Point cloud processing method based on laser radar and robot navigation method and device

Through dual echo and peak detection, fitting and classification technology, the problem of lidar being difficult to obtain accurate point clouds in the presence of reflective objects is solved, and more accurate point cloud data processing and robot navigation are achieved.

CN118918375APending Publication Date: 2024-11-08SHANGHAI TECH UNIV
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202410989104.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-23
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

In the prior art, when there is reflection phenomenon in lidar sensors, it is difficult to obtain accurate spatial point cloud data, especially in the identification of transparent glass areas.

Method used

Through double echo and peak detection, a potential reflectable plane point cloud is fitted, and a global reflectable plane map is constructed to classify the point cloud data to obtain multiple types of point cloud data.

Benefits of technology

It improves the accuracy and accuracy of point cloud data, ensures that accurate spatial point clouds can be obtained in the presence of reflective objects, and reduces safety risks in robot navigation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118918375A_ABST
    Figure CN118918375A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of point cloud processing, and discloses a point cloud processing method based on a laser radar, and a robot navigation method and device, and the method comprises the steps: obtaining point cloud data collected by the laser radar in a space; performing dual echo and peak detection according to the point cloud data so as to fit a potential reflective plane point cloud from the point cloud data; constructing and updating a global reflexible plane map according to the multi-frame potential reflexible plane; and performing point cloud classification on the point cloud data according to the global reflexible plane map to obtain multiple types of point cloud data. According to the invention, dual echo and peak detection is carried out according to the point cloud data, so that the potential reflective plane point cloud can be obtained more comprehensively; the global reflexible plane map is obtained through multi-frame reflexible plane fitting, so that the generation accuracy of the reflexible plane map is ensured, and the accuracy of point cloud classification of the global reflexible plane map based on the point cloud data is further ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of point cloud processing technology, and in particular to a laser radar-based point cloud processing method, a robot navigation method and a device. Background Art

[0002] As one of the most common sensors, LiDAR sensors can use their optical principles to obtain reliable spatial point cloud data in most cases. However, when there are objects such as mirrors or windows in the environment, the sensor is prone to adding noise points with incorrect positions in the data due to reflection phenomena. And due to the transparency of glass, a single-frame point cloud can often only obtain information about a very small part of its plane, resulting in the laser mapping algorithm being unable to identify the transparent glass area, and the transparent glass area is not included in the recognition result. In the field of mobile robots, the inaccuracy of sensor recognition results poses certain safety risks in the subsequent navigation and path planning links.

[0003] Most existing technologies rely on lidar recognition algorithms, which only consider the recognition and mapping of glass in two dimensions, and do not remove reflection points, resulting in inaccurate point cloud recognition; or rely on multi-sensor fusion algorithms to detect glass, but their range and accuracy are limited, making it difficult to achieve precise positioning; or use artificial intelligence methods to detect glass in point cloud data, but this method requires higher computing performance.

[0004] In summary, the existing technology has the problem that it is difficult to obtain accurate spatial point clouds based on the existing laser radar in the presence of reflective objects. Summary of the invention

[0005] The present invention provides a laser radar-based point cloud processing method, a robot navigation method and device, an electronic device and a computer-readable storage medium, the main purpose of which is to solve the problem that it is difficult to obtain accurate spatial point clouds based on existing laser radars in the presence of reflective objects.

[0006] To achieve the above object, the present invention provides a point cloud processing method based on laser radar, comprising:

[0007] Obtain point cloud data collected by the laser radar in space;

[0008] Performing double echo and peak detection according to the point cloud data to achieve fitting of a potential reflective plane point cloud from the point cloud data;

[0009] Construct and update a global reflective plane map based on the potential reflective planes in multiple frames;

[0010] The point cloud data is classified according to the global reflective plane map to obtain multiple types of point cloud data.

