3D map creation methods, devices and surveying equipment

Abstract

This application provides a method, apparatus, and surveying equipment for creating three-dimensional maps, relating to the field of surveying technology. The method is applied to a surveying equipment, which includes a non-repetitive scanning LiDAR. The method includes: using SLAM technology to dynamically scan with the LiDAR to obtain a global map of a target area; determining a local target area within the global map; planning multiple control points within the local target area; controlling the LiDAR to move to each control point and performing static scanning at each control point to obtain a local target map of the local target area; and stitching the local target map onto the global map to obtain a three-dimensional target map. Thus, by utilizing a non-repetitive scanning LiDAR, a single set of hardware can support a hybrid dynamic and static surveying mode, enabling efficient mapping of the global environment while simultaneously allowing for static fixed-point scanning mapping of local areas to achieve high mapping accuracy.

Description

Technical Field

[0001] This application relates to the field of surveying and mapping technology, and more specifically, to a method, apparatus and surveying equipment for creating three-dimensional maps. Background Technology

[0002] In recent years, with the rise of the metaverse trend, the importance of 3D reconstruction of the real world has been emphasized. Digital modeling of various scenarios can achieve many functions. For example, digital modeling of factories can be used for remote visual collaborative decision-making, production logistics planning, and layout planning; digital mapping of the production environment lays the foundation for future virtual factories; building building information models at construction sites allows for real-time monitoring of construction progress and verification of building models, avoiding high rework costs and project delays; and digital modeling of underground parking lots can assist in 3D navigation of parking spots, charging piles, entrances and exits, and elevator shafts.

[0003] From a first-principles perspective, current 3D reconstruction solutions mainly fall into two categories: the first is Terrestrial Laser Scanning (TLS), which performs static, fixed-point scanning and mapping; the second is Mobile Laser Scanning (MLS), which performs dynamic, mobile scanning and mapping based on SLAM (Simultaneous Localization and Mapping) technology. TLS offers high mapping accuracy, but it requires manual movement and the selection of control points based on experience. This approach results in extremely large and heavy hardware, making it a time-consuming and labor-intensive task. MLS, on the other hand, addresses the mapping efficiency issue to some extent by eliminating the need for control point selection. However, its motion estimation and odometry-based mapping principles limit its ability to achieve the same level of accuracy as TLS, making it unsuitable for applications requiring high-precision mapping, such as Building Information Modeling (BIM). Currently, the industry urgently needs a mapping solution that can guarantee high mapping efficiency across a large environment while achieving high-precision mapping in key, localized areas, while avoiding cross-device operation and enabling fully automated scanning and mapping. Summary of the Invention

[0004] This application provides a three-dimensional map creation method, apparatus, surveying equipment, and readable storage medium. It can utilize a non-repetitive scanning lidar to achieve a hybrid surveying and mapping mode that supports both dynamic and static modes using the same set of hardware. While efficiently mapping the global environment, it can also perform static fixed-point scanning mapping of local areas to achieve higher mapping accuracy.

[0005] The embodiments of this application can be implemented as follows:

[0006] In a first aspect, embodiments of this application provide a method for creating a three-dimensional map, the method being applied to a surveying device, the surveying device including a non-repeating scanning lidar, the method comprising:

[0007] Based on SLAM technology, the lidar is used to perform dynamic scanning to obtain a global map of the target area;

[0008] Identify the target local area in the global map;

[0009] Multiple static mapping control points are planned in the target local area, wherein the control points are static scanning points;

[0010] The lidar is controlled to move to each of the control points, and static scanning is performed at each of the control points to obtain a local map of the target area.

[0011] The target local map is stitched onto the global map to obtain a target 3D map.

[0012] Secondly, embodiments of this application provide a three-dimensional map creation apparatus, which is applied to a surveying and mapping equipment, the surveying and mapping equipment including a non-repetitive scanning lidar, and the apparatus includes:

[0013] The first map creation module is used to perform dynamic scanning based on SLAM technology using the LiDAR to obtain a global map of the target area;

[0014] The region determination module is used to determine the target local region in the global map;

[0015] The planning module is used to plan multiple statically mapped control points in the target local area, wherein the control points are static scanning points;

[0016] The second creation module is used to control the lidar to move to each of the control points and perform static scanning at each of the control points to obtain a target local map of the target local area.

[0017] The processing module is used to stitch the local map of the target onto the global map to obtain a three-dimensional map of the target.

[0018] Thirdly, embodiments of this application provide a surveying device, including a mobile platform, a non-repetitive scanning lidar, a processor, and a memory. The mobile platform is used to move the lidar, the lidar is used to obtain point clouds, and the memory stores machine-executable instructions that can be executed by the processor. The processor can execute the machine-executable instructions to implement the three-dimensional map creation method described in the foregoing embodiments.

[0019] Fourthly, embodiments of this application provide a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the three-dimensional map creation method as described in the foregoing embodiments.

