A fast conversion method of curved tunnel point cloud based on analytical projection transformation

Abstract

The present invention discloses a method for fast conversion of curved tunnel point clouds based on analytical projection transformation, which belongs to the field of analytical projection and comprises the following steps: firstly, performing plane linear isometric projection processing on the curved tunnel point cloud to generate a straight tunnel point cloud; secondly, performing tunnel wall semi-cylindrical isometric projection flattening processing on the tunnel wall of the straight tunnel point cloud generated in the previous step; finally, segmenting and classifying the tunnel wall appendages on the flattened tunnel wall; the method for fast conversion of curved tunnel point clouds based on analytical projection transformation provided by the present invention can effectively solve the problems of low efficiency, inaccurate segmentation results and incoherence brought about by the slicing processing method, and can effectively improve the segmentation efficiency of curved tunnel point cloud data, improve the accuracy of segmentation results and ensure data coherence.

Description

Technical Field

[0001] The invention relates to the technical field of analytical projection, and in particular to a method for quickly converting a curved tunnel point cloud based on analytical projection transformation. Background Art

[0002] Tunnels play an important role in transportation. From a social and economic perspective, tunnel safety is a key factor in ensuring transportation. However, the performance of various facilities in the tunnel (electromechanical facilities, signal facilities, ventilation facilities and other ancillary facilities) will gradually degrade during continuous use in harsh environments, posing huge safety hazards. Therefore, it is necessary to regularly divide and classify the annexes in the tunnel to ensure the normal service quality of the tunnel and the management of the annexes' components.

[0003] As a widely existing type of tunnel, curved tunnels have the characteristics of long mileage and poor lighting conditions. Based on these characteristics, 3D laser scanning technology is currently used to collect tunnel data. At present, the segmentation and classification of ancillary facilities in curved tunnels are mainly processed by slicing the curved tunnels to approximate them as straight tunnels. This method uses a certain number of slices for segmentation processing, which will cause problems such as low efficiency, inaccurate results, and incoherent segmentation results of the ancillary facilities in the curved tunnel. Therefore, how to quickly and accurately segment and classify the ancillary facilities in the tunnel has become a major challenge in the current point cloud processing of curved tunnels.

[0004] Based on this challenge, the present invention proposes a method for fast conversion of curved tunnel point clouds based on analytical projection transformation. Summary of the invention

[0005] The purpose of the present invention is to provide a method for fast conversion of curved tunnel point cloud based on analytical projection transformation, which can quickly transform the curved tunnel projection into a straight tunnel and flatten the tunnel wall by projection, thereby preparing for the subsequent segmentation and classification of the curved tunnel ancillary facilities. Compared with the current curved tunnel slicing processing method, this method can effectively improve the segmentation efficiency of curved tunnel point cloud data, improve the accuracy of segmentation results and ensure data consistency.

[0006] To achieve the above object, the present invention provides a method for fast conversion of curved tunnel point cloud based on analytical projection transformation, comprising the following steps:

[0007] Step 1, perform plane linear equidistant projection processing on the curved tunnel point cloud to generate a straight tunnel point cloud;

[0008] Step 2, performing tunnel wall semi-cylindrical equidistant projection flattening processing on the tunnel wall of the straight tunnel point cloud generated in step 1;

[0009] Step 3: Segment and classify the tunnel wall appendages on the flattened tunnel wall.

[0010] Preferably, the specific process of step 1 is as follows:

[0011] S11, using the Alpha Shapes algorithm to extract the boundary points of the curved tunnel, and then using the linear fitting method to obtain the boundary line parameters of the curved tunnel;

[0012] S12. According to the boundary line parameters fitted in S11, a number of tunnel centerline points are extracted using the tunnel centerline point automatic extraction algorithm. The linear parameters of the tunnel centerline, center C1 (a, b, 0) and radius R1 are obtained by linear fitting of the number of tunnel centerline points according to the circular curve using the least squares principle. According to the linear parameters of the tunnel centerline, the centerline point P2 (X2, Y2, Z2) corresponding to each tunnel point P1 (X1, Y1, Z1) is calculated.

