WO2025190403A1 - Procédé et appareil de détermination d'informations de sol pour dispositif automoteur, dispositif et support - Google Patents

Procédé et appareil de détermination d'informations de sol pour dispositif automoteur, dispositif et support

Info

Publication number
WO2025190403A1
WO2025190403A1 PCT/CN2025/082648 CN2025082648W WO2025190403A1 WO 2025190403 A1 WO2025190403 A1 WO 2025190403A1 CN 2025082648 W CN2025082648 W CN 2025082648W WO 2025190403 A1 WO2025190403 A1 WO 2025190403A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
ground
point
self
dimensional
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/CN2025/082648
Other languages
English (en)
Chinese (zh)
Inventor
蔡为燕
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Roborock Innovation Technology Co Ltd
Original Assignee
Beijing Roborock Innovation Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Roborock Innovation Technology Co Ltd filed Critical Beijing Roborock Innovation Technology Co Ltd
Publication of WO2025190403A1 publication Critical patent/WO2025190403A1/fr
Pending legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
    • G06V10/803—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of input or preprocessed data
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10028—Range image; Depth image; 3D point clouds
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30244—Camera pose

Definitions

  • the present disclosure relates to the technical field of control of self-propelled cleaning equipment, and in particular to a method and apparatus, equipment, and medium for determining ground information of a self-propelled cleaning equipment.
  • the purpose of the present disclosure is to provide a method and apparatus, equipment and medium for determining ground information of a self-propelled device, which are configured to enable the self-propelled device to more accurately perceive the ground environment during movement, especially to identify complex structures on the ground, such as cliffs, steps, and thresholds.
  • a method for determining ground information of a self-propelled device comprising:
  • a three-dimensional ground voxel probability map is determined based on the three-dimensional point cloud data of each frame and the pose information.
  • determining a three-dimensional ground voxel probability map based on the three-dimensional point cloud data of each frame and the pose information includes:
  • the forward path of the self-propelled device is analyzed according to the updated three-dimensional voxel probability map to determine the ground information of the self-propelled device.
  • analyzing the three-dimensional point cloud data to determine a point cloud label corresponding to each point includes:
  • the point cloud label corresponding to each point is determined.
  • determining a point cloud label corresponding to each point based on the curvature information and normal information of each point includes:
  • the point cloud label corresponding to the point is determined to be a planar point; otherwise, the point cloud label corresponding to the point is determined to be a non-planar point.
  • analyzing the three-dimensional point cloud data to determine curvature information and normal information of the current point based on the three-dimensional point cloud data of neighboring points of each point includes:
  • the curvature information is determined according to the minimum eigenvalue and the sum of all eigenvalues.
  • updating a three-dimensional voxel probability map according to the point cloud label and the pose information includes:
  • ground probability values corresponding to voxels of planar points are increased, and ground probability values corresponding to voxels of non-planar points are decreased, so as to update the three-dimensional voxel probability map.
  • analyzing the forward path of the self-propelled device according to the updated three-dimensional voxel probability map to determine ground information of the self-propelled device includes:
  • the ground information of the self-propelled device is determined according to the target detection area and the updated three-dimensional voxel probability map.
  • determining the ground information of the self-propelled device according to the target detection area and the updated three-dimensional voxel probability map includes:
  • the target detection area selecting a target voxel whose ground probability value is greater than a preset probability value from the updated three-dimensional voxel probability map;
  • ground structural features are determined.
  • the method further includes:
  • the ground posture of the self-propelled device is estimated based on the updated three-dimensional voxel probability map and the three-dimensional point cloud data of the current frame.
  • the method further includes:
  • the three-dimensional ground voxel probability map is updated according to the RGB image.
  • updating the three-dimensional ground voxel probability map according to the RGB image includes:
  • the RGB information of each point is fused into the three-dimensional ground voxel probability map to update the three-dimensional ground voxel probability map.
  • a device for determining ground information of a self-propelled device comprising:
