Tunnel chute detection method based on point cloud data

Abstract

The invention belongs to the technical field of cloud processing, and discloses a tunnel chute detection method based on point cloud data, comprising the following steps: S1, preprocessing tunnel point cloud data: S11, setting tunnel width as x and tunnel height as y, taking the larger one of x and y as a radius to construct a spherical surface, and removing point cloud outside the spherical surface; s12, setting the lowest height of the edge of the chute as h, removing point clouds with spherical surfaces below h, and obtaining top point clouds containing the chute; s2, adjusting the cloud attitude of the top point cloud; s3, intercepting the top point cloud into a strip-shaped point cloud set; and S4, projecting the strip-shaped point cloud set into a two-dimensional strip-shaped point cloud set, and extracting chute information from the two-dimensional strip-shaped point cloud set. According to the tunnel chute detection method based on the point cloud data provided by the invention, detection of chute position information, chute width information and chute depth information is completed through an automatic means; and the working efficiency of detection work and the reliability of data are improved.

Description

Technical Field

[0001] The present invention relates to the field of cloud processing technology, and in particular to a tunnel chute detection method based on point cloud data. Background Art

[0002] The overhead contact network of an electrified railway is an important component for trains to obtain electricity. The chute (or guide trough) is a device used to guide the contact wire (or contact conductor) to ensure good contact between the pantograph and the contact wire. The chute geometry includes but is not limited to the position, height, inclination angle and other parameters of the chute. These parameters are crucial to ensure that the train can safely and smoothly obtain electricity. The contact network tunnel chute geometry detection usually refers to the regular inspection or maintenance of the overhead contact network and its related facilities in the electrified railway tunnel, so as to promptly discover and repair any problems that may cause poor contact, thereby improving the safety and reliability of railway operations.

[0003] In current technology, the detection of the geometric state of the catenary tunnel chute is usually carried out manually using measuring tools, which is time-consuming, labor-intensive and inefficient. The application of extracting target data information from point cloud data is gradually becoming popular. However, in the field of catenary chutes, there is still a lack of specific solutions for directly obtaining chute information from point cloud data. Summary of the invention

[0004] The object of the present invention is to provide a tunnel chute detection method based on point cloud data, which has high automation, high work efficiency and reliable data.

[0005] To achieve this object, the present invention adopts the following technical solutions:

[0006] A tunnel chute detection method based on point cloud data comprises the following steps:

[0007] S1. Tunnel point cloud data preprocessing, S1 includes:

[0008] S11, setting the tunnel width to x, the tunnel height to y, taking the larger of x and y as the radius to construct a spherical surface, and removing the point cloud outside the spherical surface;

[0009] S12, setting the minimum height of the edge of the chute to h, removing the point cloud of the spherical surface below h, and obtaining the top point cloud including the chute;

[0010] S2. adjusting the cloud posture of the top point cloud;

[0011] S3, cutting the top point cloud into a strip point cloud set;

[0012] S4. The strip point cloud set is projected into a two-dimensional strip point cloud set, and the chute information is extracted from the two-dimensional strip point cloud set.

[0013] Preferably, S2 comprises:

[0014] S21, intercepting a target point cloud from the top point cloud;

[0015] S22, applying PCA algorithm to analyze the main direction of the target point cloud;

[0016] S23, setting the angle between the main direction and the desired direction as the rotation angle θ, and constructing a composite rotation matrix;

[0017] S24. The top point cloud adjusts the cloud posture according to the composite rotation matrix.

[0018] Preferably, the target point cloud is located in the highest point area of ​​the top point cloud.

[0019] Preferably, the S23 includes:

[0020] S231, setting the angle between the main direction and the X-axis as the rotation angle θ1, the angle between the main direction and the Y-axis as the rotation angle θ2, and the angle between the main direction and the Z-axis as the rotation angle θ3;

[0021] S232. Construct a composite rotation matrix.

[0022] Preferably, in S22, the main direction is the direction with the largest variance obtained by PCA algorithm analysis.