[0020] The 3D map creation method, apparatus, surveying equipment, and readable storage medium provided in this application include a non-repetitive scanning LiDAR in the surveying equipment. The method includes: using LiDAR to perform dynamic scanning based on SLAM technology to obtain a global map of the target area; then, determining a target local area in the global map and planning multiple static surveying control points within the target local area, which serve as static scanning points; next, controlling the LiDAR to move to each control point and performing static scanning at each control point to obtain a target local map of the target local area; finally, stitching the target local map onto the global map to obtain a target 3D map. Thus, a non-repetitive scanning LiDAR can be used to achieve a hybrid dynamic and static surveying mode using the same hardware device. While efficiently mapping the global environment, static fixed-point scanning mapping of local areas can be performed to achieve high mapping accuracy. Furthermore, during static scanning, there is no need for manual selection of control points or manual movement of the equipment used for static scanning. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is one of the block diagrams of the surveying equipment provided in the embodiments of this application;

[0023] Figure 2 This is a schematic diagram of the structure of the surveying equipment provided in the embodiments of this application;

[0024] Figure 3 A second block diagram of the surveying equipment provided in the embodiments of this application;

[0025] Figure 4 One of the flowcharts illustrating the three-dimensional map creation method provided in this application embodiment;

[0026] Figure 5 for Figure 4 A flowchart illustrating the sub-steps included in step S110;

[0027] Figure 6 for Figure 4 A flowchart illustrating the sub-steps included in step S130;

[0028] Figure 7 for Figure 4 A flowchart illustrating one of the sub-steps included in step S140;

[0029] Figure 8 for Figure 7 A flowchart illustrating the sub-steps included in neutron step S143;

[0030] Figure 9 for Figure 4 A flowchart illustrating another seed step included in step S140;

[0031] Figure 10 A second schematic flowchart illustrating the three-dimensional map creation method provided in this application embodiment;

[0032] Figure 11 One of the block diagrams of the three-dimensional map creation apparatus provided in the embodiments of this application;

[0033] Figure 12 This is a second block diagram of a three-dimensional map creation device provided in an embodiment of this application.

[0034] Icons: 100-Surveying equipment; 110-Memory; 120-Processor; 130-Communication unit; 140-Mobile platform; 150-LiDAR; 160-Fisheye camera; 200-3D map creation device; 210-First map creation module; 220-Region determination module; 230-Planning module; 240-Second creation module; 250-Processing module; 260-Quality analysis module. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0036] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0037] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0038] Currently, there are two main types of 3D reconstruction schemes: the first is Terrestrial Laser Scanning (TLS), which performs static fixed-point scanning and mapping; the second is Mobile Laser Scanning (MLS), which performs dynamic mobile scanning and mapping based on SLAM (Simultaneous Localization and Mapping) technology.

[0039] The first type of solution is represented by Leica RTC360 and FARO Focus Premium. TLS mostly uses a single-line laser, achieving full Field of View (FOV) coverage through fixed-step rotation along both horizontal and vertical axes. High laser transceiver frequencies ensure scanning efficiency and point cloud density. Because it's static scanning and mapping, there are no biases caused by pose estimation, resulting in relatively accurate mapping. However, this solution can only cover a small FOV in a short time (for example, according to publicly available data from FARO, its vertical laser scanning step is 0.009°, and its maximum scanning speed is 97Hz in the vertical direction; calculated at the maximum speed, the FOV coverage per second is 0.873° x 0.873°). Therefore, it cannot support SLAM methods based on lidar odometry, and thus cannot perform dynamic mapping; it can only perform static fixed-point scanning and mapping, stitching together overlapping point cloud regions between different control points. Because there are no suitable initial values ​​for rigid body transformations between control points (i.e., the values ​​used for coordinate system transformations before stitching), a high point cloud overlap rate must be ensured for point cloud stitching. This requires control points to be close together, resulting in a significant waste of overlapping point clouds. Furthermore, the lack of support for SLAM technology prevents the entire mapping process from being automated.

[0040] The second type of approach is represented by NAVVIS VLX and GoSLAM RS100-RTK. This approach uses repetitive scanning mechanical LiDAR (usually 16-line or 32-line LiDAR) and employs SLAM methods based on LiDAR Odometry or LiDAR-IMU Odometry to estimate motion pose, thereby creating a map in real time during motion. Because this solution uses repetitive scanning mechanical LiDAR, for example, a 32-line LiDAR only achieves a static FoV coverage of 20%. The repetitive scanning characteristic dictates that this approach must improve point cloud coverage through motion to ensure effective point cloud density during 3D mapping. Furthermore, SLAM pose estimation has relatively high errors, resulting in relatively poor dynamic mapping accuracy, generally failing to reach the mapping accuracy of TLS.

[0041] In conclusion, the industry urgently needs a surveying solution that can ensure mapping efficiency in a large environment while also achieving surveying-grade mapping in areas with high accuracy requirements. This hybrid workflow will leverage the advantages of different surveying modes, combining overall surveying efficiency and quality to provide a superior solution compared to a single mode.

[0042] However, the most readily conceivable hybrid mapping workflow is as follows: use a mobile scanning device to map the entire area; manually select control points for static scanning, and manually move the static scanning device (TLS) to the control points to complete the static mapping; finally, merge the two maps together.

[0043] Therefore, the readily conceived hybrid mapping workflow relies on two different devices, which presents a complex cross-device data processing problem. The inconsistency of scan data obtained from different devices also poses significant challenges to data processing algorithms (e.g., different lidar reflectivity definitions, point cloud densities, effective ranges, and accuracies), and errors arising from coordinate system transformations between different devices also exist. Furthermore, existing TLS devices and mobile scanning devices are all prohibitively expensive.

