and the Z axis;

[0017] Determine whether all the angles are within a threshold range;

[0018] If all the angles are within the threshold range, the point cloud data of this category is retained and confirmed as a wire point cloud;

[0019] If all the angles are not within the threshold range, the point cloud data of this category is removed from the clustering result to obtain point cloud information that meets the characteristics of the power lines.

[0020] Further, specifically, the calculation of the normal direction corresponding to each point of the same class includes the following steps:

[0021] Use the least square method to fit the wire point cloud curve;

[0022] Traverse all points on the wire point cloud curve and use K nearest neighbor or radius search to determine the neighborhood points;

[0023] The local plane is obtained by fitting the neighborhood points through principal component analysis, and the covariance matrix is ​​calculated, and the eigenvector corresponding to the minimum eigenvalue is extracted as the normal direction;

[0024] Correct the normal direction based on the normal of the viewpoint to ensure that the normal direction of each point is consistent.

[0025] Further, specifically, fitting the wire point cloud curve using the least squares method includes:

[0026] In the two-dimensional XOY plane, let the curve be represented by the polynomial f(x):

[0027]

[0028] Where i∈{0,1,2...n} is the order of the polynomial, a i are the coefficients of the polynomial orders;

[0029] Assume that the wire point cloud curve l consists of m points (x j ,y j ), j∈{0,1,...m}, the wire point cloud curve l is obtained by sampling the polynomial curve f(x), then:

[0030] y i =f(x j );

[0031] Approximate the real point y through a fitted curve f(x) j To ensure that each point is on the curve f(x) or close to the curve f(x), let the distance of each point from the curve be:

[0032] e j =y j -f(x j );

[0033] Then we need to minimize the distance from all points to the curve:

[0034]

[0035] Assume a second-order polynomial f(x) = a0 + a1x 1 +a2x 2 Fitting data (x j ,y j ),j∈{0,1,...m}, represented by Xθ=b, its least squares problem is expressed as:

[0036] minθ‖Xθ-b‖ 2 ;

[0037] If the coefficients [a0, a1, a2] T , that is, Xθ=b, the θ in it can be transformed as follows:

[0038] X T Xθ=X T b;

[0039] Assume X T Xθ=X Tb is reversible, and the fitting wire point cloud curve is obtained:

[0040] θ=(X T X) -1 X T b.

[0041] Further, specifically, the preprocessing of the point cloud data in step S1 includes filtering processing and noise reduction processing.

[0042] A computer device comprising:

[0043] processor;

[0044] A memory for storing executable instructions;

[0045] Wherein, the processor is used to read the executable instructions from the memory and execute the executable instructions to implement the power line detection method based on point cloud processing as described above.

[0046] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor implements the power line detection method based on point cloud processing as described above.

[0047] The beneficial effect of the present invention is that the power line detection method based on point cloud processing of the present invention solves the problem that the power line detection in the prior art is easily affected by the environment and has poor robustness. It has high detection accuracy for power lines, is not easily affected by the environment, and can ensure the safety of inspection work. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0049] Figure 1 It is a schematic diagram of the method flow of embodiment 1 of the present invention.

[0050] Figure 2 It is a schematic diagram of the computer device structure of the second embodiment of the present invention.

[0051] In the figure, 10 is a computer device; 1002 is a processor; 1004 is a memory; and 1006 is a transmission device. DETAILED DESCRIPTION

[0052] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0053] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0054] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0055] Example 1

[0056] The present application embodiment provides a power line detection method based on point cloud processing, such as Figure 1 As shown, the method comprises the following steps:

[0057] S1, obtain point cloud data through the radar monitoring equipment installed on the power inspection crane, and pre-process the point cloud data. The point cloud data pre-processing includes filtering and noise reduction processing, which can improve the stability and accuracy of power line detection; the filtering processing method adopts but is not limited to straight-through filtering and Gaussian filtering.