[0023] Preferably, S3 includes:

[0024] S31, intercepting an arc-shaped point cloud directly above the top point cloud;

[0025] S32, fitting a standard circle with the arc-shaped point cloud;

[0026] S33, setting the width of the strip point cloud along the circumferential direction and the distance between adjacent strip point clouds along the circumferential direction, and obtaining the strip point cloud set.

[0027] Preferably, the center of the standard circle coincides with the center of the spherical surface, and the radius of the standard circle is the same as the radius of the spherical surface.

[0028] Preferably, S4 includes:

[0029] S41, projecting the strip point cloud set onto a vertical plane to obtain a two-dimensional strip point cloud set;

[0030] S42, performing radius filtering on the two-dimensional strip point cloud set to remove noise points;

[0031] S43, segmenting the two-dimensional strip point cloud set according to a certain step length;

[0032] S44, calculating the distribution variance of each segment of the point cloud, and retaining the segments whose distribution variance is greater than a preset value;

[0033] S45, performing curve fitting on the retained segments to obtain the chute position, chute width and chute depth.

[0034] Preferably, in S45, the curve fitting applies a Gaussian function, and the form of the Gaussian function is

[0035] The chute position is equal to μ, the chute width is equal to 2σ, and the chute depth is equal to AB.

[0036] Preferably, the vertical plane is perpendicular to the setting direction of the slide groove.

[0037] Beneficial effects of the present invention:

[0038] The present invention provides a tunnel chute detection method based on point cloud data, comprising the following steps: S1, tunnel point cloud data preprocessing, S1 includes: S11, setting the tunnel width as x, the tunnel height as y, taking the larger of x and y as the radius to construct a spherical surface, and removing the point cloud outside the spherical surface; S12, setting the minimum height of the chute edge as h, removing the point cloud below h on the spherical surface, and obtaining the top point cloud containing the chute; S2, adjusting the cloud posture of the top point cloud; S3, intercepting the top point cloud into a strip point cloud set; S4, projecting the strip point cloud set into a two-dimensional strip point cloud set, and extracting chute information from the two-dimensional strip point cloud set. The present invention provides a tunnel chute detection method based on point cloud data to complete the detection of chute position information, chute width information and chute depth information through automated means, and improves the work efficiency of the detection work and the reliability of the data. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a flow chart of a tunnel chute detection method based on point cloud data provided by an embodiment of the present invention;

[0040] Figure 2 It is a point cloud data diagram of the interior of a tunnel provided by an embodiment of the present invention;

[0041] Figure 3 It is a data graph obtained by removing the point clouds outside the spherical surface from the point cloud data graph inside the tunnel provided by an embodiment of the present invention;

[0042] Figure 4 is a top point cloud data graph provided by an embodiment of the present invention;

[0043] Figure 5 is a top point cloud data diagram before posture adjustment provided by an embodiment of the present invention;

[0044] Figure 6 is a top point cloud data graph after posture adjustment provided by an embodiment of the present invention;

[0045] Figure 7 It is a cloud data graph of a strip point cloud set provided by an embodiment of the present invention;

[0046] Figure 8 is a schematic diagram of a two-dimensional strip point cloud set provided by an embodiment of the present invention;

[0047] Fig. 9 It is a partial schematic diagram of a two-dimensional strip point cloud set provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0049] In the description of the present invention, unless otherwise clearly specified and limited, the terms "connected", "connected", and "fixed" 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 an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0050] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may include that the first and second features are in direct contact, or may include that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, a first feature being "above", "above" and "above" a second feature includes that the first feature is directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. A first feature being "below", "below" and "below" a second feature includes that the first feature is directly below and obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0051] In the description of this embodiment, the terms "upper", "lower", "right", etc., directions or positional relationships are based on the directions or positional relationships shown in the drawings, and are only for the convenience of description and simplification of operation, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are only used to distinguish in the description and have no special meaning.

