WO2023005384A1 - 可移动设备的重定位方法及装置 - Google Patents

可移动设备的重定位方法及装置 Download PDF

Info

Publication number
WO2023005384A1
WO2023005384A1 PCT/CN2022/094718 CN2022094718W WO2023005384A1 WO 2023005384 A1 WO2023005384 A1 WO 2023005384A1 CN 2022094718 W CN2022094718 W CN 2022094718W WO 2023005384 A1 WO2023005384 A1 WO 2023005384A1
Authority
WO
WIPO (PCT)
Prior art keywords
identification
pose information
line segment
identification mark
sensor data
Prior art date
Application number
PCT/CN2022/094718
Other languages
English (en)
French (fr)
Inventor
刘施菲
马云飞
赖文芊
Original Assignee
北京旷视科技有限公司
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 北京旷视科技有限公司 filed Critical 北京旷视科技有限公司
Publication of WO2023005384A1 publication Critical patent/WO2023005384A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/20—Instruments for performing navigational calculations
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/02—Systems using the reflection of electromagnetic waves other than radio waves
    • G01S17/06—Systems determining position data of a target
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/88—Lidar systems specially adapted for specific applications
    • G01S17/89—Lidar systems specially adapted for specific applications for mapping or imaging
    • G01S17/894—3D imaging with simultaneous measurement of time-of-flight at a 2D array of receiver pixels, e.g. time-of-flight cameras or flash lidar

Definitions

  • the present disclosure generally relates to the technical field of positioning, and in particular, relates to a method and device for relocating a mobile device.
  • the relocation of the mobile device refers to the process of re-determining the pose of the mobile device at the initial moment of work, or when the pose (position and direction of movement) is lost.
  • Lidar relocation refers to the point cloud data currently collected by lidar and the preset point cloud map for the working environment of mobile devices.
  • the matching scores of various poses are calculated, and finally the pose whose score meets the requirements is determined as the final pose.
  • the point cloud map is pre-established.
  • there is a large difference between the point cloud map and the working environment and the positioning accuracy will be caused when the point cloud map is used for relocation.
  • similar working environments such as long corridors, there are several similar local areas in the working environment. Due to the similarity between local areas in the environment, it may cause wrong poses to be output during relocalization.
  • the present disclosure provides a method and device for relocating a movable device, so as to solve the problem in the prior art that the positioning accuracy is greatly reduced or the positioning fails during relocation.
  • an embodiment of the present disclosure provides a method for relocating a mobile device, and the method for relocating a mobile device includes:
  • the number of first identifications determined according to the sensor data is greater than or equal to a number threshold, from the second identifications recorded in the global map of the working environment, it is determined that the first identification matches the second identification mark; the first identification mark is the identification mark set in the working environment;
  • an embodiment of the present disclosure provides a relocating device for a mobile device, and the relocating device for a mobile device includes:
  • An acquisition module configured to acquire sensor data collected by the sensor in the working environment
  • a matching module configured to determine, from the second identifications recorded in the global map of the working environment, the A second identification that matches the first identification; the first identification is an identification set in the working environment;
  • a coarse positioning module configured to calculate and obtain initial pose information of the mobile device according to the correspondence between the first identification mark and the second identification mark;
  • a fine positioning module configured to determine target pose information of the mobile device according to the initial pose information, the global map, and the sensor data.
  • a computing processing device which is characterized by comprising: a memory, in which computer-readable codes are stored; one or more processors, when the computer-readable codes are stored by the When executed by one or more processors, the computing processing device executes the method for relocating a movable device.
  • a fourth aspect of the embodiments of the present disclosure provides a computer program, including computer readable codes, and when the computer readable codes run on a computing processing device, the computing processing device executes the mobile device. Relocation method.
  • a fifth aspect of the embodiments of the present application provides a computer-readable medium, in which the computer program is stored.
  • the present disclosure includes: acquiring sensor data collected by sensors in the working environment; determining from the working environment that the number of first identification marks in the working environment is greater than or equal to a preset number threshold Among the second identification marks recorded in the global map of the environment, the second identification mark matching the first identification mark is determined; the position of the identification marks arranged in the working environment is recorded in the global map; according to the first identification mark and the second identification mark Identify the corresponding relationship of the identification, and calculate the initial pose information of the mobile device; determine the target pose information of the mobile device according to the initial pose information, global map, and sensor data.
  • multiple easily identifiable identification signs can be arranged in the working environment of the mobile device, and the similarity between similar local areas can be reduced by deploying identification signs in similar working environments, thereby reducing the singularity of relocation The incidence of sexual problems.
  • rough positioning can be performed based on the identification mark, so that the process of rough positioning is not affected by environmental changes, and the positioning accuracy is improved.
  • FIG. 1 is a schematic diagram of the steps of a method for relocating a mobile device provided by an embodiment of the present disclosure
  • Fig. 2 is a schematic diagram of a working environment provided by an embodiment of the present disclosure
  • Fig. 3 is a schematic diagram of another working environment provided by an embodiment of the present disclosure.
  • FIG. 4 is a schematic diagram of specific steps of a method for relocating a mobile device provided by an embodiment of the present disclosure
  • Fig. 5 is a structural block diagram of a device for relocating a mobile device provided by an embodiment of the present disclosure
  • Fig. 6 is a block diagram of a device provided by an embodiment of the present disclosure.
  • Fig. 7 schematically shows a block diagram of a computing processing device for performing a method according to the present disclosure
  • Fig. 8 schematically shows a storage unit for holding or carrying program codes implementing the method according to the present disclosure.
  • Smart logistics uses artificial intelligence, big data, various information sensors, radio frequency identification technology, global positioning system (GPS, Global Positioning System) and other Internet of Things devices and technologies, and is widely used in material transportation, warehousing, distribution, packaging, loading and unloading and Information services and other basic activities to realize intelligent analysis and decision-making, automatic operation and high-efficiency optimized management of the material management process.
  • GPS Global Positioning System
  • the Internet of Things technology includes sensing equipment, radio frequency identification (RFID, Radio Frequency Identification) technology, laser infrared scanning, infrared induction identification, etc.
  • RFID Radio Frequency Identification
  • the Internet of Things can effectively connect the materials in the logistics with the network, and can monitor the materials in real time. Perceive environmental data such as humidity and temperature in the warehouse to ensure the storage environment of materials.
  • Perceive environmental data such as humidity and temperature in the warehouse to ensure the storage environment of materials.
  • all the data in the logistics can be sensed and collected, uploaded to the data layer of the information platform, and the data can be filtered, mined, analyzed and other operations, and finally the business process (such as transportation, storage, access, picking, packaging, distribution Picking, delivery, inventory, distribution and other links) to provide accurate data support.
  • the application direction of artificial intelligence in logistics can be roughly divided into two types: 1) those empowered by artificial intelligence technology such as unmanned trucks, automatic guided transport vehicles (AGV, Automated Guided Vehicle), autonomous mobile robots (AMR, Autonomous Mobile Robot), forklifts, shuttles, stackers, unmanned delivery vehicles, drones, service robots, robotic arms, smart terminals and other smart devices to replace part of the labor; 2) through computer vision, machine learning, operations optimization and other technologies or Algorithm-driven software systems such as transportation equipment management systems, warehouse management systems, equipment scheduling systems, and order distribution systems improve labor efficiency.
  • AMV Automated Guided Vehicle
  • AMR Autonomous Mobile Robot
  • Algorithm-driven software systems such as transportation equipment management systems, warehouse management systems, equipment scheduling systems, and order distribution systems improve labor efficiency.
  • Fig. 1 is a flow chart of the steps of a method for relocating a mobile device provided by an embodiment of the present disclosure. As shown in Fig. 1, the method may include:
  • Step 101 acquiring sensor data collected by sensors in a working environment.
  • the relocation process of a mobile device includes coarse positioning and fine positioning.
  • Coarse positioning refers to providing a rough pose of a mobile device (the pose includes the location and direction), while fine positioning can narrow down the search range of the pose based on the rough positioning results to obtain the final pose.
  • the identification mark is used to distinguish it from other objects in the working environment and provide a road sign reference.
  • the identification mark includes a reflective identification mark or an identifier.
  • a method for relocating a mobile device is applied to a mobile device with a sensor.
  • Multiple identification marks are arranged in the working environment of the mobile device, wherein the identification marks have strong characteristics of being recognized and can be Recognized by the sensor of the mobile device, the function of the identification mark is to distinguish it from other objects in the working environment, so as to provide a road sign reference.
  • the distance between similar local areas can be reduced. Similarity, reducing the occurrence of singularity problems in relocalization (that is, relocalization outputs multiple poses).
  • rough positioning can be performed based on the identification mark, so that the process of rough positioning is not affected by environmental changes, and the positioning accuracy is improved.
  • the identification mark can be set as a reflective identification mark that can reflect light, and the surface of the reflective identification mark is made of a material with a high refractive index, such as a reflective plate , reflective pillars, etc.
  • the laser beam of the lidar is emitted to the surface of the reflective identification mark, and after the reflected beam is received by the laser radar, the laser radar can calculate the distance between the reflective identification mark and the reflection intensity, so as to recognize the reflective identification mark .
  • the identification mark can be set to a more conspicuous identifier, such as a special image symbol, a specific pattern with a bright color, etc., and the camera can identify Identification, to realize the identification of the identification identification.
  • FIG. 2 and FIG. 3 uses FIG. 2 and FIG. 3 to give an example of the layout of the identification mark in the working environment where the mobile device is located.
  • FIG. 2 it shows a schematic diagram of a working environment provided by an embodiment of the present disclosure, which shows a long corridor scene 10 whose advancing direction is the X direction.
  • the long corridor scene 10 since the length of the corridor exceeds the sensor If the measurement distance is large, when the mobile device is in this scene, it can only measure the walls in the left and right directions, but it is difficult to measure in the front and rear directions, which makes it difficult for the mobile device to determine an initial value of the pose.
  • the embodiment of the present disclosure can lay out multiple identification signs 20 (at least 3) in the long corridor scene 10, so that in long corridors with similarities, the identification signs can be distinguished from similar parts and the singularity can be reduced The probability of the problem occurring.
  • FIG. 3 it shows a schematic diagram of another working environment provided by an embodiment of the present disclosure, which shows a scene 30 with local similarity.
  • this scene 30 there are regions 31 and 32, and region 31 It is very similar to area 32, that is, the layout and environment of the two are similar.