  • a first acquisition module is configured to acquire three-dimensional point cloud data of a current frame of the self-propelled device
  • an analysis module configured to analyze the three-dimensional point cloud data to determine a point cloud label corresponding to each point, wherein the point cloud label includes a planar point and a non-planar point;
  • a second acquisition module is configured to obtain current posture information of the self-propelled device
  • an updating module configured to fuse multiple frames of 3D point cloud data using an algorithm based on a 3D voxel probability map, and update the 3D voxel probability map according to the point cloud labels and the pose information;
  • the information determination module is configured to analyze the forward path of the self-propelled device according to the updated three-dimensional voxel probability map to determine the ground information of the self-propelled device.
  • a self-propelled device comprising: at least one processor; and a memory communicatively connected to the at least one processor;
  • the memory stores instructions that can be executed by the at least one processor, and the instructions are configured to execute the method described in any one of the embodiments of the first aspect.
  • a computer-readable storage medium on which computer instructions are stored.
  • the instructions are executed by a processor, the steps of the method described in any one of the embodiments of the first aspect are implemented.
  • FIG1 is a schematic structural diagram of a self-propelled device according to an exemplary embodiment.
  • FIG2 is a flow chart showing a method for determining ground information of a self-propelled device according to an exemplary embodiment.
  • FIG3 is a flowchart showing step S202 in a method for determining ground information of an autonomous vehicle according to an exemplary embodiment.
  • FIG4 is a flowchart showing updating of a three-dimensional voxel probability map in a method for determining ground information of an autonomous vehicle according to an exemplary embodiment.
  • FIG5 is a flowchart showing step S205 in a method for determining ground information of an autonomous vehicle according to an exemplary embodiment.
  • Fig. 6 is a block diagram showing a device for determining ground information of a self-propelled device according to an exemplary embodiment.
  • the disclosed embodiment provides a possible application scenario, which includes a self-propelled device 100, such as a self-propelled sweeping device, a self-propelled mopping device, a vacuum cleaner, a lawn mower, and the like.
  • a self-propelled device 100 such as a self-propelled sweeping device, a self-propelled mopping device, a vacuum cleaner, a lawn mower, and the like.
  • a household self-propelled sweeping device is taken as an example for illustration.
  • the self-propelled sweeping device can clean according to a preset route or an automatically planned route, but it will inevitably get stuck in certain places and be unable to proceed, such as a chair, a table, etc.
  • the self-propelled sweeping device can identify obstacles through a cloud server, a local server, or its own storage system, and mark the location as an obstacle location.
  • the self-propelled device can be provided with a touch-sensitive display or controlled by a mobile terminal to receive operation instructions input by the user.
  • the self-propelled device can be equipped with various sensors, such as buffers, cliff sensors, ultrasonic sensors, infrared sensors, magnetometers, accelerometers, gyroscopes, odometers and other sensing devices (the specific structure of each sensor is not introduced in detail, and any of the above sensors can be used to configure the self-propelled device).
  • the self-propelled device can also be equipped with wireless communication modules such as WIFI modules and Bluetooth modules to connect to smart terminals or servers, and receive operation instructions transmitted by the smart terminals or servers through the wireless communication modules.
  • the self-propelled device is an intelligent cleaning device, such as a self-propelled sweeping device, a self-propelled mopping device, a self-propelled floor polishing device, or a self-propelled weeding device.
  • a self-propelled sweeping device such as a self-propelled sweeping device, a self-propelled mopping device, a self-propelled floor polishing device, or a self-propelled weeding device.
  • this embodiment uses a self-propelled sweeping device as an example to describe the technical solution of the present disclosure.
  • a self-propelled device may include: a device body, a sensing system, a control system, a drive system, a cleaning system, an energy system, and a human-machine interaction system. These systems coordinate and cooperate with each other to enable the self-propelled device to move autonomously to perform cleaning functions.
  • the functional components of the self-propelled device that constitute each of these systems are integrated within the device body.
  • the device body has a roughly circular shape (both front and back), but may also have other shapes, including but not limited to a roughly D-shaped shape with a circular front and back.