[0011] In one embodiment of the present invention, performing double echo and peak detection according to the point cloud data to achieve fitting of a potential reflective plane point cloud from the point cloud data includes:

[0012] Determine the corresponding strongest echo point cloud and the last echo point cloud from the point cloud data obtained by the laser radar through double echo detection;

[0013] Perform peak detection according to the point cloud data to obtain a peak regular point cloud;

[0014] The point cloud data is screened according to the strongest echo point cloud, the last echo point cloud and the peak regularity point cloud to obtain potential reflective plane points.

[0015] In one embodiment of the present invention, the peak value detection is performed according to the point cloud data to obtain the peak regular point cloud, including:

[0016] Classify the point cloud data according to the pulse intensity to determine multiple types of intensity point clouds;

[0017] Regularity analysis is performed on the multiple types of intensity point clouds, and peak regular point clouds are screened out from the point cloud data according to the results of the regularity analysis.

[0018] In one embodiment of the present invention, the point cloud data is screened according to the strongest echo point cloud, the last echo point cloud and the peak regularity point cloud to obtain a potential reflective plane point cloud, including:

[0019] comparing a difference between the strongest echo point cloud and the last echo point cloud with a preset echo interval threshold;

[0020] Determine a potential reflective plane point cloud according to the comparison result and the strongest echo point cloud;

[0021] The peak regular point cloud identified by the laser radar under different posture information is used as a potential reflective plane point cloud.

[0022] In one embodiment of the present invention, the step of constructing and updating a global reflective plane map based on multiple frames of potential reflective planes includes:

[0023] Performing frame classification according to the potential reflective plane point cloud to obtain a set of potential reflective plane point clouds under multiple frames;

[0024] Performing plane fitting according to the reflective plane point cloud in the potential reflective plane point cloud set to obtain a reflective plane map;

[0025] Performing spatial coordinate conversion according to the pose information corresponding to the reflective plane map corresponding to each frame to obtain a reflective plane map in the same coordinate system;

[0026] The reflective plane map corresponding to each frame is updated and optimized to obtain a global reflective plane map.

[0027] In one embodiment of the present invention, performing plane fitting according to the reflective plane point cloud in the potential reflective plane point cloud set to obtain a reflective plane map includes:

[0028] Selecting a target fitting point cloud and a point cloud to be screened from the potential reflective plane point cloud set;

[0029] Generate a fitting plane according to the target fitting point cloud;

[0030] Calculating the distance between the point cloud to be screened and the fitting plane;

[0031] Update the to-be-screened point cloud and the target fitting point cloud according to the distance;

[0032] The fitting plane is updated according to the updated target fitting point, and the step of calculating the distance between the point cloud to be screened and the fitting plane is returned to until the updating of the fitting plane meets the preset updating condition, so as to obtain a reflective plane map.

[0033] In one embodiment of the present invention, the point cloud data is classified according to the global reflective plane map to obtain multiple types of point cloud data, including:

[0034] Calculating a spatial relationship between the point cloud data and the global reflective plane map;

[0035] The categories corresponding to the point cloud data are determined according to the spatial relationship to obtain multi-category point cloud data.

[0036] In order to solve the above problems, the present invention also provides a point cloud processing device based on laser radar, the device comprising:

[0037] The data acquisition module is used to obtain the point cloud data collected by the laser radar in space;

[0038] A potential reflective plane point cloud screening module, used for performing double echo and peak detection according to the point cloud data, so as to fit a potential reflective plane point cloud from the point cloud data;

[0039] A global reflective plane map construction module, used to construct and update a global reflective plane map according to the potential reflective planes in multiple frames;

[0040] The point cloud classification module is used to classify the point cloud data according to the global reflective plane map to obtain multiple types of point cloud data.

[0041] In order to solve the above problems, the present invention also provides a robot navigation method, the method comprising:

[0042] Acquire multiple types of point cloud data obtained according to the laser radar-based point cloud processing method described above;

[0043] Point cloud data collected by the laser radar in space are screened according to the multiple types of point cloud data to obtain a planning point cloud;

[0044] A planned path is generated according to the planned point cloud, and the robot is controlled to navigate according to the planned path.

