Method and device for generating outdoor large-scale scene lidar point cloud map

Abstract

The present invention relates to the technical field of point cloud data processing, and particularly relates to a method and device for generating an outdoor large-scale lidar point cloud map. The method includes: obtaining a preliminary segmentation result of the target lidar point cloud map based on a region-growing segmentation method, and performing boundary point optimization processing to obtain a target plane segmentation result of the point cloud map; calculating the target normal vectors of all the segmented planes in the target plane segmentation result, and then performing refined processing on the plane point cloud of the plane segmentation result to perform plane segmentation on the refined point cloud map, determining the target corner structure of the new target plane segmentation result, and removing the target miscellaneous points in the corner point cloud of the corner structure to generate a refined outdoor large-scale lidar point cloud map. Thus, the problem in the related art that it is impossible to increase the point cloud accuracy from the hardware and algorithm levels in a short time, which increases data noise and cannot meet the requirements for generating a point cloud map in complex scenarios, is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of point cloud data processing, and particularly relates to a method and device for generating an outdoor large-scale lidar point cloud map. Background Art

[0002] LiDAR (Light Detection And Ranging) is a high-precision active sensor and is widely used in 3D reconstruction tasks. However, due to the design defects of the lidar system itself and the errors of the 3D reconstruction algorithm, the generated 3D point cloud map often has a large amount of data noise, which is mainly manifested in the fact that the outer walls of buildings have a large thickness in outdoor large-scale scenes.

[0003] Since the lidar point cloud map is an important input for generating a 3D structure model, it is impossible to increase the point cloud accuracy from the hardware and algorithm levels in a short time, which increases the data noise and cannot meet the requirements for generating a point cloud map in complex scenarios, and urgent improvement is needed. Summary of the Invention

[0004] The present invention provides a method and device for generating an outdoor large-scale lidar point cloud map to solve the problem in the related art that since the lidar point cloud map is an important input for generating a 3D structure model, it is impossible to increase the point cloud accuracy from the hardware and algorithm levels in a short time, which increases the data noise and cannot meet the requirements for generating a point cloud map in complex scenarios.

[0005] In a first aspect embodiment of the present invention, a method for generating an outdoor large-scale lidar point cloud map is provided, including the following steps: obtaining a preliminary segmentation result of a target lidar point cloud map based on a region growing segmentation method, and performing boundary point optimization processing on the preliminary segmentation result to obtain a target plane segmentation result of the point cloud map; calculating target normal vectors of all segmentation planes in the target plane segmentation result, and using the target normal vectors to perform plane point cloud refinement processing on the plane segmentation result to generate a refined point cloud map; performing plane segmentation on the refined point cloud map to obtain a new target plane segmentation result, determining a target corner structure of the new target plane segmentation result, and removing target miscellaneous points in the corner point cloud of the corner structure to generate a refined outdoor large-scale lidar point cloud map.

[0006] Optionally, in an embodiment of the present invention, the obtaining a preliminary segmentation result of a target lidar point cloud map based on a region growing segmentation method includes: calculating local normal vectors of each scan point in the target lidar point cloud map; performing point cloud clustering on the local normal vectors of each scan point based on the region growing segmentation method to obtain the preliminary segmentation result of the target lidar point cloud map.

[0007] Optionally, in an embodiment of the present invention, the boundary point optimization process for the preliminary segmentation result includes: calculating the vertical distance and Euclidean distance from non-planar points to the segmentation plane to obtain non-planar points that meet preset conditions; dividing the non-planar points that meet the preset conditions into the segmentation plane to perform boundary point optimization on the preliminary segmentation result.

[0008] Optionally, in an embodiment of the present invention, the plane point cloud refinement process for the plane segmentation result using the target normal vector to generate a refined point cloud map includes: calculating the neighborhood centroid of the plane points in the segmentation plane; determining new plane points in the segmentation plane based on the plane points, the neighborhood centroid, and the target normal vector; generating a refined point cloud map according to the new plane points.

[0009] Optionally, in an embodiment of the present invention, determining the target corner structure of the new target plane segmentation result and removing target miscellaneous points in the corner point cloud of the corner structure to generate a refined outdoor large-scale lidar point cloud map includes: obtaining the target plane in the new target plane segmentation result; determining whether the target plane is the target corner structure; if the target plane is the target corner structure, removing the target miscellaneous points in the corner point cloud to generate the refined outdoor large-scale lidar point cloud map.

