Point cloud map increment dynamic updating method and device for environment three-dimensional modeling

Abstract

The invention discloses a point cloud map increment dynamic updating method and device for environment three-dimensional modeling, and the method comprises the following steps: obtaining front-end data, and constructing an environment map; key frame features of the environment map are extracted, loopback detection is carried out, and a sensor pose map is obtained; optimizing the sensor pose image, and correcting the pose of the key frame; the updated key frame of the sensor pose image is divided into a historical key frame and a new key frame, and after local point cloud transient updating, global point cloud transient updating and redundant historical key frame deletion processing are carried out in sequence, historical key frame data in which outdated point cloud and redundant key frames are deleted are finally output. The method has the beneficial effects that the system efficiency is improved, the storage space is saved, the map always reflects the latest state, and the application range is wide.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional modeling, and in particular to a method and device for dynamically updating a point cloud map for three-dimensional environmental modeling. Background Art

[0002] In the field of robotics, simultaneous localization and mapping (SLAM) technology is a core approach for achieving autonomous navigation and environmental perception. Traditional 3D Lidar SLAM typically assumes a static environment and lacks the ability to adapt to dynamic environments in real time, making it incapable of meeting the long-term operational needs of robots. To address the issue of map freshness, map update technology has been introduced. This technology continuously collects new 3D point cloud data from the environment and updates the original map data, maintaining consistency between the map and the real physical environment.

[0003] Existing map updates are divided into full map updates and incremental map updates. Full map update algorithms rely on periodic collection and reconstruction of the entire map area, which is inefficient, labor-intensive, and wasteful of computing resources. Existing incremental update methods typically perform pose convex optimization based on newly collected point cloud data combined with historical maps, and then directly overlay the newly collected updated point cloud on the historical map without processing the historical point cloud map. This results in the accumulation of a large number of outdated points in the point cloud map, redundant keyframes in the pose graph, reduced positioning accuracy, poor point cloud matching robustness, waste of map data storage resources, and reduced online positioning calculation efficiency. Therefore, an efficient and intelligent long-term incremental map update method is urgently needed. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for dynamically updating point cloud maps for three-dimensional environmental modeling to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: In a first aspect, an embodiment of the present invention provides a method for dynamically updating a point cloud map of a three-dimensional modeled environment, comprising the following steps: Obtain front-end data and build an environment map; Extract key frame features of the environment map, perform loop detection, and obtain the sensor pose graph; Optimize the sensor pose graph and correct the keyframe pose; The key frames after the sensor pose graph is updated are divided into historical key frames and newly created key frames. After the local point cloud transient update, global point cloud transient update, and redundant historical key frame deletion processes are carried out in sequence, the historical key frame data with outdated point clouds and redundant key frames deleted is finally output.

[0006] Preferably, the front-end data is acquired through IMU, 3D laser data, and GNSS sensor.

[0007] Preferably, the specific process of the local point cloud transient update processing is: Each laser point scan in a laser frame is regarded as a scanning ray; The dynamic attribute probability relative distance around each laser point is distributed according to a one-dimensional Gaussian distribution; The dynamic attribute probability of the laser point is updated using the Bayesian formula.

[0008] Preferably, the global point cloud transient update processing process is: Obtain high dynamic point cloud data with transient updates of local point clouds; Eliminate outdated point cloud data with a long span, including global transient updates and recall of occluded points; Determine whether there is a new local point cloud near the historical key frame point through the nearest neighbor search difference, and remove outdated points in a probabilistic update manner; At the same time, by constructing a virtual ray back-projection into the newly created keyframe, points that were mistakenly deleted due to occlusion are detected and recalled.

[0009] Preferably, the specific process of deleting redundant historical key frames is: Calculate the overlap rate between historical keyframes and newly created keyframes, and preliminarily screen out redundant keyframe nodes to be deleted; According to the different situations of nodes and edges, the deletion situations are divided into four categories. By deleting old nodes and finding new replacement nodes simultaneously, the complexity of the overall pose graph is reduced.

[0010] Preferably, the four types of deletion situations are: Delete the adjacent edges of the historical nodes; Delete historical node loop edges; Delete the history and edges related to the new node; Delete historical isolated node edges.

