Full-width structure depth detection method based on precise three dimensions

Abstract

The present application relates to the technical field of road surface detection, and provides a full-width structure depth detection method based on precise three dimensions, which comprises the following steps: receiving original road elevation data and original road grayscale data of each measuring point on the road acquired by a line scanning three-dimensional measurement sensor, wherein the measuring range of the line scanning three-dimensional measurement sensor covers the entire lane width; extracting target road elevation data and target road grayscale data within the lane range from the original road elevation data and the original road grayscale data; reconstructing road three-dimensional point cloud data based on the target road elevation data and the target road grayscale data; and determining the full-width structure depth of the road based on the reconstructed road three-dimensional point cloud data and in combination with a preset road structure depth calculation model. Since the measuring range of the line scanning three-dimensional measurement sensor covers the entire lane width and the data interference of the lane lines and their edges is removed, the final measured structure depth can comprehensively and accurately reflect the full-width structure depth condition of the road.

Description

Technical Field

[0001] The present invention relates to the technical field of road surface detection, and in particular to a full-width structural depth detection method based on precise three-dimensional (3D) technology. Background Art

[0002] The road's structural depth is a key indicator for evaluating road safety. The main existing method for assessing structural depth involves first obtaining a longitudinal profile curve at the center of the wheel track using a structural depth meter, and then inferring the road's structural depth using a theoretical model. Currently used structural depth meters typically consist of an accelerometer and a point laser ranging sensor. The point laser ranging sensor measures the distance between the vehicle and the road surface, while the accelerometer calculates the distance traveled by the vehicle's up-and-down vibrations through two integrals, theoretically determining the longitudinal elevation change of the road surface, i.e., the longitudinal profile curve.

[0003] At present, structural depth measurement only detects the structural depth of the wheel track corresponding to the longitudinal profile during driving. This measurement method cannot reflect the full-width structural depth of the road lane. In addition, the detection trajectory of the detection equipment during operation is affected by the driver's driving trajectory. Especially for sections with poor road conditions, in order to ensure their own driving comfort and safety, the driver will deliberately avoid the wheel track position with poor road conditions during driving, which will lead to inaccurate measurement results. Summary of the Invention

[0004] The present invention provides a full-width structural depth detection method based on precise three-dimensional (3D) technology, which is used to solve the problem that the structural depth measurement technology in the prior art cannot reflect the full-width structural depth of the road lane and the measurement is inaccurate.

[0005] The present invention provides a method for full-width structural depth detection based on precise three-dimensional imaging, comprising:

[0006] receiving original road elevation data and original road grayscale data of each measuring point on the road acquired by a line scanning 3D measurement sensor, wherein the measurement range of the line scanning 3D measurement sensor covers the entire lane width;

[0007] Extracting target road surface elevation data and target road surface grayscale data within the lane range from the original road surface elevation data and the original road surface grayscale data;

[0008] Reconstructing road surface three-dimensional point cloud data based on the target road surface elevation data and the target road surface grayscale data;

[0009] Based on the reconstructed three-dimensional point cloud data of the road surface and combined with the preset road surface structural depth calculation model, the full-width structural depth of the road surface is determined.

[0010] The application provides a full-width structure depth detection method based on precise three dimensions, which extracts target road surface elevation data and target road surface grayscale data in a lane range from original road surface elevation data and original road surface grayscale data, and comprises the following steps of:

[0011] Based on the original road surface elevation data, a potential lane line first area is marked by using elevation characteristics and geometric size characteristics of a road surface lane line.

[0012] Based on the original road surface grayscale data, a potential lane line second area is marked by using reflection characteristics and geometric size characteristics of a road surface lane line.

[0013] The lane line position of a current road surface is determined by combining the potential lane line first area and the potential lane line second area.

[0014] Based on the lane line position of the current road surface, the target road surface elevation data and the target road surface grayscale data in the lane range are extracted.

[0015] The application provides a full-width structure depth detection method based on precise three dimensions, which reconstructs road surface three-dimensional point cloud data based on the target road surface elevation data and the target road surface grayscale data, and comprises the following steps of:

[0016] Based on the target road surface elevation data and the target road surface grayscale data, an abnormal elevation measuring point in the lane range is determined.

[0017] Based on the target road surface elevation data of non-abnormal elevation measuring points, elevation estimation data of the abnormal elevation measuring point is estimated to generate effective road surface elevation data, wherein the effective road surface elevation data comprises the target road surface elevation data of the non-abnormal elevation measuring points and the elevation estimation data.

[0018] Based on the effective road surface elevation data, the road surface three-dimensional point cloud data is reconstructed.

[0019] The application provides a full-width structure depth detection method based on precise three dimensions, which determines an abnormal elevation measuring point in a lane range based on target road surface elevation data and target road surface grayscale data, and comprises the following steps of:

[0020] Based on the target road surface elevation data, a preliminary abnormal elevation measuring point in the lane range is determined.

[0021] Based on the target road surface grayscale data and the preliminary abnormal elevation measuring point, the abnormal elevation measuring point is determined.