[0045] In order to solve the above problems, the present invention further provides a robot navigation device, the device comprising:

[0046] A point cloud data acquisition module, which acquires multiple types of point cloud data obtained according to the above-mentioned laser radar-based point cloud processing method;

[0047] A planning point cloud screening module is used to screen the point cloud data collected by the laser radar in space according to the multiple types of point cloud data to obtain a planning point cloud;

[0048] The path planning module is used to generate a planned path according to the planned point cloud and control the robot to navigate according to the planned path.

[0049] In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising:

[0050] at least one processor; and,

[0051] a memory communicatively connected to the at least one processor; wherein,

[0052] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the steps of the above-mentioned lidar-based point cloud processing method.

[0053] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one computer program is stored. The at least one computer program is executed by a processor in an electronic device to implement the steps of the above-mentioned laser radar-based point cloud processing method.

[0054] The present invention performs double echo and peak detection based on point cloud data, and can obtain potential reflective plane point clouds more comprehensively; obtains a global reflective plane map by fitting multiple frames of reflective planes, and ensures the accuracy of the generation of the reflective plane map, thereby ensuring the accuracy of point cloud classification based on point cloud data and the global reflective plane map. Therefore, the laser radar-based point cloud processing method, robot navigation method and device, electronic device, and computer-readable storage medium proposed in the present invention can solve the problem that it is difficult to obtain accurate spatial point clouds based on existing laser radars in the presence of reflective objects. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A schematic diagram of a process flow of a laser radar-based point cloud processing method provided by an embodiment of the present invention;

[0056] Figure 2 A functional module diagram of a laser radar-based point cloud processing device provided in one embodiment of the present invention;

[0057] Figure 3 A schematic diagram of a flow chart of a robot navigation method provided by an embodiment of the present invention;

[0058] Figure 4 A functional module diagram of a robot navigation device provided by an embodiment of the present invention;

[0059] Figure 5 A schematic diagram of the structure of an electronic device for implementing the laser radar-based point cloud processing method provided in one embodiment of the present invention.

[0060] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0061] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0062] The embodiment of the present application provides a laser radar-based point cloud processing method. The execution subject of the laser radar-based point cloud processing method includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the laser radar-based point cloud processing method can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0063] Reference Figure 1FIG. 1 is a flow chart of a laser radar-based point cloud processing method according to an embodiment of the present invention. In this embodiment, the laser radar-based point cloud processing method includes:

[0064] S11. Obtain point cloud data collected by the laser radar in space.

[0065] In the embodiment of the present invention, the laser radar emits laser pulses to the target and receives echoes, calculates the distance of the target based on the round-trip time of the laser pulse and the wavelength of the laser, and cooperates with the GPS (Global Positioning System) and IMU (Inertial Measurement Unit) system to obtain its position and solid angular position, thereby collecting point cloud data of the object being measured and its surrounding environment in space.

[0066] Among them, when there are objects such as mirrors or windows in the space, the laser emitted by the lidar toward reflective surfaces such as mirrors may have return pulses based on the glass, or return pulses that pass through the glass and hit objects behind the glass, or return pulses based on reflections from the glass and hit other objects.

[0067] Therefore, the point cloud data collected by the lidar includes normal point clouds such as walls and objects in the space, point clouds of reflective planes such as glass and mirrors, and reflected point clouds formed by reflection or penetration of glass. The mirror projection of the reflective plane may form a reflected point cloud, but the corresponding point cloud does not exist in the actual indoor environment, which causes the sensor to add noise points with incorrect positions to the point cloud obtained in the space with reflective planes, making the identified point cloud data inaccurate.

[0068] S12. Perform double echo and peak detection according to the point cloud data to screen out potential reflective plane points from the point cloud data.

[0069] In the embodiments of the present invention, the laser radar system generally includes a laser transmitter and one or more receivers, which can capture the reflected laser from the target. In a typical double echo configuration, the laser radar system may have two independent receivers, each with its own time recording system to calculate the time difference of the received reflected signals.