[0011] In a second aspect, the present invention provides a device for dynamically updating a point cloud map of a three-dimensional environment model, configured to implement the method described in any one of the above embodiments, comprising: Get the module, get the front-end data, and build the environment map; The extraction module extracts key frame features of the environment map, performs loop detection, and obtains the sensor pose graph; Optimization module, which optimizes the sensor pose graph and corrects the key frame pose; The module is loaded to divide the key frames after the sensor pose graph is updated into historical key frames and newly created key frames. After the local point cloud transient update, global point cloud transient update, and redundant historical key frames deletion processes are carried out in sequence, the historical key frame data with outdated point clouds and redundant key frames deleted is finally output.

[0012] In a third aspect, the present invention provides an electronic device comprising at least one processor, wherein the processor is communicatively connected to at least one memory, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method described in any of the above embodiments.

[0013] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method described in any of the above embodiments when executed.

[0014] Compared with the prior art, the present invention has the following significant advantages: 1. Improved update efficiency: Incremental updates avoid global reconstruction. Combined with the back-end graph optimization module's precise correction of keyframe poses and the historical map loading module's enhanced matching robustness, map update time is shortened by %, significantly improving system efficiency and reducing the repeated calculations and resource waste caused by full updates.

[0015] 2. Optimize resource utilization: By deleting redundant historical keyframes, the complexity of the pose graph is effectively reduced, reducing computing resource consumption. At the same time, by eliminating outdated point clouds, storage space is saved, achieving efficient utilization of storage resources.

[0016] 3. Enhanced dynamic adaptability and matching robustness: The positioning module's loop detection combined with back-end graph optimization ensures the accuracy of keyframe poses, enabling the map to respond to environmental changes in real time. The historical map loading module further improves the robustness of point cloud matching by loading historical slice maps and fusing them with current data, ensuring that the map always reflects the actual environmental state and providing reliable support for the robot's long-term positioning and mapping.

[0017] 4. Wide range of applicable scenarios: The above technical solution is easy to integrate and can be adapted to a variety of equipment such as indoor service robots and outdoor unmanned vehicles, and has strong versatility. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] FIG1 is an overall flow chart of a method according to an embodiment of the present invention; FIG2 is a functional module diagram of a device according to an embodiment of the present invention; Figure 3 is a diagram of the technical solution for updating historical map data; Figure 4 is a schematic diagram of the principle of the local transient loading module; Figure 5 is a technical diagram of the global point cloud transient loading module; Figure 6 shows the recall scheme for occluded points that were mistakenly deleted; FIG7 is a diagram showing four types of deletion scenarios of the redundant key frame deletion module; Figure 8 This is a module diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] See also Figure 1 , the present invention provides a technical solution: A method for incremental dynamic updating of a point cloud map for three-dimensional environmental modeling comprises the following steps: S100. Obtain front-end data and build an environment map; S200. Extract key frame features of the environment map, perform loop detection, and obtain the sensor pose graph; S300. Optimize the sensor pose graph and correct the key frame pose; S400. The key frames after the sensor pose graph is updated are divided into historical key frames and newly created key frames. After the local point cloud transient update, global point cloud transient update, and redundant historical key frame deletion processes are carried out in sequence, the historical key frame data with outdated point clouds and redundant key frames deleted is finally output.

[0021] In an embodiment of the present invention, Figure 2 As shown, in order to synchronize the implementation of the above steps, a specific embodiment of the present invention also provides a point cloud map incremental dynamic update device for three-dimensional modeling of an environment, which includes an acquisition module 100, an extraction module 200, an optimization module 300 and a loading module 400.

[0022] The method of this embodiment can be used for robots to update the robot map in real time, without relying on periodic map reconstruction as in traditional technologies. This will result in the accumulation of a large number of "outdated points" in the point cloud map (such as temporary construction areas, seasonally changing leaves, and other historical points that do not exist in the new data), and the presence of redundant keyframes in the pose graph, leading to problems such as decreased positioning accuracy, poor robustness of point cloud matching, waste of storage resources, and reduced efficiency of online positioning calculations. This enables robots to achieve efficient and intelligent long-term incremental map updates.

[0023] It should be noted that the “robot” in this application is broad in scope and covers all types of automated equipment with autonomous navigation, environmental perception, and map building / updating capabilities.

[0024] In an embodiment of the present invention, the front-end data of step S100 receives newly collected IMU, 3D laser data, and GNSS sensor data through the acquisition module 100, estimates the current robot posture in real time through the state estimator, and uses the online voxel map to build a surrounding environment map.

[0025] In an embodiment of the present invention, step S200 extracts key frame features through the extraction module 200 and performs loop detection with the historical map.