[0022] The application provides a full-width structure depth detection method based on precise three dimensions, which determines a preliminary abnormal elevation measuring point in a lane range based on target road surface elevation data, and comprises the following steps of:

[0023] acquire high-frequency road elevation signals in the target road elevation data;

[0024] for any measuring point, calculate the elevation mean and the elevation variance of the high-frequency road elevation signals of all measuring points within a first preset range around the any measuring point;

[0025] for any measuring point, calculate the first abnormal elevation segmentation threshold and the second abnormal elevation segmentation threshold based on the corresponding elevation mean and the elevation variance, so that the first abnormal elevation segmentation threshold is greater than the second abnormal elevation segmentation threshold;

[0026] for any measuring point, if the corresponding high-frequency road elevation signal is greater than the first abnormal elevation segmentation threshold or less than the second abnormal elevation segmentation threshold, determine the any measuring point as the preliminary abnormal elevation measuring point.

[0027] According to the present application, a full-width structure depth detection method based on precise three-dimensional is provided, based on the target road grayscale data and the preliminary abnormal elevation measuring point, the abnormal elevation measuring point is determined, comprising:

[0028] for any preliminary abnormal elevation measuring point, calculate the grayscale mean and the grayscale variance of the target road grayscale data of all measuring points within a second preset range around the any preliminary abnormal elevation measuring point;

[0029] for any preliminary abnormal elevation measuring point, calculate the first abnormal grayscale segmentation threshold and the second abnormal grayscale segmentation threshold based on the corresponding grayscale mean and the grayscale variance, so that the first abnormal grayscale segmentation threshold is greater than the second abnormal grayscale segmentation threshold;

[0030] for any preliminary abnormal elevation measuring point, if the corresponding target road grayscale data is greater than the first abnormal elevation segmentation threshold or less than the second abnormal elevation segmentation threshold, determine the any preliminary abnormal elevation measuring point as the abnormal elevation measuring point.

[0031] According to the present application, a full-width structure depth detection method based on precise three-dimensional is provided, based on the target road elevation data of the non-abnormal elevation measuring point, the elevation estimation data of the abnormal elevation measuring point is estimated to generate effective road elevation data, comprising:

[0032] for any abnormal elevation measuring point, estimate the elevation estimation data of the any abnormal elevation measuring point based on the target road elevation data of the non-abnormal elevation measuring point within a predetermined area around the any abnormal elevation measuring point;

[0033] determine the elevation estimation data and the target road elevation data of the non-abnormal elevation measuring point as the effective road elevation data.

[0034] According to the application, a full-width construction depth detection method based on precise three-dimensional is provided, which is based on reconstructed road three-dimensional point cloud data, combined with a preset road construction depth calculation model, and determines the full-width construction depth of the road, including:

[0035] The reconstructed road three-dimensional point cloud data is divided into multiple first-level point cloud units along the driving direction;

[0036] Any first-level point cloud unit is divided into multiple second-level point cloud units along the road width direction;

[0037] For all second-level point cloud units in any first-level point cloud unit, the construction depth of all second-level point cloud units is calculated based on the preset road construction depth calculation model, and then the construction depth set of each first-level point cloud unit is obtained;

[0038] The full-width construction depth of the road is determined based on the construction depth set.

[0039] According to the application, a full-width construction depth detection method based on precise three-dimensional is provided, which is based on the construction depth set to determine the full-width construction depth of the road, including:

[0040] Based on the construction depth set, at least one of the following is determined: full-width maximum construction depth, full-width average construction depth, full-width weighted construction depth, left wheel track construction depth, right wheel track construction depth, and the construction depth set;

[0041] The full-width maximum construction depth is the maximum value of the construction depth in the construction depth set of each first-level point cloud unit;

[0042] The full-width average construction depth is the average value of the construction depth in the construction depth set of each first-level point cloud unit;

[0043] The full-width weighted construction depth is the weighted average value of the construction depth in the construction depth set of each first-level point cloud unit;

[0044] The left wheel track construction depth is the construction depth of the second-level point cloud unit corresponding to the left wheel track belt;

[0045] The right wheel track construction depth is the construction depth of the second-level point cloud unit corresponding to the right wheel track belt.

[0046] According to the application, a full-width construction depth detection method based on precise three-dimensional is provided, and the left wheel track construction depth is calculated as follows:

[0047] The position of the left wheel track belt is calculated based on the position of the lane line, the distance between the left and right wheel track belts, and the relationship that the centers of the left and right wheel track belts coincide with the center of the lane;

[0048] The set of construction depths of all the secondary point cloud units located in the left track position is determined as the left track construction depth,

[0049] The right track construction depth is calculated as follows:

[0050] The right track position is calculated based on the lane line position, the distance between the left and right track positions, and the relationship that the centers of the left and right track positions coincide with the center of the lane.