[0058] S2, performing cluster analysis on the preprocessed point cloud data based on multi-neighborhood clustering to obtain a clustering result; step S2 specifically includes the following steps:

[0059] S21, selecting any point from the preprocessed point cloud data as the current point A, obtaining the adjacent point B of the current point A, and connecting the current point A and the adjacent point B to form a line segment AB;

[0060] S22, connecting the current point A and the coordinate origin O of the radar monitoring device to form a line segment AO;

[0061] S23, calculating the distance d between the current point A and the adjacent point B, and the angle between the line segment AB and the line segment AO;

[0062] If the angle is less than the set angle threshold, or the angle is greater than the set angle threshold but the distance d is less than the set distance threshold, it is determined that the current point A and the adjacent point B belong to the same class;

[0063] S24, taking the adjacent point B as the next current point, repeating steps S21 to S23 until the cluster analysis of all points in the point cloud data is completed, and the clustering results of different point cloud classifications are obtained. The time required for cluster analysis through multi-neighborhood clustering is shorter, which improves the real-time performance of the algorithm and the efficiency of power line detection.

[0064] It should be noted that the angle threshold is set in combination with the physical properties of the target object (such as the power line), and is generally an empirical value.

[0065] S3, input the clustering results into the pre-trained machine learning model for classification, and obtain point cloud information that meets the characteristics of power lines, including: the three-dimensional coordinates and posture information of the power lines. The geometric characteristics of the power lines are analyzed by the machine learning model to obtain the common characteristics of the power lines: the power line point cloud is distributed in a linear manner, and the power lines are parallel and do not intersect each other; the material of the power line has little effect on the geometric shape of the power line, and the daily shape is mostly long straight lines with almost no bends, so it can be concluded that the bending moment of any point is almost zero, that is, the elevation is almost the same in a small local area. According to the two characteristic points of the power line point cloud linear distribution and the bending moment being almost zero, assuming that a certain point on the power line has a certain topological relationship with the points on other power lines in its field, then for each point on the power line, this topological relationship should be the same or similar, then the normal estimation of each point on the point cloud data should be the same or similar, that is, the normal direction of the laser foot point of the same power line should be very close, and the normal direction of the laser foot point of the power line should show a trend close to the Z axis, while the normal direction of the laser foot point of the non-power line presents different characteristics. For example, the normal direction of targets such as leaves is irregular; the normal vector direction of the wall point set is generally consistent, but the direction tends to be parallel to the horizontal plane; the normal direction of the pole point cloud is also generally consistent, but tends to be parallel to the horizontal plane. The point cloud data of the power line can be screened out by the normal direction. The specific implementation process includes:

[0066] The normal direction of each point in the same category in the clustering results is calculated using a pre-trained machine learning model, and the absolute value of the angle between the normal direction of each point and the Z axis (the axis perpendicular to the horizontal plane is assumed to be the Z axis);

[0067] Determine whether all angles are within the threshold range;

[0068] If all angles are within the threshold range, the point cloud data of this category is retained and confirmed as a wire point cloud;

[0069] If all angles are not within the threshold range, the point cloud data of this category is removed from the clustering results to obtain point cloud information that meets the characteristics of power lines.

[0070] In an embodiment, the calculation of the normal direction corresponding to each point of the same class includes the following steps:

[0071] Use the least square method to fit the wire point cloud curve;

[0072] Traverse all points on the wire point cloud curve and use K nearest neighbor or radius search to determine the neighborhood points;

[0073] The local plane is obtained by fitting the neighborhood points through principal component analysis, and the covariance matrix is ​​calculated, and the eigenvector corresponding to the minimum eigenvalue is extracted as the normal direction; further, it specifically includes: calculating the covariance matrix of the neighborhood point cloud, which can reflect the distribution characteristics of the point cloud in space; performing eigenvalue decomposition on the covariance matrix to obtain the eigenvalue and the corresponding eigenvector; according to the geometric characteristics, the eigenvector corresponding to the minimum eigenvalue of the covariance matrix is ​​the normal direction of the local plane. Therefore, by extracting this eigenvector, the normal information of the target point can be accurately characterized, and the accuracy of power line detection can be improved.

[0074] Correct the normal direction based on the normal of the viewpoint to ensure that the normal direction of each point is consistent.