[0052] The present embodiment provides a tunnel chute detection method based on point cloud data, which obtains the chute information of the contact network based on the point cloud data inside the tunnel, such as the chute position information, chute width information and chute depth information, so as to realize the geometric detection of the chute, and improves the work efficiency and data reliability of the chute geometry detection work through automation means.

[0053] Please refer to Figure 1 , a tunnel chute detection method based on point cloud data provided in this embodiment includes the following steps:

[0054] S1, tunnel point cloud data preprocessing;

[0055] S2, adjust the cloud posture of the top point cloud;

[0056] S3, cutting the top point cloud into a strip point cloud set;

[0057] S4. The strip point cloud set is projected into a two-dimensional strip point cloud set, and the chute information is extracted from the two-dimensional strip point cloud set.

[0058] First, the tunnel point cloud data is preprocessed to remove point clouds that are not related to the chute geometry detection and obtain the top point cloud. Secondly, the posture of the top point cloud is adjusted to make the shape of the top point cloud in the preset coordinate system more standard, which is convenient for subsequent feature processing. Furthermore, the top point cloud is segmented to obtain a strip point cloud set, and features are extracted from the strip point cloud set to obtain the chute position information, chute width information and chute depth information.

[0059] Among them, S1 includes:

[0060] S11, setting the tunnel width as x, the tunnel height as y, taking the larger of x and y as the radius to construct a spherical surface, and removing the point cloud outside the spherical surface;

[0061] S12, setting the minimum height of the edge of the chute to h, removing the point cloud of the spherical surface below h, and obtaining the top point cloud including the chute.

[0062] See also Figure 2-4 , Figure 2For the point cloud data map inside the tunnel, first set the tunnel width as x and the tunnel height as y, take the larger of x and y as the radius to construct a sphere, remove the point cloud outside the sphere, and ensure that the chute is included. Figure 3 To obtain the data graph after removing the point cloud outside the sphere, further, the minimum height of the chute edge is set to h, and the point cloud with a height less than h in the sphere is removed, and then the top point cloud containing the chute is obtained, that is, Figure 4 .

[0063] Furthermore, S2 includes:

[0064] S21, intercepting the target point cloud in the top point cloud;

[0065] S22, applying PCA algorithm to analyze the main directions of the target point cloud;

[0066] S23, setting the angle between the main direction and the desired direction as the rotation angle θ, and constructing a composite rotation matrix;

[0067] S24, the top point cloud adjusts the cloud posture according to the composite rotation matrix.

[0068] See also Figure 5 and Figure 6 , Figure 5 This is the top point cloud data map before posture adjustment. Figure 6 This is the top point cloud data diagram after posture adjustment. First, the target point cloud is intercepted in the top point cloud. The target point cloud contains enough points to represent the overall trend. Preferably, the target point cloud is located in the highest point area of ​​the top point cloud; further, the PCA algorithm is applied to analyze the selected target point cloud to find the main direction of the target point cloud. The PCA algorithm is also the principal component analysis algorithm, which is a commonly used algorithm in the prior art. It realizes data dimensionality reduction by converting the original data into a new coordinate system. It should be noted that the main direction is the direction with the largest variance. The rotation angle θ is determined according to the angle between the main direction and the desired direction, and a composite rotation matrix is ​​constructed according to the rotation angle. The rotation matrix is ​​a special transformation matrix that can keep the distance between points unchanged while changing their positions; finally, the top point cloud is transformed using the above rotation matrix to adjust its cloud posture so that the top point cloud has a more standard shape in the preset coordinate system, which is convenient for subsequent feature processing.

[0069] It should be noted that the composite rotation matrix is ​​composed of multiple rotation matrices, and the rotation axis and rotation angle of each rotation matrix are different. Therefore, S23 includes:

[0070] S231, setting the angle between the main direction and the X axis as the rotation angle θ1, the angle between the main direction and the Y axis as the rotation angle θ2, and the angle between the main direction and the Z axis as the rotation angle θ3;

[0071] S223. Construct a composite rotation matrix.