[0013] S13, calculating the projection parameters O1(d, e) of the plane linear equidistant projection of the curved tunnel according to the midline point corresponding to each tunnel point;

[0014] S14. Performing projection transformation on the coordinates of the curved tunnel point cloud.

[0015] Preferably, the specific process of step 2 is as follows:

[0016] S21, segmenting the tunnel wall and the tunnel ground of the projected straight tunnel by a straight-through filtering method;

[0017] S22, the tunnel wall of the segmented straight tunnel is linearly fitted in the shape of a circular curve to obtain the shape parameters of a standard cylinder, namely, the center of the bottom circle C2 (f, g, m) and the radius R2;

[0018] S23, using a semi-cylinder projection method to project the tunnel wall points of the straight tunnel onto a cylinder to obtain cylinder points;

[0019] S24, select a reference point for isometric projection flattening, and take the cylindrical point with the largest Y coordinate value as the reference point O2 (X0, 0, Z0) for isometric projection flattening;

[0020] S25, flattening the cylindrical points obtained in S23;

[0021] S26. Assign elevation to the flattened point cloud.

[0022] Preferably, the process of step 3 is as follows:

[0023] S31, using CSF filtering to filter out tunnel wall lining points;

[0024] S32, filtering out the lining seam point cloud on the tunnel wall by using multiple linear fitting;

[0025] S33, segmenting the contact network poles using Euclidean distance clustering;

[0026] S34. Use the connected domain algorithm to separate the contact network and the power lines.

[0027] Preferably, in S12, the specific expression for calculating the centerline point P2 (X2, Y2, Z2) corresponding to each tunnel point P1 (X1, Y1, Z1) according to the linear parameters of the tunnel centerline is as follows:

[0028]

[0029] in,

[0030]

[0031] Preferably, the projection parameter O1(d, e) is specifically a point with the largest Y coordinate value selected from the midline point set.

[0032] Preferably, the specific process of S14 is as follows: according to the point O1 (d, e) with the largest Y coordinate value selected in the midline point set in S13, the arc length between O1 (d, e) and P2 (X2, Y2, Z2) is calculated, and the midline point P2 (X2, Y2, Z2) is projected onto the Y axis according to the principle of equidistant projection, and the positional relationship between each tunnel point P1 and the corresponding midline point P2 is used to complete the projection transformation to obtain the straight tunnel wall point P4 (X3, Y3, Z3), and the calculation expression is as follows:

[0033]

[0034] Preferably, in S23, the tunnel wall point P4 (X3, Y3, Z3) of the straight tunnel is projected onto the cylinder using the semi-cylinder projection method to obtain the calculation expression of the cylindrical point P5 (X4, Y4, Z4) as follows:

[0035]

[0036] Preferably, in S25, the cylindrical point P5 (X4, Y4, Z4) is flattened by an equidistant projection method to obtain P6 (X5, Y5, Z5), and the calculation expression is as follows:

[0037]

[0038] Therefore, the present invention adopts a method for fast conversion of curved tunnel point cloud based on analytical projection transformation of the above structure, firstly, the curved tunnel projection is transformed into a straight tunnel by plane linear isometric projection method, and no slicing is required, so as to prepare for the subsequent processing of the tunnel; then, the tunnel wall point cloud is projected onto a standard cylinder by semi-cylindrical isometric projection flattening method, and the tunnel wall is flattened by isometric projection method, and finally, the tunnel wall appendages on the flattened tunnel wall are segmented and classified. The method proposed by the present invention does not require slicing of the curved tunnel, and can quickly transform the curved tunnel into a straight tunnel by projection, and secondly, the tunnel wall can be flattened by projection, which effectively solves the problem of incoherent segmentation results of tunnel ancillary facilities, and the method can process more than one million curved tunnel point cloud data in one second, and as a projection method with strict mathematical logic, the method can restore the result to the original curved tunnel point cloud data coordinates by projection inverse operation method; the method is stable to use, simple in principle, and has been verified by actual data, and is applicable to various types of curved tunnels.