[0010] An embodiment of the second aspect of the present invention provides a device for generating an outdoor large-scale lidar point cloud map, including: an acquisition module, configured to obtain a preliminary segmentation result of a target lidar point cloud map based on a region-growing segmentation method and perform boundary point optimization on the preliminary segmentation result to obtain a target plane segmentation result of the point cloud map; a processing module, configured to calculate the target normal vectors of all segmentation planes in the target plane segmentation result and use the target normal vectors to perform plane point cloud refinement on the plane segmentation result to generate a refined point cloud map; a generation module, configured to perform plane segmentation on the refined point cloud map to obtain a new target plane segmentation result, determine the target corner structure of the new target plane segmentation result, and remove target miscellaneous points in the corner point cloud of the corner structure to generate a refined outdoor large-scale lidar point cloud map.

[0011] Optionally, in an embodiment of the present invention, the obtaining module includes: a first calculation unit configured to calculate the local normal vector of each scan point in the target lidar point cloud map; a first obtaining unit configured to perform point cloud clustering on the local normal vectors of each scan point based on the region growing segmentation method to obtain the preliminary segmentation result of the target lidar point cloud map.

[0012] Optionally, in an embodiment of the present invention, the obtaining module includes: a second calculation unit configured to calculate the perpendicular distance and Euclidean distance from non-planar points to the segmentation plane to obtain non-planar points that meet preset conditions; a processing unit configured to divide the non-planar points that meet the preset conditions into the segmentation plane to perform boundary point optimization processing on the preliminary segmentation result.

[0013] Optionally, in an embodiment of the present invention, the processing module includes: a third calculation unit configured to calculate the domain centroid of planar points in the segmentation plane; a determination unit configured to determine new planar points in the segmentation plane based on the planar points, the domain centroid, and the target normal vector; a first generation unit configured to generate a refined point cloud map according to the new planar points.

[0014] Optionally, in an embodiment of the present invention, the generation module includes: a second obtaining unit configured to obtain the target plane in the new target plane segmentation result; a judgment unit configured to judge whether the target plane is the target corner structure; a second generation unit configured to, if the target plane is the target corner structure, remove the target miscellaneous points in the corner point cloud to generate a refined outdoor large-scale scene lidar point cloud map.

[0015] An embodiment of the third aspect of the present invention provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to implement the method for generating an outdoor large-scale scene lidar point cloud map as described in the above embodiments.

[0016] An embodiment of the fourth aspect of the present invention provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and when the program is executed by a processor, the method for generating an outdoor large-scale scene lidar point cloud map as described above is implemented.

[0017] An embodiment of the fifth aspect of the present invention provides a computer program product, including a computer program, and when the computer program is executed, it is used to implement the method for generating an outdoor large-scale scene lidar point cloud map as described above.

[0018] Embodiments of the present invention can obtain a preliminary segmentation result of a target lidar point cloud map based on a region growing segmentation method, and perform boundary point optimization processing to obtain a target plane segmentation result of the point cloud map; calculate the target normal vectors of all segmentation planes in the target plane segmentation result, and then perform refined processing on the plane point cloud of the plane segmentation result to segment the refined point cloud map into planes, and determine the target corner structure of the new target plane segmentation result, and remove the target noise points in the corner point cloud of the corner structure to generate a refined outdoor large-scale scene lidar point cloud map, effectively reducing data noise and meeting the requirements for generating a point cloud map in complex scenarios. Thus, the problem in the related art that it is impossible to increase the point cloud accuracy from the hardware and algorithm levels in a short time, increasing data noise and unable to meet the requirements for generating a point cloud map in complex scenarios is solved.