[0070] In complex or highly reflective environments, the dual echo mode of the lidar can distinguish and process two different reflection signals from the same target object, which may be caused by multiple reflections of the radar on the surface of the object or because the surface of the object has multiple different reflective surfaces. In highly reflective environments, the dual echo mode can help distinguish multiple reflection signals, helping to distinguish and identify objects with similar reflective characteristics.

[0071] The principle of the double echo mode is to distinguish two echo signals based on the difference in time. Since the laser pulse emitted by the lidar has a very short duration (usually in the nanosecond level), the time difference between the two return signals can be quickly measured, thereby calculating the distance difference between them, which can be used to improve the accuracy of target recognition.

[0072] Specifically, the laser radar in the present invention selects a dual echo mode to send lasers, and can receive the strongest echo and the last echo returned; wherein the strongest echo refers to the returned laser pulse with the highest intensity, and the last echo refers to the laser pulse that returns to the sensor the latest and takes the longest time.

[0073] In an actual application scenario of the present invention, if the laser emitted by the laser radar in certain directions hits a normal wall, the strongest echo returned is roughly the same as the last echo; if the laser emitted by the laser radar in certain directions hits a reflective plane, the strongest echo returned may be a pulse that directly hits the reflective plane and returns, or a pulse that passes through the reflective plane and hits an object behind and returns, and the last echo returned may be a pulse that is reflected by the reflective plane and then hits the object again, which results in a certain gap between the strongest echo returned and the last echo.

[0074] In an embodiment of the present invention, for a reflective plane, the intensity of the pulse returned by the laser radar when emitting lasers in different directions varies depending on the object hit; based on the different angles between the laser emission angle and the plane, the returned pulse will show a regular halo-type intensity variation pattern; in a certain direction, if the laser hits the reflective plane at a vertical angle of 90°, the higher the intensity of the returned pulse will be; if the laser hits the reflective plane at an angle inclined at a certain angle, the intensity of the returned pulse will be significantly lower than that of the laser hitting the reflective plane at a vertical angle of 90°, and thus the point cloud belonging to the reflective plane can be determined by the intensity variation pattern of the pulse returned by the reflective plane for lasers emitted at different angles.

[0075] In an embodiment of the present invention, performing double echo and peak detection according to the point cloud data to achieve fitting of a potential reflective plane point cloud from the point cloud data includes:

[0076] Determine the corresponding strongest echo point cloud and the last echo point cloud from the point cloud data obtained by the laser radar through double echo detection;

[0077] Perform peak detection according to the point cloud data to obtain a peak regular point cloud;

[0078] The point cloud data is screened according to the strongest echo point cloud, the last echo point cloud and the peak regularity point cloud to obtain potential reflective plane points.

[0079] Among them, the strongest echo point cloud and the last echo point cloud are the point clouds corresponding to the strongest echo and the last echo received by the laser radar.

[0080] In detail, the peak detection is performed according to the point cloud data to obtain the peak regular point cloud, including:

[0081] Classify the point cloud data according to the pulse intensity to determine multiple types of intensity point clouds;

[0082] Regularity analysis is performed on the multiple types of intensity point clouds, and peak regular point clouds are screened out from the point cloud data according to the results of the regularity analysis.

[0083] Among them, since the laser intensity returned by the laser radar when it hits different objects is different, the point cloud data can be classified according to the laser intensity; for the classified point cloud data, the point cloud corresponding to the intensity change law of the echo returned by the reflective plane is determined, and all point clouds that conform to the intensity change law are regarded as peak regularity point clouds.

[0084] Further, the point cloud data is screened according to the strongest echo point cloud, the last echo point cloud and the peak regularity point cloud to obtain a potential reflective plane point cloud, including:

[0085] comparing a difference between the strongest echo point cloud and the last echo point cloud with a preset echo interval threshold;

[0086] Determine a potential reflective plane point cloud according to the comparison result and the strongest echo point cloud;

[0087] The peak regular point cloud identified by the laser radar under different posture information is used as a potential reflective plane point cloud.