[0039] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the overall technical process of a method for rapid conversion of curved tunnel point clouds based on analytical projection transformation of the present invention;

[0041] Figure 2 It is a schematic diagram of the plane linear equidistant projection principle of an embodiment of the present invention;

[0042] Figure 3 It is a schematic diagram of the flattening principle of semi-cylinder projection according to an embodiment of the present invention;

[0043] Figure 4 Schematic diagram of the cylindrical flattening method based on isometric projection according to an embodiment of the present invention, wherein (a) represents a projection process diagram of the segmented tunnel wall point cloud, (b) represents a cylindrical projection process diagram of the tunnel wall point cloud, (c) represents a flattening process diagram of the tunnel wall point cloud, and (d) represents an elevation assignment process diagram of the flattened point cloud;

[0044] Figure 5 It is a diagram of the experimental process of the plane linear isometric projection of a curved tunnel according to an embodiment of the present invention;

[0045] Figure 6 This is a diagram of the experimental process of flattening the tunnel wall semi-cylindrical equidistant projection according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] Example

[0047] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0048] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] like Figure 1 , a method for fast conversion of curved tunnel point cloud based on analytical projection transformation, comprising the following steps:

[0050] Step 1: Perform plane linear equidistant projection processing on the curved tunnel point cloud to generate a straight tunnel point cloud; prepare for subsequent processing of the tunnel. Figure 2 The principle of the curve tunnel plane linear isometric projection method proposed in the present invention is explained. This method combines the centerline of the curve tunnel with the spatial position of the point cloud data, and transforms the curve tunnel projection into a straight tunnel using the principle of isometric projection. The specific steps are as follows:

[0051] S11, using the Alpha Shapes algorithm to extract the boundary points of the curved tunnel, and then using the linear fitting method to obtain the boundary line parameters of the curved tunnel;

[0052] S12. According to the boundary line parameters fitted in S11, a number of tunnel centerline points are extracted using the tunnel centerline point automatic extraction algorithm. The linear parameters of the tunnel centerline, center C1 (a, b, 0) and radius R1 are obtained by linear fitting of the number of tunnel centerline points according to the circular curve using the least squares principle. The centerline point P2 (X2, Y2, Z2) corresponding to each tunnel point P1 (X1, Y1, Z1) is calculated according to the linear parameters of the tunnel centerline. The specific expression for calculating the centerline point P2 (X2, Y2, Z2) corresponding to the tunnel point P1 (X1, Y1, Z1) is as follows:

[0053]

[0054] in,

[0055]

[0056] S13. Calculate the projection parameter O1(d, e) of the equidistant projection of the curved tunnel plane linear shape according to the midline point corresponding to each tunnel point; the projection parameter O1(d, e) is specifically the point with the largest Y coordinate value selected from the midline point set.

[0057] S14, perform projection transformation on the coordinates of the curved tunnel point cloud; according to the point O1(d, e) with the largest Y coordinate value selected in the midline point set in S13, calculate the arc length between O1(d, e) and P2(X2, Y2, Z2), and project the midline point P2(X2, Y2, Z2) onto the Y axis according to the principle of equidistant projection, and use the positional relationship between each tunnel point P1 and the corresponding midline point P2 to complete the projection transformation to obtain the straight tunnel wall point P4(X3, Y3, Z3), and the calculation expression is as follows:

[0058]

[0059] Step 2: Flatten the tunnel wall of the straight tunnel point cloud generated in step 1 by semi-cylindrical equidistant projection. Figure 3 As shown:

[0060] S21, segmenting the tunnel wall and the tunnel ground of the projected straight tunnel by a straight-through filtering method;

[0061] S22, the tunnel wall of the segmented straight tunnel is linearly fitted in the shape of a circular curve to obtain the shape parameters of a standard cylinder, namely, the center of the bottom circle C2 (f, g, m) and the radius R2;