  • the embodiment of the present disclosure can arrange a plurality of identification signs 20 (at least 3) in the scene 30, so that in the scene 30 with similarity, the identification signs are distinguished from similar parts, reducing the singularity problem. probability of occurrence.
  • the movable device when the movable device has no initial pose at the initial moment of work, or the movable device loses its own pose, the movable device can be controlled to perform relocation based on the positioning command, and the movable device can respond to the relocation
  • the positioning instruction is used to acquire sensor data collected by the sensor in the working environment, where the sensor data may include the quantity and position of the first identification mark recognized by the mobile device.
  • the sensor data collected by the sensor includes the positions of the 3 positioning marks.
  • the sensor data can have various types.
  • the sensor is a camera
  • the sensor data can be the image collected by the camera
  • a set of position vectors in , in some cases, point cloud data can have color information and intensity information in addition to geometric positions. Assign the corresponding points in the point cloud; the acquisition of intensity information is the echo intensity collected by the lidar.
  • Step 102 When it is determined according to the sensor data that the number of first identifiers in the working environment is greater than or equal to a preset number threshold, determine from the second identifiers recorded in the global map of the working environment A second identification that matches the first identification.
  • the global map records the positions of the identification marks arranged in the working environment; the first identification mark is the identification mark set in the working environment.
  • the method of rough positioning can be determined according to the number of first identification marks recognized by the sensor.
  • the number is greater than or equal to the preset number threshold (such as 3), if the number of identified first identification marks is less than the preset number threshold, then the identification marks cannot be used for rough positioning, and other rough positioning methods can be selected.
  • the preset number threshold such as 3
  • the identification marks can be used for rough positioning, and the rough positioning method is to combine the identified two first identification marks into an identification group , is matched with the group consisting of two second identifications recorded in the preset global map, so as to determine the matching pair consisting of the position of the first identification and the position of the second identification, and finally according to these matching pairs , based on the least squares algorithm, to solve the initial pose information.
  • the position of the first identification mark recognized by the sensor can reflect the approximate area where the mobile device is located.
  • the identification group composed of the second identification identification is matched, and the matching pair of the first identification identification in the sensor coordinate system and the corresponding second identification identification in the map coordinate system can be obtained.
  • the matching process between the identification groups reflects the sensor coordinate system. Identifies a match between the local location represented by the group and a location similar to the local location in the map coordinate system.
  • the matching pair reflects the corresponding relationship between similar positions in the two coordinate systems. Based on the matching pair and the least squares algorithm, the equation for solving the initial pose information can be obtained, thereby obtaining the initial pose information.
  • the global map is a map preset according to the working environment of the mobile device.
  • the global map can be constructed according to the sensor data of the working environment (such as environmental photos, environmental point cloud data).
  • the global map can be updated according to actual needs, such as rebuilding the global map every preset time, or updating the global map when the working environment changes greatly.
  • Step 103 Calculate and obtain initial pose information of the mobile device according to the correspondence relationship between the first identification mark and the second identification mark.
  • a matching pair includes a correspondence between a first identification mark and a second identification mark. Since a matching pair reflects the correspondence between similar positions in two coordinate systems, all matching pairs obtained , which reflects the corresponding relationship between the local position in the sensor coordinate system and the corresponding similar position in the map coordinate system. After obtaining this matching correspondence, you can construct the initial pose information based on the least squares algorithm Equation, so as to obtain the initial pose information.
  • the least square method is a mathematical tool widely used in many disciplines of data processing such as error estimation, uncertainty, system identification, prediction, and forecasting.
  • the unknown data can be easily obtained by using the least square method, and the sum of squares of the errors between the obtained data and the actual data can be minimized.
  • the first identification and the second identification Identify the corresponding relationship, construct the least squares equation, and solve the least squares equation, and use the obtained solution as the initial pose information.
  • the initial pose information is an initial value of the pose obtained by the mobile device through rough positioning, which reflects the approximate position and direction of the mobile device. Since its accuracy is not enough for the navigation of the mobile device, it needs to be followed up. Precise positioning is performed based on the initial pose information to obtain accurate target pose information.
  • Step 104 Determine target pose information of the mobile device according to the initial pose information, the global map, and the sensor data.
  • the mobile device can further perform fine positioning, that is, based on the initial pose information, the sensor data collected by the sensor is matched with the global map, so as to solve the final
  • the target pose information of the target pose information can be the precise pose obtained by the mobile device after fine positioning, which reflects the precise position and orientation of the mobile device, and the mobile device can use the target pose information as the current navigation pose, for the follow-up work.
  • a search window can be set near the initial pose in the global map, and multiple candidate poses can be set in the search window according to the search step.
  • the sensor is a lidar sensor
  • the combination of each candidate pose can obtain the laser point cloud transformed by the candidate pose, and match the laser point cloud transformed by the candidate pose with the global map to obtain the score of the candidate pose, with the highest score
  • the least squares equation for solving the target pose information can be constructed to obtain the target pose information.
  • a method for relocating a mobile device includes: acquiring sensor data collected by a sensor in a working environment; In the case of a preset quantity threshold, from the second identifications recorded in the global map of the working environment, determine the second identification that matches the first identification; the global map records the identifications arranged in the working environment Position: Calculate the initial pose information of the mobile device according to the correspondence between the first identification mark and the second identification mark; determine the target pose information of the mobile device according to the initial pose information, global map, and sensor data.
  • multiple easily identifiable identification signs can be arranged in the working environment of the mobile device, and the similarity between similar local areas can be reduced by deploying identification signs in similar working environments, thereby reducing the singularity of relocation The incidence of sexual problems.
  • rough positioning can be performed based on the identification mark, so that the process of rough positioning is not affected by environmental changes, and the positioning accuracy is improved.
  • Fig. 4 is a flowchart of specific steps of a method for relocating a mobile device provided by an embodiment of the present disclosure. As shown in Fig. 4, the method may include:
  • Step 201 acquiring sensor data collected by sensors in the working environment.
  • the senor includes a lidar sensor
  • the sensor data includes laser point cloud data
  • the identification mark includes a reflective identification mark
  • the global map includes a grid map constructed according to the laser point cloud data.
  • the senor may be a laser radar sensor, and according to the working characteristics of the laser radar, the identification mark can be set as a reflective identification mark that can reflect light, and the surface of the reflective identification mark adopts a high refractive index materials, such as reflective plates, reflective columns, etc., the laser beam of the laser radar is emitted to the surface of the reflective identification mark, and after the reflected beam is received by the laser radar, the distance between the laser radar and the reflective identification mark and the reflection intensity can be calculated , so as to recognize the reflective identification mark.
  • the global map can be a raster map constructed from laser point cloud data.
  • the grid map is also called the occupancy grid map (Occupancy Grid Map).
  • the occupancy grid map is to divide the working environment into grids of a certain size. The sum of the probabilities is 1, and the occupancy ratio of the grid (occupancy probability/free probability) is stored in each grid.
  • the laser point cloud will fall into different grids, so the process of building a grid map is actually updating the probability of each grid being occupied.
  • the probability of the grid within its field of view will be updated. The more times a grid is hit by the laser, the higher the probability that it is occupied.
  • Step 202 When it is determined according to the sensor data that the number of first identifiers in the working environment is greater than or equal to a preset number threshold, determine from the second identifiers recorded in the global map of the working environment A second identification that matches the first identification.
  • the global map records the positions of the identification marks arranged in the working environment; the first identification mark is the identification mark set in the working environment.
  • step 102 For this step, reference may be made to the above-mentioned step 102 for details, which will not be repeated here.
  • step 202 may specifically include:
  • Sub-step 2021 respectively calculating the similarity between the first line segment between the two first identification marks and the second line segment between the two second identification marks.
  • Sub-step 2022 determine the first target line segment and the second target line segment whose similarity is greater than or equal to the similarity threshold, wherein the first identification marks at both ends of the first target line segment and the second target line segment at both ends of the second target line segment There is a one-to-one correspondence between the two identification marks.
  • the embodiment of the present disclosure can perform rough positioning based on the set Li in the sensor coordinate system and the set Mi in the map coordinate system.
  • the second line segment formed by the identification is matched to determine the matching pair consisting of the position of the first identification identification and the position of the second identification identification, and finally according to these matching pairs, based on the least squares algorithm, the initial pose information is solved.
  • the first identification mark and the corresponding second identification mark can be determined by calculating the similarity between the first line segment formed by the two first identification marks and the second line segment formed by the two second identification marks The degree of matching between them is based on the first target line segment and the second target line segment whose similarity is greater than or equal to the preset similarity threshold, and can be selected from the identification marks that constitute the first target line segment and the second target line segment. A corresponding relationship formed by an identification mark and a second identification mark.
  • each of the first line segments is a line segment between the first reference identification and any other first identification
  • the first reference identification is a plurality of first identifications determined according to the sensor data.
  • each second line segment is a line segment between the second reference identification identification and any other second identification identification
  • the second reference identification is a plurality of second identifications recorded in the global map
  • the sub-step 2021 includes:
  • Sub-step A1 according to the absolute value of the difference between the first length of the first line segment and the second length of the second line segment, and the smaller value of the first length and the second length, Calculate the similarity between the first line segment and the second line segment respectively.
  • the establishment process of the first line segment and the second line segment is firstly described through the following example:
  • a first reflective mark La can be randomly extracted from the set Li
  • a second reflective mark Mm can be randomly extracted from the set Mi
  • ⁇ La, Mm ⁇ can be used as a matching pair.
  • the first reflective mark La As the reference first identification mark
  • the second reflective mark Mm as the reference second identification mark.
  • a first reflective mark Lb can be randomly selected from the remaining first reflective marks of the set Li
  • a second reflective mark Mn can be randomly selected from the remaining second reflective marks of the set Mm
  • the reference first reflective mark La A first line segment ⁇ La, Lb ⁇ is established with the first reflective mark Lb
  • a second line segment ⁇ Mm, Mn ⁇ is established based on the reference second reflective mark Mm and the second reflective mark Mn.
  • the similarity between the first line segment ⁇ La, Lb ⁇ and the second line segment ⁇ Mm, Mn ⁇ can be calculated.
  • the benchmark first can be further removed from the set Li.