[0075] In an embodiment, fitting the wire point cloud curve using the least squares method includes:

[0076] In the two-dimensional XOY plane, let the curve be represented by the polynomial f(x):

[0077]

[0078] where i∈{0,1,2...n} is the order of the polynomial, a i are the coefficients of the polynomial orders;

[0079] Assume that the wire point cloud curve l consists of m points (x j ,y j ), j∈{0,1,...m}, the wire point cloud curve l is obtained by sampling the polynomial curve f(x), then:

[0080] y i =f(x j );

[0081] Approximate the real point y through a fitted curve f(x) jTo ensure that each point is on the curve f(x) or close to the curve f(x), let the distance of each point from the curve be:

[0082] e j =y j -f(x j );

[0083] Then we need to minimize the distance from all points to the curve:

[0084]

[0085] Assume a second-order polynomial f(x) = a0 + a1x 1 +a2x 2 Fitting data (x j ,y j ), j∈{0,1,...m}, this is actually a least squares solution problem of a non-homogeneous linear equation system Ax=b. In order to avoid ambiguity with a and x above, it is expressed as Xθ=b. The least squares problem is expressed as:

[0086] minθ‖Xθ-b‖ 2 ;

[0087] If the coefficients [a0, a1, a2] T , that is, Xθ=b, the θ in it can be transformed as follows:

[0088] X T Xθ=X T b;

[0089] Assume X T Xθ=X T b is reversible, and the fitting wire point cloud curve is obtained:

[0090] θ=(X T X) -1 X T b.

[0091] It should be noted that the power line point cloud as a whole still has inherent characteristics. For example, if the ratio of the length to the width of the power line is greater than 1, it can be preliminarily determined whether it meets the characteristics of the wire. In an embodiment, the data set for training the machine learning model includes: data in which the ratio of the length to the width of the power line is greater than 1, and the angle between the normal direction and the z-axis is within a threshold range. The data set is input into the machine learning model, and the machine learning model is trained to obtain a trained machine learning model.

[0092] To sum up, the power line detection method based on point cloud processing of the present invention solves the problem that the power line detection in the prior art is easily affected by the environment and has poor robustness. It has high detection accuracy for power lines, is not easily affected by the environment, and can ensure the safety of inspection work.

[0093] Example 2

[0094] An embodiment of the present application provides a computer device, which includes a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by the processor to implement a power line detection method based on point cloud processing as provided in the above method embodiment.

[0095] Figure 2 The hardware structure diagram of a device for implementing a power line detection method based on point cloud processing provided in an embodiment of the present application is shown. The device may participate in constituting or include the apparatus or system provided in an embodiment of the present application. Figure 2 As shown, the computer device 10 may include one or more processors 1002 (the processor may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 1004 for storing data, and a transmission device 1006 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 2 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 2 More or fewer components as shown, or with Figure 2 Different configurations shown.

[0096] It should be noted that the one or more processors and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer device 10 (or mobile device). As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0097] The memory 1004 can be used to store software programs and modules of application software, such as a program instruction / data storage device corresponding to a power line detection method based on point cloud processing in an embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 1004, that is, implementing the above-mentioned method. The memory 1004 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 1004 may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the computer device 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0098] The transmission device 1006 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer device 10. In one example, the transmission device 1006 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 1006 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0099] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer device 10 (or mobile device).

[0100] Example 3

[0101] An embodiment of the present application also provides a computer-readable storage medium, which can be set in a server to store at least one instruction or at least one program related to a power line detection method based on point cloud processing in a method embodiment. The at least one instruction or the at least one program is loaded and executed by the processor to implement a power line detection method based on point cloud processing provided in the above method embodiment.

[0102] Optionally, in this embodiment, the storage medium may be located in at least one of the multiple network servers of the computer network. Optionally, in this embodiment, the storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0103] Example 4

[0104] The embodiment of the present invention further provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes a power line detection method based on point cloud processing provided in the above various optional embodiments.