[0051] The set of construction depths of all the secondary point cloud units located in the right track position is determined as the right track construction depth,

[0052] The full-width weighted construction depth is calculated as follows:

[0053] The weight w of the construction depth corresponding to each secondary point cloud unit in the set of construction depths of each primary point cloud unit is calculated according to the following formula: i

[0054]

[0055] DIS i = min(|x i -x L |,|x i -x R |)

[0056] where DIS i is the minimum value of the distance between the position corresponding to the construction depth of the i-th secondary point cloud unit in the road width direction and the left and right track positions, x i is the position of the i-th secondary point cloud unit in the road width direction, x L and x R are the positions of the left and right track positions in the road width direction, respectively, where i = 1, 2, …, n, and n is the number of secondary point cloud units in each primary point cloud unit.

[0057] The full-width weighted construction depth of each primary point cloud unit is calculated according to the following formula:

[0058]

[0059] where SMTD i is the construction depth of the i-th secondary point cloud unit in each primary point cloud unit.

[0060] ​The application provides a full-width construction depth detection method based on precise three dimensions. BRIEF DESCRIPTION OF DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0062] Figure 1 Fig. 1 is one of flow diagrams of the full-width construction depth detection method based on precise three dimensions provided by the application;

[0063] Figure 2 Fig. 2 is another of flow diagrams of the full-width construction depth detection method based on precise three dimensions provided by the application;

[0064] Figure 3 Fig. 3 is a third of flow diagrams of the full-width construction depth detection method based on precise three dimensions provided by the application;

[0065] Figure 4 Fig. 4 is a fourth of flow diagrams of the full-width construction depth detection method based on precise three dimensions provided by the application;

[0066] Figure 5 Fig. 5 is a fifth of flow diagrams of the full-width construction depth detection method based on precise three dimensions provided by the application;

[0067] Figure 6 Fig. 6 is a sixth of flow diagrams of the full-width construction depth detection method based on precise three dimensions provided by the application;

[0068] Figure 7 Figure 7 is a schematic diagram of a seventh flow of a full-width construction depth detection method based on precise three dimensions according to the present application

[0069] Figure 8 Figure 1 is a schematic diagram of a structure of a full-width construction depth detection device based on precise three dimensions according to the present application.

[0070] Figure 9 Figure 6 is a schematic diagram of a structure of an electronic device according to the present application. DETAILED DESCRIPTION

[0071] In order to make the objectives, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0072] The full-width construction depth detection method based on precise three dimensions provided by the embodiments of the present application comprises the following steps. Figure 1 As shown in the figure, the method comprises the following steps.

[0073] Step S110: receiving original road elevation data and original road grayscale data of each measuring point on the road acquired by a line scanning three-dimensional measuring sensor, wherein the measuring range of the line scanning three-dimensional measuring sensor covers the entire lane width. Specifically, the line scanning three-dimensional measuring sensor comprises a laser and a high-speed three-dimensional camera, and the line scanning three-dimensional measuring sensor is installed on a vehicle-mounted platform. During the measuring process, the line scanning three-dimensional measuring sensor continuously collects the elevation information and grayscale information of the road surface, i.e. the original road elevation data and the original road grayscale data, in the road direction. The original road elevation data and the original road grayscale data can be acquired simultaneously by receiving the data returned by the line scanning three-dimensional measuring sensor. In order to reflect the full-width construction depth of the road surface, it is required that the measuring range of the line scanning three-dimensional measuring sensor covers the entire lane in the road width direction, and the collection interval of the original road elevation data and the original road grayscale data in the road width direction is less than or equal to 5mm, and the collection interval in the driving direction is less than or equal to 5mm. One or more sets of line scanning three-dimensional measuring sensors can be arranged on the vehicle according to the road width, so that the total coverage width can reach full-lane coverage.

[0074] Step S120: extracting target road elevation data and target road grayscale data within the lane range from the original road elevation data and the original road grayscale data, i.e. removing the original road elevation data and the original road grayscale data outside the road edge lane line or even the lane line, so that the data used for three-dimensional point cloud reconstruction only contains the data within the lane, thereby enabling the final measurement result to accurately reflect the full-width construction depth of the road surface.

[0075] Step S130: reconstructing road surface three-dimensional point cloud data based on the target road surface elevation data and the target road surface grayscale data.

[0076] Step S140: determining road surface full-width texture depth based on the reconstructed road surface three-dimensional point cloud data and a preset road surface texture depth calculation model.

[0077] In the full-width texture depth detection method based on precise three-dimension of the embodiment, the measurement range of the line scanning three-dimensional measurement sensor covers the entire lane width, and the original road surface elevation data and the original road surface grayscale data are data covering the entire lane width, which will not avoid the track band position with poor road conditions. Moreover, the target road surface elevation data and the target road surface grayscale data within the lane range are used to reconstruct the road surface three-dimensional point cloud data, and the data interference of the lane lines and other non-lane areas of the lane edges is removed. Therefore, the final measured texture depth can fully and accurately reflect the road surface full-width texture depth condition.