  • a line of the first identification mark Lb can be formed according to the position coordinates of La and Lb A line segment line1; Since the second line segment ⁇ Mm, Mn ⁇ records the position coordinates of the reference second identification mark Mm and the second identification mark Mn in the map coordinate system, a second line segment can be formed according to the position coordinates of Mm and Mn line2. Moreover, since the absolute lengths of a line segment in different coordinate systems are the same, the similarity between the first line segment line1 and the second line segment line2 can be calculated. The same applies to other first line segments and second line segments other than the example.
  • the calculation of the similarity between the first line segment line1 and the second line segment line2 can refer to the following formula 1:
  • S is the similarity between the first line segment line1 and the second line segment line2
  • d is the first length of the first line segment line1
  • d' is the second length of the second line segment line2
  • ⁇ is the maximum tolerance length difference
  • Substep 2022 includes:
  • Sub-step B1 forming a corresponding relationship between the reference first identification mark and the reference second identification mark.
  • Sub-step B2 combining the other first identification mark except the reference first identification mark among the two first identification marks forming the first line segment of the target with the two second identification marks forming the second line segment of the target Another second identification mark except the reference second identification mark constitutes another corresponding relationship.
  • the first line segment ⁇ La, Lb ⁇ and the second line segment ⁇ Mm, Mn ⁇ it is assumed that the first line segment ⁇ La, Lb ⁇ and the second line segment ⁇ Mm, Mn ⁇ If the similarity is greater than the preset similarity threshold, in addition to forming a corresponding relationship ⁇ La, Mm ⁇ between the reference first identification mark La and the reference second identification mark Mm, the first line segment ⁇ La, Lb ⁇ in another first identification mark Lb and another second identification mark Mn in the second line segment ⁇ Mm, Mn ⁇ form another corresponding relationship ⁇ Lb, Mn ⁇ . The same is true for other target first line segments and target second line segments other than the example.
  • the conditions for terminating the matching include: all matches between the first identification and the second identification are completed, the matching duration is longer than or It is equal to the preset time length, and the obtained number of corresponding relationships is greater than or equal to any one of the corresponding relationship number thresholds.
  • a matching termination condition for example, when all matches between the first identification and the second identification are completed, the matching is terminated; or the matching duration is greater than or equal to the preset
  • the matching can be terminated in order to avoid affecting the user experience due to too long matching time; or according to actual needs, when the number of obtained corresponding relationships is greater than or equal to the threshold of the number of matching pairs, it is considered that the obtained corresponding relationships can be roughly Locating, matching terminated.
  • Step 203 Calculate and obtain initial pose information of the mobile device according to the correspondence relationship between the first identification mark and the second identification mark.
  • step 103 For this step, reference may be made to the above-mentioned step 103 for details, which will not be repeated here.
  • step 203 may specifically determine the initial pose information of the mobile device according to the corresponding relationship and the conversion relationship from the sensor coordinate system of the mobile device to the map coordinate system of the global map.
  • the specific implementation process includes:
  • Sub-step 2031 Construct the first solution of the initial pose information according to the corresponding relationship, the least squares solution algorithm, and the transformation relationship from the sensor coordinate system of the mobile device to the map coordinate system of the global map Equation expression.
  • Sub-step 2032 calculate the initial pose information of the mobile device.
  • the navigation of the mobile device needs to be implemented based on the map coordinate system. Since the sensor data collected by the sensor of the mobile device is in the sensor coordinate system, it can be based on the least squares algorithm, correspondence, and The conversion relationship from the sensor coordinate system to the map coordinate system is used to construct the first solution equation expression for solving the initial pose information.
  • Li is the position coordinate of the identification mark in the sensor coordinate system, defined as (Xli, Yli);
  • Mi is the position coordinate of the identification mark in the map coordinate system, defined as (Xmi, Ymi).
  • the initial pose information to be solved is the position (x, y) and angle ⁇ of the mobile device in the map. Since there is a matching relationship between Li and Mi, the coordinates of Li are converted from the sensor coordinate system to the map coordinate system.
  • the result obtained should be Mi, from which the following formula 2 can be obtained.
  • three unknown quantities (x, y, ⁇ ) can be solved, and three or more equations can be constructed based on the least squares solution algorithm, that is, the position (x, y) and angle of the mobile device on the map can be obtained ⁇ , so that the initial pose information can be obtained
  • Step 204 Determine target pose information of the mobile device according to the initial pose information, the global map, and the sensor data.
  • step 104 For this step, reference may be made to the above-mentioned step 104 for details, which will not be repeated here.
  • step 204 may specifically include:
  • Sub-step 2041 Determine a local search area in the global map according to the initial pose information.
  • Sub-step 2042 using the sensor data, performing a pose matching operation in the local search area, and obtaining the target pose information according to the pose information with the highest matching degree.
  • the initial pose information is obtained based on rough positioning for further fine positioning, and the target pose information for guiding the mobile device to navigate can be obtained.
  • the fine positioning process can be described as: given an initial position According to the window size and search step length, multiple candidate poses are determined, and an optimal pose is obtained through matching, so that the probability of sensor data occurrence is maximized.
  • a search window w can be set around the initial pose. Given the size of the search window and the search step, multiple expressions of candidate poses in the map coordinate system of the global map can be obtained. The number of candidate pose information is related to the size of the search window and the search step.
  • the sensor data is defined as ⁇ hk ⁇ .
  • hk refers to the position coordinates of the kth laser point in the lidar coordinate system.
  • T ⁇ the corresponding candidate pose information
  • ⁇ x and ⁇ y are the coordinates of the mobile device in the map coordinate system, and ⁇ is the direction angle of the mobile device. Since the fine positioning process can be described as a search window near the given initial pose, according to the window size and The search step determines multiple candidate poses, and obtains an optimal pose through matching, which maximizes the probability of sensor data occurrence, so it can be summarized as the following nonlinear optimization problem:
  • w is a search window
  • Mnearest(T ⁇ hk) is the occupancy probability of the grid cell closest to T ⁇ hk in the global map.
  • the sensor data is hk
  • the initial pose information is T ⁇ .
  • the senor includes a lidar sensor
  • the sensor data includes laser point cloud data
  • the identification mark includes a reflective identification mark
  • the global map includes a grid map constructed according to the laser point cloud data
  • Sub-step C1 according to the sensor data, the size of the search window and the search step, determine a plurality of candidate pose information for the local search area.
  • Sub-step C2 In the local search area, determine the grid hit by the laser point cloud data corresponding to each candidate pose information.
  • Sub-step C3 calculate the average occupancy probability of all the grids that are hit, and use the average occupancy probability as the matching degree of the candidate pose information.
  • Sub-step C4 obtaining the target pose information according to the candidate pose information with the highest matching degree.
  • the score calculation is the average of the occupancy probabilities of all laser point clouds hitting a grid in the global map. Therefore, the higher the score, the higher the probability of laser point cloud appearance, and the closer the candidate pose information is to the real pose.
  • a strategy to speed up the search can be used in the specific implementation called branch and bound.
  • the discrete area corresponding to each pose expression in the global map can be determined first, and the discrete area reflects the position hit by the laser point cloud, and then the average occupancy probability of all grids contained in the discrete area can be calculated as (reflecting the average occupancy probability of all laser point clouds hitting the grid in the global map under the candidate pose), as the score of the candidate pose information corresponding to the pose expression.
  • the obtained candidate pose information with the highest score is the accuracy of the grid resolution level
  • the accuracy of the grid resolution level considering the limited accuracy caused by the resolution of the map grid, if you want to further optimize the accuracy, you need to perform Interpolation (the output of the M smooth function is the probability that the grid is occupied, which is a number within (0, 1)), and the interpolation algorithm can specifically be a bicubic interpolation algorithm. Through the optimization of interpolation, it can provide better accuracy than raster resolution.
  • This part of the equivalent least squares problem formula is as follows: the probability that the grid is occupied is the largest, which is equivalent to the minimum probability that the grid is not occupied. Based on the second solution equation expression constructed by the least squares algorithm, the target pose information can be obtained.
  • the M smooth function is an interpolation function
  • the sensor data is hk
  • the initial pose information is T ⁇ .
  • the method may further include:
  • Step 205 Construct a local map according to the sensor data when the number of the first identification marks is less than the number threshold.
  • the preset number threshold such as 3
  • another parallel solution can be carried out at this time, that is, rough positioning based on grayscale matching.
  • Gray-scale matching is the process of matching the gray-scale local image and the gray-scale global image to solve the pose. Different poses have different scores, and the higher the score, the greater the value of the pose.
  • the sensor data can be used to build a local map first, that is, the mobile device can build a local map for the current local environment.
  • the laser point cloud data (equivalent to multi-frame laser) obtained at a certain angle is constructed as a local map; 2.
  • Step 206 converting the local map into a local grayscale image, and converting the global map into a global grayscale image.
  • the embodiment of the present disclosure can convert the local map constructed from sensor data into a local grayscale image, that is, project the occupancy ratio stored in the grid of the local map to [ 0, 255] to get a local grayscale image.
  • the embodiments of the present disclosure can convert the global map into a global grayscale image, that is, project the occupancy ratio stored in the grid of the global map to the range of [0, 255], Get a global grayscale image.
  • Step 207 Perform matching calculation on the local grayscale image and the global grayscale image to obtain at least one piece of pose information and a matching score corresponding to each pose information.
  • grayscale matching can be performed on each position of the grayscale local image and the global grayscale image.
  • the matching process will obtain at least one pose information and a matching score corresponding to each pose information.
  • the matching score reflects How well the pose information matches the global map.
  • grayscale matching is to calculate the similarity between two images by using some similarity measure.
  • Commonly used grayscale-based matching methods include: mean absolute difference algorithm, absolute error sum algorithm, error sum of squares algorithm, average error sum of squares algorithm, normalized product correlation algorithm, etc.
  • Step 208 Use the pose information with the highest matching score as the initial pose information.
  • the pose information with the largest matching score can be considered as the pose that best matches the global map.
  • step 204 can be executed to perform subsequent the precise positioning process.
  • a method for relocating a mobile device includes: acquiring sensor data collected by a sensor in a working environment; In the case of a preset quantity threshold, from the second identifications recorded in the global map of the working environment, determine the second identification that matches the first identification; the global map records the identifications arranged in the working environment Position: Calculate the initial pose information of the mobile device according to the correspondence between the first identification mark and the second identification mark; determine the target pose information of the mobile device according to the initial pose information, global map, and sensor data.
  • multiple easily identifiable identification signs can be arranged in the working environment of the mobile device, and the similarity between similar local areas can be reduced by deploying identification signs in similar working environments, thereby reducing the singularity of relocation The incidence of sexual problems.