  • the perception system includes a line laser module located above or to the side of the device body.
  • the control system's main control unit is connected to the line laser module and controls the autonomous device's functions based on the line laser module's perception results.
  • the specific location of the line laser module within the device body is not limited.
  • it can be, but is not limited to, the front, rear, left, right, top, middle, or bottom of the device body.
  • the line laser module can be positioned in the middle, top, or bottom of the device body in the height direction.
  • a line laser module is disposed on the front side of the device body; the front side is the side toward which the device body faces during the forward movement of the mobile device.
  • the control system is provided on a circuit board inside the main body of the machine, and includes a computing processor, such as a central processing unit and an application processor, that communicates with non-temporary memories, such as hard disks, flash memories, and random access memories.
  • the application processor utilizes a positioning algorithm, such as Simultaneous Localization And Mapping (SLAM), based on obstacle information fed back by a laser ranging device to draw a real-time map of the environment in which the self-propelled device is located.
  • SLAM Simultaneous Localization And Mapping
  • control system combines the distance information and speed information fed back by sensors, cliff sensors, magnetometers, accelerometers, gyroscopes, odometers, and other sensor devices provided on the buffer to comprehensively determine the current working state and location of the sweeper, as well as the sweeper's current posture, such as crossing a threshold, getting on a carpet, being on a cliff, being stuck above or below, having a full dust box, being picked up, etc.
  • Specific next-step action strategies are also given for different situations, so that the operation of the self-propelled device better meets the owner's requirements and provides a better user experience.
  • the drive system can steer the self-propelled device across the ground based on a drive command with distance and angle information (e.g., x, y, and ⁇ components).
  • the drive system includes a drive wheel module that can simultaneously control the left and right wheels.
  • the optional drive wheel module includes a left drive wheel module and a right drive wheel module, respectively.
  • the left and right drive wheel modules are opposed along a transverse axis defined by the main body.
  • the self-propelled device may include one or more driven wheels, including but not limited to universal wheels.
  • the drive wheel module includes a running wheel and a drive motor, as well as control circuitry for controlling the drive motor.
  • the drive wheel module may also be connected to a circuit for measuring the drive current and an odometer.
  • the drive wheel module can be detachably connected to the main body to facilitate disassembly and maintenance.
  • the drive wheel may have a biased drop-down suspension system that is movably fastened, for example, rotatably attached, to the main body of the self-propelled device and receives a spring bias that biases it downward and away from the main body of the self-propelled device.
  • the spring bias allows the drive wheel to maintain contact and traction with the ground with a certain ground force, while the cleaning elements of the self-propelled device 100 also contact the ground with a certain pressure.
  • the cleaning system can be a dry cleaning system and/or a wet cleaning system.
  • a dry cleaning system the main cleaning function comes from the cleaning system composed of a roller brush, a dust box, a fan, an air outlet, and the connecting parts between the four.
  • the roller brush which has a certain interference with the ground, sweeps up the garbage on the ground and rolls it to the front of the dust suction port between the roller brush and the dust box.
  • the air with suction generated by the fan and passing through the dust box is then sucked into the dust box.
  • the dry cleaning system may also include a side brush 152 with a rotating shaft, the rotating shaft is at a certain angle relative to the ground, so as to be configured to move debris to the roller brush area of the cleaning system.
  • the energy system includes rechargeable batteries, such as nickel-metal hydride batteries and lithium batteries.
  • the rechargeable batteries can be connected to a charging control circuit, a battery pack charging temperature detection circuit, and a battery undervoltage monitoring circuit.
  • the charging control circuit, battery pack charging temperature detection circuit, and battery undervoltage monitoring circuit are then connected to the microcontroller control circuit.
  • the host is charged by connecting to a charging station via charging electrodes located on the side or bottom of the device. If dust adheres to the exposed charging electrodes, the accumulated charge during charging can cause the plastic surrounding the electrodes to melt and deform, or even deform the electrodes themselves, preventing normal charging.