[0078] As shown in FIG. 10, in some embodiments, step S120 specifically comprises: Figure 2

[0079] Step S210: based on the original road surface elevation data, using the elevation feature and the geometric size feature of the road surface lane line, marking a potential lane line first area. A lane on a road is usually defined by the lane lines on both sides, and the geometric size (i.e. width and height) of the lane line conforms to the relevant standards of the road, and the elevation of the lane line is higher than that of the normal road surface. Therefore, the lane line first area (i.e. the area of the two lane lines) can be marked by the original road surface elevation data, the elevation feature and the geometric size feature of the lane line.

[0080] Step S220: based on the original road surface grayscale data, using the reflection feature and the geometric size feature of the road surface lane line, marking a potential lane line second area. The lane line is usually white or yellow, which is different from the reflection feature of the gray road surface. Therefore, the lane line second area can be marked by the original road surface grayscale data, the reflection feature and the geometric size feature of the lane line.

[0081] ​Step S230: Determine the lane line position of the current road surface by combining the first potential lane line region and the second potential lane line region. The two different lane line regions are marked by steps S210 and S220 respectively, and the results obtained by different methods can be verified with each other. For example, if the two different lane line regions completely overlap or the overlap rate is as high as 80% or more, it is considered that the lane line regions marked by the two methods are relatively accurate. When determining the lane line position of the current road surface, the maximum region covered by the two lane line regions (i.e., the union of the two lane line regions) can be used as the lane line region, i.e., the position of the lane line. Using the maximum region covered by the two lane line regions as the lane line region can exclude abnormal points on the edge of the lane as much as possible, so that the final measurement result is more accurate.

[0082] Step S240: Extract target road surface elevation data and target road surface grayscale data within the lane range based on the lane line position of the current road surface. Specifically, since the lane line position of the current road surface is determined, i.e., the region of the lane line is determined, the lane range can be determined according to the region between the inner edges of the left lane line region and the right lane line region, so as to extract the target road surface elevation data and the target road surface grayscale data within the lane range.

[0083] In step S130, the target road surface elevation data and the target road surface grayscale data can be directly used to reconstruct the three-dimensional point cloud data of the road surface, but there may be some abnormal data in the target road surface elevation data and the target road surface grayscale data within the lane range. The detection points corresponding to the abnormal data are abnormal measurement points. For road surfaces (e.g., asphalt road surfaces), the road surface aggregates are relatively dense, and the shapes of the aggregates are different. For positions where the profiles between aggregates change sharply and the gaps are deep, or for some road crack regions, some measurement points may be invalid (3D cameras cannot observe the light signals returned by the corresponding measurement points), and the measurement values of these measurement points are obviously abnormal. Such measurement points usually exhibit characteristics such as high frequency and local mutation. If such measurement points are not removed, the accuracy of the measurement result will be significantly affected.

[0084] Therefore, in some embodiments, the data of the abnormal measurement points can be processed in a manner as shown in Figure 3 to form valid data, and then the three-dimensional point cloud data of the road surface is reconstructed by using the valid data. Specifically, step S130 includes:

[0085] Step S310: Determine the abnormal elevation measurement points within the lane range based on the target road surface elevation data and the target road surface grayscale data.

[0086] Step S320: estimating elevation estimation data of the abnormal elevation measuring points based on the target road elevation data of the non-abnormal elevation measuring points, to generate valid road elevation data, which includes the target road elevation data of the non-abnormal elevation measuring points and the elevation estimation data.

[0087] Step S330: reconstructing road three-dimensional point cloud data based on the valid road elevation data.

[0088] As shown in Figure 4 , step S310 specifically includes:

[0089] Step S410: determining preliminary abnormal elevation measuring points within the lane range based on the target road elevation data.

[0090] Step S420: determining the abnormal elevation measuring points based on the target road grayscale data and the preliminary abnormal elevation measuring points.

[0091] In some embodiments, as shown in Figure 5 , step S410 specifically includes:

[0092] Step S510: obtaining high-frequency road elevation signals in the target road elevation data. Wherein, for each measuring point, the target road elevation data collected by the line scanning three-dimensional measuring sensor includes two parts of low-frequency signals and high-frequency signals, in this embodiment, the high-frequency signal part is extracted, and the high-frequency signal can be obtained by filtering (such as high-pass filtering) or frequency domain transformation (such as Fourier transform, wavelet transform, etc.).

[0093] Step S520: for any measuring point, calculating the elevation mean and the elevation variance of the high-frequency road elevation signals of all measuring points within a first preset range around the any measuring point, wherein the measuring points in the first preset range can be the measuring points within 5-20 rows and 5-20 columns around the any measuring point.

[0094] Step S530: for any measuring point, calculating a first abnormal elevation segmentation threshold and a second abnormal elevation segmentation threshold based on the corresponding elevation mean and the elevation variance, so that the first abnormal elevation segmentation threshold is greater than the second abnormal elevation segmentation threshold. Specifically, the calculation methods of the first abnormal elevation segmentation threshold T1 and the second abnormal elevation segmentation threshold T2 are as follows:

[0095] T1 = A h +k1*S h

[0096] T2 = A h -k2*S h

[0097] Wherein, A h is the elevation mean, and S hfor the height variance, k1 and k2 are respectively the first coefficient and the second coefficient, both of which are greater than 0. The values of T1 and T2 can be adjusted by adjusting the values of k1 and k2, so as to determine the screening range of the preliminary abnormal height measuring points.