[0062] S23, using the semi-cylinder projection method to project the tunnel wall point of the straight tunnel onto the cylinder to obtain the cylinder point; the calculation expression for projecting the tunnel wall point P4 (X3, Y3, Z3) of the straight tunnel onto the cylinder to obtain the cylinder point P5 (X4, Y4, Z4) is as follows;

[0063]

[0064] S24, select a reference point for isometric projection flattening, and take the cylindrical point with the largest Y coordinate value as the reference point O2 (X0, 0, Z0) for isometric projection flattening;

[0065] S25, flatten the cylindrical point obtained in S23; flatten the cylindrical point P5 (X4, Y4, Z4) by using the equidistant projection method to obtain P6 (X5, Y5, Z5), and the calculation expression is as follows:

[0066]

[0067] S26. Assign elevation to the flattened point cloud.

[0068] Step 3: Segment and classify the tunnel wall appendages on the flattened tunnel wall. The process is as follows:

[0069] S31, using CSF filtering to filter out tunnel wall lining points;

[0070] The flattened results of the tunnel wall mainly include the tunnel wall lining points, various attachments and other noise points. Among them, the tunnel wall lining points are located at the bottom of the flattened plane. In order to accurately and quickly segment and classify the tunnel wall attachments, the lining points are filtered out using the CSF cloth filter from bottom to top, and the filtered results are divided into two parts: non-lining points and lining points. Among them, the non-lining points mainly include various attachments, lining seams and other noise points.

[0071] S32, filtering out the lining seam point cloud on the tunnel wall by using multiple linear fitting;

[0072] Since the lining seam point cloud is evenly distributed on non-lining points and is closely connected with the appendages, it will cause the problem of inaccurate classification of appendages. Therefore, the lining seam point cloud is filtered out by combining the linear features and spatial position features of the lining seam and using multiple linear fitting methods.

[0073] S33, segmenting the contact network poles using Euclidean distance clustering;

[0074] S34. Use the connected domain algorithm to separate the contact network and the power lines.

[0075] After filtering out the lining seam point cloud, a point feature-based method is used to segment and classify non-lining points. First, the Euclidean distance clustering method is used to segment the contact network poles. Second, the connectivity threshold algorithm is used to segment and classify the transmission lines and contact networks.

[0076] Figure 4 The figure is a schematic diagram of the cylindrical flattening method based on equidistant projection. Among them, (a) shows the projection process of the tunnel wall point cloud after segmentation, (b) shows the cylindrical projection process of the tunnel wall point cloud, (c) shows the flattening process of the tunnel wall point cloud, and (d) shows the elevation assignment process of the flattened point cloud. Through the cylindrical flattening method based on equidistant projection, the tunnel wall point cloud can be projected and flattened, and the elevation of the flattened tunnel wall point cloud can be assigned. On this basis, various attachments on the tunnel wall can be segmented as a whole, effectively solving the problems of incoherent segmentation results and low efficiency caused by the slicing processing method. Figure 5 and Figure 6 The process of converting a curved tunnel into a straight tunnel and the process of flattening from tunnel wall to tunnel wall are shown respectively.