  • the human-machine interaction system includes buttons on the main unit panel for users to select functions; it may also include a display screen and/or indicator lights and/or a speaker to show the user the current machine status or function options; and it may also include a mobile client program.
  • the mobile client can display a map of the device's environment and the device's location, providing users with a richer and more user-friendly set of functions.
  • FIG2 is a flow chart showing a method for determining ground information of a self-propelled device according to an exemplary embodiment.
  • the method for determining ground information of a self-propelled device includes:
  • Step S201 Acquire the three-dimensional point cloud data of the current frame of the self-propelled device.
  • the 3D point cloud data observed by the sensor at the current moment can be obtained, and noise can be filtered based on an intensity threshold, and downsampling can be performed.
  • the intensity threshold can be set to A. If the intensity value of the 3D point cloud data is greater than A, it will be filtered and deleted; otherwise, the 3D point cloud data will be retained.
  • the sources of obtaining three-dimensional point cloud data include but are not limited to depth cameras, area array laser sensors, etc.
  • Step S202 obtaining the current posture information of the self-propelled device
  • the current position information of the self-propelled device can be obtained from the positioning module of the self-propelled device, wherein the position information may include the position coordinates x, y, the orientation angle theta, etc. of the self-propelled device on the ground.
  • Step S203 determining a three-dimensional ground voxel probability map based on the three-dimensional point cloud data of each frame and the pose information.
  • step S203 includes:
  • Step S2031 Analyze the three-dimensional point cloud data to determine a point cloud label corresponding to each point, wherein the point cloud label includes a planar point and a non-planar point;
  • step S2031 includes:
  • Step S301 analyzing the three-dimensional point cloud data to determine curvature information and normal information of the current point based on the three-dimensional point cloud data of the neighboring points of each point;
  • step S301 includes:
  • Step S3011 determining points whose distance to each point is within a preset distance value range as neighborhood points;
  • a certain number of adjacent points are selected as neighborhood points. These points can be selected using a distance metric or by searching for all points within a sphere with a fixed radius. For example, all points around each point whose distance to it is less than or equal to A can be determined as its corresponding neighborhood points.
  • Step S3012 Determine the covariance matrix corresponding to each point and its neighboring points.
  • a covariance matrix is calculated using point P and its neighboring points.
  • the covariance matrix is a second-order statistic based on the deviation of the coordinates of point P and its neighboring points. Specifically, the covariance of the coordinates of point P and each neighboring point can be calculated separately to form a covariance matrix.
  • Step S3013 performing eigendecomposition on the covariance matrix, and determining normal information according to the eigenvector corresponding to the minimum eigenvalue of the covariance matrix;
  • the eigenvectors of the covariance matrix represent the main directions of data distribution, while the eigenvalues represent the variance of the data in these directions. Perform eigendecomposition on the covariance matrix to find the eigenvalues and eigenvectors.
  • the eigenvector corresponding to the minimum eigenvalue of the covariance matrix is usually selected as the normal vector of the point because it represents the direction in which the data changes the least, that is, the normal direction of the plane.
  • Step S3014 Determine the curvature information according to the minimum eigenvalue and the sum of all eigenvalues.
  • the curvature can be defined using the minimum eigenvalue divided by the sum of all eigenvalues:
  • ⁇ 1 , ⁇ 2 , ⁇ 3 are the three eigenvalues of the covariance matrix.
  • Step S302 Determine the point cloud label corresponding to each point based on the curvature information and normal information of each point.
  • determining a point cloud label corresponding to each point based on the curvature information and normal information of each point includes:
  • the point cloud label corresponding to the point is determined to be a planar point; otherwise, the point cloud label corresponding to the point is determined to be a non-planar point.
  • each point is distinguished as a plane point or an edge point based on its curvature.
  • Points that 1) fall within a ground height threshold, 2) have a curvature less than a preset curvature value, or 3) have a normal vector with a ground normal angle less than a preset angle are marked as ground points.
  • Points that are above or below the ground height threshold are marked as non-ground points.
  • the ground height threshold range and preset angle can be set as needed, allowing the autonomous vehicle to meet different needs and usage environments.