  • rough positioning can be performed based on the identification mark, so that the process of rough positioning is not affected by environmental changes, and the positioning accuracy is improved.
  • Fig. 5 is a structural block diagram of a relocating device for a mobile device provided by an embodiment of the present disclosure, which is applied to a mobile device with sensors, and multiple identification signs are arranged in the working environment of the mobile device, as shown in Fig. 5 As shown, the relocating device of the mobile device includes:
  • the device can include:
  • An acquisition module 301 configured to acquire sensor data collected by the sensor in the working environment
  • the matching module 302 is configured to determine, from the second identifications recorded in the global map of the working environment, the number of the first identifications determined according to the sensor data is greater than or equal to the number threshold, A second identification that matches the first identification; the first identification is an identification set in the working environment;
  • the matching module 302 includes:
  • a similarity submodule configured to calculate the similarity between the first line segment between the two first identification marks and the second line segment between the two second identification marks;
  • each of the first line segments is a line segment between the first reference identification and any other first identification
  • the first reference identification is a plurality of first identifications determined according to the sensor data.
  • Each of the second line segments is a line segment between the second reference identification mark and any other second identification mark, and the second reference mark is one of the plurality of second identification marks recorded in the global map.
  • the similarity submodule includes:
  • a similarity unit configured to be based on the absolute value of the difference between the first length of the first line segment and the second length of the second line segment, and the smaller of the first length and the second length value, and calculate the similarity between the first line segment and the second line segment.
  • the matching pair sub-module is used to determine the target first line segment and the target second line segment whose similarity is greater than or equal to the similarity threshold, wherein the first identification marks at both ends of the target first line segment and the target second line segment There is a one-to-one correspondence between the second identification marks at both ends.
  • matching submodules include:
  • a first combining unit configured to form a corresponding relationship between the reference first identification mark and the reference second identification mark
  • the second combining unit is configured to combine the other first identification mark except the reference first identification mark among the two first identification marks forming the first line segment of the target with the two second identification marks forming the second line segment of the target Among the two identification marks, another second identification mark except the reference second identification mark constitutes another corresponding relationship.
  • a coarse positioning module 303 configured to calculate and obtain initial pose information of the mobile device according to the corresponding relationship between the first identification mark and the second identification mark;
  • the coarse positioning module 303 includes:
  • the first construction submodule is configured to determine the initial pose information of the movable device according to the corresponding relationship and the conversion relationship from the sensor coordinate system of the movable device to the map coordinate system of the global map.
  • the fine positioning module 304 is configured to determine target pose information of the movable device according to the initial pose information, the global map, and the sensor data.
  • the fine positioning module 304 includes:
  • a search area submodule configured to determine a local search area in the global map through the initial pose information
  • the fine positioning sub-module is used to use the sensor data to perform a pose matching operation in the local search area, and obtain the target pose information according to the pose information with the highest matching degree.
  • the senor includes a laser radar sensor, the sensor data includes laser point cloud data, the identification mark includes a reflective identification mark, and the global map includes a grid map constructed according to the laser point cloud data;
  • Modules include:
  • a candidate unit configured to determine a plurality of candidate pose information for the local search area according to the sensor data, the size of the search window and the search step;
  • the hit area unit is used to determine the grid hit by the laser point cloud data corresponding to each candidate pose information in the local search area;
  • a matching degree unit configured to calculate the average occupancy probability of all the grids hit, and use the average occupancy probability as the matching degree of the candidate pose information
  • a determining unit configured to obtain the target pose information according to the candidate pose information with the highest matching degree.
  • the conditions for terminating the matching include: all matches between the first identification and the second identification are completed, the matching duration is longer than or It is equal to the preset time length, and the obtained number of corresponding relationships is greater than or equal to any one of the corresponding relationship number thresholds.
  • the device also includes:
  • a local map module configured to construct a local map according to the sensor data when the number of the first identification marks is less than a threshold value
  • a grayscale module configured to convert the local map into a local grayscale, and convert the global map into a global grayscale
  • a matching calculation module configured to perform matching calculation on the local grayscale image and the global grayscale image to obtain at least one pose information and a matching score corresponding to each pose information
  • a scoring module configured to use the pose information with the largest matching score as the initial pose information.
  • the embodiment of the present disclosure provides a device for relocating mobile equipment, including: acquiring sensor data collected by sensors in the working environment; determining the number of first identification marks in the working environment according to the sensor data is greater than or equal In the case of a preset quantity threshold, from the second identifications recorded in the global map of the working environment, determine the second identification that matches the first identification; the global map records the identifications arranged in the working environment Position: Calculate the initial pose information of the mobile device according to the correspondence between the first identification mark and the second identification mark; determine the target pose information of the mobile device according to the initial pose information, global map, and sensor data.
  • multiple easily identifiable identification signs can be arranged in the working environment of the mobile device, and the similarity between similar local areas can be reduced by deploying identification signs in similar working environments, thereby reducing the singularity of relocation The incidence of sexual problems.
  • rough positioning can be performed based on the identification mark, so that the process of rough positioning is not affected by environmental changes, and the positioning accuracy is improved.
  • an embodiment of the present disclosure also provides an apparatus, specifically referring to FIG.
  • the processor 610 implements various processes of the embodiment of the method for relocating a mobile device in the foregoing embodiments, and can achieve the same technical effect. To avoid repetition, details are not repeated here.
  • An embodiment of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, each process of the foregoing embodiment of the method for relocating a mobile device is implemented, and The same technical effect can be achieved, so in order to avoid repetition, details will not be repeated here.
  • the computer-readable storage medium can be a read-only memory (Read-Only Memory, referred to as ROM), a random access memory (Random Access Memory, referred to as RAM), a magnetic disk or an optical disk, etc.
  • An embodiment of the present disclosure also provides a computer program, which can be stored in the cloud or on a local storage medium.
  • the computer program is run by a computer or a processor, it is used to execute the corresponding steps of the relocation method of the mobile device according to the embodiment of the present disclosure, and is used to realize the corresponding steps in the relocation device of the mobile device according to the embodiment of the present disclosure. module.
  • the device embodiments described above are only illustrative, and the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed to multiple network elements. Part or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. It can be understood and implemented by those skilled in the art without any creative efforts.
  • the various component embodiments of the present disclosure may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof.
  • a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all functions of some or all components in the computing processing device according to the embodiments of the present disclosure.
  • DSP digital signal processor
  • the present disclosure can also be implemented as an apparatus or apparatus program (eg, computer program and computer program product) for performing a part or all of the methods described herein.
  • Such a program realizing the present disclosure may be stored on a computer-readable medium, or may have the form of one or more signals.
  • Such a signal may be downloaded from an Internet site, or provided on a carrier signal, or provided in any other form.
  • FIG. 7 illustrates a computing processing device that may implement methods according to the present disclosure.
  • the computing processing device conventionally includes a processor 1010 and a computer program product or computer readable medium in the form of memory 1020 .
  • Memory 1020 may be electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read Only Memory), EPROM, hard disk, or ROM.
  • the memory 1020 has a storage space 1030 for program code 1031 for performing any method steps in the methods described above.
  • the storage space 1030 for program codes may include respective program codes 1031 for respectively implementing various steps in the above methods. These program codes can be read from or written into one or more computer program products.
  • These computer program products comprise program code carriers such as hard disks, compact disks (CDs), memory cards or floppy disks.
  • Such a computer program product is typically a portable or fixed storage unit as described with reference to FIG. 8 .
  • the storage unit may have storage segments, storage spaces, etc. arranged similarly to the memory 1020 in the computing processing device of FIG. 7 .
  • the program code can eg be compressed in a suitable form.
  • the storage unit includes computer readable code 1031', i.e. code readable by, for example, a processor such as 1010, which code, when executed by a computing processing device, causes the computing processing device to perform the above-described methods. each step.
  • connection should be interpreted in a broad sense, for example, it can be a fixed connection or a detachable connection , or integrally connected; may be mechanically connected, may also be electrically connected; may be directly connected, may also be indirectly connected through an intermediary, and may be internal communication between two components.
  • the disclosed systems, devices and methods may be implemented in other ways.
  • the device embodiments described above are only illustrative.
  • the division of the units is only a logical function division.
  • multiple units or components can be combined or May be integrated into another system, or some features may be ignored, or not implemented.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
  • each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit.
  • the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium executable by a processor.
  • the technical solution of the present disclosure is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including Several instructions are used to make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods described in various embodiments of the present disclosure.
  • the aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-OnlyMemory), random access memory (RAM, RandomAccessMemory), magnetic disk or optical disk and other media that can store program codes.
  • references herein to "one embodiment,” “an embodiment,” or “one or more embodiments” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure. Additionally, please note that examples of the word “in one embodiment” herein do not necessarily all refer to the same embodiment.
  • any reference signs placed between parentheses shall not be construed as limiting the claim.
  • the word “comprising” does not exclude the presence of elements or steps not listed in a claim.
  • the word “a” or “an” preceding an element does not exclude the presence of a plurality of such elements.
  • the disclosure can be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means can be embodied by one and the same item of hardware.
