Point cloud data filtering method and device for coal handling in coal yards

Abstract

The present disclosure provides a method and device for filtering point cloud data for coal counting in a coal yard. The method includes: obtaining three-dimensional point cloud data of a coal pile; triangulating the projection of the three-dimensional point cloud data of the coal pile on a horizontal plane; calculating the overall variance of the actual heights of the three vertex vectors in each triangular mesh; calculating the actual absolute value of the slope between the three vertex vectors in each triangular mesh; determining whether there is abnormal point cloud data in the triangular mesh based on the overall variance of the actual heights and the actual absolute value of the slope of each triangular mesh; performing correction filtering on the triangular mesh with abnormal point cloud data; and reconstructing a three-dimensional model of the coal pile based on the height information of each triangular mesh and each corresponding point cloud data after correction filtering. The method disclosed in the present disclosure can reduce the coal counting data error caused by abnormal point cloud data during the process of laser counting in a coal yard, effectively preserve the details of the on-site fuel pile, improve data accuracy, and improve coal counting management in the coal yard.

Description

Technical Field

[0001] The present disclosure belongs to the technical field of coal panning in coal yards, and particularly relates to a point cloud data filtering method and device for coal panning in coal yards. Background Art

[0002] Fuel management in coal-fired power plants has always been a critical issue. Coal accounts for 60%-70% of a plant's operating costs. Improving coal yard management can significantly reduce these costs. Existing coal yard fuel management relies heavily on manually creating coal type storage diagrams based on inflow and outflow information. This method is inaccurate and inefficient. Manually creating coal type storage information in coal yards can only roughly construct coal type storage information based on experience and visual observation of on-site conditions. This information diagram significantly affects subsequent fuel storage and retrieval. For example, different coal quality test data can significantly affect boiler combustion. Therefore, how to more accurately complete the informatization and digitization of coal type distribution data in coal yards through automated means has become a major research direction.

[0003] During the laser coal counting process, coal pushers or bucket wheel stacker arms often operate above the coal yard, causing the laser scanner to scan the operating equipment during the scanning process, resulting in abnormal data and affecting the accuracy of the inventory data in the material yard.

[0004] Existing point cloud data filtering algorithms include voxel filtering algorithms, statistical filtering algorithms, radius filtering algorithms, etc., which are relatively common point cloud filtering algorithms. They can eliminate some noise data to a certain extent, but they will also eliminate some details of the entire model, making the model smooth and blurred, and the error of the coal pan data calculated based on these data will increase. Summary of the Invention

[0005] The present disclosure aims to solve at least one of the technical problems existing in the prior art and provide a method and device for filtering point cloud data of coal in a coal yard.

[0006] In one aspect of the present disclosure, a method for filtering point cloud data for coal distribution in a coal yard is provided, the method comprising:

[0007] Obtain three-dimensional point cloud data of the coal pile;

[0008] Performing triangulation on the projection of the three-dimensional point cloud data of the coal pile on a horizontal plane to obtain a plurality of triangulated meshes;

[0009] Calculating the total variance of the actual heights of the three vertex vectors in each of the triangle meshes;

[0010] Calculating the actual slope absolute value between three vertex vectors in each of the triangle meshes;

[0011] Determining whether there is abnormal point cloud data in the triangular mesh according to the actual height overall variance and the actual slope absolute value of each triangular mesh;

[0012] Performing correction filtering processing on each triangular mesh containing the abnormal point cloud data;

[0013] The three-dimensional model of the coal pile is reconstructed based on the height information of each triangular mesh after correction and filtering and the corresponding point cloud data.

[0014] In some embodiments, determining whether abnormal point cloud data exists in the triangular mesh according to the actual height overall variance and the actual slope absolute value of each triangular mesh includes:

[0015] Comparing the actual overall height variance of the triangular mesh with a preset overall height variance threshold, and comparing the actual absolute value of the slope of each vertex vector of the triangular mesh with a preset absolute slope threshold;

[0016] If the actual height overall variance of the triangular mesh is greater than the height overall variance threshold, and the actual slope absolute value of each vertex vector of the triangular mesh is greater than the slope absolute threshold, it is determined that the triangular mesh has abnormal point cloud data.