  • Step S2032 fusing multiple frames of 3D point cloud data, and updating a 3D voxel probability map based on the point cloud labels and the pose information;
  • Step S2033 Analyze the forward path of the self-propelled device according to the updated three-dimensional voxel probability map to determine the ground information of the self-propelled device.
  • RGB information of the ground environment can also be collected by the RGB camera and fused into the updated 3D ground voxel probability map. This fusion method can provide richer details such as color and texture for the ground information, thus making the environmental perception more comprehensive.
  • A) Determine the extrinsic matrix of the 3D point cloud acquisition sensor and RGB camera. This matrix can be obtained through calibration.
  • the RGB information of the point (the color value of the corresponding pixel position on the corresponding RGB image) is obtained, and then updated into the 3D ground voxel probability map.
  • updating the 3D voxel probability map according to the point cloud label and the pose information includes:
  • Step S401 converting the three-dimensional point cloud data into a world coordinate system based on the pose information
  • Step S402 determining the voxel grid coordinates corresponding to each point based on the voxel grid origin and resolution;
  • Step S403 According to the point cloud labels, the ground probability values corresponding to the voxels of the planar points are increased, and the ground probability values corresponding to the voxels of the non-planar points are decreased, so as to update the three-dimensional voxel probability map.
  • a fixed value for increase when increasing the ground probability value, can be preset, such as a, and the ground probability value corresponding to the voxel of the planar point can be directly added to a to serve as the updated ground probability value.
  • a fixed value for decrease can also be preset, such as b, and the ground probability value corresponding to the voxel of the non-planar point can be subtracted from b to serve as the updated ground probability value.
  • the ground probability value interval to which the ground probability value of the voxel of the point belongs can also be determined. Different increase or decrease values can be set for different ground probability value intervals, so that the specific value of increase and the specific value of decrease are determined according to the ground probability value interval to which they belong.
  • Step S205 analyzing the forward path of the self-propelled device according to the updated three-dimensional voxel probability map to determine the ground information of the self-propelled device.
  • step S205 includes:
  • Step S501 determining a forward path of the self-propelled device according to the current posture information of the self-propelled device
  • the current position information of the self-propelled device is obtained from its positioning system, including the position coordinates x, y of the self-propelled device on the ground and the heading angle theta, thereby determining the path of the self-propelled device.
  • Step S502 determining a target detection area corresponding to the self-propelled device based on the forward path of the self-propelled device and its size, mobility, and application scenario;
  • the size and range of the forward detection area are preset based on the size, mobility, and application scenario of the autonomous vehicle.
  • the preset forward detection area for a small household autonomous sweeper is the width of the vehicle horizontally and the distance the autonomous vehicle travels per second vertically.
  • the detection area needs to be expanded to identify ground structure features in advance.
  • Step S503 traverse the updated three-dimensional voxel probability map within the target detection area, and select target voxels whose ground probability value is greater than a preset probability value;
  • the voxels in the updated three-dimensional voxel probability map are screened by presetting probability values, so that target voxels that meet the requirements can be screened out.
  • Step S504 mapping the target voxel into the ground plane map, and determining the height of the mapping point on the ground plane map according to the height value of the target voxel to obtain a ground information map;
  • the voxel probability map is traversed, and those voxels whose ground probability values exceed a specific threshold are selected and mapped to the ground plane map (i.e., the XY plane map of the ground, that is, the top view of the ground).
  • the ground refers to the ground within the detection area of the self-propelled device.
  • the height value of the voxel determines the height of its mapping point on the XY plane, forming an XY plane height map (similar to a depth map). By extracting the edge contours of the height map, line segments depicting areas with significant height changes can be obtained. By analyzing these line segments, different ground structure features can be identified.
  • Step S505 determining ground structural features according to the ground information map, wherein the ground structural features include cliffs, steps, and thresholds.