[0088] Among them, if the difference between the strongest echo point cloud and the last echo point cloud is less than the echo interval threshold, it means that the area hit by the laser emitted by the lidar in the corresponding direction is a normal object area; if the difference between the strongest echo point cloud and the last echo point cloud is greater than or equal to the echo interval threshold, it means that the area hit by the laser emitted by the lidar in the corresponding direction is a reflective plane, and the point cloud of the strongest echo returned in this direction can be used as the reflective plane point cloud.

[0089] In detail, when the laser radar collects point cloud data in space, it can rely on mobile devices such as robots to realize point cloud collection in different spatial positions, that is, the laser radar has its corresponding point cloud data in different frames and different postures. The present invention can obtain potential reflective plane point clouds by screening the point cloud data in different postures, wherein the point cloud contains three-dimensional spatial position coordinates, corresponding posture and other information.

[0090] The present invention can determine potential reflective plane point clouds in point cloud data through the two methods of double echo and peak detection, thereby improving the comprehensiveness of point cloud data screening and expanding the scope of potential reflective plane point cloud screening; wherein, potential reflective plane point clouds may include reflective plane point clouds, reflection point clouds, etc., and there may also be normal point clouds that are misidentified due to errors in double echo and peak detection. Therefore, the potential reflective plane point clouds can be plane fitted to further achieve preparatory identification of the reflection plane point clouds.

[0091] S13, constructing and updating a global reflective plane map according to the potential reflective planes in multiple frames.

[0092] In an embodiment of the present invention, since the laser radar can be located at different spatial positions to transmit the radar when collecting point clouds in space, the potential reflective plane point clouds collected by the corresponding transmission frames at each spatial position can be summarized, and the point clouds can be plane-fitted and spatially transformed according to the posture corresponding to the transmission frame to obtain a global reflective plane map.

[0093] In the embodiment of the present invention, the step of constructing and updating a global reflective plane map based on multiple frames of potential reflective planes includes:

[0094] Performing frame classification according to the potential reflective plane point cloud to obtain a set of potential reflective plane point clouds under multiple frames;

[0095] Performing plane fitting according to the reflective plane point cloud in the potential reflective plane point cloud set to obtain a reflective plane map;

[0096] Performing spatial coordinate conversion according to the pose information corresponding to the reflective plane map corresponding to each frame to obtain a reflective plane map in the same coordinate system;

[0097] The reflective plane map corresponding to each frame is updated and optimized to obtain a global reflective plane map.

[0098] Among them, by using the posture information corresponding to each frame of the reflective plane map, each frame of the reflective plane map can be transferred to the same coordinate system, which is convenient for the data unification of the subsequent global plane fitting.

[0099] In detail, due to the difference in posture information, the reflective plane map corresponding to each frame may be different; the present invention optimizes the map in a global situation through the reflective plane maps corresponding to all frames, taking into account the influence of the laser radar emission conditions in different postures and the influence of the collection range of different spatial positions, to ensure that the obtained global reflective plane map is more comprehensive and accurate.

[0100] Furthermore, performing plane fitting according to the reflective plane point cloud in the potential reflective plane point cloud set to obtain a reflective plane map includes:

[0101] Selecting a target fitting point cloud and a point cloud to be screened from the potential reflective plane point cloud set;

[0102] Generate a fitting plane according to the target fitting point cloud;

[0103] Calculating the distance between the point cloud to be screened and the fitting plane;

[0104] Update the to-be-screened point cloud and the target fitting point cloud according to the distance;

[0105] The fitting plane is updated according to the updated target fitting point, and the step of calculating the distance between the point cloud to be screened and the fitting plane is returned to until the updating of the fitting plane meets the preset updating condition, so as to obtain a reflective plane map.

[0106] Among them, RANSAC (Random Sample Consensus) can be used to perform point cloud plane fitting.

[0107] Specifically, three point clouds are selected from the set of potential reflective plane point clouds as target fitting point clouds, and the remaining point clouds are selected as point clouds to be screened. A plane model can be constructed based on these three target fitting point clouds; then the distances from all point clouds to be screened to the fitting plane are calculated, and the point clouds to be screened are marked according to the set distance threshold. Points less than the distance threshold can be marked as inliers, and the inliers and the initially selected target fitting point clouds are updated. The point clouds required for the fitting plane are re-determined and the plane is fitted, thereby realizing the update of the fitting plane. Finally, when the updated target fitting point cloud is greater than a preset number or when the number of iterations for updating the fitting plane reaches a preset number of iterations, the iterative update can be stopped to obtain a reflective plane map.