[0072] Specifically, the rotation operation is performed with the Z axis, Y axis and X axis as the rotation axes, and the main directions are oriented to the X axis, Y axis and Z axis in turn. For example, the top point cloud is first rotated around the Z axis by an angle of θ1, and further rotated around the Y axis by an angle of θ2. At this time, since the rotation around the Z axis has been performed, it is actually rotated around the new Y axis. Further, the top point cloud is rotated around the X axis by an angle of θ3. Similarly, it is actually rotated around the new Z axis.

[0073] The composite rotation matrix is ​​expressed as:

[0074] R total =R x (-θ 3 )·R y (-θ 2 )·R z (-θ 1 )

[0075] It should be noted that the multiplication order of the rotation matrix is ​​executed from right to left, and the rightmost rotation occurs first. Therefore, the composite rotation matrix provided in this embodiment first rotates the Z axis, then rotates the Y axis based on the result, and finally rotates the X axis based on the above two results.

[0076] See also Figure 7 , Figure 7 It is a cloud data graph of a strip point cloud collection. Specifically, S3 includes:

[0077] S31, intercepting an arc point cloud just above the top point cloud;

[0078] S32, fitting a standard circle with the arc-shaped point cloud;

[0079] S33, setting the width of the strip point cloud along the circumferential direction and the distance between adjacent strip point clouds along the circumferential direction, and obtaining a strip point cloud set.

[0080] First, a segment of arc-shaped point cloud is selected from the top point cloud, and the arc-shaped point cloud is located directly above the scanning instrument, and the circumferential direction of the tunnel is approximately the circumferential direction of the arc-shaped point cloud; secondly, a standard circle is fitted with the arc-shaped point cloud using the least squares method or other fitting methods, and the center and radius of the standard circle are recorded. In this embodiment, the center of the standard circle coincides with the center of the above-mentioned sphere, and the radius of the standard circle is the same as the radius of the above-mentioned sphere; further, the required width of the strip point cloud along the circumferential direction and the distance between adjacent strip point clouds along the circumferential direction are set, and the angle step Δθ is calculated. The angle step is the center angle corresponding to the distance between adjacent strip point clouds along the circumferential direction, and the starting angle and the ending angle of each segment of the strip point cloud intercepted along the circumferential direction in the top point cloud are calculated until the circumference of the entire top point cloud is covered.

[0081] Optionally, D is the distance between adjacent strip point clouds along the circumferential direction, and R is the radius of the standard circle.

[0082] Furthermore, S4 includes:

[0083] S41, projecting the strip point cloud set onto a vertical plane to obtain a two-dimensional strip point cloud set;

[0084] S42, performing radius filtering on the two-dimensional strip point cloud set to remove noise points;

[0085] S43, segmenting the two-dimensional strip point cloud set according to a certain step length;

[0086] S44, calculating the distribution variance of each segment of the point cloud, and retaining the segments whose distribution variance is greater than a preset value;

[0087] S45. Perform curve fitting on the retained segments to obtain the chute position, chute width, and chute depth.

[0088] First, a vertical plane perpendicular to the setting direction of the slideway is selected as the projection surface, and the strip point cloud set is projected onto the projection surface to obtain a two-dimensional strip point cloud set, such as Figure 8 As shown in , the difficulty of subsequent processing is reduced; further, for each point in the two-dimensional strip point cloud set, a fixed search radius is set, and only a sufficient number of neighboring points within the search radius are retained to filter out noise points; further, the two-dimensional strip point cloud set is segmented according to a certain step size, and the distribution variance of the point cloud is calculated in each segment. A large distribution variance indicates that the segment has a significant geometric change, that is, it represents the existence of a chute, so the segment with a large distribution variance is retained, and the segment with a small distribution variance is ignored, such as Fig. 9 As shown; finally, the retained segments are subjected to curve fitting to obtain the chute position, chute width and chute depth.

