A tunnel three-dimensional laser point cloud data denoising interpolation method, system and device

Abstract

The application provides a tunnel three-dimensional laser point cloud data denoising interpolation method, system and device, relates to the application technical field of mountain tunnel three-dimensional laser scanning, and the method comprises the following steps: obtaining initial point cloud data of excavation, primary support and secondary lining section, calculating the azimuth angle and overbreak and underbreak value of each point in the point cloud data, removing noise through the difference method and the mean value method, performing interpolation processing based on the improved Kriging interpolation method, and calculating parameters such as overbreak and underbreak value, overbreak and underbreak area and underbreak intrusion position number. Through the method provided by the application, the noise point group and discrete noise points can be effectively removed, the precision and accuracy of tunnel section overbreak and underbreak detection are improved, the misjudgment of underbreak and intrusion position is avoided, the point cloud data at the missing position can be automatically completed, the precision of overbreak and underbreak detection is further improved, and the calculation result is more suitable for engineering practice.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional laser scanning application technology for mountain tunnels, and in particular to a method, system, and device for denoising and interpolating three-dimensional laser point cloud data of tunnels. Background Technology

[0002] Currently, mountain tunnels are mostly constructed using mining methods, which inevitably lead to over-excavation and under-excavation during the blasting excavation process. In order to effectively control over-excavation and avoid under-excavation, thereby saving construction costs and improving the safety of the lining structure, it is necessary to conduct over-excavation and under-excavation detection on the tunnel cross-section.

[0003] With the advancement of technology, three-dimensional laser scanning technology has been applied to the detection of over-excavation and under-excavation of tunnel cross sections. This technology obtains the three-dimensional coordinates, reflectivity, and texture information of each point on the tunnel surface through laser ranging, and can quickly acquire point cloud data inside the tunnel. Compared with the traditional point-by-point and surface-by-surface measurement method used in detection, the three-dimensional laser scanning technology has the characteristics of convenience, high detection accuracy, and wide detection range. At the same time, it can improve construction efficiency and construction quality, and facilitate construction project management.

[0004] Due to the presence of structures such as steel reinforcement mesh, ventilation and water pipes, and equipment chambers within tunnels, the scanned point cloud data contains some noise. This may lead to inaccurate over-excavation and under-excavation calculations, as well as misjudgments of under-excavation and encroachment locations, severely impacting the accuracy and precision of the results. Most existing research focuses on denoising large-scale discrete noise points such as airborne dust, while research on removing noise clusters from steel reinforcement mesh, water pipes, ventilation ducts, and chambers within tunnels is relatively limited. This results in inaccurate cross-sectional over-excavation and under-excavation detection parameters, potentially leading to misjudgments of the number and location of under-excavation and encroachment points, hindering refined cross-sectional detection of mountain tunnels. Therefore, there is an urgent need to research a more practical method for refined denoising and interpolation of 3D laser scanning point cloud data for mountain tunnels. Summary of the Invention

[0005] The purpose of this invention is to provide a method for denoising and interpolating three-dimensional laser point cloud data in tunnels, so as to solve at least one of the above-mentioned technical problems existing in the prior art.

[0006] To solve the above-mentioned technical problems, the present invention provides a method for denoising and interpolating three-dimensional laser point cloud data of tunnels, which includes acquiring initial point cloud data, establishing a rectangular coordinate system with the center of the arch arc as the origin, calculating the azimuth angle and over- or under-excavation value of each point in the point cloud data, and calculating the over- or under-excavation detection parameters of the cross section after denoising and interpolation processing.

[0007] The cross-sectional over- or under-excavation detection parameters include over- or under-excavation values, number of under-excavation encroachments, and over- or under-excavation area.

[0008] The above methods can effectively remove noise and improve the accuracy of cross-section over-excavation and under-excavation detection; at the same time, the point cloud data at the missing locations can be automatically completed, further improving the accuracy of over-excavation and under-excavation detection.

[0009] Furthermore, the denoising includes difference-based denoising, which includes: sorting the points according to the magnitude of the azimuth angle, calculating the azimuth angle difference and over-dig / under-dig value difference between adjacent points, and performing group denoising when the azimuth angle difference is less than a first preset threshold and the over-dig / under-dig value difference is greater than a second preset threshold.

[0010] The above denoising methods are used to remove noise clusters.

[0011] Preferably, the specific method for group denoising includes: determining the neighborhood of each point based on the first preset threshold, calculating the average over-drilling and under-drilling value of the points in the neighborhood, dividing the neighborhood into two groups based on the comparison between each point in the neighborhood and the average value, distinguishing noise points and normal points according to the number of points in each group, with points in a minority group being noise points and points in a majority group being normal points.

[0012] Furthermore, the method for determining the neighborhood includes: determining the neighborhood range of a point based on a certain point by a ratio multiple of the first preset threshold.

[0013] Furthermore, the method for determining the second preset threshold includes: calculating the maximum and minimum values ​​of the difference between the over-drilling and under-drilling values, and using the ratio of the difference between the maximum and the minimum values ​​as the second preset threshold, which facilitates the determination of a suitable second preset threshold.