[0019] Additional aspects and advantages of the present invention will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and / or additional aspects and advantages of the present invention will become apparent and easy to understand from the following description of the embodiments in conjunction with the drawings, wherein:

[0021] Figure 1 is a flowchart of a method for generating an outdoor large-scale scene lidar point cloud map according to an embodiment of the present invention;

[0022] Figure 2 is a schematic diagram of the principle of a method for generating an outdoor large-scale scene lidar point cloud map according to a specific embodiment of the present invention;

[0023] Figure 3 is a schematic diagram of a plane segmentation result based on region growing according to a specific embodiment of the present invention;

[0024] Figure 4 is a schematic diagram of a plane segmentation result based on region growing and boundary point optimization according to a specific embodiment of the present invention;

[0025] Figure 5 is a schematic diagram of a point cloud map before refined processing of plane point cloud according to a specific embodiment of the present invention;

[0026] Figure 6 is a schematic diagram of a point cloud map after refined processing of plane point cloud according to a specific embodiment of the present invention;

[0027] Figure 7 is a schematic diagram of a point cloud map before detail optimization based on building structure regularization according to a specific embodiment of the present invention;

[0028] Figure 8Schematic diagram of a point cloud map after detail optimization based on the regularization of building structures according to a specific embodiment of the present invention;

[0029] Figure 9 Structural schematic diagram of a device for generating an outdoor large - scale lidar point cloud map according to an embodiment of the present invention;

[0030] Figure 10 Structural schematic diagram of an electronic device according to an embodiment of the present invention. Detailed implementation manners

[0031] The embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements with the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention, and should not be construed as limiting the present invention.

[0032] The method and device for generating an outdoor large - scale lidar point cloud map according to an embodiment of the present invention will be described below with reference to the accompanying drawings. In view of the problem in the related technology mentioned in the above - mentioned background technology that it is impossible to increase the point cloud accuracy from the hardware and algorithm levels in a short time, which increases data noise and cannot meet the requirements for generating a point cloud map in complex scenarios, the present invention provides a method for generating an outdoor large - scale lidar point cloud map. In this method, based on the region - growing segmentation method, a preliminary segmentation result of the target lidar point cloud map can be obtained, and boundary point optimization processing is performed to obtain the target plane segmentation result of the point cloud map; calculate the target normal vectors of all segmentation planes in the target plane segmentation result, and then perform plane point cloud refinement processing on the plane segmentation result to segment the refined point cloud map, determine the target corner structure of the new target plane segmentation result, and remove the target miscellaneous points in the corner point cloud of the corner structure to generate a refined outdoor large - scale lidar point cloud map, effectively reducing data noise and meeting the requirements for generating a point cloud map in complex scenarios. Thus, the problem in the related technology that it is impossible to increase the point cloud accuracy from the hardware and algorithm levels in a short time, which increases data noise and cannot meet the requirements for generating a point cloud map in complex scenarios, is solved.

[0033] Specifically, Figure 1 Flow schematic diagram of a method for generating an outdoor large - scale lidar point cloud map provided by an embodiment of the present invention.

[0034] As Figure 1 shown, the method for generating an outdoor large - scale lidar point cloud map includes the following steps:

[0035] In step S101, based on the region-growing segmentation method, obtain the preliminary segmentation result of the target lidar point cloud map, and perform boundary point optimization on the preliminary segmentation result to obtain the target plane segmentation result of the point cloud map.

[0036] It can be understood that the embodiments of the present invention can obtain the preliminary segmentation result of the target lidar point cloud map based on the region-growing segmentation method. For example, as Figure 2 shown, where the red scan points are non-planar scan points. The embodiments of the present invention can adopt a parallel computing strategy to calculate the local normal vector of each scan point in the point cloud map, and then use the region-growing point cloud clustering method to cluster the scan points with similar local normal vectors to obtain the preliminary segmentation result of the outdoor large-scale scene lidar point cloud map. Then, the boundary points of the preliminary segmentation result can be optimized, that is, by calculating the vertical distance and Euclidean distance from the non-planar points to the segmentation plane, some non-planar points, such as the point cloud at the wall corner, are attributed to the segmentation plane to obtain the target plane segmentation result of the point cloud map, effectively improving the executability of the refined outdoor large-scale scene lidar point cloud map.

[0037] Among them, in an embodiment of the present invention, based on the region-growing segmentation method, obtaining the preliminary segmentation result of the target lidar point cloud map includes: calculating the local normal vector of each scan point in the target lidar point cloud map; based on the region-growing segmentation method, performing point cloud clustering on the local normal vectors of each scan point to obtain the preliminary segmentation result of the target lidar point cloud map.