[0017] In some embodiments, determining whether abnormal point cloud data exists in the triangular mesh includes:

[0018] The vertices of the triangular mesh containing abnormal point cloud data are sorted according to their heights, and the vertex corresponding to the maximum value is the abnormal point cloud data.

[0019] In some embodiments, performing correction filtering on each triangular mesh containing abnormal point cloud data includes:

[0020] The abnormal point cloud data in the triangulated network is corrected and filtered using the remaining two normal point cloud data in the triangulated network.

[0021] In some embodiments, the use of the remaining two normal point cloud data in the triangulated network to perform correction filtering on the abnormal point cloud data in the triangulated network includes:

[0022] The abnormal point cloud data is corrected and filtered according to the following relationship:

[0023] NoisePoint′=(x n ,y n ,(z i +z j ) / 2)

[0024] Among them, NoisePoint′ is the abnormal point cloud data after correction, x n is the x-axis coordinate of the abnormal point cloud data before correction, y n is the y-axis coordinate of the abnormal point cloud data before correction, z i is the z-axis coordinate of one of the normal point cloud data in the triangular mesh where the abnormal point cloud data is located, z j It is the z-axis coordinate of another normal point cloud data in the triangular mesh where the abnormal point cloud data is located.

[0025] Another aspect of the present disclosure provides a point cloud data filtering device for coal distribution in a coal yard, the device comprising:

[0026] An acquisition module, used to acquire three-dimensional point cloud data of the coal pile;

[0027] A meshing module, configured to perform triangular meshing on the projection of the three-dimensional point cloud data of the coal pile on a horizontal plane to obtain a plurality of triangular meshes;

[0028] a calculation module, configured to calculate the total variance of the actual heights of the three vertex vectors in each of the triangular meshes; and

[0029] The calculation module is further configured to calculate the actual absolute value of the slope between the three vertex vectors in each of the triangular meshes;

[0030] a processing module, configured to determine whether there is abnormal point cloud data in the triangular mesh according to the actual height overall variance and the actual slope absolute value of each triangular mesh;

[0031] A correction module, configured to perform correction filtering on each triangular mesh containing abnormal point cloud data;

[0032] A module is established for reconstructing a three-dimensional model of the coal pile based on the height information of each triangular mesh after correction and filtering and the corresponding point cloud data.

[0033] In some embodiments, the processing module is further configured to:

[0034] Comparing the actual overall height variance of the triangular mesh with a preset overall height variance threshold, and comparing the actual absolute value of the slope of each vertex vector of the triangular mesh with a preset absolute slope threshold;

[0035] If the actual height overall variance of the triangular mesh is greater than the height overall variance threshold, and the actual slope absolute value of each vertex vector of the triangular mesh is greater than the slope absolute threshold, it is determined that the triangular mesh has abnormal point cloud data.

[0036] In some embodiments, the processing module is further configured to:

[0037] The vertices of the triangular mesh containing abnormal point cloud data are sorted according to their heights, and the vertex corresponding to the maximum value is the abnormal point cloud data.

[0038] In some embodiments, the correction module is further configured to:

[0039] The abnormal point cloud data in the triangulated network is corrected and filtered using the remaining two normal point cloud data in the triangulated network.

[0040] In some embodiments, the correction module performs correction filtering on the abnormal point cloud data according to the following relationship:

[0041] NoisePoint′=(x n ,y n ,(z i +z j ) / 2)

[0042] Among them, NoisePoint′ is the abnormal point cloud data after correction, x n is the x-axis coordinate of the abnormal point cloud data before correction, y n is the y-axis coordinate of the abnormal point cloud data before correction, z i is the z-axis coordinate of one of the normal point cloud data in the triangular mesh where the abnormal point cloud data is located, z j It is the z-axis coordinate of another normal point cloud data in the triangular mesh where the abnormal point cloud data is located.