[0014] Furthermore, the denoising also includes mean-based denoising, which is set after the difference-based denoising. The mean-based denoising includes: after the difference-based denoising, sorting the remaining points according to the size of the azimuth angle, determining the neighborhood of each point based on the proportion multiple of the first preset threshold, calculating the average over-drilling and under-drilling value of the points in the neighborhood, calculating the difference between each point and the average value, and if the difference is greater than the third preset threshold, then the current point is determined to be a noise point.

[0015] Furthermore, the third preset threshold can be a proportional multiple of the second preset threshold.

[0016] Preferably, the third preset threshold is 0.5 times the second preset threshold.

[0017] The above denoising methods can effectively remove discrete noise points.

[0018] Furthermore, the interpolation process includes: sorting the denoised points according to their azimuth angles, calculating the azimuth angle difference between adjacent points, and when the azimuth angle difference is greater than a fourth preset threshold, solving for the azimuth angle and over- or under-dig value of the interpolation point between the adjacent points, and adding the interpolation point to the point cloud data.

[0019] Preferably, the method for solving the interpolation process can be an interpolation calculation method based on, for example, Kriging interpolation.

[0020] The above interpolation process is used to automatically and smoothly complete the point cloud data at the missing locations, further improving the accuracy of over-excavation and under-excavation detection.

[0021] Furthermore, the over-excavation and under-excavation areas are calculated using the angle integration method. The specific steps include: sorting the point cloud data according to the azimuth angle, calculating the difference in azimuth angle between adjacent points, calculating the sector area micro-element enclosed by the adjacent points and the origin based on the sector area formula, filtering the micro-element values ​​greater than or equal to 0 and summing them to obtain the over-excavation area, filtering the micro-element values ​​less than 0 and summing them to obtain the under-excavation area, and summing the over-excavation area and the under-excavation area to obtain the over-excavation and under-excavation area.

[0022] On the other hand, the present invention also provides a tunnel three-dimensional laser point cloud data denoising interpolation system, including a data receiving module, a data processing module, and a result generation module:

[0023] The data receiving module is used to receive initial point cloud data;

[0024] The data processing module includes a preprocessing unit, a noise reduction unit, an interpolation unit, and an over- or under-dig calculation unit.

[0025] The preprocessing unit, based on the initial point cloud data, establishes a rectangular coordinate system with the center of the arch arc as the origin, calculates the azimuth angle and over- or under-dig values ​​of each point, and then sends them to the denoising unit.

[0026] The denoising unit removes noise clusters using the difference method based on the azimuth difference and the over-drilling / under-drilling difference between two adjacent points; it removes discrete noise points using the mean method based on the mean of the over-drilling / under-drilling values ​​within the neighborhood of each point; and it sends the noise-removed point cloud data to the interpolation unit.

[0027] The interpolation unit calculates the azimuth difference between adjacent points based on the point cloud data after noise removal, and then solves the azimuth and over- or under-dig values ​​of the interpolation points. After adding the interpolation points to the point cloud data, it sends them to the over- or under-dig calculation unit.

[0028] The over-excavation and under-excavation calculation unit calculates the over-excavation and under-excavation values ​​and the number of under-excavation encroachments based on the interpolated point cloud data; it calculates the over-excavation and under-excavation area using the angle integration method and sends the calculation results to the result generation module.

[0029] The result generation module is used to output the point cloud data after noise reduction and interpolation processing, as well as the cross-section over-excavation and under-excavation detection parameters.

[0030] In another aspect, the present invention also provides a tunnel three-dimensional laser point cloud data denoising interpolation device for the above-mentioned tunnel three-dimensional laser point cloud data denoising interpolation method, the device including a processor, a memory, and a bus, the memory storing instructions and data that can be read by the processor; the processor is used to call the instructions and data in the memory; the bus connects the various functional components to transmit information.

[0031] By adopting the above technical solution, the present invention has the following beneficial effects:

[0032] This invention provides a method, system, and device for denoising and interpolating three-dimensional laser point cloud data of tunnels. It can effectively remove noise clusters and discrete noise points, improve the accuracy and precision of over-excavation and under-excavation detection of mountain tunnel cross-sections, and avoid misjudgment of under-excavation, encroachment locations, and number of locations. It can also automatically complete the point cloud data at missing locations, further improving the accuracy of over-excavation and under-excavation detection. Attached Figure Description

[0033] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0034] Figure 1 A flowchart of a tunnel three-dimensional laser point cloud data denoising interpolation method provided in an embodiment of the present invention;

[0035] Figure 2 This is a schematic diagram of the initial point cloud data of a two-dimensional cross-section provided in an embodiment of the present invention;

[0036] Figure 3 for Figure 1 Flowchart of noise reduction using the median difference method;

[0037] Figure 4 This is a schematic diagram of the difference method denoising provided in an embodiment of the present invention;

[0038] Figure 5 for Figure 1 Flowchart of the median-means denoising method;

[0039] Figure 6 This is a schematic diagram of the mean-based noise reduction method provided in an embodiment of the present invention;

[0040] Figure 7 for Figure 1 Interpolation process flowchart;

[0041] Figure 8 This is a schematic diagram of the interpolation method provided in an embodiment of the present invention;

[0042] Figure 9 A schematic diagram of angle integration provided for an embodiment of the present invention;

[0043] Figure 10 This is the original point cloud data map;

[0044] Figure 11 This is a point cloud data image processed using the tunnel laser point cloud denoising interpolation method provided in this embodiment of the invention;

[0045] Figure 12 This is a result image generated using a conventional cross-section over-excavation and under-excavation detection method;

[0046] Figure 13 The image shows the result generated using the tunnel 3D laser point cloud data denoising and interpolation method provided in this embodiment of the invention.