[0108] S14. Classifying the point cloud data according to the global reflective plane map to obtain multiple types of point cloud data.

[0109] The present invention can divide the point cloud into normal point clouds that are not affected by reflection, reflective surface point clouds from highly reflective objects such as glass or mirrors, reflected point clouds that are finally received by the sensor after mirror reflection, and transmission point clouds from obstacles on the other side of a transmissive plane such as glass.

[0110] In the embodiment of the present invention, the point cloud data is classified according to the global reflective plane map to obtain multiple types of point cloud data, including:

[0111] Calculating a spatial relationship between the point cloud data and the global reflective plane map;

[0112] The categories corresponding to the point cloud data are determined according to the spatial relationship to obtain multi-category point cloud data.

[0113] Specifically, based on the point cloud data and the laser radar, a transmission line is constructed to determine whether there is an intersection when the transmission line extends to the global reflective plane map, that is, to determine the spatial relationship between the point cloud and the global reflective plane map.

[0114] Among them, if there is an intersection between the emission line and the global reflective plane map, it is considered to belong to the reflective plane point cloud; if the point cloud is located in front of the global reflective plane map, it is judged whether it belongs to a normal point cloud based on the distance from the global reflective plane map. When the distance is greater than or equal to the preset distance threshold, it is considered to belong to a normal point cloud. When the distance is less than the preset distance threshold, it is considered to belong to a reflective plane point cloud; if the point cloud is located on the back of the global reflective plane map, it is considered to belong to a reflection point cloud or a transmission point cloud.

[0115] The present invention can remove the reflection point cloud finally received by the sensor after mirror reflection and the transmission point cloud from obstacles on the other side of a transmissive plane such as glass, or remove the reflection point cloud finally received by the sensor after mirror reflection.

[0116] The present invention performs double echo and peak detection based on point cloud data, and can obtain potential reflective plane point clouds more comprehensively; obtains a global reflective plane map by fitting multiple frames of reflective planes, and ensures the accuracy of the generation of the reflective plane map, thereby ensuring the accuracy of point cloud classification based on point cloud data and the global reflective plane map. Therefore, the laser radar-based point cloud processing method proposed in the present invention can solve the problem that it is difficult to obtain accurate spatial point clouds based on existing laser radars in the presence of reflective objects.

[0117] like Figure 2 , is a functional module diagram of a laser radar-based point cloud processing device 100 provided in one embodiment of the present invention. According to the functions implemented, the laser radar-based point cloud processing device may include a data acquisition module 101, a potential reflective plane point cloud screening module 102, a global reflective plane map construction module 103, and a point cloud classification module 104. The module of the present invention may also be referred to as a unit, which refers to a series of computer program segments that can be executed by a processor of a laser radar-based point cloud processing device and can complete fixed functions, which are stored in the memory of the laser radar-based point cloud processing device.

[0118] In this embodiment, the functions of each module / unit are as follows:

[0119] The data acquisition module 101 is used to obtain point cloud data collected by the laser radar in space;

[0120] The potential reflective plane point cloud screening module 102 is used to perform double echo and peak detection according to the point cloud data, so as to fit the potential reflective plane point cloud from the point cloud data;

[0121] The global reflective plane map construction module 103 is used to construct and update the global reflective plane map according to the potential reflective planes in multiple frames;

[0122] The point cloud classification module 104 is used to classify the point cloud data according to the global reflective plane map to obtain multiple types of point cloud data.

[0123] In detail, each module described in the laser radar-based point cloud processing device 100 described in the embodiment of the present invention adopts the same technical means as the laser radar-based point cloud processing method described in the accompanying drawings when used, and can produce the same technical effects, which will not be repeated here.