[0098] Step S540: for any measuring point, if the corresponding high-frequency road surface height signal of the measuring point is greater than the first abnormal height segmentation threshold or less than the second abnormal height segmentation threshold, the measuring point is determined as the preliminary abnormal height measuring point, that is, the corresponding any measuring point outside the interval [T2, T1] is determined as the preliminary abnormal height measuring point.

[0099] In some embodiments, as shown in FIG. 4, step S420 specifically includes: Figure 6

[0100] Step S610: for any preliminary abnormal height measuring point, the gray mean value and the gray variance of the target road surface gray data of all measuring points within a second preset range around the preliminary abnormal height measuring point are calculated. The measuring points in the second preset range can be the measuring points within 5-100 rows and 5-100 columns around the any measuring point.

[0101] Step S620: for any preliminary abnormal height measuring point, the first abnormal gray segmentation threshold and the second abnormal gray segmentation threshold are calculated based on the corresponding gray mean value and the gray variance, so that the first abnormal gray segmentation threshold is greater than the second abnormal gray segmentation threshold. Specifically, the calculation method of the first abnormal gray segmentation threshold T3 and the second abnormal gray segmentation threshold T4 is as follows:

[0102] T3 = k g1 *A g +k3*S g

[0103] T4 = k g2 *A g -k4*S g

[0104] wherein A g is the gray mean value, S g is the gray variance, k3, k4, k g1 and k g2 are respectively the third coefficient, the fourth coefficient, the fifth coefficient and the sixth coefficient, k3, k4, k g1 and k g2 are all greater than 0. The values of T3 and T4 can be adjusted by adjusting the values of k3, k4, k g1 and k g2 , so as to determine the screening range of the abnormal height measuring points.

[0105] ​Step S630: For any preliminary abnormal elevation measuring point, if the corresponding target road surface gray data of the preliminary abnormal elevation measuring point is greater than the first abnormal elevation segmentation threshold or less than the second abnormal elevation segmentation threshold, the preliminary abnormal elevation measuring point is determined as the abnormal elevation measuring point, that is, the corresponding preliminary abnormal elevation measuring point outside the interval [T4, T3] is determined as the abnormal elevation measuring point.

[0106] In some embodiments, step S320 specifically comprises:

[0107] For any abnormal elevation measuring point, the elevation estimation data of the abnormal elevation measuring point is estimated based on the target road surface elevation data of the non-abnormal elevation measuring points in a predetermined area around the abnormal elevation measuring point (for example, in a range of 5-10 rows and 5-10 columns around the measuring point). Specifically, the elevation estimation data of the position corresponding to the abnormal elevation measuring point can be estimated by triangulation or interpolation.

[0108] The elevation estimation data and the target road surface elevation data of the non-abnormal elevation measuring points are determined as the effective road surface elevation data.

[0109] In this embodiment, the target road surface elevation data of the abnormal elevation measuring point is estimated based on the target road surface elevation data of the normal measuring points in a predetermined area around the abnormal elevation measuring point, so that the target road surface elevation data of the abnormal elevation measuring point is regressed to the normal value in the predetermined area, thereby obtaining the effective road surface elevation data. The road surface three-dimensional point cloud data reconstructed by the effective road surface elevation data is more accurate, and the finally measured full-width pavement texture depth is also more accurate.

[0110] In some embodiments, as shown in FIG. 13, step S140 specifically comprises: Figure 7

[0111] Step S710: The reconstructed road surface three-dimensional point cloud data is divided into a plurality of first-level point cloud units along the driving direction, that is, the reconstructed road surface three-dimensional point cloud data is divided into a plurality of first-level point cloud units along the road length direction, and each first-level point cloud unit covers the entire road width. Specifically, the reconstructed road surface three-dimensional point cloud data can be divided into a plurality of first-level point cloud units along the driving direction according to the length requirement of the texture depth calculation (for example, 0.3 meters, 1 meter, 10 meters, 20 meters, 100 meters, or 1000 meters, etc.).

[0112] Step S720: Any first-level point cloud unit is divided into a plurality of second-level point cloud units along the road width direction, that is, each first-level point cloud unit is divided into a plurality of second-level point cloud units along the road width direction.

[0113] ​Step S730: For all the secondary point cloud units in any of the primary point cloud units, calculate the structure depths of all the secondary point cloud units based on the preset road surface structure depth calculation model, and then obtain a structure depth set SMTD of each primary point cloud unit, denoted as: {SMTD1, SMTD2, …, SMTD n} for the structure depth SMTD of the primary point cloud unit. i max SMTDi for the structure depth of the i-th secondary point cloud unit in the primary point cloud unit, i = 1, 2, …, n.

[0114] Step S740: Determine the full-width structure depth of the road surface based on the structure depth set.