[0047] Figure 14 A diagram of a tunnel 3D laser point cloud data denoising interpolation system provided in an embodiment of the present invention. Detailed Implementation

[0048] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] The terms "first," "second," and "third," etc., in the specification, embodiments, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, such as including a series of steps or modules. A method, system, product, or apparatus is not necessarily limited to those steps or modules explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or apparatuses. "And / or" is used to indicate the selection of one or both of the two objects to which it is connected.

[0050] The present invention will be further explained below with reference to specific embodiments.

[0051] like Figure 1 As shown in the figure, this embodiment provides a method for denoising and interpolating three-dimensional laser point cloud data in tunnels, which includes the following steps:

[0052] Step 1: Obtain initial point cloud data. The point cloud data comes from the tunnel excavation, initial support, and secondary lining sections. After dividing the overall point cloud by mileage, extract the section point cloud data, and then convert the coordinate system into two-dimensional plane coordinates of the point cloud data to obtain the initial two-dimensional section point cloud data, such as... Figure 2 As shown in the figure, the dashed line is the inner contour line of the tunnel design, the scattered points are the point cloud data obtained by scanning, and a Cartesian coordinate system is established with the center O1 of the arch arc as the origin.

[0053] Step 2: Calculate the azimuth angle α of each point in the point cloud data. i and over- or under-dig value D i The formula is as follows:

[0054] α i =arctan((y i -y k ) / (x i -x k ));

[0055]

[0056] Where, α i The azimuth angle is specifically the angle between the line connecting the i-th data point to the center of the arc containing that data point and the positive x-axis, based on α. i Different quadrants in the coordinate system require conversion and adjustment to adjust α. i The value range is set to [-90°, 270°), which ensures that the start and end positions of the data points are at the center point of the cross-section invert arch. The center point of the cross-section invert arch is located at... Figure 2 The lowest point of the dashed line of the interrupted surface contour is used to facilitate the calculation of subsequent azimuth difference values. The value range of i is i = 1, 2, 3, ..., n, where n represents the number of data points, x i and y i Represents the two-dimensional coordinates of the data points, x k y k and R k This represents the x and y coordinates and radius of the arc corresponding to the center of the circle containing the data point;

[0057] Step 3: Based on the azimuth angle and over- or under-mining values, denoise the point cloud data using the difference method and the mean method:

[0058] The difference-based denoising method includes the following steps: Figure 3 As shown:

[0059] Step 311: Sort all data points in ascending order according to the azimuth angle;

[0060] Step 312: Calculate the difference β between the azimuth angles of adjacent data points. i and the first preset threshold δ1:

[0061] β i =α i+1 -α i ;

[0062] The first preset threshold δ1 is usually taken as 1 to 3°. Generally speaking, the larger the value of the first preset threshold δ1, the more noise will be removed in the subsequent process, which may also cause misjudgment and remove some non-noise data points. The smaller the value of the first preset threshold δ1, the fewer noise will be removed in the subsequent process, which can preserve the original data to the maximum extent, but the denoising effect will be poor.

[0063] Step 313: Calculate the difference d between the over-mining and under-mining values ​​of the i-th and (i+1)-th data points. i ;

[0064] d i =D i+1 -D i ;

[0065] Select d i The maximum value in is denoted as d. max Select d i The minimum value in is denoted as d. min The second preset threshold δ2 is calculated using the following formula;

[0066] δ2=(d max -d min ) / k1;

[0067] Where k1 represents the noise reduction coefficient, which usually ranges from 3 to 5. Generally speaking, the larger the value of k1, the more noise will be removed, but it may also cause misjudgment and remove some non-noise data points. The smaller the value of k1, the fewer noise will be removed, which can preserve the original data to the maximum extent, but the noise reduction effect will be worse.

[0068] Step 314: Judge the i-th data point: when |β i |<δ1 and|d i When |>δ2, perform group denoising operation, the specific steps are as follows:

[0069] Step 3141: Based on the azimuth angle α i And the first preset threshold δ1 is used to find the neighborhood of the i-th data point, the neighborhood range being (α) i –0.5δ1, α i+0.5δ1), calculate the average value of the over-mining and under-mining values ​​of the data points in the neighborhood, denoted as d. mean1 ;

[0070] Step 3142: Filter the neighborhood containing values ​​greater than d. mean1 The data points form a set A1, and the set contains n1 data points; filter the neighborhood where the number of data points is less than d. mean1 The data points form a set A2, and the set contains n2 data points;

[0071] Step 3143: If n1 > n2, it means that the set A2 contains fewer data points. The set A2 is a noise set. Noise judgment is performed as follows: If the i-th data point is in A2, the i-th data point is regarded as noise and removed, and the (i+1)-th data point is retained; if the i-th data point is in A1, the (i+1)-th data point is regarded as noise and removed, and the i-th data point is retained.