[0038] In the actual execution process, the embodiments of the present invention can first calculate the local normal vector of each scan point in the lidar point cloud map. The calculation method is as follows:

[0039]

[0040] Among them, is the current scan point, is the neighborhood scan point of the current scan point, is the total number of neighborhood points (usually set to 50), is the local normal vector of the scan point to be solved.

[0041] Then, using the local normal vector of the scan point as the input, perform region-growing point cloud clustering, which is implemented using the RegionGrown class of PCL. The specific parameter settings are as follows:

[0042]

[0043] Thus, it is possible to obtain Figure 3The plane segmentation result based on region growing as shown, where the non-planar point clouds at the wall corners are correctly assigned to the corresponding segmented planes, effectively improving the calculation efficiency, segmentation accuracy and quality, and at the same time reducing the influence of noise.

[0044] Optionally, in an embodiment of the present invention, boundary point optimization processing is performed on the preliminary segmentation result, including: calculating the vertical distance and Euclidean distance from the non-planar points to the segmented plane to obtain non-planar points that meet the preset conditions; dividing the non-planar points that meet the preset conditions into the segmented plane to perform boundary point optimization processing on the preliminary segmentation result.

[0045] As a possible implementation manner, the embodiments of the present invention can process the scanned points at the wall corners that are not assigned to the plane. For each non-planar scanned point , calculate its vertical distance from other segmented planes through the following calculation formula, and sort the vertical distances, and retain the top ten segmented planes with the closest vertical distances for subsequent processing. The calculation formula is as follows:

[0046]

[0047] where is the total number of neighborhood points, is the vertical distance from this non-planar scanned point to the plane , is the local normal vector of the scanned point to be solved.

[0048] For the top ten segmented planes with the closest vertical distances , calculate the Euclidean distance from to the ten planes respectively through the following formula:

[0049]

[0050] where is the non-planar scanned point, is the Euclidean distance from the non-planar scanned point to the plane among the ten planes, is the plane point of the plane .

[0051] Assign to the segmented plane with the smallest , and finally the updated segmented plane can be obtained, where Figure 4The planar segmentation result after boundary point optimization is shown, so that the embodiments of the present invention can perform boundary point optimization processing on the preliminary segmentation result to obtain the target planar segmentation result of the point cloud map, significantly improving the segmentation accuracy and quality of the point cloud map.

[0052] In step S102, calculate the target normal vectors of all the segmentation planes in the target planar segmentation result, and use the target normal vectors to perform refined processing on the planar point cloud of the planar segmentation result to generate a refined point cloud map.

[0053] It can be understood that the embodiments of the present invention can calculate the target normal vectors of all the segmentation planes in the target planar segmentation result, and use the target normal vectors to perform refined processing on the planar point cloud of the planar segmentation result. For example, as Figure 2 shown, the embodiments of the present invention can perform neighborhood search on each planar point of the segmentation plane, and then calculate the neighborhood center point, and compress the planar point along the plane normal vector towards the neighborhood center, so as to achieve the effect of thinning the plane, generate a refined point cloud map, and effectively reduce data noise.

[0054] Optionally, in an embodiment of the present invention, using the target normal vector to perform refined processing on the planar point cloud of the planar segmentation result to generate a refined point cloud map includes: calculating the neighborhood centroid of the planar points in the segmentation plane; determining new planar points in the segmentation plane based on the planar points, the neighborhood centroid, and the target normal vector; and generating a refined point cloud map according to the new planar points.

[0055] In some embodiments, the embodiments of the present invention can perform refined operation on the planar point cloud of the updated segmentation plane obtained in the above steps For the segmentation plane , first calculate the neighborhood centroid of its planar point according to the following formula, that is:

[0056]

[0057] where is set to 150.

[0058] Then, use principal component analysis to calculate the normal vector of the segmentation plane , and update the planar point according to the following formula, that is:

[0059]

[0060] where is the neighborhood centroid of the planar point , is the segmentation plane Normal vector of the

[0061] After performing the above operations on the plane points of all the segmentation planes, the refinement of the plane point cloud based on local normal is completed. As shown in Figure 5 and Figure 6 , the effects of the lidar point cloud map before and after performing the refinement of the plane point cloud based on local normal are shown respectively.