[0089] This embodiment provides a curve fitting method. Specifically, the curve fitting applies a Gaussian function, and the form of the Gaussian function is:

[0090]

[0091] Among them, μ is the mean of the function, that is, the center position of the curve peak, σ is the standard deviation, that is, the width of the curve, A is the maximum value of the function, that is, the peak value of the curve, and B is the curve offset. By finding the optimal values ​​of μ, σ, A and B in the Gaussian function, we can get the chute position, chute width and chute depth. The chute position is equal to μ, the chute width is equal to 2σ, and the chute depth is equal to AB.

[0092] In other feasible embodiments, other curve fitting methods may also be selected, which are not specifically limited here.

[0093] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, various obvious changes, readjustments and substitutions can be made without departing from the protection scope of the present invention. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A tunnel chute detection method based on point cloud data, characterized in that: The following steps are involved: S1. Tunnel point cloud data preprocessing, S1 includes: S11, setting the tunnel width to x, the tunnel height to y, taking the larger of x and y as the radius to construct a spherical surface, and removing the point cloud outside the spherical surface; S12, setting the minimum height of the edge of the chute to h, removing the point cloud of the spherical surface below h, and obtaining the top point cloud including the chute; S2. adjusting the cloud posture of the top point cloud; S3, cutting the top point cloud into a strip point cloud set; S4. The strip point cloud set is projected into a two-dimensional strip point cloud set, and the chute information is extracted from the two-dimensional strip point cloud set.

2. The tunnel chute detection method based on point cloud data according to claim 1 is characterized in that: The S2 includes: S21, intercepting a target point cloud from the top point cloud; S22, applying PCA algorithm to analyze the main direction of the target point cloud; S23, setting the angle between the main direction and the desired direction as the rotation angle θ, and constructing a composite rotation matrix; S24. The top point cloud adjusts the cloud posture according to the composite rotation matrix.

3. The tunnel chute detection method based on point cloud data according to claim 2 is characterized in that: The target point cloud is located in the highest point area of ​​the top point cloud.

4. The tunnel chute detection method based on point cloud data according to claim 2 is characterized in that: The S23 includes: S231, setting the angle between the main direction and the X-axis as the rotation angle θ1, the angle between the main direction and the Y-axis as the rotation angle θ2, and the angle between the main direction and the Z-axis as the rotation angle θ3; S232. Construct a composite rotation matrix.

5. The tunnel chute detection method based on point cloud data according to claim 2 is characterized in that: In S22, the main direction is the direction with the largest variance obtained by PCA algorithm analysis.

6. The tunnel chute detection method based on point cloud data according to claim 1 is characterized in that: The S3 includes: S31, intercepting an arc-shaped point cloud directly above the top point cloud; S32, fitting a standard circle with the arc-shaped point cloud; S33, setting the width of the strip point cloud along the circumferential direction and the distance between adjacent strip point clouds along the circumferential direction, and obtaining the strip point cloud set.

7. The tunnel chute detection method based on point cloud data according to claim 6 is characterized in that: The center of the standard circle coincides with the center of the spherical surface, and the radius of the standard circle is the same as the radius of the spherical surface.

8. The tunnel chute detection method based on point cloud data according to claim 1 is characterized in that: The S4 includes: S41, projecting the strip point cloud set onto a vertical plane to obtain the two-dimensional strip point cloud set; S42, performing radius filtering on the two-dimensional strip point cloud set to remove noise points; S43, segmenting the two-dimensional strip point cloud set according to a certain step length; S44, calculating the distribution variance of each segment of the point cloud, and retaining the segments whose distribution variance is greater than a preset value; S45, performing curve fitting on the retained segments to obtain the chute position, chute width and chute depth.

9. The tunnel chute detection method based on point cloud data according to claim 8 is characterized in that: In S45, the curve fitting applies a Gaussian function, and the form of the Gaussian function is The chute position is equal to μ, the chute width is equal to 2σ, and the chute depth is equal to AB.

10. The tunnel chute detection method based on point cloud data according to claim 8, characterized in that: The vertical plane is perpendicular to the setting direction of the slide groove.

Images