[0026] In an embodiment of the present invention, the process of extracting key frame features is as follows: The system receives real-time data from 3D LiDAR (10 frames / second), IMU (100Hz), and GNSS (10Hz), and estimates the robot's real-time pose through point cloud registration and multi-sensor fusion. When the robot moves 0.5 meters or experiences a significant angular change in its pose, a keyframe is generated. Features such as edge points and planar points are extracted from this keyframe for subsequent matching.

[0027] In an embodiment of the present invention, key frame features include geometric features of the point cloud (such as edges and planes) and global descriptors (used for matching with historical maps).

[0028] In an embodiment of the present invention, the loop detection process includes: Extract the global descriptor from the current keyframe (used to characterize the global features of the point cloud of this frame); The descriptor is matched with the features of the key frames in the historical map, and the matching threshold is set to 0.8; If a loop is detected, that is, the current key frame successfully matches a historical key frame, the loop edge is recorded.

[0029] In an embodiment of the present invention, step S300 specifically includes: The pose graph is constructed using keyframes as "nodes" and two types of "edges" as constraints. These two types of edges include odometry edges and loop edges. Specifically, odometry edges are based on the relative poses between adjacent keyframes calculated by the front-end odometry module, forming constraint relationships between adjacent nodes. Specifically, loop edges are obtained by loop detection in the extraction module 200 and represent the pose constraint relationship between the current keyframe and previous keyframes.

[0030] On this basis, the optimization module 300 uses the nonlinear least squares method to perform global optimization on the constructed pose graph. By minimizing the errors of the constraint edges in the pose graph, the cumulative errors caused by sensor drift, etc. are eliminated, and finally the corrected key frame pose is output to achieve accurate correction of the global pose, and then the map is calibrated, providing a reliable pose association basis for subsequent historical map data updates.

[0031] In an embodiment of the present invention, the loading module 400 is composed of a local point cloud transient loading module, a global point cloud transient loading module, and a redundant historical keyframe deletion module. In a specific embodiment, the local point cloud transient loading module, the global point cloud transient loading module, and the redundant historical keyframe deletion module can be integrated into a dynamic slice map loader. During the point cloud data loading process, the map calibrated by the optimization module 300 is corrected and then loaded into the acquisition module 100.

[0032] In the embodiment of the present invention, the specific processing process of step S400 is: like Figure 3 As shown in the figure, the corrected key frame pose is divided into historical key frames and newly created key frames, and enters the local point cloud transient loading module, the global point cloud transient loading module, and the redundant historical key frame deletion module in sequence, and finally outputs the historical key frame data with outdated point clouds and redundant key frames deleted.

[0033] Specifically, the "transient" in this application refers to the dynamic attribute probability of the point cloud, while the "local transient" refers to the dynamic probability of the point cloud in a short period of time (during a single data collection process), and the "global transient" refers to the dynamic probability of the point cloud in a longer time span. The local transient update principle is as follows: Figure 4 shown.

[0034] Specifically, each laser point scan in a laser frame is considered a scanning ray, where the black dot is the laser scanning origin, the white dot is the free point f sampled by the ray as it passes through the free space along the way, and the gray dot is the occupied point o where the ray finally hits the object surface. After aggregating the key frames into a local map, the ɛl around each observation point oi or free point fi follows a one-dimensional Gaussian distribution relative to the distance (x): Where σ is the standard deviation, the initial value of ɛl is 0.5, and each point in Oi or Fi retrieves the k nearest neighbor points and updates the ɛl of these neighboring points using the Bayesian formula: The global point cloud transient loading module technical solution is as follows: Figure 5 The global point cloud transient update refers to the process of filtering out the highly dynamic point cloud through local transients in the first step and then removing the outdated point cloud in the previous time span.

[0035] In an embodiment of the present invention, in a global transient update, local point cloud maps are constructed for historical key frames and newly created key frames respectively, and the differences in neighbor searches are used to determine which historical key frame points have no newly created local point clouds nearby, thereby gradually updating through a probabilistic update method, and ultimately preliminarily treating points greater than a set threshold as outdated points.

[0036] There is an anomaly here: when creating a new keyframe map, due to some obstructions in the current scan, some historical observation data is in the blind spot of the current new scan. These historical observation points will also be regarded as outdated points, but in fact these points are not outdated points.

[0037] In the embodiment of the present invention, for the abnormal phenomenon processing method, such as Figure 6 As shown in the figure, a 3D voxel map is constructed based on the historical keyframes and the new keyframe point cloud. When an outdated point is detected, a virtual ray is constructed and back-projected into the new keyframe. When it is detected that an occupied voxel in the new frame blocks the virtual ray, the outdated point can be determined to be an erroneously deleted point due to occlusion.