[0044] As the above analysis shows, existing TLS devices and mobile scanning devices cannot be compatible with the other mode. In principle, the only solution that can achieve a single device using a hybrid static and dynamic mapping mode is to use a non-repeating scanning LiDAR. Therefore, this application provides a mapping device solution that uses a non-repeating scanning LiDAR to construct a global map ranging from coarse to fine. Optionally, the solution provided in this application may also include an omnidirectional fisheye camera for image acquisition to obtain a color map.

[0045] Existing surveying solutions mostly employ multiple cameras with different perspectives for the visual component, requiring multiple calibrations and image stitching. This not only introduces more sources of error but also increases system cost, size, and integration difficulty. In contrast, the solution provided in this application introduces a fisheye camera into the surveying system, resulting in a lightweight, simple, easily integrated, and low-cost system.

[0046] The shortcomings of the above solutions are the result of the inventors' practice and careful research. Therefore, the discovery process of the above problems and the solutions proposed by the inventors in the embodiments of this application below should be considered as contributions made by the inventors to this application.

[0047] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0048] Please refer to Figure 1 , Figure 1This is one of the block diagrams of a surveying device 100 provided in an embodiment of this application. The surveying device 100 may be, but is not limited to, a robot. The surveying device 100 may include a memory 110, a processor 120, and a communication unit 130. The memory 110, processor 120, and communication unit 130 are electrically connected directly or indirectly to each other to achieve data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines.

[0049] The memory 110 is used to store programs or data. The memory 110 may be, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc.

[0050] The processor 120 is used to read / write data or programs stored in the memory 110 and execute corresponding functions. For example, the memory 110 stores a 3D map creation device 200, which includes at least one software function module that can be stored in the memory 110 in the form of software or firmware. The processor 120 executes various functional applications and data processing by running the software programs and modules stored in the memory 110, such as the 3D map creation device 200 in this embodiment, thereby implementing the 3D map creation method in this embodiment.

[0051] The communication unit 130 is used to establish a communication connection between the 110 and other communication terminals through the network, and to send and receive data through the network.

[0052] It should be understood that, Figure 1 The structure shown is only a schematic diagram of the surveying equipment 100. The surveying equipment 100 may also include a... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown. Figure 1 The components shown can be implemented using hardware, software, or a combination thereof.

[0053] In this embodiment, as Figure 2 and Figure 3As shown, the surveying equipment 100 may further include: a mobile platform 140 (i.e., Figure 3 The system includes a Mobile Platform (MPP) and a non-repeating scanning LiDAR 150. The LiDAR 150 can be an omnidirectional radar. The LiDAR 150 can be mounted on the mobile platform 140, which can move the LiDAR 150. The LiDAR 150 is used to obtain point clouds through scanning, and then a map can be created based on the point clouds.

[0054] Among them, such as Figure 3 As shown, the lidar 150 integrates an IMU (Inertial Measurement Unit), whose measurement data is fused with lidar data for more accurate motion estimation (pose estimation).

[0055] Optionally, the lidar 150 can be mounted on a rotating unit, such as a two-dimensional gimbal mount, so that the rotating unit can drive the lidar 150 to rotate, thereby changing the detection angle.

[0056] Optionally, in cases where it is necessary to create a color map, such as Figure 2 and Figure 3 As shown, the mapping device 100 may further include a fisheye camera 160. The fisheye camera 160 can be installed in the same way as the lidar 150. The fisheye camera 160 can be an omnidirectional camera. The fisheye camera 160 is used to take pictures when the lidar 150 acquires point clouds to obtain environmental images. The point clouds can then be colored based on the environmental images to obtain a color map.

[0057] As one possible implementation, the mapping device 100 utilizes an omnidirectional sensor combination comprised of a Livox Mid-360 LiDAR and a fisheye camera. The Livox Mid-360 is a 4-line LiDAR, the first to combine non-repeating scanning capabilities with a 360° horizontal field of view. Its omnidirectional field of view provides sufficient information to support SLAM methods, and its non-repeating scanning characteristic allows the field of view coverage to approach 100% over time, thus also enabling its use for fixed-point static mapping. The fisheye camera, with its 360° horizontal and 70° vertical field of view, effectively matches the LiDAR's field of view, ensuring color information in all point clouds. The entire sensor combination is very small, with the LiDAR and fisheye camera measuring only 6.5x6.5x6.5cm and 5x5x10cm respectively, allowing for integration onto any mobile platform, such as a mobile chassis, a robotic dog, or a drone. As one possible example, the sensor array is rigidly connected to a worm gear gimbal and fixed as a whole to a mobile chassis. The LiDAR, IMU, and fisheye camera can be synchronized via software-triggered TTL (Transistor-Transistor Logic) signals.

[0058] Typical hybrid mapping workflows require the simultaneous use of mobile scanning equipment and TLS (Transient Tracking) equipment, leveraging the advantages of different devices to achieve better results. However, in this embodiment, a static and dynamic hybrid mapping workflow can be achieved using only a single set of hardware. Furthermore, compared to readily conceivable hybrid mapping solutions using TLS + mobile scanning equipment, the mapping equipment 100 provided in this embodiment can operate in both coarse and fine modes simultaneously, with significantly reduced costs. Moreover, by using only one set of hardware, this embodiment avoids the problems associated with hybrid mapping using TLS + mobile scanning equipment, including issues related to cross-device data processing, data inconsistencies between different devices, and coordinate system transformations between different devices. Additionally, the introduction of a fisheye camera reduces the cost and complexity of the vision-based approach and improves system integration. By using an omnidirectional non-repetitive scanning LiDAR and an omnidirectional fisheye camera to form an omnidirectional sensor combination, mapping efficiency can be significantly improved.