[0124] Reference Figure 3 , which is a flow chart of a robot navigation method provided by an embodiment of the present invention.

[0125] In this embodiment, the robot navigation method includes:

[0126] S21, obtaining multiple types of point cloud data obtained according to a laser radar-based point cloud processing method;

[0127] S22, performing point cloud screening on the point cloud data collected by the laser radar in space according to the multiple types of point cloud data to obtain a planning point cloud;

[0128] S23: Generate a planned path according to the planned point cloud, and control the robot to navigate according to the planned path.

[0129] like Figure 4 , which is a functional module diagram of a robot navigation device 200 provided by an embodiment of the present invention. According to the functions to be implemented, the robot navigation device may include a point cloud data acquisition module 201, a planning point cloud screening module 202, and a path planning module 203.

[0130] In this embodiment, the functions of each module / unit are as follows:

[0131] The point cloud data acquisition module 201 acquires multiple types of point cloud data obtained by the laser radar-based point cloud processing method according to any one of claims 1 to 7;

[0132] The planning point cloud screening module 202 is used to screen the point cloud data collected by the laser radar in space according to the multiple types of point cloud data to obtain a planning point cloud;

[0133] The path planning module 203 is used to generate a planned path according to the planned point cloud, and control the robot to navigate according to the planned path.

[0134] like Figure 5 , is a schematic diagram of the structure of an electronic device for implementing a laser radar-based point cloud processing method provided by an embodiment of the present invention.

[0135] The electronic device 300 may include a processor 301 , a memory 302 , a communication bus 303 , and a communication interface 304 , and may also include a computer program stored in the memory 302 and executable on the processor 301 , such as a laser radar-based point cloud processing program.

[0136] In some embodiments, the processor 301 may be composed of an integrated circuit, for example, it may be composed of a single packaged integrated circuit, or it may be composed of multiple packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and a combination of various control chips.

[0137] In some embodiments, the memory 302 may be an internal storage unit of the electronic device, such as a mobile hard disk of the electronic device. In other embodiments, the memory 302 may also be an external storage device of the electronic device. The memory 302 may be used not only to store application software and various data installed in the electronic device, such as the code of the point cloud processing program based on the laser radar, but also to temporarily store data that has been output or is to be output.

[0138] The communication bus 303 can be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 302 and at least one processor 301, etc.

[0139] The communication interface 304 is used for communication between the electronic device and other devices, and includes a network interface and a user interface.

[0140] Figure 5 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 5 The structure shown does not constitute a limitation on the electronic device 300 , and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0141] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0142] The laser radar-based point cloud processing program stored in the memory 302 in the electronic device 300 is a combination of multiple instructions. When running in the processor 301, the above-mentioned laser radar-based point cloud processing method can be implemented.

[0143] Specifically, the specific implementation method of the processor 301 for the above instructions can refer to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.

[0144] Furthermore, if the modules / units integrated in the electronic device 300 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. The computer-readable storage medium can be volatile or non-volatile.

[0145] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor of an electronic device, it can implement the above-mentioned laser radar-based point cloud processing method.

[0146] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0147] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0148] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0149] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0150] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A point cloud processing method based on laser radar, It is characterized in that The method comprises: Obtain point cloud data collected by the laser radar in space; Performing double echo and peak detection according to the point cloud data to achieve fitting of a potential reflective plane point cloud from the point cloud data; Construct and update a global reflective plane map based on the potential reflective planes in multiple frames; The point cloud data is classified according to the global reflective plane map to obtain multiple types of point cloud data.

2. The laser radar-based point cloud processing method according to claim 1, It is characterized in that The performing of double echo and peak detection according to the point cloud data to achieve fitting of a potential reflective plane point cloud from the point cloud data includes: Determine the corresponding strongest echo point cloud and the last echo point cloud from the point cloud data obtained by the laser radar through double echo detection; Perform peak detection according to the point cloud data to obtain a peak regular point cloud; The point cloud data is screened according to the strongest echo point cloud, the last echo point cloud and the peak regularity point cloud to obtain potential reflective plane points.