  • ground segments depicting areas with significant height changes can be obtained.
  • different ground structures can be identified based on different ground structure characteristics. For example, for a cliff, the height value on one side of the edge segment is the ground height, and the other side is a null value (the sensor cannot obtain the point cloud of the cliff part, so there is no value); for steps, the height value on one side of the edge segment is the ground height, and the other side is a value below the ground height; for thresholds, there are two parallel edge segments, and the height value of their middle position falls within the preset threshold range of the threshold height. Based on the above-mentioned ground structure characteristics, the specific ground structure can be determined, thereby more accurately perceiving the complex ground environment.
  • the method further includes:
  • the ground posture of the self-propelled device is estimated based on the updated three-dimensional voxel probability map and the three-dimensional point cloud data of the current frame.
  • the two are aligned to calculate the roll angle and pitch angle of the self-propelled device, thereby estimating the ground posture of the self-propelled device.
  • the point cloud registration may use an ICP (Iterative Closest Point) algorithm, which optimizes the transformation matrix (including rotation and translation) between two sets of point clouds through an iterative process:
  • ICP Intelligent Closest Point
  • Initialization Select the 3D point cloud of the current frame as the source point cloud and the reconstructed 3D ground voxel probability map as the target. Use the pose of the self-propelled vehicle obtained by its positioning system as the initial estimated transformation.
  • the source point cloud is transformed into the target coordinate system. For each point in the transformed source point cloud, the closest grid in the target voxel grid with a ground probability greater than a certain threshold is found. The center point of this grid and the source point cloud point constitute the closest point pair. This step is typically accelerated using a KD tree or other spatial data structure.
  • attitude angles Extract the rotational component from the final transformation matrix and convert it into Euler angles or other angular representations. These angles are the roll and pitch angles of the autonomous vehicle, representing the attitude of the autonomous vehicle relative to the ground.
  • Fig. 6 is a block diagram showing a device for determining ground information of a self-propelled device according to an exemplary embodiment.
  • a ground information determination device 60 for a self-propelled device comprising:
  • a first acquisition module 61 is configured to acquire three-dimensional point cloud data of each frame of the self-propelled device
  • a second acquisition module 62 is configured to obtain current posture information of the self-propelled device
  • the information determination module 63 is configured to determine a three-dimensional ground voxel probability map based on the three-dimensional point cloud data of each frame and the posture information.
  • the information determination module 63 includes:
  • a label determination unit configured to analyze the three-dimensional point cloud data to determine a point cloud label corresponding to each point, wherein the point cloud label includes a planar point and a non-planar point;
  • a fusion unit configured to fuse multiple frames of three-dimensional point cloud data and update a three-dimensional voxel probability map according to the point cloud labels and the pose information
  • the path analysis unit is configured to analyze the forward path of the self-propelled device according to the updated three-dimensional voxel probability map to determine the ground information of the self-propelled device.
  • the analysis module 62 includes:
  • an analyzing unit configured to analyze the three-dimensional point cloud data to determine curvature information and normal information of a current point based on the three-dimensional point cloud data of neighboring points of each point;
  • the label determination unit is configured to determine the point cloud label corresponding to each point according to the curvature information and normal information of each point.
  • the label determination unit is configured to:
  • the point cloud label corresponding to the point is determined to be a planar point; otherwise, the point cloud label corresponding to the point is determined to be a non-planar point.
  • the analysis unit is configured to:
  • the curvature information is determined according to the minimum eigenvalue and the sum of all eigenvalues.
  • the update module includes:
  • a coordinate conversion unit configured to convert the three-dimensional point cloud data into a world coordinate system based on the pose information
  • a coordinate determining unit configured to determine the voxel grid coordinates corresponding to each point based on the voxel grid origin and the resolution