[0059] Please refer to Figure 3 and Figure 4 , Figure 4 This is one of the flowcharts illustrating a three-dimensional map creation method provided in this application embodiment. The method can be applied to the aforementioned surveying equipment 100, which includes a non-repeating scanning LiDAR. The specific flow of the three-dimensional map creation method is described in detail below. In this embodiment, the method may include steps S110 to S150.

[0060] Step S110: Based on SLAM technology, the lidar is used to perform dynamic scanning to obtain a global map of the target area.

[0061] In this embodiment, the target area is the region where a global map needs to be constructed, and the specific area can be determined based on the actual situation. A coarse global map of the target area can be obtained based on SLAM technology.

[0062] Step S120: Determine the target local area in the global map.

[0063] Step S130: Plan multiple statically mapped control points in the target local area.

[0064] Optionally, in this embodiment, the target local area is an area requiring higher-precision mapping, and therefore static mapping is performed in this area. The target local area can be a region of interest manually specified by the user in the global map, or a local area determined through other means. Since a global map has already been constructed, the approximate environment of the target local area in the global map can be determined from the global map. Furthermore, based on the information about the target local area determined from the global map, control point planning can be performed in the target local area to determine multiple static mapping control points (i.e., Figure 3 (Planning for Viewpoint in ROI). Here, the control point is the location where the LiDAR performs static scanning; that is, the control point is a static scanning point, or a scanning viewpoint.

[0065] Step S140: Control the lidar to move to each of the control points and perform static scanning at each of the control points to obtain a local map of the target area.

[0066] When multiple control points are identified, the mapping equipment can move the lidar to each control point and perform static mapping (i.e., stationary scanning) at each control point to obtain point clouds. Then, based on the obtained point clouds, fine mapping in ROI is performed to construct a target local map of the target local area.

[0067] It is understood that the target local map is a map obtained through static scanning, therefore the accuracy of the target local map is higher than that of the global map, and the target local map is a high-resolution mapping map.

[0068] Step S150: The target local map is stitched onto the global map to obtain a target 3D map.

[0069] Having obtained a coarse global map and a fine target local map, the target local map can be stitched onto the global map. In this way, a global map ranging from coarse to fine can be obtained, serving as the target 3D map corresponding to the target area.

[0070] In this embodiment, an omnidirectional lidar with non-repetitive scanning is used to enable a single hardware device to support a hybrid mapping mode that combines dynamic and static mapping. This allows for efficient mapping of the entire environment while simultaneously enabling static point-to-point mapping of local areas, achieving high mapping accuracy. Furthermore, it allows for automatic exploration and hybrid mapping of unfamiliar environments without prior maps.

[0071] Optionally, the lidar can be an omnidirectional lidar. This provides sufficient information to support mapping using SLAM methods while simultaneously improving the efficiency of global map creation.

[0072] Optionally, as a possible implementation, the mapping device may further include a fisheye camera, which can be an omnidirectional fisheye camera, for example, a camera with a 360° horizontal field of view and a 70° vertical field of view. This fisheye camera and the aforementioned lidar can be used to... Figure 5 The method shown obtains a colored global map, which includes not only environmental location information but also color information.

[0073] Please refer to Figure 5 , Figure 5 for Figure 4 A flowchart illustrating the sub-steps included in step S110. In this embodiment, step S110 may include sub-steps S111 to S112.

[0074] Sub-step S111: The point cloud is obtained by dynamic scanning using the lidar, and the fisheye camera is controlled to acquire images while the lidar is performing dynamic scanning to obtain a first environmental image.

[0075] In this embodiment, while obtaining point clouds for dynamic mapping using a lidar, the fisheye camera can be simultaneously controlled to acquire a first environmental image. Optionally, the orientation of the lidar when acquiring the point cloud can be consistent with the orientation of the fisheye camera when acquiring the image, so as to acquire environmental information from the same angle.

[0076] Sub-step S112: Based on the calibrated intrinsic parameters of the fisheye camera and the extrinsic parameters of the lidar relative to the fisheye camera, colorize the point cloud obtained during the dynamic scanning process of the first environmental image to obtain a global map with color information.

[0077] The intrinsic parameters of the fisheye camera and the extrinsic parameters of the lidar relative to the fisheye camera can be pre-calibrated or obtained through calibration when colorization is required. The specific calibration method can be determined according to actual needs and is not specifically limited here.

[0078] Based on the intrinsic parameters of the fisheye camera and the extrinsic parameters of the lidar relative to the fisheye camera, the correspondence between each pixel in the simultaneously acquired first environmental image and the points in the point cloud can be determined. Then, based on this correspondence, each point in the point cloud can be colored. Next, a map can be constructed from the colored point cloud to obtain a global map with color information.

[0079] Once a global map is created, a target local area can be delineated within it as the region requiring higher-precision mapping. Static scanning can then be performed on this target local area, and a local map of that area can be obtained through point cloud scanning, stitching, and other methods.