[0062] In step S103, the refined point cloud map is subjected to plane segmentation to obtain a new target plane segmentation result, and the target corner structure of the new target plane segmentation result is determined, and the target miscellaneous points in the corner point cloud of the corner structure are removed to generate a refined outdoor large-scale scene lidar point cloud map.

[0063] It can be understood that the embodiments of the present invention can perform plane segmentation on the refined point cloud map to obtain a new target plane segmentation result, and determine the target corner structure of the new target plane segmentation result. For example, as shown in Figure 2 , based on the detailed optimization of the building structure regularization, for the refined point cloud map, first perform the plane segmentation of the above steps to obtain a new target plane segmentation result. Then, determine the corner structure through orthogonal inspection and Euclidean distance, and finally identify and remove the miscellaneous points in the corner point cloud, so as to generate a refined outdoor large-scale scene lidar point cloud map, greatly improving the accuracy of device fusion mapping, reducing data noise, and meeting the requirements for generating a refined point cloud map in an outdoor large-scale scene.

[0064] Optionally, in an embodiment of the present invention, determining the target corner structure of the new target plane segmentation result and removing the target miscellaneous points in the corner point cloud of the corner structure to generate a refined outdoor large-scale scene lidar point cloud map includes: obtaining the target plane in the new target plane segmentation result; determining whether the target plane is a target corner structure; if the target plane is a target corner structure, removing the target miscellaneous points in the corner point cloud to generate a refined outdoor large-scale scene lidar point cloud map.

[0065] In some embodiments, the embodiments of the present invention first perform plane segmentation based on region growth and boundary point optimization on the point cloud map that has been refined by plane point cloud to obtain a new plane segmentation result , for the new plane segmentation result , two planes are sequentially extracted from all the planes, and it is determined whether they are corner structures according to orthogonality and Euclidean distance. For the plane pair , each plane in is retrieved in turn, and the orthogonality of and is calculated according to the following formula:

[0066]

[0067] Sort the orthogonality and select the five planes with the smallest planes as candidate planes , and then calculate according to the following formula and the Euclidean distance of each plane in

[0068]

[0069] where represents and the candidate plane in the Euclidean distance of represents the plane points of is the number of plane points, is the plane the plane points of is the number of plane points.

[0070] If the candidate plane with the smallest and the plane orthogonality is less than a given threshold (set to 0.17 here), then it is considered that and constitute a corner structure.

[0071] For each retrieved corner plane pair and , first regularize its building structure. Process each plane of the plane pair separately. For one of the planes , divide its plane points into a positive point set and a negative point set according to the following formula:

[0072]

[0073] where refers to the centroid of the plane .

[0074] First process the point set with more points in the positive point set and the negative point set . Assume the point set has more points. For each point in update according to the following formula:

[0075]

[0076] Among them, is a point in the point set; is the distance threshold for judging the distance of a point from another plane, which is set to 0.2 meters here; is the distance that the point in the point set moves towards another plane, which is set to 0.1 meters here.

[0077] After the above operations are completed for both of the two planes facing the corner plane, finally, noise points are removed. For the updated plane pairs and are processed separately. For one of the planes , the plane points are still divided into a positive point set and a negative point set according to the following formula:

[0078]

[0079] Here, it is assumed that the point set with fewer points is , and the point set is regarded as a noise point set, and the points in it are removed from the point cloud map, that is, the detailed optimization operation based on the building structure regularization is completed. Figure 7 and Figure 8 are the point cloud maps before and after processing respectively. Figure 7 is the point cloud map before the detailed optimization based on the building structure regularization, and there are obvious vacancies at the wall corners. Figure 8 is the point cloud map after the detailed optimization based on the building structure regularization, and the vacancies at the wall corners are filled.

[0080] According to the method for generating an outdoor large - scale scene lidar point cloud map proposed in the embodiment of the present invention, a preliminary segmentation result of the target lidar point cloud map can be obtained based on the region - growing segmentation method, and boundary point optimization processing is performed to obtain the target plane segmentation result of the point cloud map; calculate the target normal vectors of all the segmented planes in the target plane segmentation result, and then perform refined processing on the plane point cloud of the plane segmentation result to segment the refined point cloud map into planes, and determine the target corner structure of the new target plane segmentation result, and remove the target noise points in the corner point cloud of the corner structure to generate a refined outdoor large - scale scene lidar point cloud map, effectively reducing data noise and meeting the requirements for generating a point cloud map in complex scenarios. Thus, the problem in the related technology that it is impossible to increase the point cloud accuracy from the hardware and algorithm levels in a short time, increasing data noise and unable to meet the requirements for generating a point cloud map in complex scenarios is solved.