  • the updating unit is configured to increase the ground probability value corresponding to the voxel of the planar point and decrease the ground probability value corresponding to the voxel of the non-planar point according to the point cloud label, so as to update the three-dimensional voxel probability map.
  • the information determination module includes:
  • a path determination unit configured to determine a forward path of the self-propelled device according to current position information of the self-propelled device
  • an area determination unit configured to determine a target detection area corresponding to the self-propelled device based on a forward path of the self-propelled device and its size, mobility, and application scenario;
  • the information determining unit is configured to determine the ground information of the self-propelled device according to the target detection area and the updated three-dimensional voxel probability map.
  • the information determining unit is configured to:
  • the target detection area selecting a target voxel whose ground probability value is greater than a preset probability value from the updated three-dimensional voxel probability map;
  • ground structural features are determined.
  • the device further comprises:
  • the posture estimation module is configured to estimate the ground posture of the self-propelled device based on the updated three-dimensional voxel probability map and the three-dimensional point cloud data of the current frame.
  • the device further comprises:
  • an image determination module configured to determine an RGB image of the ground on which the self-propelled device is currently located
  • the updating module is further configured to update the three-dimensional ground voxel probability map according to the RGB image.
  • the update module is further configured to:
  • the RGB information of each point is fused into the three-dimensional ground voxel probability map to update the three-dimensional ground voxel probability map.
  • a self-propelled device comprising: at least one processor; and a memory communicatively connected to the at least one processor;
  • the memory stores instructions that can be executed by the at least one processor, and the instructions are set to be configured to execute the method described in any one of the embodiments of the first aspect.
  • a computer-readable storage medium on which computer instructions are stored.
  • the instructions are executed by a processor, the steps of the method described in any one of the embodiments of the first aspect are implemented.
  • the computer-readable medium described in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof.
  • a computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination thereof.
  • Computer-readable storage media may include, but are not limited to, an electrical connection having one or more wires, 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 thereof.
  • a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or component.
  • a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof.
  • a computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program configured for use by or in conjunction with an instruction execution system, apparatus, or device.
  • Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wireline, optical fiber cable, RF, or any suitable combination thereof.
  • each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the above-mentioned module, program segment, or a part of code contains one or more executable instructions configured to implement the specified logical function.
  • the functions marked in the box can also occur in a different order than the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved.
  • each box in the block diagram or flowchart, and the combination of the boxes in the block diagram or flowchart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
  • the units involved in the embodiments described in this disclosure may be implemented in software or hardware, and the units described may also be provided in a processor. In some cases, the names of these units do not constitute limitations on the units themselves.
  • the present disclosure further provides a computer-readable medium, which may be included in the electronic device described in the above embodiments, or may exist independently and not incorporated into the electronic device.
  • the computer-readable medium carries one or more programs, which, when executed by the electronic device, enable the electronic device to implement the centralized cargo delivery method described in the above embodiments.
  • modules or units of a device configured to perform an action are mentioned in the above detailed description, this division is not mandatory.
  • the features and functions of two or more modules or units described above can be embodied in one module or unit.
  • the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.
  • the technical solution according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.
  • a non-volatile storage medium which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.
  • a computing device which can be a personal computer, a server, a mobile terminal, or a network device, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Computing Systems (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