[0115] Specifically, this step includes:

[0116] Determine at least one of the full-width maximum structure depth, the full-width average structure depth, the full-width weighted structure depth, the left track structure depth, the right track structure depth, and the structure depth set based on the structure depth set.

[0117] The full-width maximum structure depth is the maximum value of the structure depths in the structure depth set of each primary point cloud unit: SMTD max = max ({SMTD1, SMTD2, …, SMTD n}).

[0118] The full-width average structure depth is the average value of the structure depths in the structure depth set of each primary point cloud unit:

[0119]

[0120] The full-width weighted structure depth is the weighted average value of the structure depths in the structure depth set of each primary point cloud unit.

[0121] The left track structure depth is the structure depth of the secondary point cloud unit corresponding to the left track band.

[0122] The right track structure depth is the structure depth of the secondary point cloud unit corresponding to the right track band.

[0123] Specifically, the left track structure depth is calculated as follows:

[0124] Calculate the position of the left track band based on the position of the lane line, the distance between the left and right track bands, and the relationship that the centers of the left and right track bands coincide with the center of the lane.

[0125] Determine the set of structure depths of all the secondary point cloud units located in the position of the left track band as the left track structure depth: SMTD Left = {SMTD ii=L, L is the ID of the secondary point cloud unit at the left track position.

[0126] Specifically, the right track construction depth is calculated as follows:

[0127] The right track position is calculated based on the lane line position, the distance between the left and right track bands, and the relationship that the centers of the left and right track bands coincide with the center of the lane.

[0128] The set formed by the construction depths of all secondary point cloud units at the right track position is determined as the right track construction depth: SMTD Right = {SMTD i i=R, R is the ID of the secondary point cloud unit at the right track position.

[0129] Specifically, the full-width weighted construction depth is calculated as follows:

[0130] The weight w of the construction depth corresponding to each secondary point cloud unit in the construction depth set of each primary point cloud unit is calculated according to the following formula: i :

[0131]

[0132] DIS i = min(|x i -x L |,|x i -x R |)

[0133] wherein, DIS i is the minimum value of the distance between the position corresponding to the construction depth of the i-th secondary point cloud unit and the left track band and the distance between the position corresponding to the construction depth of the i-th secondary point cloud unit and the right track band in the road width direction, x i is the position of the i-th secondary point cloud unit in the road width direction, x L and x R are the positions of the left track band and the right track band in the road width direction, respectively, wherein i=1, 2, …, n, and n is the number of secondary point cloud units in each primary point cloud unit.

[0134] The full-width weighted construction depth of each primary point cloud unit is calculated according to the following formula:

[0135]

[0136] wherein, SMTD i is the construction depth of the i-th secondary point cloud unit in each primary point cloud unit.

[0137] In actual application, one or more of the above full-width construction depths can be selected according to different application scenarios.

[0138] The following describes a full-width structure depth detection device based on precise three dimensions provided by the present application. The full-width structure depth detection device based on precise three dimensions described below can be referred to in correspondence with the full-width structure depth detection method based on precise three dimensions described above.

[0139] The full-width structure depth detection device based on precise three dimensions provided by the present application, as shown in Figure 8 includes:

[0140] The data receiving module 810 is configured to receive original road elevation data and original road grayscale data of each measurement point on the road acquired by a line scanning three-dimensional measurement sensor, and the measurement range of the line scanning three-dimensional measurement sensor covers the entire lane width.

[0141] The data extraction module 820 is configured to extract target road elevation data and target road grayscale data within the lane range from the original road elevation data and the original road grayscale data.

[0142] The data reconstruction module 830 is configured to reconstruct road three-dimensional point cloud data based on the target road elevation data and the target road grayscale data.

[0143] The structure depth determination module 840 is configured to determine the full-width structure depth of the road based on the reconstructed road three-dimensional point cloud data and in combination with a preset road structure depth calculation model.

[0144] Figure 9 An example of a schematic diagram of the physical structure of an electronic device is shown in Figure 9 The electronic device can include a processor 910, a communications interface 920, a memory 930, and a communications bus 940, wherein the processor 910, the communications interface 920, and the memory 930 can communicate with each other through the communications bus 940. The processor 910 can invoke the logical instructions in the memory 930 to execute a full-width structure depth detection method based on precise three dimensions, which includes:

[0145] The data receiving module 810 is configured to receive original road elevation data and original road grayscale data of each measurement point on the road acquired by a line scanning three-dimensional measurement sensor, and the measurement range of the line scanning three-dimensional measurement sensor covers the entire lane width.

[0146] The data extraction module 820 is configured to extract target road elevation data and target road grayscale data within the lane range from the original road elevation data and the original road grayscale data.

[0147] The data reconstruction module 830 is configured to reconstruct road three-dimensional point cloud data based on the target road elevation data and the target road grayscale data.

[0148] Based on the reconstructed road surface three-dimensional point cloud data, a preset road surface structure depth calculation model is combined to determine the full-width structure depth of the road surface.