[0072] Step 3144: If n1≤n2, it means that the set A1 contains a small number of data points. The set A1 is a noise set. Noise judgment is performed as follows: if the i-th data point is in A1, the i-th data point is regarded as noise and removed, and the (i+1)-th data point is retained; if the i-th data point is in A2, the (i+1)-th data point is regarded as noise and removed, and the i-th data point is retained.

[0073] Step 315: Repeat step 314 for all data points i = 1, 2, 3, ..., n, until all data points have been judged;

[0074] The above-described interpolation method for denoising can effectively remove noise clusters from the original point cloud data, such as... Figure 4 As shown in the Chinese box;

[0075] The mean-based denoising method includes the following steps: Figure 5 As shown:

[0076] Step 321: Sort the point cloud data after denoising by the interpolation method in ascending order of azimuth angle;

[0077] Step 322: Based on the azimuth angle α i And the first preset threshold δ1 is used to find the neighborhood of the i-th data point, the neighborhood range being (α) i –2δ1, α i +2δ1), calculate the average value of the over-mining and under-mining values ​​of the data points in the neighborhood, denoted as d. mean2 ;

[0078] Step 323: Based on the over-mining and under-mining value D i The average value d of the over-mining and under-mining values ​​of the data points in the neighborhoodmean2 The difference is determined by comparing it with a third preset threshold, which can be 0.5 times the second preset threshold. If |D i -d mean2 If |>0.5δ2, then the current data point is considered to be discrete noise and is removed;

[0079] Step 324: Repeat steps 322 to 323 for all data points i = 1, 2, 3, ..., n, until all data points have been judged.

[0080] The above-described mean-based denoising method can effectively remove discrete noise points from point cloud data, such as... Figure 6 As shown in the Chinese box.

[0081] Step 4: Perform interpolation on the denoised point cloud data, and calculate the point cloud data at the missing location based on the weighted average of neighboring point cloud data, such as... Figure 7 As shown, the specific steps are as follows:

[0082] Step 41: Sort the denoised point cloud data in ascending order of azimuth angle;

[0083] Step 42: Calculate the azimuth difference β between the i-th and (i+1)-th data points. i And determine the fourth preset threshold δ3 of the azimuth difference in the interpolation algorithm. The value range of the fourth preset threshold δ3 is 0.5 to 1°. Generally speaking, the smaller the value of δ3, the more interpolation points will be inserted, and the smoother the connection of the point cloud data will be.

[0084] Step 43: Judge the i-th data point. If |β i If |>δ3, then calculate α i and α i+1 The azimuth angle α of the j-th interpolation point j The over- or under-excavation value d of the interpolation point is calculated based on the Kriging interpolation method. j The formula is as follows:

[0085] α j =α i+j ×δ3 / k3;

[0086]

[0087] Where k3 represents the interpolation coefficient, with a value ranging from 50 to 100. The smaller the value of k3, the more interpolation points are inserted, and the smoother the connection of the point cloud data will be; α i and D i As mentioned earlier, α represents the azimuth and over- or under-drilling value of the i-th data point. i+1 and D i+1This represents the azimuth and over- or under-dig value of the (i+1)th data point.

[0088] Step 44, α i and α i+1 All interpolation points calculated within the range are added to the denoised point cloud data;

[0089] Step 45: For all data points in the range i = 1, 2, 3, ..., n-1, repeat steps 43 to 44 until all data points have been judged and interpolation has been performed.

[0090] The above interpolation method can be used to supplement denoised point cloud data or point cloud data with missing points, making the point cloud data graph smooth and complete. Figure 8 As shown in the middle ellipse.

[0091] Step 5: Calculate the over-excavation and under-excavation areas using angle integration. This avoids the problem that the y-value may not be unique when integrating x in conventional area calculations. Figure 9 As shown, the specific steps are as follows:

[0092] Step 51: Sort the interpolated point cloud data in ascending order of azimuth angle;

[0093] Step 52: Calculate the difference β between the azimuth angles of the i-th and (i+1)-th data points. i :

[0094] β i =α i+1 -α i ;

[0095] Step 53: Calculate the micro-element s of the over-dig and under-dig area corresponding to the sector formed by the i-th and (i+1)-th data points and the center of the circle. i :

[0096]

[0097] Step 54: Repeat step 53 for data points i = 1, 2, 3, ..., n-1 until all over-excavation and under-excavation area micro-elements have been calculated;

[0098] Step 55: Filter the s calculated in steps 53-54 i Find the values ​​greater than or equal to 0 in the range, sum the values, and obtain the over-excavated area, denoted as S1;

[0099] Step 56: Filter the s calculated in steps 53-54 i Find the values ​​less than 0 in the range, sum them up, and obtain the under-excavated area, denoted as S2;

[0100] Step 57: Sum the over-excavated area and the under-excavated area to obtain the over-excavated and under-excavated area S.