L'invention concerne un procédé et un appareil de détermination d'informations de sol pour un dispositif automoteur, un dispositif et un support. Le procédé consiste à : acquérir des données de nuage de points tridimensionnel de trames d'un dispositif automoteur ; acquérir des informations de pose actuelle du dispositif automoteur ; et déterminer un graphe de probabilité de voxel de sol tridimensionnel sur la base des données de nuage de points tridimensionnel des trames et des informations de pose.
PCT/CN2025/082648 2024-03-15 2025-03-14 Procédé et appareil de détermination d'informations de sol pour dispositif automoteur, dispositif et support Pending WO2025190403A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202410304291.7A CN118411403A (zh) 2024-03-15 2024-03-15 自行走设备的地面信息确定方法及装置、设备和介质
CN202410304291.7 2024-03-15

Publications (1)

Publication Number Publication Date
WO2025190403A1 true WO2025190403A1 (fr) 2025-09-18

Family

ID=92000109

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2025/082648 Pending WO2025190403A1 (fr) 2024-03-15 2025-03-14 Procédé et appareil de détermination d'informations de sol pour dispositif automoteur, dispositif et support

Country Status (2)

Country Link
CN (1) CN118411403A (fr)
WO (1) WO2025190403A1 (fr)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118411403A (zh) * 2024-03-15 2024-07-30 北京石头创新科技有限公司 自行走设备的地面信息确定方法及装置、设备和介质
CN119916808A (zh) * 2025-03-31 2025-05-02 深圳市元鼎智能创新有限公司 水池自动清洁装置以及控制方法及其计算机存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190147646A1 (en) * 2017-11-10 2019-05-16 International Business Machines Corporation Creating a three-dimensional map utilizing retrieved rfid tag information
CN112859859A (zh) * 2021-01-13 2021-05-28 中南大学 一种基于三维障碍物体素对象映射的动态栅格地图更新方法
CN116465393A (zh) * 2023-04-27 2023-07-21 北京石头创新科技有限公司 基于面阵激光传感器的同步定位和建图方法及装置
CN116608847A (zh) * 2023-04-27 2023-08-18 北京石头创新科技有限公司 基于面阵激光传感器和图像传感器的定位和建图方法
CN118411403A (zh) * 2024-03-15 2024-07-30 北京石头创新科技有限公司 自行走设备的地面信息确定方法及装置、设备和介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190147646A1 (en) * 2017-11-10 2019-05-16 International Business Machines Corporation Creating a three-dimensional map utilizing retrieved rfid tag information
CN112859859A (zh) * 2021-01-13 2021-05-28 中南大学 一种基于三维障碍物体素对象映射的动态栅格地图更新方法
CN116465393A (zh) * 2023-04-27 2023-07-21 北京石头创新科技有限公司 基于面阵激光传感器的同步定位和建图方法及装置
CN116608847A (zh) * 2023-04-27 2023-08-18 北京石头创新科技有限公司 基于面阵激光传感器和图像传感器的定位和建图方法
CN118411403A (zh) * 2024-03-15 2024-07-30 北京石头创新科技有限公司 自行走设备的地面信息确定方法及装置、设备和介质

Also Published As

Publication number Publication date
CN118411403A (zh) 2024-07-30

Similar Documents

Publication Publication Date Title
JP7671391B2 (ja) 障害物検出方法および装置、自走式ロボット並びに記憶媒体
US12390069B2 (en) Obstacle avoidance method and device for self-walking robot, robot, and storage medium
US20250107681A1 (en) Method and apparatus for constructing map of working region for robot, robot, and medium
CN111990930B (zh) 一种测距方法、装置、机器人和存储介质
US20230200613A1 (en) Obstacle avoidance method for robot, obstacle avoidance device for robot, and non-transitory computer readable storage medium
WO2025190403A1 (fr) Procédé et appareil de détermination d'informations de sol pour dispositif automoteur, dispositif et support
WO2017101721A1 (fr) Dispositif de nettoyage automatique et procédé de nettoyage
CN112205937A (zh) 一种自动清洁设备控制方法、装置、设备和介质
CN113625700B (zh) 自行走机器人控制方法、装置、自行走机器人和存储介质
WO2025082173A1 (fr) Procédé et appareil d'étalonnage de robot automoteur, et dispositif et support
WO2024140195A1 (fr) Procédé et appareil d'évitement d'obstacle de dispositif automoteur basés sur un laser linéaire, et dispositif et support
WO2024140376A1 (fr) Procédé de détection d'état de glissement, dispositif et support de stockage
US20240212182A1 (en) Distance measurement method and device, and robot and storage medium

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 25769778

Country of ref document: EP

Kind code of ref document: A1