  • the use of the words first, second, and third, etc. does not indicate any order. These words can be interpreted as names.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Electromagnetism (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Automation & Control Theory (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

一种可移动设备的重定位方法及装置，包括：获取传感器在工作环境中采集的传感器数据（101）；在根据传感器数据确定的第一识别标识的数量大于或等于数量阈值的情况下，从工作环境的全局地图中所记录的第二识别标识中，确定与第一识别标识匹配的第二识别标识（102）；根据第一识别标识与第二识别标识的对应关系，计算得到可移动设备的初始位姿信息（103）；根据初始位姿信息、全局地图、传感器数据，确定可移动设备的目标位姿信息（104）。通过在相似工作环境中部署识别标识可以降低相似局部区域的之间相似性，从而降低重定位的奇异性问题的发生几率。针对时常发生较大变化的工作场景，使得粗定位的过程不受环境变化的影响，提高了定位精度。

Description

可移动设备的重定位方法及装置
本公开要求在2021年07月29日提交中国专利局、公开号为202110867220.4发明名称为“可移动设备的重定位方法及装置”的中国专利公开的优先权，其全部内容通过引用结合在本公开中。
技术领域
本公开一般地涉及定位技术领域，特别是涉及一种可移动设备的重定位方法及装置。
背景技术
可移动设备的重定位是指可移动设备在工作初始时刻，或在丢失位姿(位置和移动方向)的情况下，重新确定位姿的过程。
在目前，可移动设备可以采用激光雷达实现重定位，激光雷达重定位是指，将激光雷达当前采集的点云数据，与针对可移动设备的工作环境预设的点云地图进行各种位姿的匹配，从而计算各种位姿的匹配得分，最终将得分满足需求的位姿确定为最终位姿。
但是，目前方案中，点云地图是预先建立的，在工作环境发生大比例变动的情况下，点云地图与工作环境之间存在较大差异，采用点云地图进行重定位时会造成定位精度大幅下降或者定位失败，另外，在相似工作环境中，如长走廊、工作环境存在若干个相似的局部区域，由于环境中局部区域之间的相似性，可能会导致重定位时输出错误位姿。
发明内容
本公开提供一种可移动设备的重定位方法及装置，以便解决现有技术中重定位时会造成定位精度大幅下降或者定位失败的问题。
为了解决上述技术问题，本公开是这样实现的：
第一方面，本公开实施例提供了一种可移动设备的重定位方法，所述可移动设备的重定位方法包括：
获取传感器在工作环境中采集的传感器数据；
在根据所述传感器数据确定的第一识别标识的数量大于或等于数量阈值的情况下，从所述工作环境的全局地图中所记录的第二识别标识中，确定与所述第一识别标识匹配的第二识别标识；所述第一识别标识为所述工作环境中设置的识别标识；
根据所述第一识别标识与所述第二识别标识的对应关系，计算得到所述可移动设备的初始位姿信息；
根据所述初始位姿信息、所述全局地图、所述传感器数据，确定所述可移动设备的目标位姿信息。
第二方面，本公开实施例提供了一种可移动设备的重定位装置，所述可移动设备的重定位装置包括：
获取模块，用于获取传感器在工作环境中采集的传感器数据；
匹配模块，用于在根据所述传感器数据确定的第一识别标识的数量大于或等于数量阈值的情况下，从所述工作环境的全局地图中所记录的第二识别标识中，确定与所述第 一识别标识匹配的第二识别标识；所述第一识别标识为所述工作环境中设置的识别标识；
粗定位模块，用于根据所述第一识别标识与所述第二识别标识的对应关系，计算得到所述可移动设备的初始位姿信息；
精定位模块，用于根据所述初始位姿信息、所述全局地图、所述传感器数据，确定所述可移动设备的目标位姿信息。
本公开实施例的第三方面，提供了一种计算处理设备，其特征在于，包括：存储器，其中存储有计算机可读代码；一个或多个处理器，当所述计算机可读代码被所述一个或多个处理器执行时，所述计算处理设备执行所述的可移动设备的重定位方法。
本公开实施例的第四方面，提供了一种计算机程序，包括计算机可读代码，当所述计算机可读代码在计算处理设备上运行时，所述计算处理设备执行所述的可移动设备的重定位方法。
本申请实施例的第五方面，提供了一种计算机可读介质，其中存储了所述的计算机程序。
在本公开实施例中，本公开包括：获取传感器在工作环境中采集的传感器数据；在根据传感器数据确定工作环境中的第一识别标识的数量大于或等于预设数量阈值的情况下，从工作环境的全局地图中所记录的第二识别标识中，确定与第一识别标识匹配的第二识别标识；全局地图中记录有工作环境中布置的识别标识的位置；根据第一识别标识与第二识别标识的对应关系，计算得到可移动设备的初始位姿信息；根据初始位姿信息、全局地图、传感器数据，确定可移动设备的目标位姿信息。在本公开中，可以在可移动设备的工作环境中布置多个易于被识别的识别标识，通过在相似工作环境中部署识别标识可以降低相似局部区域的之间相似性，从而降低重定位的奇异性问题的发生几率。针对时常发生较大变化的工作场景，由于场景中识别标识的位置固定不变，则可以基于识别标识进行粗定位，使得粗定位的过程不受环境变化的影响，提高了定位精度。
上述说明仅是本公开技术方案的概述，为了能够更清楚了解本公开的技术手段，而可依照说明书的内容予以实施，并且为了让本公开的上述和其它目的、特征和优点能够更明显易懂，以下特举本公开的具体实施方式。
附图说明
为了更清楚地说明本公开实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图是本公开的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本公开实施例提供的一种可移动设备的重定位方法的步骤示意图；
图2是本公开实施例提供的一种工作环境的示意图；
图3是本公开实施例提供的另一种工作环境的示意图；
图4是本公开实施例提供的一种可移动设备的重定位方法的具体步骤示意图；
图5是本公开实施例提供的一种可移动设备的重定位装置的结构框图；
图6是本公开实施例提供的一种装置的框图；
图7示意性地示出了用于执行根据本公开的方法的计算处理设备的框图；
图8示意性地示出了用于保持或者携带实现根据本公开的方法的程序代码的存储单元。
具体实施例
为使本公开实施例的目的、技术方案和优点更加清楚，下面将结合本公开实施例中的附图，对本公开实施例中的技术方案进行描述。
随着物联网、人工智能、大数据等智能化技术的发展，利用这些智能化技术对传统物流业进行转型升级的需求愈加强劲，智慧物流(ILS，Intelligent Logistics System)成为物流领域的研究热点。智慧物流利用人工智能、大数据以及各种信息传感器、射频识别技术、全球定位系统(GPS，Global Positioning System)等物联网装置和技术，广泛应用于物料的运输、仓储、配送、包装、装卸和信息服务等基本活动环节，实现物料管理过程的智能化分析决策、自动化运作和高效率优化管理。物联网技术包括传感设备、射频识别(RFID，Radio Frequency Identification)技术、激光红外扫描、红外感应识别等，物联网能够将物流中的物料与网络实现有效连接，并可实时监控物料，还可感知仓库的湿度、温度等环境数据，保障物料的储存环境。通过大数据技术可感知、采集物流中所有数据，上传至信息平台数据层，对数据进行过滤、挖掘、分析等作业，最终对业务流程(如运输、入库、存取、拣选、包装、分拣、出库、盘点、配送等环节)提供精准的数据支持。人工智能在物流中的应用方向可以大致分为两种：1)以人工智能技术赋能的如无人卡车、自动导引运输车(AGV，Automated Guided Vehicle)、自主移动机器人(AMR，Autonomous Mobile Robot)、叉车、穿梭车、堆垛机、无人配送车、无人机、服务机器人、机械臂、智能终端等智能设备代替部分人工；2)通过计算机视觉、机器学习、运筹优化等技术或算法驱动的如运输设备管理系统、仓储管理、设备调度系统、订单分配系统等软件系统提高人工效率。随着智慧物流的研究和进步，该项技术在众多领域展开了应用，例如零售及电商、电子产品、烟草、医药、工业制造、鞋服、纺织、食品等领域。
图1是本公开实施例提供的一种可移动设备的重定位方法的步骤流程图，如图1所示，该方法可以包括：
步骤101、获取传感器在工作环境中采集的传感器数据。
在具体应用中，可移动设备的重定位过程包括粗定位和精定位。粗定位是指提供一个可移动设备的大致的位姿(位姿包括所处位置和方向)，精定位则可以根据粗定位结果，缩小位姿搜索范围，得到最终位姿。
可选的，识别标识用于在所述工作环境中区分于其他物体，提供路标参考。
可选的，所述识别标识包括反光识别标识、或标识符。
在本公开提供的一种可移动设备的重定位方法应用于具有传感器的可移动设备，可移动设备的工作环境中布置有多个识别标识，其中，识别标识具有较强的被识别特性，可以被可移动设备的传感器所识别到，识别标识的作用是在工作环境中区分于其他物体，达到提供一个路标参考的作用，如通过在相似工作环境中部署识别标识可以降低相似局部区域的之间相似性，降低重定位的奇异性问题(即重定位输出多个位姿)的发生几率。针对时常发生较大变化的工作场景，由于场景中识别标识的位置固定不变，则可以基于识别标识进行粗定位，使得粗定位的过程不受环境变化的影响，提高了定位精度。
一种实现情况下，若传感器为激光雷达，则根据激光雷达的工作特性，可以将识别标识设定为可反光的反光识别标识，反光识别标识的表面采用高折射率的材料，如，反光板、反光柱等，激光雷达的激光光束发射至反光识别标识表面，反射回来的光束被激光雷达接收到后，激光雷达可以计算与反光识别标识之间的距离以及反射强度，从而识别到反光识别标识。