[0101] Step 6: Based on the above results, output refined cross-sectional over-excavation and under-excavation detection parameters, which include information such as over-excavation area, under-excavation area, maximum over-excavation value, maximum under-excavation value, and number of under-excavation encroachments.

[0102] The above methods can effectively remove noise and improve the accuracy of cross-section over-excavation and under-excavation analysis; at the same time, the point cloud data at the missing locations can be automatically completed, further improving the accuracy of over-excavation and under-excavation analysis.

[0103] Specifically, based on the original point cloud data in real-world conditions, the map exhibits a wavy, disjointed, and incomplete shape due to noise clusters and discrete noise points generated by steel mesh, water pipes, ventilation ducts, equipment chambers, etc. within the tunnel, as well as data loss caused by factors such as occlusion, scanning, or construction. Examples of these features are shown below. Figure 10 As shown in a, b, and c; through the above-mentioned refined denoising and interpolation processing, noise clusters and discrete noise points generated by factors such as steel mesh, water pipes, air ducts, and equipment chambers were effectively removed, and interpolated data points were scientifically supplemented, making the original point cloud data map complete and smooth, as shown in the figures below. Figure 11 As shown in a, b, and c, after... Figure 10 and Figure 11 A comparison of the three example images shows that the method of this invention reduces the impact of noise on cross-section over-excavation and under-excavation analysis, achieves accurate analysis of cross-section over-excavation and under-excavation, and makes the measured point cloud data map line smoother and more in line with engineering reality.

[0104] like Figure 12 As shown in Figures a and b, the conventional analysis method is used to detect the over-excavation and under-excavation parameters of two types of raw point cloud data:

[0105] Figure 12 Figure a shows the over-excavated area: 8.7269 square meters; under-excavated area: -0.0123 square meters; maximum over-excavation value: 0.8875 meters; maximum under-excavation value: -0.1496 meters; number of under-excavation sites: 7.

[0106] Figure 12 Figure b shows the over-excavated area: 5.312 square meters; under-excavated area: -0.0359 square meters; maximum over-excavation value: 0.4575 meters; maximum under-excavation value: -0.1496 meters; number of under-excavation sites: 3.

[0107] The results of refined over-mining and under-mining analysis of the two types of original point cloud data using the method of the embodiments of the present invention are as follows: Figure 13 As shown in a and b:

[0108] Figure 13 Figure a shows the over-excavated area: 8.991 square meters; under-excavated area: none; maximum over-excavation value: 0.8875 meters; maximum under-excavation value: none; number of under-excavation locations: none.

[0109] Figure 13 Figure b shows the over-excavated area: 5.4382 square meters; under-excavated area: -0.0003 square meters; maximum over-excavation value: 0.4575 meters; maximum under-excavation value: -0.0082 meters; number of under-excavation sites: 1.

[0110] After Figure 12 and Figure 13 A comparison of the two example diagrams reveals that, due to the influence of noise, conventional over-excavation and under-excavation calculations show under-excavation phenomena in both cross-sections. However, in actual engineering projects, there is no under-excavation, leading to misjudgments of the under-excavation location and number of under-excavation sites. At the same time, there are also discrepancies in the calculation results of the over-excavation and under-excavation areas. The results obtained by the method of this invention are more accurate and more in line with the actual engineering situation.

[0111] In particular, the method of this invention has been applied to multiple mountain tunnels under construction. After testing and statistical analysis of a scanning section of about 300 meters and nearly 800 cross-sections, the results show that the method of this invention can achieve an efficiency of more than 98% in removing cross-section noise. It can effectively remove noise clusters and discrete noise points, while preserving the original point cloud to the maximum extent, and truly and comprehensively reflecting the over-excavation and under-excavation of the tunnel cross-section.

[0112] In summary, the embodiments of the present invention can accurately perform cross-sectional over-excavation and under-excavation analysis, reduce the adverse effects of noise on the calculation of over-excavation and under-excavation areas, avoid misjudgment of under-excavation locations and the number of under-excavation points, and truly reflect the cross-sectional over-excavation and under-excavation situation, thereby achieving refined calculation of cross-sectional over-excavation and under-excavation detection.

[0113] On the other hand, embodiments of the present invention also provide a tunnel three-dimensional laser point cloud data denoising interpolation system, such as... Figure 14 As shown, it includes a data receiving module, a data processing module, and a result generation module:

[0114] The data receiving module is used to receive initial point cloud data;

[0115] The data processing module includes a preprocessing unit, a noise reduction unit, an interpolation unit, and an over- or under-dig calculation unit.

[0116] The preprocessing unit, based on the initial point cloud data, establishes a rectangular coordinate system with the center of the arch arc as the origin, calculates the azimuth angle and over- or under-dig values ​​of each point, and then sends them to the denoising unit.