[0081] Secondly, a device for generating an outdoor large - scale scene lidar point cloud map proposed in the embodiment of the present invention is described with reference to the accompanying drawings.

[0082] Figure 9 It is a block diagram of a device for generating an outdoor large - scene lidar point cloud map according to an embodiment of the present invention.

[0083] As Figure 9 shown, the device 10 for generating an outdoor large - scene lidar point cloud map includes: an acquisition module 100, a processing module 200, and a generation module 300.

[0084] Specifically, the acquisition module 100 is configured to obtain a preliminary segmentation result of the target lidar point cloud map based on a region - growing segmentation method, and perform boundary point optimization processing on the preliminary segmentation result to obtain a target plane segmentation result of the point cloud map.

[0085] The processing module 200 is configured to calculate the target normal vectors of all segmentation planes in the target plane segmentation result, and use the target normal vectors to perform refined processing on the plane point cloud of the plane segmentation result to generate a refined point cloud map.

[0086] The generation module 300 is configured to perform plane segmentation on the refined point cloud map to obtain a new target plane segmentation result, determine the target corner structure of the new target plane segmentation result, and remove target miscellaneous points in the corner point cloud of the corner structure to generate a refined outdoor large - scene lidar point cloud map.

[0087] Optionally, in an embodiment of the present invention, the acquisition module 100 includes: a first calculation unit and a first acquisition unit.

[0088] Among them, the first calculation unit is configured to calculate the local normal vector of each scan point in the target lidar point cloud map.

[0089] The first acquisition unit is configured to perform point cloud clustering on the local normal vectors of each scan point based on a region - growing segmentation method to obtain a preliminary segmentation result of the target lidar point cloud map.

[0090] Optionally, in an embodiment of the present invention, the acquisition module 100 includes: a second calculation unit and a processing unit.

[0091] Among them, the second calculation unit is configured to calculate the vertical distance and Euclidean distance from non - planar points to the segmentation plane to obtain non - planar points that meet preset conditions.

[0092] The processing unit is configured to divide the non - planar points that meet the preset conditions into the segmentation plane to perform boundary point optimization processing on the preliminary segmentation result.

[0093] Optionally, in an embodiment of the present invention, the processing module 200 includes: a third calculation unit, a determination unit, and a first generation unit.

[0094] Among them, the third calculation unit is used to calculate the domain centroid of the plane points in the segmentation plane.

[0095] The determination unit is used to determine new plane points in the segmentation plane based on the plane points, the domain centroid, and the target normal vector.

[0096] The first generation unit is used to generate a refined point cloud map according to the new plane points.

[0097] Optionally, in an embodiment of the present invention, the generation module 300 includes: a second acquisition unit, a judgment unit, and a second generation unit.

[0098] Among them, the second acquisition unit is used to acquire the target plane in the new target plane segmentation result.

[0099] The judgment unit is used to judge whether the target plane is a target corner structure.

[0100] The second generation unit is used to remove the target miscellaneous points in the corner point cloud if the target plane is a target corner structure, so as to generate a refined outdoor large-scale scene lidar point cloud map.

[0101] It should be noted that the foregoing explanation of the embodiment of the method for generating an outdoor large-scale scene lidar point cloud map also applies to the device for generating an outdoor large-scale scene lidar point cloud map in this embodiment, and will not be repeated here.