[0105] It should be noted that the above-mentioned sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above-mentioned specific embodiments of the present application are described. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0106] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment and storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0107] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0108] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A power line detection method based on point cloud processing, characterized in that: The method The following steps are involved: S1, acquiring point cloud data through a radar monitoring device installed on a power inspection crane, and preprocessing the point cloud data; S2, clustering analysis is performed on the preprocessed point cloud data based on multi-neighborhood clustering to obtain clustering results; S3, inputting the clustering results into a pre-trained machine learning model for classification to obtain point cloud information that meets the characteristics of the power lines, wherein the point cloud information includes: three-dimensional coordinates and posture information of the power lines.

2. The power line detection method based on point cloud processing according to claim 1, characterized in that: The step S2 specifically includes the following steps: S21, selecting any point from the preprocessed point cloud data as the current point A, obtaining the adjacent point B of the current point A, and connecting the current point A and the adjacent point B to form a line segment AB; S22, connecting the current point A and the coordinate origin O of the radar monitoring device to form a line segment AO; S23, calculating the distance d between the current point A and the adjacent point B, and the angle between the line segment AB and the line segment AO; If the angle is less than the set angle threshold, or the angle is greater than the set angle threshold but the distance d is less than the set distance threshold, it is determined that the current point A and the adjacent point B belong to the same class; S24, taking the adjacent point B as the next current point, repeating steps S21 to S23 until the clustering analysis of all points in the point cloud data is completed to obtain the clustering result.

3. The power line detection method based on point cloud processing according to claim 1, characterized in that: The step S3 specifically comprises the following steps: The pre-trained machine learning model is used to calculate the normal direction of each point in the same category in the clustering results, and the absolute value of the angle between the normal direction of each point and the Z axis; Determine whether all the angles are within a threshold range; If all the angles are within the threshold range, the point cloud data of this category is retained and confirmed as a wire point cloud; If all the angles are not within the threshold range, the point cloud data of this category is removed from the clustering result to obtain point cloud information that meets the characteristics of the power lines.

4. The power line detection method based on point cloud processing according to claim 3, characterized in that: The calculation of the normal direction for each point of the same class includes the following steps: Use the least square method to fit the wire point cloud curve; Traverse all points on the wire point cloud curve and use K nearest neighbor or radius search to determine the neighborhood points; The local plane is obtained by fitting the neighborhood points through principal component analysis, and the covariance matrix is ​​calculated, and the eigenvector corresponding to the minimum eigenvalue is extracted as the normal direction; Correct the normal direction based on the normal of the viewpoint to ensure that the normal direction of each point is consistent.

5. The power line detection method based on point cloud processing according to claim 4, characterized in that: The least square method is used to fit the wire point cloud curve, including: In the two-dimensional XOY plane, let the curve be represented by the polynomial f(x): Where i∈{0,1,2...n} is the order of the polynomial, a i are the coefficients of the polynomial orders; Assume that the wire point cloud curve l consists of m points (x j ,y j ), j∈{0,1,...m}, the wire point cloud curve l is obtained by sampling the polynomial curve f(x), then: y i =f(x j ); Approximate the real point y through a fitted curve f(x) j To ensure that each point is on the curve f(x) or close to the curve f(x), let the distance of each point from the curve be: e j =y j -f(x j ); Then we need to minimize the distance from all points to the curve: Assume a second-order polynomial f(x) = a0 + a1x 1 +a2x 2 Fitting data (x j ,y j ),j∈{0,1,...m}, represented by Xθ=b, its least squares problem is expressed as: minθ‖Xθ-b‖ 2 ; If the coefficients [a0, a1, a2] T , that is, Xθ=b, the θ in it can be transformed as follows: X T Xθ=X T b; Assume X T Xθ=X T b is reversible, and the fitting wire point cloud curve is obtained: θ=(X T X) -1 X T b。 6. The power line detection method based on point cloud processing according to claim 1, characterized in that: The preprocessing of the point cloud data in step S1 includes filtering and noise reduction.

7. A computer device, characterized in that: include: processor; A memory for storing executable instructions; The processor is used to read the executable instructions from the memory and execute the executable instructions to implement the power line detection method based on point cloud processing as described in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements the power line detection method based on point cloud processing according to any one of claims 1 to 6.

Images