[0038] At this point, dynamic updates at the point cloud level have been completed. Finally, the redundant keyframe removal module removes redundant keyframes. Because certain areas are frequently updated in actual map updates, the pose optimization graph can contain a large number of redundant nodes over time, increasing the time required for graph optimization and leading to the accumulation of duplicate point clouds.

[0039] In an embodiment of the present invention, a solution for removing redundant key frames includes: Calculate the overlap ratio between historical keyframes and newly created keyframes to preliminarily screen out redundant keyframe nodes to be deleted. The overlap ratio between historical keyframes and newly created keyframes is calculated based on the overlap ratio of the voxel grid. The formula is as follows: Initially screen out redundant keyframe nodes to be deleted. Based on the differences between the nodes and edges to be deleted, there are four types of deletion situations, see Figure 7 Shown, including: Delete the adjacent edges of the historical nodes; Delete historical node loop edges; Delete the history and edges related to the new node; Delete historical isolated node edges.

[0040] By deleting old nodes and simultaneously finding new replacement nodes, redundant historical keyframes are removed after each update, reducing the complexity of the overall pose graph. The loading module 400 loads the point cloud data after deleting redundant historical keyframes into the acquisition module 100 to complete the incremental update of the map.

[0041] In a specific embodiment of the present invention, the robot is equipped with an incremental dynamic update device for a point cloud map of a three-dimensional modeled environment provided by an embodiment of the present invention. During the operation of the robot, the acquisition module receives real-time data collected by 3D lidar (10 frames / second), IMU (100Hz) and GNSS (10Hz), and generates a key frame every 0.5 meters or significant angle change by locating the robot's walking distance.

[0042] In a specific embodiment of the present invention, the extraction module 200 extracts the global descriptor of the key frame and matches it with the key frame features of the historical map, and the matching threshold is set to 0.8.

[0043] In a specific embodiment of the present invention, the optimization module 300 loads a local point cloud slice with a radius of 20 meters according to the current posture, and converts it into a voxel map with a voxel size of 0.1 meters.

[0044] In a specific embodiment of the present invention, in the local point cloud transient loading module, 10 nearest neighbor points are retrieved for each point, the search radius is 0.2 meters, the initial value of the dynamic probability is 0.5, the standard deviation is 0.1 meters, and points with a dynamic probability greater than 0.9 are marked as high dynamic points.

[0045] In a specific embodiment of the present invention, in the global point cloud transient loading module, the voxel size of the voxel map constructed by the historical and newly created keyframe point clouds is 0.2 meters, the five nearest neighbor points are retrieved for the historical point cloud points, the search half distance is 0.5 meters, the initial value of the global dynamic probability is 0.5, the standard deviation is 0.2 meters, and the points with a probability greater than 0.95 are preliminarily marked as outdated points.

[0046] In a specific embodiment of the present invention, in the redundant historical key frame deletion module, the calculation formula for the overlap ratio is overlap_ratio, and the overlap ratio threshold is 0.85.

[0047] In a specific embodiment of the present invention, steps S100 to S300 constitute the incremental mapping phase, and step S400 constitutes the historical map data update phase. The incremental mapping and historical map data update phases begin by loading a historical 3D point cloud map and pose graph from local storage or a cloud server at system startup. For the first run, an empty map is initialized. The vehicle is equipped with a 3D lidar, IMU, and GNSS to collect environmental data in real time. The lidar generates 10 frames of point cloud per second, the IMU provides 100Hz high-frequency gesture feedback, and the GNSS provides a 10Hz global position reference.

[0048] In a specific embodiment of the present invention, in actual testing, the method was run in a 200m×250m office park for 6 months, completing a total of six rounds of updates, processing 500GB of data, and reducing storage space from 1.2TB to 780GB, saving about 35%. The positioning accuracy was maintained within 0.1 meter, and the robustness was significantly improved in dynamic environments such as construction areas and vegetation changes.

[0049] See also Figure 8 , Figure 8A schematic diagram of an electronic device 20 that can implement an embodiment of the present invention is shown, and the electronic device is intended to represent various forms of control devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0050] The electronic device 20 includes at least one processor 21 and a memory, such as a read-only memory (ROM) 22 and a random access memory (RAM) 23, communicatively connected to the at least one processor 21. The memory stores a computer program that can be executed by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 22 or the computer program loaded from the storage unit 28 into the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 20 can also be stored in the RAM 23. The processor 21, ROM 22, and RAM 23 are interconnected via a bus 24. An input / output (I / O) interface 25 is also connected to the bus 24.