[0043] The disclosed method and device for filtering point cloud data for coal counting in a coal yard can significantly reduce the coal counting data errors caused by abnormal point cloud data during the process of laser coal counting in the coal yard. While retaining the details of the on-site fuel pile to the greatest extent, the abnormal point cloud data can be filtered out, thereby improving data accuracy and enhancing coal counting management in the coal yard. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A schematic diagram of a three-dimensional model obtained by reconstructing a coal pile using existing technology;

[0045] Figure 2 A schematic diagram of gridding three-dimensional point cloud data of a coal pile according to an embodiment of the present disclosure;

[0046] Figure 3 A schematic diagram of three-dimensional gridding of abnormal point cloud data included in point cloud data according to another embodiment of the present disclosure;

[0047] Figure 4This is a flow chart of a method for filtering point cloud data of coal in a coal yard according to another embodiment of the present disclosure;

[0048] Figure 5 A schematic diagram of a three-dimensional model obtained by reconstructing a coal pile according to another embodiment of the present disclosure;

[0049] Figure 6 This is a structural schematic diagram of a point cloud data filtering device for coal panning in a coal yard according to another embodiment of the present disclosure. DETAILED DESCRIPTION

[0050] In order to enable those skilled in the art to better understand the technical solutions of the present disclosure, the present disclosure is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0051] The following first describes the background of the present disclosure.

[0052] The coal pile in the coal yard is relatively flat, and the reconstructed 3D model is also relatively flat. When a coal pusher or bucket wheel stacker and reclaimer arm is operating above the coal yard, the scanned point cloud data will show many discrete outliers. After the 3D model is reconstructed, many burrs will appear, which seriously does not conform to the current fuel distribution situation on site. Figure 1 shown.

[0053] Point cloud data refers to a set of vectors in a three-dimensional coordinate system, including the spatial coordinate information (x, y, z) of each point. The point cloud data of the coal yard is projected on the horizontal plane to form a triangulated mesh using a triangulated mesh algorithm, such as Figure 2 shown.

[0054] After the point cloud data is projected and triangulated on the horizontal plane, it can be reconstructed into a 3D model based on the grid information and the height information of each point. The noise data contained in the point cloud data is displayed as follows when it is 3D gridded. Figure 3 shown.

[0055] according to Figure 3 It can be seen that NoisePoint represents the abnormal point cloud data often collected when on-site laser scanning collects data. Points p1, p2, p3, and p4 are normal point cloud data (generally normal fuel distribution data). The characteristics of abnormal point cloud data are protrusions and outliers. In the reconstructed three-dimensional model, they appear as protrusions and burrs. Based on this, the present disclosure mainly involves how to eliminate data errors caused by abnormal point cloud data.

[0056] like Figure 4 As shown, the embodiment of the present disclosure relates to a point cloud data filtering method S100 for coal distribution in a coal yard, and the method S100 includes:

[0057] S110. Acquire three-dimensional point cloud data of the coal pile.

[0058] Specifically, in this step, a laser scanning device can be used to obtain the three-dimensional point cloud data of the entire coal pile, such as Figure 2 As shown in the figure, during the laser scanning process, the spatial position coordinates of all points can be obtained, and p1(x1,y1,z1), p2(x2,y2,z2), p3(x3,y3,z3), p4(x4,y4,z4) are set, NoisePoint=(x n ,y n ,z n ).

[0059] S120 , performing triangular meshing on the projection of the three-dimensional point cloud data of the coal pile on a horizontal plane to obtain a plurality of triangular meshes.

[0060] Specifically, in this step, Figure 2 and Figure 3 As shown, the top line combination of the triangle surface can be known through the triangular meshing algorithm, such as Figure 3 It can be seen that (p1, p2, NoisePoint), (p3, p2, NoisePoint), (p3, p4, NoisePoint), (p2, p4, NoisePoint).

[0061] S130: Calculate the overall variance of the actual heights of the three vertex vectors in each of the triangular meshes.