[0149] In addition, the logical instructions in the memory 930 described above can be implemented in the form of a software function unit and sold or used as a stand-alone product, and can be stored in a computer-readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0150] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program can be executed by a processor to enable a computer to execute the precision three-dimensional based full-width structure depth detection method provided by the above-mentioned methods, the method comprising:

[0151] Receiving original road surface elevation data and original road surface grayscale data of each measurement point on the road acquired by a line scanning three-dimensional measurement sensor, and the measurement range of the line scanning three-dimensional measurement sensor covers the entire lane width.

[0152] Extracting target road surface elevation data and target road surface grayscale data within the lane range from the original road surface elevation data and the original road surface grayscale data.

[0153] Reconstructing road surface three-dimensional point cloud data based on the target road surface elevation data and the target road surface grayscale data.

[0154] Based on the reconstructed road surface three-dimensional point cloud data, a preset road surface structure depth calculation model is combined to determine the full-width structure depth of the road surface.

[0155] In yet another aspect, the present application also provides a non-transitory computer readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the precision three-dimensional based full-width structure depth detection method provided by the above-mentioned methods, the method comprising:

[0156] The original road surface elevation data and the original road surface grayscale data of each measuring point on the road are received, and a measuring range of the line scanning three-dimensional measuring sensor covers the entire lane width.

[0157] Target road surface elevation data and target road surface grayscale data within the lane range are extracted from the original road surface elevation data and the original road surface grayscale data.

[0158] Based on the target road surface elevation data and the target road surface grayscale data, the road surface three-dimensional point cloud data is reconstructed.

[0159] Based on the reconstructed road surface three-dimensional point cloud data, the full-width road surface texture depth is determined in combination with a preset road surface texture depth calculation model.

[0160] The device embodiments described above are only illustrative, wherein the units illustrated as separate components can or can not be physically separated, and the components illustrated as units can or can not be physical units, i.e., can be located in one place or distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. Those skilled in the art can understand and implement it without creative labor.

[0161] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be realized by means of software and the necessary general hardware platform, and of course, it can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0162] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for full-width structural depth detection based on precise three-dimensional, characterized in that: include: receiving original road elevation data and original road grayscale data of each measuring point on the road acquired by a line scanning 3D measurement sensor, wherein the measurement range of the line scanning 3D measurement sensor covers the entire lane width; Extract target road surface elevation data and target road surface grayscale data within the lane range from the original road surface elevation data and the original road surface grayscale data. The lane range refers to the range after removing the lane lines on both sides of the road and the area outside the lane lines; Reconstructing road surface three-dimensional point cloud data based on the target road surface elevation data and the target road surface grayscale data; Based on the reconstructed 3D point cloud data of the road surface and the preset road surface structural depth calculation model, the full-width structural depth of the road surface is determined; Extracting target road surface elevation data and target road surface grayscale data within the lane range from the original road surface elevation data and the original road surface grayscale data, including: Based on the original road surface elevation data, using elevation features and geometric size features of road surface lane lines, marking a potential first lane line area; Based on the original road surface grayscale data, using the reflective characteristics and geometric size characteristics of the road surface lane line, marking a potential second lane line area; Determining a lane line position on a current road surface by combining the potential first lane line region and the potential second lane line region; Based on the lane line position of the current road surface, the target road surface elevation data and target road surface grayscale data within the lane range are extracted.

2. The method for full-width structural depth detection based on precise three-dimensional analysis according to claim 1, characterized in that: Reconstructing road surface three-dimensional point cloud data based on the target road surface elevation data and the target road surface grayscale data includes: Determining abnormal elevation measurement points within a lane range based on the target road surface elevation data and the target road surface grayscale data; estimating elevation data of abnormal elevation measurement points based on target road surface elevation data of non-abnormal elevation measurement points to generate effective road surface elevation data, the effective road surface elevation data including: target road surface elevation data of non-abnormal elevation measurement points and the elevation estimation data; Reconstruct road surface three-dimensional point cloud data based on the effective road surface elevation data.

3. The method for full-width structural depth detection based on precise three-dimensional analysis according to claim 2, characterized in that: Determining abnormal elevation measurement points within a lane range based on the target road surface elevation data and the target road surface grayscale data includes: Determining preliminary abnormal elevation measurement points within a lane range based on the target road surface elevation data; The abnormal elevation measurement points are determined based on the target road surface grayscale data and the preliminary abnormal elevation measurement points.

4. The method for full-width structural depth detection based on precise three-dimensional analysis according to claim 3, characterized in that: Based on the target road elevation data, preliminary abnormal elevation measurement points within the lane range are determined, including: Acquiring a high-frequency road surface elevation signal from the target road surface elevation data; For any measuring point, calculate the elevation mean and elevation variance of the high-frequency road surface elevation signals of all measuring points within a first preset range around the measuring point; For any measuring point, a first abnormal elevation segmentation threshold and a second abnormal elevation segmentation threshold are calculated based on the corresponding elevation mean and elevation variance, so that the first abnormal elevation segmentation threshold is greater than the second abnormal elevation segmentation threshold; For any measuring point, if the high-frequency road surface elevation signal corresponding to the measuring point is greater than the first abnormal elevation segmentation threshold or less than the second abnormal elevation segmentation threshold, the measuring point is determined to be the preliminary abnormal elevation measuring point.