另一种实现情况下，若传感器为摄像头，则可以将识别标识设定为较显眼的标识符，如，特殊图像符号，具有鲜艳颜色的特定图案等，摄像头可以基于采集的图像中包含的识别标识，实现对识别标识的识别。
以下通过图2和图3对可移动设备所处的工作环境中，识别标识的布局方式进行举例。
参照图2，其示出了本公开实施例提供的一种工作环境的示意图，其示出了一条前进方向为X方向的长走廊场景10，长走廊场景10中，由于走廊的长度超出了传感器的测量距离，则可移动设备处于这种场景时，仅能对左右方向的墙壁进行测量，而难以在前后方向上进行测量，导致可移动设备难以确定一个位姿初值，使得在将采集的传感器数据与全局地图匹配时，会输出多个位姿结果，造成重定位奇异性问题。为了解决这个问题，本公开实施例可以在长走廊场景10中，布局多个识别标识20(至少3个)，使得在具有相似性的长走廊中，通过识别标识区别于相似部分，降低奇异性问题的发生几率。
参照图3，其示出了本公开实施例提供的另一种工作环境的示意图，其示出了一种具有局部相似性的场景30，该场景30中，存在区域31和区域32，区域31和区域32非常相似，即二者的布局、环境相似，可移动设备处于这种场景时，易造成重定位奇异性问题。为了解决这个问题，本公开实施例可以在场景30中，布局多个识别标识20(至少3个)，使得在具有相似性的场景30中，通过识别标识区别于相似部分，降低奇异性问题的发生几率。
在该步骤中，可移动设备在工作初始时刻无初始的位姿，或可移动设备丢失自身位姿的情况下，可以基于定位指令控制可移动设备进行重定位，可移动设备可以响应于重定位的定位指令，获取传感器在工作环境中采集的传感器数据，其中，传感器数据可以包括可移动设备所识别到的第一识别标识的数量和位置。
如，场景中布置了10个定位标识，可移动设备在某一时刻通过传感器识别到3个定 位标识，则传感器采集的传感器数据中包含着3个定位标识的位置。
其中，传感器数据可以具有多种类型，在传感器为摄像头时，传感器数据可以为摄像头采集的图像；在传感器为激光雷达时，传感器数据可以为点云数据，点云数据是指在一个三维坐标系统中的一组位置向量的集合，一些情况下，点云数据除了具有几何位置以外，还可以有颜色信息和强度信息，颜色信息通常是通过相机获取彩色影像，然后将对应位置的像素的颜色信息赋予点云中对应的点；强度信息的获取是激光雷达采集到的回波强度。
步骤102、在根据所述传感器数据确定工作环境中的第一识别标识的数量大于或等于预设数量阈值的情况下，从所述工作环境的全局地图中所记录的第二识别标识中，确定与所述第一识别标识匹配的第二识别标识。
所述全局地图中记录有所述工作环境中布置的识别标识的位置；第一识别标识为所述工作环境中设置的识别标识。
在本公开实施例中，可以针对传感器所识别到的第一识别标识的数量，来决定粗定位的方式，若要通过利用识别标识来进行粗定位，则需要所识别到的第一识别标识的数量大于或等于预设数量阈值(如3个)，若所识别到的第一识别标识的数量小于预设数量阈值，则无法利用识别标识来进行粗定位，则可以选取其他的粗定位方式，如灰度匹配定位。
进一步的，在第一识别标识的数量大于或等于预设数量阈值的情况下，可以利用识别标识来进行粗定位，这种粗定位方式是将由识别到的两个第一识别标识组成的标识组，与由预设的全局地图中记录的两个第二识别标识组成的标识组进行匹配，从而确定由第一识别标识的位置和第二识别标识的位置组成的匹配对，最后根据这些匹配对，基于最小二乘算法，求解初始位姿信息。
具体的，传感器所识别到的第一识别标识的位置，可以反映可移动设备所处的大致区域，通过将传感器坐标系下两个第一识别标识组成的标识组，与地图坐标系下两个第二识别标识组成的标识组进行匹配，可以得到传感器坐标系下第一识别标识与地图坐标系下对应的第二识别标识的匹配对，标识组之间的匹配过程，反映了传感器坐标系下标识组代表的局部位置，与地图坐标系下与该局部位置相似的位置之间的匹配。匹配对即反映了两个坐标系下相似位置之间的对应关系，基于匹配对和最小二乘算法，可以得到用于求解初始位姿信息的方程式，从而求得初始位姿信息。
其中，全局地图是根据可移动设备的工作环境预先设置的地图，在可移动设备执行任务之前，可以根据工作环境的传感器数据(如环境照片，环境点云数据)，构建全局地图。全局地图可以根据实际需求进行更新，如每隔预设时间重新构建一次全局地图，或，在工作环境发生较大变化的情况下，更新全局地图。
步骤103、根据所述第一识别标识与所述第二识别标识的对应关系，计算得到所述可移动设备的初始位姿信息。
在本公开实施例中，匹配对包括一个第一识别标识与一个第二识别标识的对应关系，由于匹配对反映了两个坐标系下相似位置之间的对应关系，则所得到的所有匹配对，反映了传感器坐标系下的局部位置，与地图坐标系下对应的相似位置之间的对应关系，得到这种匹配的对应关系后，可以基于最小二乘算法构建用于求解初始位姿信息的方程式，从而求得初始位姿信息。其中，最小二乘法是一种在误差估计、不确定度、系统辨识及预测、预报等数据处理诸多学科领域得到广泛应用的数学工具，其通过最小化误差的平方和寻找数据的最佳函数匹配，利用最小二乘法可以简便地求得未知的数据，并使得这些求得的数据与实际数据之间误差的平方和为最小，在本公开实施例中，可以根据第一识别标识与第二识别标识的对应关系，构建最小二乘方程式，并通过求解最小二乘方程式，将求得的解作为初始位姿信息。初始位姿信息是可移动设备通过粗定位得到一个位姿初始值，其反映了可移动设备所处的大致位置和方向，由于其精度还不足以满足可移动设备的导航使用，因此后续还需要基于该初始位姿信息进行精定位，从而得到精确的目标位姿信息。
步骤104、根据所述初始位姿信息、所述全局地图、所述传感器数据，确定所述可移动设备的目标位姿信息。
在获得了粗定位的初始位姿信息结果后，可移动设备可以进一步进行精定位，即以初始位姿信息为基础，进行传感器采集的传感器数据与全局地图的匹配，从而求解得到可移动设备最终的目标位姿信息，目标位姿信息可以为可移动设备在精定位后得到的精确位姿，其反映了可移动设备的精确位置和方向，可移动设备可以以目标位姿信息为当前的导航位姿，进行后续的工作。
具体的，可以在全局地图中初始位姿附近设置一个搜索窗口，搜索窗口中按照搜索步长可以设定有多个候选位姿，在传感器为激光雷达传感器的情况下，基于初始位姿信息和每个候选位姿的组合，可以得到经过候选位姿变换过的激光点云，将该经过候选位姿变换过的激光点云与全局地图进行匹配，可以得到该候选位姿的得分，得分最高的候选位姿再经过插值和平滑函数的优化后，可以构建用于求解目标位姿信息的最小二乘方程，从而求解得到目标位姿信息。
综上，本公开实施例提供的一种可移动设备的重定位方法，包括：获取传感器在工作环境中采集的传感器数据；在根据传感器数据确定工作环境中的第一识别标识的数量大于或等于预设数量阈值的情况下，从工作环境的全局地图中所记录的第二识别标识中，确定与第一识别标识匹配的第二识别标识；全局地图中记录有工作环境中布置的识别标识的位置；根据第一识别标识与第二识别标识的对应关系，计算得到可移动设备的初始位姿信息；根据初始位姿信息、全局地图、传感器数据，确定可移动设备的目标位姿信息。在本公开中，可以在可移动设备的工作环境中布置多个易于被识别的识别标识，通过在相似工作环境中部署识别标识可以降低相似局部区域的之间相似性，从而降低重定位的奇异性问题的发生几率。针对时常发生较大变化的工作场景，由于场景中识别标识 的位置固定不变，则可以基于识别标识进行粗定位，使得粗定位的过程不受环境变化的影响，提高了定位精度。
图4是本公开实施例提供的一种可移动设备的重定位方法的具体步骤流程图，如图4所示，该方法可以包括：
步骤201、获取传感器在工作环境中采集的传感器数据。
该步骤具体可以参照上述步骤101，此处不再赘述。
可选的，所述传感器包括激光雷达传感器，所述传感器数据包括激光点云数据，所述识别标识包括反光识别标识，所述全局地图包括根据激光点云数据构建的栅格地图。
在本公开实施例的一种实现方式下，传感器可以为激光雷达传感器，则根据激光雷达的工作特性，可以将识别标识设定为可反光的反光识别标识，反光识别标识的表面采用高折射率的材料，如，反光板、反光柱等，激光雷达的激光光束发射至反光识别标识表面，反射回来的光束被激光雷达接收到后，激光雷达可以计算与反光识别标识之间的距离以及反射强度，从而识别到反光识别标识。全局地图则可以为根据激光点云数据构建的栅格地图。
其中，栅格地图也称为占据栅格地图(Occupancy Grid Map)占据栅格地图是将工作环境划分成一定大小的栅格，每个栅格有两种状态：占用和空闲，两种状态的概率之和为1，每个栅格内存储有栅格的占据比(占用概率/空闲概率)。建立栅格地图的过程中，激光点云会落入不同栅格内，因此建立栅格地图的过程其实是在更新每个栅格被占用的概率。具体的，激光雷达传感器每次发射光束扫描环境，其视野范围内的栅格的概率都会被更新，一个栅格被激光击中次数越多，表示它被占用的概率越高。
步骤202、在根据所述传感器数据确定工作环境中的第一识别标识的数量大于或等于预设数量阈值的情况下，从所述工作环境的全局地图中所记录的第二识别标识中，确定与所述第一识别标识匹配的第二识别标识。
其中，所述全局地图中记录有所述工作环境中布置的识别标识的位置；所述第一识别标识为所述工作环境中设置的识别标识。
该步骤具体可以参照上述步骤102，此处不再赘述。
可选的，步骤202具体可以包括：
子步骤2021、分别计算两个所述第一识别标识间的第一线段，与两个所述第二识别标识间的第二线段之间的相似度。
子步骤2022、确定相似度大于或等于相似度阈值的目标第一线段和目标第二线段，其中，所述目标第一线段两端的第一识别标识与所述目标第二线段两端的第二识别标识一一对应。
在本公开实施例中，假设工作环境中布置了m个识别标识，传感器所扫描到的n个第一识别标识构成的集合为Li＝{L1、L2....Ln}，全局地图中记录的多个第二识别标识构成的集合为Mi＝{M1、M2....Mm}。本公开实施例可以基于传感器坐标系下的集合Li和 地图坐标系下的集合Mi进行粗定位，粗定位的方式是将由两个第一识别标识组成的第一线段，与两个第二识别标识组成的第二线段进行匹配，从而确定由第一识别标识的位置和第二识别标识的位置组成的匹配对，最后根据这些匹配对，基于最小二乘算法，求解初始位姿信息。