[0080] As one possible implementation method, it can be achieved through Figure 6 The method shown determines the location of control points used for static scanning. Please refer to... Figure 6 , Figure 6 for Figure 4 A flowchart illustrating the sub-steps included in step S130. In this embodiment, step S130 may include sub-steps S131 to S132.

[0081] Sub-step S131: Obtain the horizontal area of ​​the effective scanning area of ​​the lidar.

[0082] Sub-step S132: Based on the preset overlap range and the horizontal area, with the goal of maximizing the coverage of the horizontal area of ​​the target local region and minimizing the number of control points, determine multiple control points.

[0083] In this embodiment, when automatically selecting static mapping control points in the target local area, the target local area in the global map can be transformed into a top-down two-dimensional map. The area of ​​this two-dimensional map is the area of ​​the horizontal region of the target local area. On the horizontal plane, the effective scanning area of ​​the lidar can be considered as a circle with radius R. Given a preset overlap range between circles, a control point selection planning problem can be constructed to maximize the coverage of the horizontal region of the target local area and minimize the number of control points (i.e., minimize the scanning time), thereby solving for the positions of multiple control points.

[0084] The preset overlap range is a pre-set overlap range. The overlap within this range refers to the overlap between horizontal regions of the effective scanning area of ​​the LiDAR, specifically the overlap between the circles corresponding to the control point. The maximum value of this preset overlap range is used to limit the overlap of the identified control points to be as small as possible, thereby reducing the number of scans required to create a local target map and improving scanning efficiency. The minimum value of this preset overlap range is the overlap required to ensure the point clouds can be stitched together; the specific value can be set based on actual conditions or experience.

[0085] Once control points are identified, the surveying equipment can automatically move to each control point for static scanning. Utilizing the non-repetitive scanning characteristics of lidar, a dense point cloud is obtained through time accumulation, and then a local map of the target is obtained based on the dense point cloud.

[0086] As one possible implementation method, it can be achieved through Figure 7 The local map of the target is obtained as shown. Please refer to... Figure 7 , Figure 7 for Figure 4 A flowchart illustrating one of the sub-steps included in step S140. In this embodiment, step S140 may include sub-steps S141 to S143.

[0087] Sub-step S141: Control the lidar at each control point to scan at different angles to obtain point clouds corresponding to different angles.

[0088] Sub-step S142: For each control point, the point clouds corresponding to different angles of the control point are stitched together to obtain the point cloud corresponding to the control point.

[0089] In this embodiment, at each control point, the lidar needs to be rotated multiple times to different angles to acquire point clouds corresponding to different angular velocities. The point clouds scanned at different angles from the same control point can be stitched together based on the rotation angle to obtain a near 360° x 300° full-field-of-view point cloud for that control point. For example, the gimbal housing of the lidar can be rotated multiple times to drive the lidar to the corresponding angle; then, using the theoretical rigid body transformation corresponding to the gimbal rotation (i.e., the gimbal rotation angle) as the initial value, ICP (Iterative Closest Point) stitching is performed on the point clouds scanned at different angles to obtain a near 360° x 300° full-field-of-view point cloud.

[0090] Optionally, a preset scanning duration can be set. Once the scanning of this duration is completed at a control point, it can be determined that the scanning at that control point has been completed, and the scan can proceed to the next control point.

[0091] Sub-step S143 involves stitching together the point clouds corresponding to different control points to obtain the target local map.

[0092] Please refer to Figure 8 , Figure 8 for Figure 7 A flowchart illustrating the sub-steps included in neutron step S143. In this embodiment, sub-step S143 may include sub-steps S1431 to S1432.

[0093] Sub-step S1431: For the point clouds corresponding to different control points, based on the relative pose estimation results between the two control points, perform preliminary stitching on the point clouds corresponding to the two control points to obtain preliminary stitching results.

[0094] Sub-step S1432: Perform fine registration on the preliminary stitching result to obtain the target local map.

[0095] In this embodiment, the mapping device may further include a pose estimation unit. When performing global map creation, the relative pose of the mapping device between two locations calculated by this pose estimation unit can be used to complete the global map creation. This pose estimation unit may include an odometry device. That is, the pose estimation unit is also used for pose estimation when obtaining the global map based on SLAM technology.

[0096] After obtaining a 360°x300° point cloud at each control point, the pose estimate provided by the odometry when the surveying equipment moves between two control points can be used as the initial value for stitching the point clouds scanned from the two control points. Then, ICP is used for further fine registration, resulting in a surveying-grade detailed map of the ROI area with higher accuracy than SLAM dynamic mapping. The pose estimate provided by the odometry is the relative pose estimate of the surveying equipment between the two positions calculated by the pose estimation unit.

[0097] Typical point cloud stitching algorithms, such as the ICP algorithm, rely heavily on an accurate initial transformation for further optimization. Without a good initial value, it's necessary to ensure a high degree of overlap between the two point clouds to be stitched. This is achieved through coarse registration methods to obtain initial values ​​before fine stitching. Existing static scanning mapping schemes lack motion estimation capabilities, thus lacking accurate relative poses between two control points. Therefore, existing static scanning mapping schemes require that these two control points not be too far apart; otherwise, the accuracy of point cloud stitching cannot be guaranteed, significantly reducing the efficiency of static scanning. Furthermore, manually selecting control points based on experience cannot guarantee an appropriate spacing, as improper selection of scan points (too few points failing to cover the entire area, or too low overlap) frequently leads to mapping failures.