[0117] The denoising unit removes noise clusters using the difference method based on the azimuth difference and the over-drilling / under-drilling difference between two adjacent points; it removes discrete noise points using the mean method based on the mean of the over-drilling / under-drilling values ​​within the neighborhood of each point; and it sends the noise-removed point cloud data to the interpolation unit.

[0118] The interpolation unit calculates the azimuth difference between adjacent points based on the point cloud data after noise removal, and then solves the azimuth and over- or under-dig values ​​of the interpolation points. After adding the interpolation points to the point cloud data, it sends them to the over- or under-dig calculation unit.

[0119] The over-excavation and under-excavation calculation unit calculates the over-excavation and under-excavation values ​​and the number of under-excavation encroachments based on the interpolated point cloud data; it calculates the over-excavation and under-excavation area using the angle integration method and sends the calculation results to the result generation module.

[0120] The result generation module is used to output the denoised and interpolated point cloud data and cross-sectional over-excavation and under-excavation detection parameters. These parameters include over-excavation area, under-excavation area, maximum over-excavation value, maximum under-excavation value, and the number of under-excavation encroachments. Furthermore, this invention also provides a tunnel three-dimensional laser point cloud data denoising and interpolation device for the aforementioned tunnel three-dimensional laser point cloud data denoising and interpolation method. The device includes a processor, a memory, and a bus.

[0121] The memory stores instructions and data that can be read by the processor;

[0122] The processor is used to retrieve instructions and data from the memory, and the specific process is as follows:

[0123] Step 1: Obtain initial point cloud data. The point cloud data comes from the tunnel excavation, initial support and secondary lining sections. After the overall point cloud is cut by mileage, the section point cloud data is extracted and then the coordinate system is transformed to obtain the two-dimensional section initial point cloud data.

[0124] Step 2: Calculate the azimuth angle α of each point in the point cloud data. i and over- or under-dig value D i The formula is as follows:

[0125] α i =arctan((y i -y k ) / (x i -x k ));

[0126]

[0127] Where, α i The azimuth angle is specifically the angle between the line connecting the i-th data point to the center of the arc containing that data point and the positive x-axis, based on α.i Different quadrants in the coordinate system require conversion and adjustment to adjust α. i The value range of is set to [-90°, 270°), which ensures that the starting and ending positions of the data points are at the center point of the cross-section invert arch, facilitating the calculation of subsequent azimuth difference values. The value range of i is i = 1, 2, 3, ..., n, where n represents the number of data points, x i and y i Represents the two-dimensional coordinates of the data points, x k y k and R k This represents the x and y coordinates and radius of the arc corresponding to the center of the circle containing the data point;

[0128] Step 3: Based on the azimuth angle and over- or under-mining values, denoise the point cloud data using the difference method and the mean method:

[0129] The difference-based denoising method includes the following steps:

[0130] Step 311: Sort all data points in ascending order according to the azimuth angle;

[0131] Step 312: Calculate the difference β between the azimuth angles of adjacent data points. i and the first preset threshold δ1:

[0132] β i =α i+1 -α i ;

[0133] The first preset threshold δ1 is typically taken as 1 to 3°;

[0134] Step 313: Calculate the difference d between the over-mining and under-mining values ​​of the i-th and (i+1)-th data points. i ;

[0135] d i =D i+1 -D i ;

[0136] Select d i The maximum value in is denoted as d. max Select d i The minimum value in is denoted as d. min The second preset threshold δ2 is calculated using the following formula;

[0137] δ2=(d max -d min ) / k1;

[0138] Wherein, k1 represents the noise reduction coefficient, which usually takes a value of 3 to 5;

[0139] Step 314: Judge the i-th data point: when |β i |<δ1 and|d i When |>δ2, a data point grouping denoising operation is performed. The specific method of grouping denoising includes determining the neighborhood of each point based on the first preset threshold δ1, calculating the average over-mining and under-mining value of the points in the neighborhood, dividing the neighborhood into two groups based on the comparison between each point in the neighborhood and the average value, and distinguishing noise points and normal points according to the number of points in each group. Points in a minority group are noise points, and points in a majority group are normal points.

[0140] Step 315: Repeat step 314 for all data points i = 1, 2, 3, ..., n, until all data points have been judged.

[0141] The mean-based noise reduction method includes the following steps:

[0142] Step 321: Sort the point cloud data after denoising by the interpolation method in ascending order of azimuth angle;

[0143] Step 322: Based on the azimuth angle α i And the first preset threshold δ1 is used to find the neighborhood of the i-th data point, the neighborhood range being (α) i –2δ1, α i +2δ1), calculate the average value of the over-mining and under-mining values ​​of the data points in the neighborhood, denoted as d. mean2 ;

[0144] Step 323: Based on the over-mining and under-mining value D i The average value d of the over-mining and under-mining values ​​of the data points in the neighborhood mean2 The difference is determined by comparing it with a third preset threshold, which can be 0.5 times the second preset threshold. If |D i -d mean2 If |>0.5δ2, then the current data point is considered to be discrete noise and is removed;

[0145] Step 324: Repeat steps 322 to 323 for all data points i = 1, 2, 3, ..., n, until all data points have been judged.