在该步骤中，可以通过计算两个第一识别标识构成的第一线段与两个第二识别标识构成的第二线段之间的相似度，来确定第一识别标识与对应第二识别标识之间的匹配程度，基于相似度大于或等于预设相似度阈值的目标第一线段和目标第二线段，可以从构成目标第一线段和目标第二线段的识别标识中，选取由第一识别标识与第二识别标识构成的对应关系。
可选的，每个所述第一线段为第一基准识别标识与任一其他第一识别标识之间的线段，所述第一基准标识是根据所述传感器数据确定的多个第一识别标识之一；每个所述第二线段为第二基准识别标识与任一其他第二识别标识之间的线段，所述第二基准标识是所述全局地图中所记录的多个第二识别标识之一，所述子步骤2021包括：
子步骤A1、根据所述第一线段的第一长度和所述第二线段的第二长度的差值的绝对值，以及所述第一长度和所述第二长度中的较小值，分别计算所述第一线段与所述第二线段之间的相似度。
在本公开实施例中，对第一线段和第二线段的建立过程，先通过以下示例进行说明：
假设工作环境中布置了m个识别标识，传感器所扫描到的n个第一识别标识构成的集合为Li＝{L1、L2....Ln}，全局地图中记录的多个第二识别标识构成的集合为Mi＝{M1、M2....Mm}。
第一步可以从集合Li中随机提取一个第一反光标识La，从集合Mi中随机提取一个第二反光标识Mm，并将{La，Mm}作为一个匹配对，同时可以将第一反光标识La作为基准第一识别标识，以及将第二反光标识Mm作为基准第二识别标识。
第二步可以从集合Li的剩余第一反光标识中随机选取一个第一反光标识Lb，从集合Mm的剩余第二反光标识中随机选取一个第二反光标识Mn，并基于基准第一反光标识La和第一反光标识Lb建立一个第一线段{La，Lb}，以及基于基准第二反光标识Mm和第二反光标识Mn建立一个第二线段{Mm，Mn}。之后可以计算第一线段{La，Lb}和第二线段{Mm，Mn}之间的相似度，在计算完这两个标识组的相似度后，可以进一步从集合Li中除了基准第一反光标识La、第一反光标识Lb之外的剩余第一反光标识中随机选取一个第一反光标识Lc，从集合Mm中除了基准第二反光标识Mm、第二反光标识Mn之外剩余的第二反光标识中随机选取一个第二反光标识Mo，然后建立第一线段{La，Lc}，以及第二线段{Mm，Mo}，再进行第一线段{La，Lc}以及第二线段{Mm，Mo}之间的相似度计算。之后重复上述建立标识组的操作，直至达到终止条件。
本公开实施例中，由于第一线段{La，Lb}记录了传感器坐标系下基准第一识别标识La和第一识别标识Lb的位置坐标，则根据La、Lb的位置坐标可以构成一条第一线段line1； 由于第二线段{Mm，Mn}记录了地图坐标系下基准第二识别标识Mm和第二识别标识Mn的位置坐标，则根据Mm、Mn的位置坐标可以构成一条第二线段line2。并且，由于一条线段在不同坐标系下的绝对长度是相同的，因此可以计算第一线段line1和第二线段line2的相似度。对于示例之外的其他第一线段和第二线段同理。
具体的，第一线段line1和第二线段line2的相似度计算可以参照下述公式1：
Figure PCTCN2022094718-appb-000001
其中，S为第一线段line1和第二线段line2的相似度，d为第一线段line1的第一长度，d’为第二线段line2的第二长度，δ为最大容忍长度差值，可以根据实际需求进行取值。min(d，d’)是指取d和d’中的最小值。
子步骤2022包括：
子步骤B1、将所述基准第一识别标识与所述基准第二识别标识组成一个所述对应关系。
子步骤B2、将构成所述目标第一线段的两个第一识别标识中除了基准第一识别标识的另一个第一识别标识，与构成所述目标第二线段的两个第二识别标识中除了基准第二识别标识的另一个第二识别标识组成另一个所述对应关系。
在本公开实施例中，针对上述对第一线段{La，Lb}和第二线段{Mm，Mn}示例，假设第一线段{La，Lb}和第二线段{Mm，Mn}的相似度大于预设相似度阈值，则除了将其中的基准第一识别标识La和基准第二识别标识Mm构成一个对应关系{La，Mm}之外，还可以将第一线段{La，Lb}中的另一个第一识别标识Lb与第二线段{Mm，Mn}中的另一个第二识别标识Mn组成另一个对应关系{Lb，Mn}。对于示例之外的其他目标第一线段和目标第二线段同理。
可选的，在确定与所述第一识别标识匹配的第二识别标识的过程中，终止匹配的条件包括：完成了所有第一识别标识和第二识别标识之间的匹配、匹配时长大于或等于预设时长、得到的所述对应关系的数量大于或等于对应关系数量阈值中的任意一种。
在本公开实施例中，确定匹配对的过程中，存在匹配终止条件，如，在完成了所有第一识别标识和第二识别标识之间的匹配时，终止匹配；或匹配时长大于或等于预设时长时，为了避免匹配时长过长而影响到用户体验，可以终止匹配；或按照实际需求，在得到的对应关系的数量大于或等于匹配对数量阈值时，认为得到的对应关系已可以进行粗定位，匹配终止。
步骤203、根据所述第一识别标识与所述第二识别标识的对应关系，计算得到所述可移动设备的初始位姿信息。
该步骤具体可以参照上述步骤103，此处不再赘述。
可选的，步骤203具体可以通过根据所述对应关系，以及所述可移动设备的传感器坐标系到所述全局地图的地图坐标系的转换关系，确定所述可移动设备的初始位姿信息的方式实现，具体实现过程包括：
子步骤2031、根据所述对应关系、最小二乘求解算法，以及所述可移动设备的传感器坐标系到所述全局地图的地图坐标系的转换关系，构建所述初始位姿信息的第一求解方程表达式。
子步骤2032、根据所述第一求解方程表达式，计算得到所述可移动设备的初始位姿信息。
在本公开实施例中，可移动设备的导航需要基于地图坐标系进行实现，由于可移动设备的传感器采集的传感器数据是传感器坐标系下的，则可以基于最小二乘求解算法、对应关系、以及传感器坐标系到地图坐标系的转换关系，构建用于求解初始位姿信息的第一求解方程表达式。
例如，假设步骤202得到了三组对应关系{Li，Mi}，i＝1、2、3。Li是传感器坐标系下的识别标识的位置坐标，定义为(Xli，Yli)；Mi是地图坐标系下的识别标识的位置坐标，定义为(Xmi，Ymi)。要求解的初始位姿信息是可移动设备在地图中的位置(x，y)以及角度θ，由于Li和Mi之间存在匹配关系，则将Li的坐标由传感器坐标系转换到地图坐标系，得到的结果应该是Mi，由此可以得到以下公式2。
Figure PCTCN2022094718-appb-000002
其中，
Figure PCTCN2022094718-appb-000003
表示传感器坐标系到地图坐标系的转换关系。
Figure PCTCN2022094718-appb-000004
基于
Figure PCTCN2022094718-appb-000005
和公式2，可以求解三个未知量(x，y，θ)，基于最小二乘求解算法构建三个及以上的方程，即求得可移动设备在地图中的位置(x，y)以及角度θ，从而能够得到初始位姿信息
步骤204、根据所述初始位姿信息、所述全局地图、所述传感器数据，确定所述可移动设备的目标位姿信息。
该步骤具体可以参照上述步骤104，此处不再赘述。
可选的，步骤204具体可以包括：
子步骤2041、通过所述初始位姿信息，在所述全局地图中确定局部搜索区域。
子步骤2042、采用所述传感器数据，在所述局部搜索区域中进行位姿匹配操作，并 根据匹配度最大的位姿信息得到所述目标位姿信息.
在本公开实施例中，基于粗定位得到初始位姿信息进行进一步的精定位，可以得到用于指导可移动设备进行导航的目标位姿信息，可以将精定位过程描述为：给定一个初始位姿附近的搜索窗口，按照窗口大小及搜索步长确定多个候选位姿，通过匹配得到一个最优位姿，使得传感器数据出现的概率最大化。
在该步骤中，可以在初始位姿附近设置一个搜索窗口w。给定搜索窗口大小和搜索步长，可以得到多个处于所述全局地图的地图坐标系下的候选位姿的表达式。候选位姿信息的数量与搜索窗口的大小和搜索步长有关。
在该步骤中，定义传感器数据为{hk}，在传感器为激光雷达传感器时，hk是指第k个激光点在激光雷达坐标系下的位置坐标，对于任意一个候选位姿信息Tε，得到对应的处于全局地图的地图坐标系下的激光点云位置坐标表达式如下：
Figure PCTCN2022094718-appb-000006
其中，εx，εy分别是可移动设备在地图坐标系下的坐标，εθ是可移动设备的方向角，由于精定位过程可以被描述为给定一个初始位姿附近的搜索窗口，按照窗口大小及搜索步长确定多个候选位姿，通过匹配得到一个最优位姿，使得传感器数据出现的概率最大化，因此可以将其总结为如下非线性优化问题：
Figure PCTCN2022094718-appb-000007
其中，w为一个搜索窗口，Mnearest(Tεhk)是全局地图中离Tεhk最近的栅格单元的占用概率。传感器数据为hk，初始位姿信息为Tε。
可选的，所述传感器包括激光雷达传感器，所述传感器数据包括激光点云数据，所述识别标识包括反光识别标识，所述全局地图包括根据激光点云数据构建的栅格地图，子步骤2042具体可以包括：
子步骤C1、根据所述传感器数据、搜索窗口的大小以及搜索步长，确定针对所述局部搜索区域的多个候选位姿信息。
子步骤C2、在所述局部搜索区域中，确定每个所述候选位姿信息对应的激光点云数据所击中的栅格。
子步骤C3、计算击中的所有的栅格的占用概率平均值，并将所述占用概率平均值作为所述候选位姿信息的匹配度。
子步骤C4、根据匹配度最大的候选位姿信息得到所述目标位姿信息。
进一步的，需要从多个候选位姿信息中选取一个最佳的候选位姿信息，每个候选位姿信息与全局地图匹配会计算一个得分。最佳的候选位姿信息的评价标准是得分最高。 在传感器为激光雷达传感器的情况下，得分计算是所有激光点云击中全局地图中栅格的占用概率平均值。所以得分越高，激光点云出现概率越高，候选位姿信息越接近真实位姿。这部分在具体实现时可以用到一个加速搜索的策略叫分支定界。
在该步骤中，可以先确定每个位姿表达式在全局地图中对应的离散区域，离散区域反映了激光点云击中的位置，之后可以将离散区域包含的所有栅格的占用概率平均值(反映了该候选位姿下所有激光点云击中全局地图中栅格的占用概率均值)，作为该位姿表达式对应的候选位姿信息的得分。