3. The laser radar-based point cloud processing method according to claim 2, It is characterized in that The step of performing peak detection according to the point cloud data to obtain a peak regular point cloud comprises: Classify the point cloud data according to the pulse intensity to determine multiple types of intensity point clouds; Regularity analysis is performed on the multiple types of intensity point clouds, and peak regular point clouds are screened out from the point cloud data according to the results of the regularity analysis.

4. The laser radar-based point cloud processing method according to claim 2, It is characterized in that The step of screening the point cloud data according to the strongest echo point cloud, the last echo point cloud and the peak regularity point cloud to obtain a potential reflective plane point cloud includes: comparing a difference between the strongest echo point cloud and the last echo point cloud with a preset echo interval threshold; Determine a potential reflective plane point cloud according to the comparison result and the strongest echo point cloud; The peak regular point cloud identified by the laser radar under different posture information is used as a potential reflective plane point cloud.

5. The laser radar-based point cloud processing method according to claim 1, It is characterized in that The step of constructing and updating a global reflective plane map according to multiple frames of potential reflective planes includes: Performing frame classification according to the potential reflective plane point cloud to obtain a set of potential reflective plane point clouds under multiple frames; Performing plane fitting according to the reflective plane point cloud in the potential reflective plane point cloud set to obtain a reflective plane map; Performing spatial coordinate conversion according to the pose information corresponding to the reflective plane map corresponding to each frame to obtain a reflective plane map in the same coordinate system; The reflective plane map corresponding to each frame is updated and optimized to obtain a global reflective plane map.

6. The laser radar-based point cloud processing method according to claim 5, It is characterized in that The step of performing plane fitting according to the reflective plane point cloud in the potential reflective plane point cloud set to obtain a reflective plane map includes: Selecting a target fitting point cloud and a point cloud to be screened from the potential reflective plane point cloud set; Generate a fitting plane according to the target fitting point cloud; Calculating the distance between the point cloud to be screened and the fitting plane; Update the to-be-screened point cloud and the target fitting point cloud according to the distance; The fitting plane is updated according to the updated target fitting point, and the step of calculating the distance between the point cloud to be screened and the fitting plane is returned to until the updating of the fitting plane meets the preset updating condition, so as to obtain a reflective plane map.

7. The laser radar-based point cloud processing method according to claim 1, It is characterized in that The point cloud data is classified according to the global reflective plane map to obtain multiple types of point cloud data, including: Calculating a spatial relationship between the point cloud data and the global reflective plane map; The categories corresponding to the point cloud data are determined according to the spatial relationship to obtain multi-category point cloud data.

8. A point cloud processing device based on laser radar, It is characterized in that The device comprises: The data acquisition module is used to obtain the point cloud data collected by the laser radar in space; A potential reflective plane point cloud screening module, used for performing double echo and peak detection according to the point cloud data, so as to fit a potential reflective plane point cloud from the point cloud data; A global reflective plane map construction module, used to construct and update a global reflective plane map according to the potential reflective planes in multiple frames; The point cloud classification module is used to classify the point cloud data according to the global reflective plane map to obtain multiple types of point cloud data.

9. A robot navigation method, It is characterized in that The method comprises: Acquire multiple types of point cloud data obtained by the laser radar-based point cloud processing method according to any one of claims 1 to 7; Point cloud data collected by the laser radar in space are screened according to the multiple types of point cloud data to obtain a planning point cloud; A planned path is generated according to the planned point cloud, and the robot is controlled to navigate according to the planned path.

10. A robot navigation device, It is characterized in that The device comprises: A point cloud data acquisition module, which acquires multiple types of point cloud data obtained by the laser radar-based point cloud processing method according to any one of claims 1 to 7; A planning point cloud screening module is used to screen the point cloud data collected by the laser radar in space according to the multiple types of point cloud data to obtain a planning point cloud; The path planning module is used to generate a planned path according to the planned point cloud and control the robot to navigate according to the planned path.

Citation Information

Patent Citations

  • Robot instant localization and mapping method and system based on multiple information sources

    CN113432600A