[0051] Multiple components in the electronic device 20 are connected to the I / O interface 25, including: an input unit 26, such as a keyboard, a mouse, etc.; an output unit 27, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 29, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 29 allows the electronic device 20 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0052] The processor 21 may be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the processor 21 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 21 performs the various methods and processes described above.

[0053] In some embodiments, the methods of the above embodiments may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as the storage unit 28. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 20 via the ROM 22 and / or the communication unit 29. When the computer program is loaded into the RAM 23 and executed by the processor 21, one or more steps of the method described above may be performed. Alternatively, in other embodiments, the processor 21 may be configured to perform the methods of the above embodiments in any other appropriate manner (e.g., by means of firmware).

[0054] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0055] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0056] In the context of the present invention, computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage medium can include but is not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage medium can be a machine-readable signal medium. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0057] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0058] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0059] A computing system may include clients and servers. The clients and servers are generally remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0060] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for incremental dynamic updating of point cloud maps for three-dimensional environmental modeling, characterized in that: The following steps are involved: Obtain front-end data and build an environment map; Extract key frame features of the environment map, perform loop detection, and obtain the sensor pose graph; Optimize the sensor pose graph and correct the keyframe pose; The key frames after the sensor pose graph is updated are divided into historical key frames and newly created key frames. After the local point cloud transient update, global point cloud transient update, and redundant historical key frame deletion processes are carried out in sequence, the historical key frame data with outdated point clouds and redundant key frames deleted is finally output.

2. The method for incremental dynamic updating of a point cloud map for three-dimensional environmental modeling according to claim 1, characterized in that: The front-end data is obtained through IMU, 3D laser data, and GNSS sensors.

3. The method for incremental dynamic updating of a point cloud map for three-dimensional environmental modeling according to claim 2, characterized in that: The specific process of the local point cloud transient update processing is as follows: Each laser point scan in a laser frame is regarded as a scanning ray; The dynamic attribute probability relative distance around each laser point is distributed according to a one-dimensional Gaussian distribution; The dynamic attribute probability of the laser point is updated using the Bayesian formula.

4. The method for incremental dynamic updating of a point cloud map for three-dimensional environmental modeling according to claim 3, characterized in that: The global point cloud transient update processing process is as follows: Obtain high dynamic point cloud data with transient updates of local point clouds; Eliminate outdated point cloud data with a long span, including global transient updates and recall of occluded points; Determine whether there is a new local point cloud near the historical key frame point through the nearest neighbor search difference, and remove outdated points in a probabilistic update manner; At the same time, by constructing a virtual ray back-projection into the newly created keyframe, points that were mistakenly deleted due to occlusion are detected and recalled.

5. The method for incremental dynamic updating of a point cloud map for three-dimensional environmental modeling according to claim 4, characterized in that: The specific process of deleting redundant historical key frames is as follows: Calculate the overlap rate between historical keyframes and newly created keyframes, and preliminarily screen out redundant keyframe nodes to be deleted; According to the different situations of nodes and edges, the deletion situations are divided into four categories. By deleting old nodes and finding new replacement nodes simultaneously, the complexity of the overall pose graph is reduced.

6. The method for incremental dynamic updating of a point cloud map for three-dimensional environmental modeling according to claim 5, characterized in that: The four types of deletion situations are: Delete the adjacent edges of the historical nodes; Delete historical node loop edges; Delete the history and edges related to the new node; Delete historical isolated node edges.

7. A device for dynamically updating point cloud maps for three-dimensional environmental modeling, used to implement the method according to any one of claims 1 to 6, characterized in that: include: Get the module, get the front-end data, and build the environment map; The extraction module extracts key frame features of the environment map, performs loop detection, and obtains the sensor pose graph; Optimization module, which optimizes the sensor pose graph and corrects the key frame pose; The module is loaded to divide the key frames after the sensor pose graph is updated into historical key frames and newly created key frames. After the local point cloud transient update, global point cloud transient update, and redundant historical key frames deletion processes are carried out in sequence, the historical key frame data with outdated point clouds and redundant key frames deleted is finally output.

8. An electronic device, characterized in that: The method comprises at least one processor, wherein the processor is communicatively connected to at least one memory, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method according to any one of claims 1 to 6 when the instructions are executed.

Images