[0102] According to the device for generating an outdoor large-scale scene lidar point cloud map provided by the embodiment of the present invention, based on the region-growing segmentation method, the preliminary segmentation result of the target lidar point cloud map can be obtained, and the boundary point optimization process can be performed to obtain the target plane segmentation result of the point cloud map; calculate the target normal vectors of all the segmentation planes in the target plane segmentation result, and then perform refined processing on the plane point cloud of the plane segmentation result, so as to perform plane segmentation on the refined point cloud map, and determine the target corner structure of the new target plane segmentation result, and remove the target miscellaneous points in the corner point cloud of the corner structure, so as to generate a refined outdoor large-scale scene lidar point cloud map, effectively reducing data noise and meeting the requirements for generating a point cloud map in complex scenarios. Thus, the problem in the related technology that the point cloud accuracy cannot be increased from the hardware and algorithm levels in a short time, the data noise is increased, and the requirements for generating a point cloud map in complex scenarios cannot be met is solved.

[0103] Figure 10 It is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. The electronic device may include:

[0104] A memory 1001, a processor 1002, and a computer program stored in the memory 1001 and executable on the processor 1002.

[0105] When the processor 1002 executes the program, it implements the method for generating an outdoor large-scale lidar point cloud map provided in the above embodiments.

[0106] Furthermore, the electronic device further includes:

[0107] A communication interface 1003 for communication between the memory 1001 and the processor 1002.

[0108] The memory 1001 is used to store a computer program executable on the processor 1002.

[0109] The memory 1001 may include a high-speed RAM memory, and may also include non-volatile memory, such as at least one disk memory.

[0110] If the memory 1001, the processor 1002, and the communication interface 1003 are implemented independently, the communication interface 1003, the memory 1001, and the processor 1002 can be interconnected through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 10 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0111] Optionally, in a specific implementation, if the memory 1001, the processor 1002, and the communication interface 1003 are integrated on a chip, the memory 1001, the processor 1002, and the communication interface 1003 can communicate with each other through an internal interface.

[0112] The processor 1002 may be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention.

[0113] This embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the method for generating an outdoor large-scale lidar point cloud map as described above is implemented.

[0114] This embodiment also provides a computer program product, including a computer program. When the computer program is executed, it is used to implement the method for generating an outdoor large-scale lidar point cloud map as described above.

[0115] In the description of this specification, the descriptions with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or N embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0116] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0117] Any process or method description in the flowchart or described in other ways herein can be understood as representing a module, segment, or part of code including one or N executable instructions for implementing a customized logical function or process. The scope of the preferred embodiments of the present invention includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in the reverse order according to the involved functions, rather than in the order shown or discussed, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0118] The logic and / or steps represented in the flowchart or otherwise described herein can, for example, be considered as a definable sequence list of executable instructions for implementing logical functions, which can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in conjunction with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection part (electronic device) having one or N wirings, a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then stored in a computer memory.

[0119] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0120] Those of ordinary skill in the art of this technology can understand that all or part of the steps carried by the method of the above embodiments can be completed by a program instructing relevant hardware, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0121] In addition, each functional unit in various embodiments of the present invention may be integrated into a processing module, may exist individually in physical form for each unit, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0122] The above-mentioned storage medium may be a read-only memory, a magnetic disk, an optical disc, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for generating an outdoor large-scale lidar point cloud map, characterized in that, It includes the following steps: Based on the region-growing segmentation method, obtain the preliminary segmentation result of the target lidar point cloud map, and perform boundary point optimization processing on the preliminary segmentation result to obtain the target plane segmentation result of the point cloud map. Among them, the step of obtaining the preliminary segmentation result of the target lidar point cloud map based on the region-growing segmentation method includes: calculating the local normal vector of each scan point in the target lidar point cloud map; based on the region-growing segmentation method, performing point cloud clustering on the local normal vectors of each scan point to obtain the preliminary segmentation result of the target lidar point cloud map. Among them, the calculation method of the local normal vector of each scan point is as follows: Among them, is the current scanning point, is the neighborhood scanning point of the current scanning point, is the total number of neighborhood points, is the local normal vector of the scanning point to be solved; Calculate the target normal vectors of all segmentation planes in the target plane segmentation result, and use the target normal vectors to perform refined processing on the plane point cloud of the plane segmentation result to generate a refined point cloud map. Among them, the step of using the target normal vectors to perform refined processing on the plane point cloud of the plane segmentation result to generate a refined point cloud map includes: calculating the neighborhood centroid of the plane points in the segmentation plane; based on the plane points, the neighborhood centroid, and the target normal vectors, determine the new plane points in the segmentation plane; generate a refined point cloud map according to the new plane points. Among them, the calculation method of the neighborhood centroid of the plane points in the segmentation plane is as follows: Among them, is set to 150, is the centroid of the domain of the planar point , and is the dividing plane; Next, the principal component analysis is used to calculate the normal vector of the segmentation plane and the plane points are updated using the following formula to obtain new plane points. The calculation method is as follows: ​ Among them, is a new planar point, is the centroid of the neighborhood of the planar point , and is the normal vector of the partitioning plane . Perform plane segmentation on the refined point cloud map to obtain a new target plane segmentation result, determine the target corner structure of the new target plane segmentation result, and remove the target miscellaneous points in the corner point cloud of the corner structure to generate a refined outdoor large-scale scene lidar point cloud map.