[0077] Therefore, the present invention adopts a method for fast conversion of curved tunnel point cloud based on analytical projection transformation. The method first transforms the curved tunnel projection into a straight tunnel through a plane linear isometric projection method; secondly, the tunnel wall is projected and flattened through a semi-cylindrical isometric projection method, and finally the tunnel wall appendages on the flattened tunnel wall are segmented and classified. The method mentioned in the present invention can quickly project the curved tunnel into a straight tunnel and project and flatten the tunnel wall. It solves the problems of low segmentation efficiency, inaccurate results and incoherent segmentation results of the current commonly used technologies.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for fast conversion of curved tunnel point cloud based on analytical projection transformation, characterized in that: The following steps are involved: Step 1, perform plane linear equidistant projection processing on the curved tunnel point cloud to generate a straight tunnel point cloud; Step 2, performing tunnel wall semi-cylindrical equidistant projection flattening processing on the tunnel wall of the straight tunnel point cloud generated in step 1; Step 3, segmenting and classifying the tunnel wall appendages on the flattened tunnel wall; The specific process of step 1 is as follows: S11, using the Alpha Shapes algorithm to extract the boundary points of the curved tunnel, and then using the linear fitting method to obtain the boundary line parameters of the curved tunnel; S12. According to the boundary line parameters fitted in S11, a number of tunnel centerline points are extracted using the tunnel centerline point automatic extraction algorithm. The linear parameters of the tunnel centerline, center C1 (a, b, 0) and radius R1 are obtained by linear fitting of the number of tunnel centerline points according to the circular curve using the least squares principle. The centerline point P2 (X2, Y2, Z2) corresponding to each tunnel point P1 (X1, Y1, Z1) is calculated according to the linear parameters of the tunnel centerline. S13, calculating the projection parameters O1(d, e) of the plane linear equidistant projection of the curved tunnel according to the midline point corresponding to each tunnel point; S14, performing projection transformation on the coordinates of the curved tunnel point cloud; The specific process of step 2 is as follows: S21, segmenting the tunnel wall and the tunnel ground of the projected straight tunnel by a straight-through filtering method; S22, the tunnel wall of the segmented straight tunnel is linearly fitted in the shape of a circular curve to obtain the shape parameters of a standard cylinder, namely, the center of the bottom circle C2 (f, g, m) and the radius R2; S23, using a semi-cylinder projection method to project the tunnel wall points of the straight tunnel onto a cylinder to obtain cylinder points; S24, select a reference point for isometric projection flattening, and take the cylindrical point with the largest Y coordinate value as the reference point O2 (X0, 0, Z0) for isometric projection flattening; S25, flattening the cylindrical points obtained in S23; S26. Assign elevation to the flattened point cloud.

2. According to the method of claim 1, the method is characterized in that: The process for step 3 is as follows: S31, using CSF filtering to filter out tunnel wall lining points; S32, filtering out the lining seam point cloud on the tunnel wall by using multiple linear fitting; S33, segmenting the contact network poles using Euclidean distance clustering; S34. Use the connected domain algorithm to separate the contact network and the power lines.

3. The method for rapid conversion of curved tunnel point cloud based on analytical projection transformation according to claim 2 is characterized in that: In S12, the specific expression for calculating the centerline point P2 (X2, Y2, Z2) corresponding to each tunnel point P1 (X1, Y1, Z1) according to the linear parameters of the tunnel centerline is as follows: ； in, 。 4. The method for rapid conversion of curved tunnel point cloud based on analytical projection transformation according to claim 3 is characterized in that: The projection parameter O1(d,e) is specifically the point with the largest Y coordinate value selected from the midline point set.

5. The method for rapid conversion of curved tunnel point cloud based on analytical projection transformation according to claim 4 is characterized in that: The specific process of S14 is as follows: according to the point O1(d, e) with the largest Y coordinate value selected in the midline point set in S13, the arc length between O1(d, e) and P2(X2, Y2, Z2) is calculated, and the midline point P2(X2, Y2, Z2) is projected onto the Y axis according to the principle of equidistant projection. The projection transformation is completed by using the positional relationship between each tunnel point P1 and the corresponding midline point P2 to obtain the straight tunnel wall point P4(X3, Y3, Z3). The calculation expression is as follows: 。 6. The method for rapid conversion of curved tunnel point cloud based on analytical projection transformation according to claim 5 is characterized in that: In S23, the tunnel wall point P4 (X3, Y3, Z3) of the straight tunnel is projected onto the cylinder using the semi-cylinder projection method to obtain the cylindrical point P5 (X4, Y4, Z4). The calculation expression is as follows: 。 7. The method for rapid conversion of curved tunnel point cloud based on analytical projection transformation according to claim 6 is characterized by: In S25, the cylindrical point P5 (X4, Y4, Z4) is flattened by the equidistant projection method to obtain P6 (X5, Y5, Z5). The calculation expression is as follows: 。

Images