[0062] Specifically, in this step, the actual height variance of the three vertex vectors in each triangle mesh is calculated as follows:

[0063] heightAvg=(z1+z2+z3) / 3

[0064] height 2 =((z1-heightAvg) 2 +(z2-heightAvg) 2 +(z3-heightAvg) 2 ) / 3

[0065] S140: Calculate the actual absolute value of the slope between the three vertex vectors in each of the triangular meshes.

[0066] Specifically, in this step, the actual slope absolute value between the three vertex vectors in each triangle mesh is calculated according to the following relationship:

[0067]

[0068]

[0069]

[0070] S150 , determining whether there is abnormal point cloud data in the triangular mesh according to the actual height overall variance and the actual slope absolute value of each triangular mesh.

[0071] Specifically, in this step, the actual overall height variance of the triangle mesh may be compared with a preset overall height variance threshold, and the actual absolute value of the slope of each vertex vector of the triangle mesh may be compared with a preset absolute slope threshold.

[0072] If the actual height overall variance of the triangular mesh is greater than the height overall variance threshold, and the actual slope absolute value of each vertex vector of the triangular mesh is greater than the slope absolute threshold, it is determined that the triangular mesh has abnormal point cloud data.

[0073] In this step, the height overall variance threshold and slope absolute threshold can be calculated by analyzing the equipment parameters of the abnormal point cloud data caused by the laser coal mining site.

[0074] More specifically, after determining that a triangular mesh contains abnormal point cloud data, it is necessary to determine which vertex in the triangular mesh is the abnormal point cloud data. As an example, the vertices of the triangular mesh containing the abnormal point cloud data can be sorted by height, and the vertex corresponding to the maximum value is the abnormal point cloud data.

[0075] S160: Perform correction filtering processing on each triangular mesh containing the abnormal point cloud data.

[0076] Specifically, in this step, due to the high density of point cloud data during the laser coal processing, the area of ​​the triangular surface reconstructed by triangulation will basically not exceed 1m 3 , in the actual fuel coal pile site, 1m 3 The regional fuel distribution is relatively uniform and the model is relatively flat, so the protruding abnormal point cloud data can be corrected by the other two normal point cloud data.

[0077] Specifically, the abnormal point cloud data is corrected and filtered according to the following relationship:

[0078] NoisePoint′=(x n ,y n ,(z i +z j ) / 2)

[0079] Among them, NoisePoint′ is the abnormal point cloud data after correction, x n is the x-axis coordinate of the abnormal point cloud data before correction, yn is the y-axis coordinate of the abnormal point cloud data before correction, z i is the z-axis coordinate of one of the normal point cloud data in the triangular mesh where the abnormal point cloud data is located, z j It is the z-axis coordinate of another normal point cloud data in the triangular mesh where the abnormal point cloud data is located.

[0080] S170 , reconstructing a three-dimensional model of the coal pile based on the height information of each triangular mesh after correction and filtering and the corresponding point cloud data.

[0081] Specifically, in this step, the three-dimensional model of the coal pile is reconstructed based on the height information of each triangular mesh and each corresponding point cloud data after correction and filtering. The three-dimensional model is as follows: Figure 5 As shown. Figure 5 It can be seen that the burrs on the reconstructed 3D model disappear, and the details of the on-site fuel pile can be retained to the greatest extent.

[0082] The disclosed point cloud data filtering method for coal counting in a coal yard can significantly reduce the coal counting data errors caused by abnormal point cloud data during the process of laser coal counting in the coal yard. While retaining the details of the on-site fuel pile to the greatest extent, it can filter out abnormal point cloud data, improve data accuracy, and enhance coal counting management in the coal yard.

[0083] Another aspect of the present disclosure is Figure 6 As shown, a point cloud data filtering device 100 for coal in a coal field is provided. The device 100 can be applied to the method described above. For details, please refer to the relevant descriptions above and will not be described here. The device 100 includes:

[0084] The acquisition module 110 is used to acquire three-dimensional point cloud data of the coal pile.