[0098] In this scheme, sub-steps S131 and S132 can automatically plan control points with appropriate intervals, and the relative pose required for point cloud stitching can be obtained through the pose estimation unit, eliminating the need to calculate the relative pose based on the overlap of point clouds between different control points. This reduces the number of control points, improves the speed of static mapping, and ensures the quality of the mapping.

[0099] Alternatively, it can also be done through Figure 9 The method shown utilizes a fisheye camera and LiDAR to obtain a colored local map of the target. This fisheye camera can be an omnidirectional fisheye camera. Please refer to... Figure 9 , Figure 9 for Figure 4 A flowchart illustrating another sub-step included in step S140. In this embodiment, step S140 may include sub-steps S145 to S146.

[0100] In sub-step S145, the lidar is used to perform static scanning at each of the control points to obtain a point cloud, and while the lidar is used to perform static scanning at each of the control points, the fisheye camera is controlled to acquire images to obtain a second environmental image.

[0101] In this embodiment, while obtaining point clouds for static mapping using a lidar, the fisheye camera can be simultaneously controlled to acquire a second environmental image. Optionally, the orientation of the lidar when acquiring the point cloud can be consistent with the orientation of the fisheye camera when acquiring the image, so as to acquire environmental information from the same angle.

[0102] Sub-step S146: Based on the calibrated intrinsic parameters of the fisheye camera and the extrinsic parameters of the lidar relative to the fisheye camera, colorize the point cloud obtained during the static scanning process of the second environmental image to obtain a target local map with color information.

[0103] Based on the intrinsic parameters of the fisheye camera and the extrinsic parameters of the lidar relative to the fisheye camera, the correspondence between each pixel in the simultaneously acquired second environmental image and the points in the point cloud can be determined. Then, based on this correspondence, each point in the point cloud can be colored. Next, a map can be constructed from the colored point cloud to obtain a local target map with color information.

[0104] The process involves first coloring the point clouds acquired simultaneously based on the second environmental image; then, stitching together the colored point clouds corresponding to different angular velocities at the same control point; and finally, stitching together the colored point clouds corresponding to different control points to obtain a colored local map of the target. It is understood that, except for the colored portion, the process of obtaining the target local map with color information is the same as sub-steps S141 to S143, and will not be elaborated upon here.

[0105] Having obtained the target local map and global map, ICP stitching can be used to stitch the detailed target local map onto a coarse global map to obtain a target 3D map. The target local map and global map used to stitch the target 3D map can both be in color, or only one of them can be in color; the specific settings can be determined based on actual needs.

[0106] Please refer to Figure 10 , Figure 10 This is a second schematic flowchart illustrating the three-dimensional map creation method provided in this application embodiment. In this embodiment, after step S150, the method may further include step S160.

[0107] Step S160: Analyze the quality of the target 3D map to obtain the map quality.

[0108] In this embodiment, after obtaining the target 3D map through steps S110 to S150, the quality of the target 3D map can be analyzed to obtain map quality. The specific analysis method used can be set according to actual needs. If the map quality is poor, steps S110 to S150 can be re-executed to reconstruct the target 3D map, or a better quality target 3D map can be obtained through data supplementation, etc.

[0109] Optionally, a first normal vector for the walls in the target 3D map and a second normal vector for the ground in the target 3D map can be calculated, and it can be determined whether the first normal vector and the second normal vector are perpendicular. Specifically, the first normal vector of a wall and the second normal vector of the ground at a single location can be calculated, and then it can be determined whether the corresponding first normal vector and the second normal vector are perpendicular. Alternatively, for multiple locations, the perpendicularity of the corresponding first normal vector and the second normal vector at each location can be determined separately. If there are cases where they are not perpendicular, the map quality can be determined to be unqualified, i.e., poor. For example, the normal vectors of the main walls and the ground can be calculated, and it can be determined whether their included angle is 90°.

[0110] Optionally, by comparison, walls in the target 3D map whose length meets a first preset length requirement can be selected, and then the third normal vector at different locations on these walls can be calculated. The first preset length requirement can be set according to actual conditions, such as the longest wall in the target 3D map, or one of the first few longest walls, etc. Next, it can be determined whether the third normal vectors at these multiple locations are parallel. If they are not parallel, the map quality can be determined to be unqualified.

[0111] Alternatively, by comparison, ground surfaces in the target 3D map whose length meets the second preset length requirement can be selected, and then the fourth normal vector at different locations on these ground surfaces can be calculated. The second preset length requirement can be set according to actual conditions, such as the longest ground surface in the target 3D map, or one of the first few longest ground surfaces. Next, it can be determined whether the fourth normal vectors at these multiple locations are parallel. If they are not parallel, the map quality can be determined to be unqualified.

[0112] That is, the normal vectors at different locations on the wall and / or the ground can be calculated to determine whether they are parallel. If they are not parallel, it indicates that the quality of the target 3D map is poor.

[0113] The quality of the fine point cloud in the local map of the target in the target 3D map can also be analyzed. If the quality of the fine point cloud is poor, it means that the quality of the target 3D map is poor.