[0146] Step 4: Perform interpolation on the denoised point cloud data, and calculate the point cloud data at the missing location based on the weighted average of neighboring point cloud data. The specific steps are as follows:

[0147] Step 41: Sort the denoised point cloud data in ascending order of azimuth angle;

[0148] Step 42: Calculate the azimuth difference β between the i-th and (i+1)-th data points. iAnd determine the fourth preset threshold δ3 of the azimuth difference in the interpolation algorithm, wherein the value of the fourth preset threshold δ3 is in the range of 0.5 to 1°;

[0149] Step 43: Judge the i-th data point. If |β i If |>δ3, then calculate α i and α i+1 The azimuth angle α of the j-th interpolation point j The over- or under-mining value d of the interpolation point is calculated based on interpolation algorithms known in the art, such as Kriging interpolation. j The formula is as follows:

[0150] α j =α i+j ×δ3 / k3;

[0151]

[0152] Where k3 represents the interpolation coefficient, with a value ranging from 50 to 100. The smaller the value of k3, the more interpolation points are inserted, and the smoother the connection of the point cloud data will be; α i and D i As mentioned earlier, α represents the azimuth and over- or under-drilling value of the i-th data point. i+1 and D i+1 This represents the azimuth and over- or under-dig value of the (i+1)th data point.

[0153] Step 44, α i and α i+1 All interpolation points calculated within the range are added to the denoised point cloud data;

[0154] Step 45: For all data points in the range i = 1, 2, 3, ..., n-1, repeat steps 43 to 44 until all data points have been judged and interpolation has been performed.

[0155] Step 5: Calculate the over-excavation and under-excavation areas using angle integration. This avoids the problem that the y-value may not be unique when integrating x in conventional area calculations. The specific steps are as follows:

[0156] Step 51: Sort the interpolated point cloud data in ascending order of azimuth angle;

[0157] Step 52: Calculate the difference β between the azimuth angles of the i-th and (i+1)-th data points. i :

[0158] β i =α i+1 -α i ;

[0159] Step 53: Calculate the micro-element s of the over-dig and under-dig area corresponding to the sector formed by the i-th and (i+1)-th data points and the center of the circle. i :

[0160]

[0161] Step 54: Repeat step 53 for data points i = 1, 2, 3, ..., n-1 until all over-excavation and under-excavation area micro-elements have been calculated;

[0162] Step 55: Filter the s calculated in steps 53-54 i Find the values ​​greater than or equal to 0 in the range, sum the values, and obtain the over-excavated area, denoted as S1;

[0163] Step 56: Filter the s calculated in steps 53-54 i Find the values ​​less than 0 in the range, sum them up, and obtain the under-excavated area, denoted as S2;

[0164] Step 57: Sum the over-excavated area and the under-excavated area to obtain the over-excavated and under-excavated area S.

[0165] Step 6: Based on the above results, output refined cross-sectional over-excavation and under-excavation detection parameters, which include information such as over-excavation area, under-excavation area, maximum over-excavation value, maximum under-excavation value, and number of under-excavation encroachments.

[0166] The bus is used to connect various functional components and transmit information.

[0167] In another implementation, this solution can be implemented using a device, which may include corresponding modules that perform one or more steps in the various embodiments described above. A module may be one or more hardware modules specifically configured to perform the corresponding step, or implemented by a processor configured to perform the corresponding step, or stored in a computer-readable medium for implementation by a processor, or implemented through some combination thereof.

[0168] The processor executes the various methods and processes described above. For example, the method implementations in this scheme can be implemented as software programs tangibly contained in a machine-readable medium, such as memory. In some implementations, part or all of the software program can be loaded and / or installed via memory and / or a communication interface. When the software program is loaded into memory and executed by the processor, one or more steps of the methods described above can be performed. Alternatively, in other implementations, the processor can be configured to execute one of the methods described above by any other suitable means (e.g., by means of firmware).

[0169] This device can be implemented using a bus architecture. A bus architecture can include any number of interconnect buses and bridges, depending on the specific application of the hardware and overall design constraints. The bus connects various circuits, including one or more processors, memory, and / or hardware modules. The bus can also connect various other circuits such as peripherals, voltage regulators, power management circuitry, external antennas, etc.

[0170] Buses can be Industry Standard Architecture (ISA) buses, Peripheral Component Interconnect (PCI) buses, or Extended Industry Standard Component (EISA) buses, etc. Buses can be divided into address buses, data buses, control buses, etc.