[0085] The meshing module 120 is used to perform triangular meshing on the projection of the three-dimensional point cloud data of the coal pile on the horizontal plane to obtain a plurality of triangular meshes.

[0086] The calculation module 130 is used to calculate the total variance of the actual heights of the three vertex vectors in each triangle mesh. The calculation module 130 is also used to calculate the absolute value of the actual slope between the three vertex vectors in each triangle mesh.

[0087] The processing module 140 is configured to determine whether abnormal point cloud data exists in the triangular mesh according to the actual height overall variance and the actual slope absolute value of each triangular mesh.

[0088] The correction module 150 is configured to perform correction filtering on each of the triangular meshes containing the abnormal point cloud data.

[0089] The establishment module 160 is used to reconstruct the three-dimensional model of the coal pile according to the height information of each triangular mesh and each corresponding point cloud data after the correction and filtering processing.

[0090] The point cloud data filtering device disclosed in the present invention for coal counting in a coal yard can significantly reduce the coal counting data errors caused by abnormal point cloud data during the process of laser counting in the coal yard. On the basis of retaining the details of the on-site fuel pile to the greatest extent, it can filter out abnormal point cloud data, improve data accuracy, and improve coal counting management in the coal yard.

[0091] In some embodiments, the processing module 140 is further configured to:

[0092] The actual overall height variance of the triangle mesh is compared with a preset overall height variance threshold, and the actual absolute value of the slope of each vertex vector of the triangle mesh is compared with a preset absolute slope threshold.

[0093] If the actual height overall variance of the triangular mesh is greater than the height overall variance threshold, and the actual slope absolute value of each vertex vector of the triangular mesh is greater than the slope absolute threshold, it is determined that the triangular mesh has abnormal point cloud data.

[0094] In some embodiments, the processing module 140 is further configured to:

[0095] The vertices of the triangular mesh containing abnormal point cloud data are sorted according to their heights, and the vertex corresponding to the maximum value is the abnormal point cloud data.

[0096] In some embodiments, the correction module 150 is further configured to:

[0097] The abnormal point cloud data in the triangulated network is corrected and filtered using the remaining two normal point cloud data in the triangulated network.

[0098] In some embodiments, the correction module 150 performs correction filtering on the abnormal point cloud data according to the following relationship:

[0099] NoisePoint′=(x n ,y n ,(z i +z j ) / 2)

[0100] Among them, NoisePoint′ is the abnormal point cloud data after correction, x n is the x-axis coordinate of the abnormal point cloud data before correction, y n is the y-axis coordinate of the abnormal point cloud data before correction, z iis the z-axis coordinate of one of the normal point cloud data in the triangular mesh where the abnormal point cloud data is located, z j It is the z-axis coordinate of another normal point cloud data in the triangular mesh where the abnormal point cloud data is located.

[0101] It is understood that the above embodiments are merely exemplary embodiments for illustrating the principles of the present disclosure, and the present disclosure is not limited thereto. Those skilled in the art may make various modifications and improvements without departing from the spirit and substance of the present disclosure, and such modifications and improvements are also considered to be within the scope of protection of the present disclosure.

Claims

1. A point cloud data filtering method for coal in a coal field, characterized in that: The method comprises: Obtain three-dimensional point cloud data of the coal pile; Performing triangulation on the projection of the three-dimensional point cloud data of the coal pile on a horizontal plane to obtain a plurality of triangulated meshes; Calculating the total variance of the actual heights of the three vertex vectors in each of the triangle meshes; Calculating the actual slope absolute value between three vertex vectors in each of the triangle meshes; Determining whether there is abnormal point cloud data in the triangular mesh according to the actual height overall variance and the actual slope absolute value of each triangular mesh; Performing correction filtering processing on each triangular mesh containing the abnormal point cloud data; Reconstructing a three-dimensional model of the coal pile based on the height information of each triangular mesh and each corresponding point cloud data after correction and filtering processing; The determining whether there is abnormal point cloud data in the triangular mesh according to the actual height overall variance and the actual slope absolute value of each triangular mesh includes: Comparing the actual overall height variance of the triangular mesh with a preset overall height variance threshold, and comparing the actual absolute value of the slope of each vertex vector of the triangular mesh with a preset absolute slope threshold; If the actual height overall variance of the triangular mesh is greater than the height overall variance threshold, and the actual slope absolute value of each vertex vector of the triangular mesh is greater than the slope absolute threshold, it is determined that the triangular mesh has abnormal point cloud data.