2. The method for generating an outdoor large-scale scene lidar point cloud map according to claim 1, wherein The boundary point optimization processing of the preliminary segmentation result includes: Calculate the vertical distance and Euclidean distance from the non-planar points to the segmentation plane to obtain non-planar points that meet the preset conditions; Divide the non-planar points that meet the preset conditions into the segmentation plane to perform boundary point optimization processing on the preliminary segmentation result.

3. The method for generating an outdoor large-scale scene lidar point cloud map according to claim 1, wherein The step of determining the target corner structure of the new target plane segmentation result and removing the target miscellaneous points in the corner point cloud of the corner structure to generate a refined outdoor large-scale scene lidar point cloud map includes: Obtain the target plane in the new target plane segmentation result; Judge whether the target plane is the target corner structure; If the target plane is the target corner structure, then remove the target miscellaneous points in the corner point cloud to generate the refined outdoor large-scale scene lidar point cloud map.

4. An apparatus for generating an outdoor large-scale lidar point cloud map, characterized in that, It includes: An acquisition module, configured to obtain a preliminary segmentation result of a target lidar point cloud map based on a region-growing segmentation method, and perform boundary point optimization processing on the preliminary segmentation result to obtain a target plane segmentation result of the point cloud map. The obtaining of the preliminary segmentation result of the target lidar point cloud map based on the region-growing segmentation method includes: calculating the local normal vector of each scan point in the target lidar point cloud map; based on the region-growing segmentation method, performing point cloud clustering on the local normal vectors of each scan point to obtain the preliminary segmentation result of the target lidar point cloud map. The calculation method of the local normal vector of each scan point is as follows: Among them, is the current scanning point, is the neighborhood scanning point of the current scanning point, is the total number of neighborhood points, is the local normal vector of the scanning point to be solved; A processing module, configured to calculate the target normal vectors of all segmentation planes in the target plane segmentation result, and use the target normal vectors to perform refined processing on the plane point cloud of the plane segmentation result to generate a refined point cloud map. The performing of refined processing on the plane point cloud of the plane segmentation result by using the target normal vectors to generate a refined point cloud map includes: calculating the neighborhood centroid of the plane points in the segmentation plane; determining new plane points in the segmentation plane based on the plane points, the neighborhood centroid, and the target normal vectors; generating a refined point cloud map according to the new plane points. The calculation method of the neighborhood centroid of the plane points in the segmentation plane is as follows: Among them, is set to 150, is the centroid of the domain of the planar point , and is the dividing plane; Next, the principal component analysis is used to calculate the normal vector of the segmentation plane and the plane points are updated using the following formula to obtain new plane points. The calculation method is as follows: ​ Among them, is a new planar point, is the centroid of the neighborhood of the planar point , and is the normal vector of the dividing plane . A generation module, configured to perform plane segmentation on the refined point cloud map to obtain a new target plane segmentation result, determine the target corner structure of the new target plane segmentation result, and remove target noise points in the corner point cloud of the corner structure to generate a refined outdoor large-scale scene lidar point cloud map.

5. An electronic device, characterized in that, Comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor. The processor executes the program to implement the method for generating an outdoor large-scale scene lidar point cloud map according to any one of claims 1-3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program is executed by the processor to be used for implementing the method for generating an outdoor large-scale scene lidar point cloud map according to any one of claims 1-3.

7. A computer program product, comprising a computer program, characterized in that, The computer program is executed by the processor to be used for implementing the method for generating an outdoor large-scale scene lidar point cloud map according to any one of claims 1-3.

Images