[0171] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for denoising and interpolating three-dimensional laser point cloud data of tunnels, characterized in that, This includes acquiring initial point cloud data, establishing a rectangular coordinate system with the center of the arch arc as the origin, calculating the azimuth angle and over- or under-excavation value of each point in the point cloud data, and calculating the cross-sectional over- or under-excavation detection parameters after denoising and interpolation processing; The calculated cross-sectional over-excavation and under-excavation detection parameters include the calculation of the over-excavation and under-excavation area using the angle integration method, specifically including: Step 51: Sort the interpolated point cloud data in ascending order of azimuth angle; Step 52, calculate the first... Azimuth of data points and the Azimuth of data points The difference : ； Step 53, calculate the first... and the The micro-element of the over- and under-excavation area corresponding to the sector formed by the data points and the center of the circle. : ； in, Indicates the first Over-mining and under-mining values ​​for each data point; This represents the radius of the circle whose center is located on the arc containing the data point; Step 54: Data points Repeat step 53 until all over- and under-excavation area micro-elements have been calculated; Step 55: Filter the results calculated in steps 53-54. The values ​​greater than or equal to 0 are summed to obtain the over-excavated area, denoted as . ; Step 56: Filter the results calculated in steps 53-54. The values ​​less than 0 are summed to obtain the under-excavated area, denoted as . ; Step 57: Sum the over-excavated area and the under-excavated area to obtain the over-excavated and under-excavated areas. .

2. The method for denoising and interpolating three-dimensional laser point cloud data in tunnels according to claim 1, characterized in that, The denoising includes difference denoising, which includes: sorting the points according to the size of the azimuth angle, calculating the azimuth angle difference and over-dig / under-dig value difference between adjacent points, and performing group denoising when the azimuth angle difference is less than a first preset threshold and the over-dig / under-dig value difference is greater than a second preset threshold.

3. The method for denoising and interpolating three-dimensional laser point cloud data in tunnels according to claim 2, characterized in that, The specific method for grouped denoising includes determining the neighborhood of each point based on the first preset threshold, calculating the average over-drilling and under-drilling value of points in the neighborhood, dividing the neighborhood into two groups based on the comparison between each point in the neighborhood and the average value, and distinguishing between noise points and normal points according to the number of points in each group. Points in a minority group are noise points, and points in a majority group are normal points.

4. The method for denoising and interpolating three-dimensional laser point cloud data in tunnels according to claim 3, characterized in that, The method for determining the neighborhood includes: determining the neighborhood range of a point based on a certain point by a ratio multiple of the first preset threshold.

5. The method for denoising and interpolating three-dimensional laser point cloud data in tunnels according to claim 2, characterized in that, The method for determining the second preset threshold includes: calculating the maximum and minimum values ​​of the difference between the over-dig and under-dig values ​​respectively, and using the ratio of the difference between the maximum and the minimum values ​​as the second preset threshold.

6. The method for denoising and interpolating three-dimensional laser point cloud data in tunnels according to claim 2, characterized in that, The denoising also includes mean denoising, which is set after the difference denoising. The mean denoising includes: after the difference denoising, sorting the remaining points according to the size of the azimuth angle, determining the neighborhood of each point based on the proportion multiple of the first preset threshold, calculating the average over-drilling and under-drilling value of the points in the neighborhood, calculating the difference between each point and the average value, and if the difference is greater than the third preset threshold, then the current point is determined to be a noise point.

7. The method for denoising and interpolating three-dimensional laser point cloud data in tunnels according to claim 1, characterized in that, The interpolation process includes: sorting the denoised points according to their azimuth angles, calculating the azimuth angle difference between adjacent points, and when the azimuth angle difference is greater than a fourth preset threshold, solving for the azimuth angle and over- or under-dig value of the interpolation point between the adjacent points, and adding the interpolation point to the point cloud data.

8. A tunnel three-dimensional laser point cloud data denoising interpolation system employing the tunnel three-dimensional laser point cloud data denoising interpolation method as described in any one of claims 1-7, characterized in that, It includes a data receiving module, a data processing module, and a result generation module: The data receiving module is used to receive initial point cloud data; The data processing module includes a preprocessing unit, a noise reduction unit, an interpolation unit, and an over- or under-dig calculation unit. The preprocessing unit, based on the initial point cloud data, establishes a rectangular coordinate system with the center of the arch arc as the origin, calculates the azimuth angle and over- or under-dig values ​​of each point, and then sends them to the denoising unit. The noise reduction unit removes noise clusters based on the azimuth difference and over- or under-dig value difference between two adjacent points using the difference method. Based on the average over- and under-mining values ​​within the neighborhood of each point, discrete noise points are removed using the mean method. The noise-removed point cloud data is sent to the interpolation unit; The interpolation unit calculates the azimuth difference between adjacent points based on the point cloud data after noise removal, and then solves the azimuth and over- or under-dig values ​​of the interpolation points. After adding the interpolation points to the point cloud data, it sends them to the over- or under-dig calculation unit. The over-excavation and under-excavation calculation unit calculates the over-excavation and under-excavation values ​​and the number of under-excavation encroachments based on the interpolated point cloud data; it calculates the over-excavation and under-excavation area using the angle integration method and sends the calculation results to the result generation module. The result generation module is used to output the point cloud data after noise reduction and interpolation processing, as well as the cross-section over-excavation and under-excavation detection parameters.

9. A tunnel three-dimensional laser point cloud data denoising interpolation device for the tunnel three-dimensional laser point cloud data denoising interpolation method according to any one of claims 1-7, characterized in that, The device includes a processor, a memory, and a bus. The memory stores instructions and data that can be read by the processor. The processor is used to call the instructions and data in the memory. The bus connects the various functional components to transmit information.

Images