2. The method according to claim 1, characterized in that The determining whether abnormal point cloud data exists in the triangular mesh includes: The vertices of the triangular mesh containing abnormal point cloud data are sorted according to their heights, and the vertex corresponding to the maximum value is the abnormal point cloud data.

3. The method according to claim 1 or 2, characterized in that The corrective filtering process is performed on each triangular mesh containing abnormal point cloud data, including: The abnormal point cloud data in the triangular mesh is corrected and filtered using the remaining two normal point cloud data in the triangular mesh.

4. The method according to claim 3, characterized in that The method of using the remaining two normal point cloud data in the triangular mesh to perform correction filtering on the abnormal point cloud data in the triangular mesh includes: The abnormal point cloud data is corrected and filtered according to the following relationship: in, is the abnormal point cloud data after correction, x n is the x-axis coordinate of the abnormal point cloud data before correction, y n is the y-axis coordinate of the abnormal point cloud data before correction, z i is the z-axis coordinate of one of the normal point cloud data in the triangular mesh where the abnormal point cloud data is located, z j It is the z-axis coordinate of another normal point cloud data in the triangular mesh where the abnormal point cloud data is located.

5. A point cloud data filtering device for coal in a coal field, characterized in that: The device comprises: An acquisition module, used to acquire three-dimensional point cloud data of the coal pile; A meshing module, configured to perform triangular meshing on the projection of the three-dimensional point cloud data of the coal pile on a horizontal plane to obtain a plurality of triangular meshes; a calculation module, configured to calculate the total variance of the actual heights of the three vertex vectors in each of the triangular meshes; and The calculation module is further configured to calculate the actual absolute value of the slope between the three vertex vectors in each of the triangular meshes; a processing module, configured to determine whether there is abnormal point cloud data in the triangular mesh according to the actual height overall variance and the actual slope absolute value of each triangular mesh; A correction module, configured to perform correction filtering on each triangular mesh containing abnormal point cloud data; Establishing a module for reconstructing a three-dimensional model of the coal pile based on the height information of each triangular mesh and each corresponding point cloud data after correction and filtering; The processing module is further specifically used to: compare the actual overall height variance of the triangular mesh with a preset overall height variance threshold, and compare the actual absolute value of the slope of each vertex vector of the triangular mesh with a preset absolute slope threshold; if the actual overall height variance of the triangular mesh is greater than the overall height variance threshold, and the actual absolute value of the slope of each vertex vector of the triangular mesh is greater than the absolute slope threshold, then it is determined that the triangular mesh has abnormal point cloud data.

6. The device according to claim 5, characterized in that The processing module is further configured to: The vertices of the triangular mesh containing abnormal point cloud data are sorted according to their heights, and the vertex corresponding to the maximum value is the abnormal point cloud data.

7. The device according to claim 5 or 6, characterized in that The correction module is further configured to: The abnormal point cloud data in the triangular mesh is corrected and filtered using the remaining two normal point cloud data in the triangular mesh.

8. The device according to claim 7, characterized in that The correction module performs correction filtering on the abnormal point cloud data according to the following relationship: in, is the abnormal point cloud data after correction, x n is the x-axis coordinate of the abnormal point cloud data before correction, y n is the y-axis coordinate of the abnormal point cloud data before correction, z i is the z-axis coordinate of one of the normal point cloud data in the triangular mesh where the abnormal point cloud data is located, z j It is the z-axis coordinate of another normal point cloud data in the triangular mesh where the abnormal point cloud data is located.

Images