[0114] Optionally, it can be determined whether the coverage of the selected control points to the target local area meets preset requirements, that is, whether the selected control points in the target local area can cover the entire target local area well. This determination method can be as follows: When performing ICP stitching on the detailed map of the target local area and the global coarse map, the average nearest point distance can be used as a criterion. If the detailed map of the target local area fails to cover the target local area well, the uncovered area will result in a larger average nearest point distance. Therefore, a threshold can be set to determine whether the coverage of the target local area (selection of control points) during detailed mapping is qualified.

[0115] That is, when stitching the target local map to the global map using the ICP stitching method, the nearest point distance for each point can be calculated. Then, the average of the nearest point distances for each point is calculated as the mean nearest point distance. Next, it is determined whether the mean nearest point distance is greater than a preset distance. If it is greater, it can be determined that the coverage of the control point to the target local area does not meet the preset requirements, and the quality of the target 3D map is poor.

[0116] The quality of the map can be determined as unqualified if at least one of the three methods mentioned above is included in the result and no result is obtained.

[0117] The following example uses a colored 3D map of the target target, combined with... Figure 3 An example is given to illustrate the target's 3D map.

[0118] First, based on SLAM technology, the aforementioned LiDAR is used for dynamic scanning, and a coarse global map of the target area is obtained through Global Coarse Mapping. Simultaneously, a camera can be used to acquire a first environmental image during the dynamic scanning process to colorize the global map.

[0119] The ROI (Region of Interest) is determined through ROI selection. Then, control points for static mapping (Planning for Viewpoints in ROI) are selected within the target local area. The mapping equipment is moved to each control point to collect point clouds from different angles (Stationary Scanning). Next, the point clouds from different control points are stitched together to complete fine mapping within the ROI, obtaining a mapping-grade fine map with higher accuracy than SLAM dynamic mapping. Furthermore, during the static mapping process, a second environmental image can be simultaneously acquired using a camera to colorize the fine map of the ROI.

[0120] Finally, the detailed map of the ROI is stitched onto the coarse map of the global environment using the ICP algorithm. Based on the first environmental image obtained during dynamic scanning and the second environmental image obtained during static mapping, coloring (Map Registration & Colorization) is performed to obtain the target 3D map. Alternatively, the point cloud can be colored first, and then the colored global and local maps can be obtained based on the colored point cloud; or the map can be built first based on the point cloud and then colored. The specific execution order can be determined based on the actual situation.

[0121] It can perform mapping quality inspection on the target 3D map to determine the map quality.

[0122] To perform the corresponding steps in the above embodiments and various possible methods, an implementation of a 3D map creation device 200 is given below. Optionally, the 3D map creation device 200 can adopt the above-described... Figure 1 The diagram shows the component structure of the surveying equipment 100. Please refer to... Figure 11 , Figure 11This is one of the block diagrams illustrating the 3D map creation device 200 provided in this application embodiment. It should be noted that the 3D map creation device 200 provided in this embodiment has the same basic principle and technical effects as those in the above embodiments. For the sake of brevity, any parts not mentioned in this embodiment can be referred to the corresponding content in the above embodiments. The 3D map creation device 200 can be applied to a surveying equipment 100, which includes a non-repetitive scanning LiDAR. The 3D map creation device 200 may include: a first map creation module 210, a region determination module 220, a planning module 230, a second creation module 240, and a processing module 250.

[0123] The first map creation module 210 is used to perform dynamic scanning based on SLAM technology using the LiDAR to obtain a global map of the target area.

[0124] The region determination module 220 is used to determine the target local region in the global map.

[0125] The planning module 230 is used to plan multiple static mapping control points in the target local area, wherein the control points are static scanning points.

[0126] The second creation module 240 is used to control the lidar to move to each of the control points and perform static scanning at each of the control points to obtain a target local map of the target local area.

[0127] The processing module 250 is used to stitch the target local map onto the global map to obtain a target 3D map.

[0128] Please refer to Figure 12 , Figure 12 This is a second block diagram of the 3D map creation device 200 provided in an embodiment of this application. The 3D map creation device 200 may further include a quality analysis module 260.

[0129] The quality analysis module 260 is used to analyze the quality of the target 3D map and obtain the map quality.

[0130] Optionally, the above modules can be stored in the form of software or firmware. Figure 1 The memory 110 shown is either stored in or embedded in the operating system (OS) of the surveying equipment 100, and can be used by... Figure 1 The processor 120 executes the program. Meanwhile, the data and program code required to execute the above modules can be stored in the memory 110.

[0131] This application also provides a readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned three-dimensional map creation method.

[0132] In summary, the 3D map creation method, apparatus, surveying equipment, and readable storage medium provided in this application include a non-repetitive scanning LiDAR in the surveying equipment. The method comprises: using LiDAR to perform dynamic scanning based on SLAM technology to obtain a global map of the target area; then, determining a target local area in the global map and planning multiple static surveying control points in the target local area, which are static scanning points; next, controlling the LiDAR to move to each control point and performing static scanning at each control point to obtain a target local map of the target local area; finally, stitching the target local map onto the global map to obtain a target 3D map. Thus, a non-repetitive scanning LiDAR can be used to achieve a hybrid dynamic and static surveying mode supported by the same hardware device. While efficiently mapping the global environment, static fixed-point scanning mapping of local areas can be performed to achieve high mapping accuracy. Furthermore, during static scanning, there is no need for manual selection of control points or manual movement of the equipment used for static scanning.