5. The method for full-width structural depth detection based on precise three-dimensional analysis according to claim 3, characterized in that: Determining the abnormal elevation measurement point based on the target road surface grayscale data and the preliminary abnormal elevation measurement point includes: For any preliminary abnormal elevation measurement point, calculate the grayscale mean and grayscale variance of the target road surface grayscale data of all measurement points within a second preset range around the preliminary abnormal elevation measurement point; For any preliminary abnormal elevation measurement point, a first abnormal grayscale segmentation threshold and a second abnormal grayscale segmentation threshold are calculated based on the corresponding grayscale mean and grayscale variance, so that the first abnormal grayscale segmentation threshold is greater than the second abnormal grayscale segmentation threshold; For any preliminary abnormal elevation measurement point, if the corresponding target road surface grayscale data is greater than a first abnormal elevation segmentation threshold or less than a second abnormal elevation segmentation threshold, the any preliminary abnormal elevation measurement point is determined to be the abnormal elevation measurement point.

6. The method for full-width structural depth detection based on precise three-dimensional analysis according to claim 2, characterized in that: Based on the target road surface elevation data of the non-abnormal elevation measurement points, the elevation estimation data of the abnormal elevation measurement points are estimated to generate effective road surface elevation data, including: For any abnormal elevation measurement point, estimating elevation data of the abnormal elevation measurement point based on target road surface elevation data of non-abnormal elevation measurement points within a predetermined area around the abnormal elevation measurement point; The elevation estimation data and the target road surface elevation data of the non-abnormal elevation measurement point are determined as the valid road surface elevation data.

7. The method for full-width structural depth detection based on precise three-dimensional imaging according to any one of claims 1 to 6, characterized in that: Based on the reconstructed 3D pavement point cloud data and the preset pavement structural depth calculation model, the full-width pavement structural depth is determined, including: Divide the reconstructed road surface 3D point cloud data into multiple first-level point cloud units along the driving direction; Dividing any of the first-level point cloud units into a plurality of second-level point cloud units along the road width direction; For all the secondary point cloud units in any of the primary point cloud units, the structural depths of all the secondary point cloud units are calculated based on a preset road surface structural depth calculation model, thereby obtaining a structural depth set for each of the primary point cloud units; The full-width construction depth of the road surface is determined based on the construction depth set.

8. The method for full-width structural depth detection based on precise three-dimensional analysis according to claim 7, characterized in that: Determining the full-width construction depth of the road surface based on the construction depth set includes: Determine, based on the structural depth set: at least one of: a full-width maximum structural depth, a full-width average structural depth, a full-width weighted structural depth, a left wheel track structural depth, a right wheel track structural depth, and the structural depth set; The maximum structural depth of the entire image is the maximum structural depth in the structural depth set of each first-level point cloud unit; The full-width average structural depth is the average structural depth in the structural depth set of each first-level point cloud unit; The full-width weighted structural depth is the weighted average of the structural depths in the structural depth set of each first-level point cloud unit; The left wheel track construction depth is the construction depth of the secondary point cloud unit corresponding to the left wheel track band; The right wheel track construction depth is the construction depth of the secondary point cloud unit corresponding to the right wheel track band.

9. The method for full-frame structure depth detection based on precise three-dimensional analysis according to claim 8, characterized in that: The left wheel track construction depth is calculated as follows: The left wheel track position is calculated based on the lane line position, the distance between the left and right wheel tracks, and the coincidence of the left and right wheel track centers with the lane center. The set of construction depths of all secondary point cloud units located at the left wheel track zone is determined as the left wheel track construction depth. The right wheel track construction depth is calculated as follows: The right wheel track position is calculated based on the lane line position, the distance between the left and right wheel tracks, and the coincidence of the left and right wheel track centers with the lane center. The set of construction depths of all secondary point cloud units located at the right wheel track zone is determined as the right wheel track construction depth. The full-width weighted structural depth is calculated as follows: The weight w of the corresponding structural depth of each secondary point cloud unit in the structural depth set of each primary point cloud unit is calculated according to the following formula: i : DIS i =min(|x i -x L |,|x i -x R |) Among them, DIS i is the minimum value of the distance between the position corresponding to the construction depth of the i-th secondary point cloud unit and the left wheel track belt and the right wheel track belt in the road width direction, x i is the position of the i-th secondary point cloud unit in the road width direction, x L and x R are the positions of the left and right wheel tracks in the road width direction, respectively, where i = 1, 2, …, n, and n is the number of secondary point cloud units in each primary point cloud unit; The full-width weighted structural depth of each first-level point cloud unit is calculated according to the following formula: Among them, SMTD i is the construction depth of the i-th secondary point cloud unit in each primary point cloud unit.

Images