[0133] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0134] In addition, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0135] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0136] The above description is merely an optional embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A three-dimensional map creating method characterized by comprising: The method is applied to a mapping device including a non-repeating scanning laser radar, and the method includes: based on SLAM technology, dynamic scanning is performed by using the laser radar to obtain a global map of a target region; a target local region in the global map is determined; a plurality of static mapping control points in the target local region are planned, wherein the control points are static scanning points; the laser radar is controlled to move to each control point and perform static scanning at each control point to obtain a target local map of the target local region; the target local map is spliced on the global map to obtain a target three-dimensional map; the plurality of static mapping control points in the target local region are planned, including: a horizontal area of an effective scanning area of the laser radar is obtained; a plurality of control points are determined according to a preset overlap range and the horizontal area, with the goal of maximizing the coverage of the horizontal area of the target local region and minimizing the number of control points, wherein the overlap in the preset overlap range is the overlap between the horizontal areas of the effective scanning area of the laser radar.

2. The method of claim 1, wherein, the static scanning at each control point to obtain the target local map of the target local region includes: the laser radar is controlled to scan at different angles at each control point to obtain point clouds corresponding to different angles; for each control point, the point clouds corresponding to different angles of the control point are spliced to obtain a point cloud corresponding to the control point; the point clouds corresponding to different control points are spliced to obtain the target local map.

3. The method of claim 2, wherein, the mapping device further includes a pose estimation unit, and the splicing of the point clouds corresponding to different control points to obtain the target local map includes: for the point clouds corresponding to different control points, the point clouds corresponding to two control points are preliminarily spliced according to a relative pose estimation result between the two control points to obtain a preliminary splicing result, wherein the relative pose estimation result is calculated by the pose estimation unit, and the pose estimation unit is also used for pose estimation when the global map is obtained based on SLAM technology; the preliminary splicing result is precisely registered to obtain the target local map.

4. The method according to any one of claims 1 to 3, characterized in that, the laser radar is an omnidirectional laser radar, the mapping device further includes an omnidirectional fisheye camera, the global map is a colored map, and the dynamic scanning by using the laser radar to obtain the global map of the target region includes: point clouds are obtained by dynamic scanning by using the laser radar, and the fisheye camera is controlled to capture images to obtain a first environment image during the dynamic scanning by using the laser radar; based on the first environment image, the point clouds obtained during the dynamic scanning are colored according to the calibrated intrinsic parameters of the fisheye camera and the extrinsic parameters of the laser radar relative to the fisheye camera to obtain a global map with color information.

5. The method according to any one of claims 1 to 3, characterized in that, The laser radar is an omnidirectional laser radar, the mapping device further comprises an omnidirectional fisheye camera, the target local map is a colored map, and the static scanning at each control point to obtain the target local map of the target local area comprises: performing static scanning at each control point by using the laser radar to obtain a point cloud, and controlling the fisheye camera to perform image acquisition when performing static scanning at each control point by using the laser radar to obtain a second environment image; based on the second environment image, coloring the point cloud obtained in the static scanning process according to the calibrated intrinsic parameters of the fisheye camera and the extrinsic parameters of the laser radar relative to the fisheye camera, to obtain a target local map with color information.

6. The method according to any one of claims 1 to 3, characterized in that, The method further comprises: analyzing the quality of the target three-dimensional map to obtain a map quality.

7. The method of claim 6, wherein, When the analysis result obtained includes no, it is determined that the map quality is unqualified, wherein the analysis is performed by at least one of the following ways: calculating a first normal vector of a wall surface in the target three-dimensional map and a second normal vector of a ground surface in the target three-dimensional map, and determining whether the first normal vector and the second normal vector are perpendicular; determining whether third normal vectors at different positions of a wall surface with a length satisfying a first preset length requirement in the target three-dimensional map are parallel, and / or determining whether fourth normal vectors at different positions of a ground surface with a length satisfying a second preset length requirement in the target three-dimensional map are parallel; determining whether the coverage of the selected control points on the target local area satisfies a preset requirement.

8. A three-dimensional map creating apparatus characterized by comprising: The device is applied to a mapping device, and the mapping device comprises a non-repetitive scanning laser radar, and the device comprises: a first map creation module configured to perform dynamic scanning by using the laser radar based on a SLAM technology to obtain a global map of a target area; a region determination module configured to determine a target local area in the global map; a planning module configured to plan a plurality of static mapping control points in the target local area, wherein the control points are static scanning points; a second creation module configured to control the laser radar to move to each of the control points and perform static scanning at each of the control points to obtain a target local map of the target local area; a processing module configured to splice the target local map on the global map to obtain a target three-dimensional map; the planning of the plurality of static mapping control points in the target local area comprises: obtaining a horizontal area of an effective scanning area of the laser radar; determining a plurality of control points according to a preset overlap range and the horizontal area, with the goal of maximizing the coverage of the horizontal area of the target local area and minimizing the number of control points, wherein the overlap in the preset overlap range is the overlap between the horizontal areas of the effective scanning area of the laser radar.

9. A mapping device characterized by A mobile platform, a non-repeating scanning lidar, a processor and a memory, the mobile platform for moving the lidar, the lidar for obtaining a point cloud, the memory storing machine executable instructions executable by the processor, the processor executable to implement the three-dimensional map creation method of any one of claims 1-7.

Images