在本公开实施例中，由于得到的得分最大的候选位姿信息是栅格分辨率级别的精度，考虑到地图栅格的分辨率导致的精度有限，若希望更近一步优化精度，则需要进行插值(M smooth函数该函数的输出结果为栅格被占用的概率，是(0，1)以内的数)，插值算法具体可以为双三次插值算法。通过插值的优化，能够提供比栅格分辨率更好的精度。这部分等价的最小二乘问题式子如下：希望栅格被占据概率最大，等价于栅格不被占用的概率最小。基于最小二乘算法构建的第二求解方程表达式，可以得到目标位姿信息。
Figure PCTCN2022094718-appb-000008
其中，M smooth函数为插值函数，传感器数据为hk，初始位姿信息为Tε。
可选的，在步骤201之后，所述方法还可以包括：
步骤205、在所述第一识别标识的数量小于数量阈值的情况下，根据所述传感器数据构建局部地图。
在本公开实施例中，若传感器识别到的第一识别标识的数量小于预设数量阈值(如3个)的情况下，则由于识别到的第一识别标识的数量太少，难以进行基于识别标识的粗定位过程，此时可以进行并行的另一种方案，即基于灰度匹配进行粗定位。
灰度匹配是利用灰度局部图和灰度全局图进行匹配，从而求解出位姿的过程，不同位姿具有不同的得分，得分越高的位姿的价值越大。
在该步骤中，可以先将传感器数据构建局部地图，即可移动设备为当前所处的局部环境构建一张局部地图，现有几种建立方式：1、将可移动设备上的激光雷达传感器旋转一定的角度得到的激光点云数据(相当于多帧激光)，构建为一个局部地图；2、将可移动设备上的激光雷达传感器转360°得到的激光点云(相当于多帧激光)，构建为一个局部地图，这种方式的扫描范围最大，使得后续匹配的成功率越大；3、将可移动设备定位姿得到的激光点云(即一帧激光)，构建为一个局部地图。
步骤206、将所述局部地图转换为局部灰度图，以及将所述全局地图转换为全局灰度图。
在该步骤中，基于灰度匹配需要利用局部灰度图的需求，本公开实施例可以将传感器数据构建局部地图转换为局部灰度图，即将局部地图的栅格内存储的占据比投影至[0， 255]的范围内，得到一张局部灰度图。基于灰度匹配需要利用全局灰度图的需求，本公开实施例可以将全局地图转换为全局灰度图，即将全局地图的栅格内存储的占据比投影至[0，255]的范围内，得到一张全局灰度图。
步骤207、将所述灰度局部图与所述全局灰度图进行匹配计算，得到至少一个位姿信息以及每个位姿信息对应的匹配得分。
在本公开实施例中，可以将灰度局部图与全局灰度图的各个位置进行灰度匹配，匹配过程会得到至少一个位姿信息以及每个位姿信息对应的匹配得分，匹配得分反映了位姿信息与全局地图的匹配程度。其中，灰度匹配是通过利用某种相似性度量，计算两幅图像之间的相似度。常用的基于灰度的匹配方法包括：平均绝对差算法、绝对误差和算法、误差平方和算法、平均误差平方和算法、归一化积相关算法等。
步骤208、将所述匹配得分最大的位姿信息作为所述初始位姿信息。
在本公开实施例中，匹配得分最大的位姿信息可以认为是最与全局地图匹配的位姿，在将匹配得分最大的位姿信息作为初始位姿信息之后，可以执行步骤204，以进行后续的精定位过程。
综上，本公开实施例提供的一种可移动设备的重定位方法，包括：获取传感器在工作环境中采集的传感器数据；在根据传感器数据确定工作环境中的第一识别标识的数量大于或等于预设数量阈值的情况下，从工作环境的全局地图中所记录的第二识别标识中，确定与第一识别标识匹配的第二识别标识；全局地图中记录有工作环境中布置的识别标识的位置；根据第一识别标识与第二识别标识的对应关系，计算得到可移动设备的初始位姿信息；根据初始位姿信息、全局地图、传感器数据，确定可移动设备的目标位姿信息。在本公开中，可以在可移动设备的工作环境中布置多个易于被识别的识别标识，通过在相似工作环境中部署识别标识可以降低相似局部区域的之间相似性，从而降低重定位的奇异性问题的发生几率。针对时常发生较大变化的工作场景，由于场景中识别标识的位置固定不变，则可以基于识别标识进行粗定位，使得粗定位的过程不受环境变化的影响，提高了定位精度。
图5是本公开实施例提供的一种可移动设备的重定位装置的结构框图，应用于具有传感器的可移动设备，所述可移动设备的工作环境中布置有多个识别标识，如图5所示，，所述可移动设备的重定位装置包括：
该装置可以包括：
获取模块301，用于获取传感器在工作环境中采集的传感器数据；
匹配模块302，用于在根据所述传感器数据确定的第一识别标识的数量大于或等于数量阈值的情况下，从所述工作环境的全局地图中所记录的第二识别标识中，确定与所述第一识别标识匹配的第二识别标识；所述第一识别标识为所述工作环境中设置的识别标识；
可选的，匹配模块302包括：
相似度子模块，用于分别计算两个所述第一识别标识间的第一线段，与两个所述第二识别标识间的第二线段之间的相似度；
可选的，每个所述第一线段为第一基准识别标识与任一其他第一识别标识之间的线段，所述第一基准标识是根据所述传感器数据确定的多个第一识别标识之一；
每个所述第二线段为第二基准识别标识与任一其他第二识别标识之间的线段，所述第二基准标识是所述全局地图中所记录的多个第二识别标识之一。
可选的，相似度子模块包括：
相似度单元，用于根据所述第一线段的第一长度和所述第二线段的第二长度的差值的绝对值，以及所述第一长度和所述第二长度中的较小值，计算所述第一线段与所述第二线段之间的相似度。
匹配对子模块，用于确定相似度大于或等于相似度阈值的目标第一线段和目标第二线段，其中，所述目标第一线段两端的第一识别标识与所述目标第二线段两端的第二识别标识一一对应。
可选的，匹配对子模块包括：
第一组合单元，用于将所述基准第一识别标识与所述基准第二识别标识组成一个所述对应关系；
第二组合单元，用于将构成所述目标第一线段的两个第一识别标识中除了基准第一识别标识的另一个第一识别标识，与构成所述目标第二线段的两个第二识别标识中除了基准第二识别标识的另一个第二识别标识组成另一个所述对应关系。
粗定位模块303，用于根据所述第一识别标识与所述第二识别标识的对应关系，计算得到所述可移动设备的初始位姿信息；
可选的，粗定位模块303包括：
第一构建子模块，用于根据所述对应关系，以及所述可移动设备的传感器坐标系到所述全局地图的地图坐标系的转换关系，确定所述可移动设备的初始位姿信息。
精定位模块304，用于根据所述初始位姿信息、所述全局地图、所述传感器数据，确定所述可移动设备的目标位姿信息。
可选的，精定位模块304包括：
搜索区域子模块，用于通过所述初始位姿信息，在所述全局地图中确定局部搜索区域；
精定位子模块，用于采用所述传感器数据，在所述局部搜索区域中进行位姿匹配操作，并根据匹配度最大的位姿信息得到所述目标位姿信息。
可选的，所述传感器包括激光雷达传感器，所述传感器数据包括激光点云数据，所述识别标识包括反光识别标识，所述全局地图包括根据激光点云数据构建的栅格地图；精定位子模块包括：
候选单元，用于根据所述传感器数据、搜索窗口的大小以及搜索步长，确定针对所 述局部搜索区域的多个候选位姿信息；
击中区域单元，用于在所述局部搜索区域中，确定每个所述候选位姿信息对应的激光点云数据所击中的栅格；
匹配度单元，用于计算击中的所有的栅格的占用概率平均值，并将所述占用概率平均值作为所述候选位姿信息的匹配度；
确定单元，用于根据匹配度最大的候选位姿信息得到所述目标位姿信息。
可选的，在确定与所述第一识别标识匹配的第二识别标识的过程中，终止匹配的条件包括：完成了所有第一识别标识和第二识别标识之间的匹配、匹配时长大于或等于预设时长、得到的所述对应关系的数量大于或等于对应关系数量阈值中的任意一种。
可选的，所述装置还包括：
局部地图模块，用于在所述第一识别标识的数量小于数量阈值的情况下，根据所述传感器数据构建局部地图；
灰度图模块，用于将所述局部地图转换为局部灰度图，以及将所述全局地图转换为全局灰度图；
匹配计算模块，用于将所述灰度局部图与所述全局灰度图进行匹配计算，得到至少一个位姿信息以及每个位姿信息对应的匹配得分；
得分模块，用于将所述匹配得分最大的位姿信息作为所述初始位姿信息。
综上，本公开实施例提供的一种可移动设备的重定位装置，包括：获取传感器在工作环境中采集的传感器数据；在根据传感器数据确定工作环境中的第一识别标识的数量大于或等于预设数量阈值的情况下，从工作环境的全局地图中所记录的第二识别标识中，确定与第一识别标识匹配的第二识别标识；全局地图中记录有工作环境中布置的识别标识的位置；根据第一识别标识与第二识别标识的对应关系，计算得到可移动设备的初始位姿信息；根据初始位姿信息、全局地图、传感器数据，确定可移动设备的目标位姿信息。在本公开中，可以在可移动设备的工作环境中布置多个易于被识别的识别标识，通过在相似工作环境中部署识别标识可以降低相似局部区域的之间相似性，从而降低重定位的奇异性问题的发生几率。针对时常发生较大变化的工作场景，由于场景中识别标识的位置固定不变，则可以基于识别标识进行粗定位，使得粗定位的过程不受环境变化的影响，提高了定位精度。
另外，本公开实施例还提供一种装置，具体可以参照图6，该装置600包括处理器610，存储器620以及存储在存储器620上并可在处理器610上运行的计算机程序，该计算机程序被处理器610执行时实现上述实施例的可移动设备的重定位方法实施例的各个过程，且能达到相同的技术效果，为避免重复，这里不再赘述。
本公开实施例还提供一种计算机可读存储介质，计算机可读存储介质上存储有计算机程序，该计算机程序被处理器执行时实现上述的可移动设备的重定位方法实施例的各个过程，且能达到相同的技术效果，为避免重复，这里不再赘述。其中，所述的计算机 可读存储介质，可以为只读存储器(Read-Only Memory，简称ROM)、随机存取存储器(Random Access Memory，简称RAM)、磁碟或者光盘等。
本公开实施例还提供了一种计算机程序，该计算机程序可以存储在云端或本地的存储介质上。在该计算机程序被计算机或处理器运行时用于执行本公开实施例的可移动设备的重定位方法的相应步骤，并且用于实现根据本公开实施例的可移动设备的重定位装置中的相应模块。
以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性的劳动的情况下，即可以理解并实施。
本公开的各个部件实施例可以以硬件实现，或者以在一个或者多个处理器上运行的软件模块实现，或者以它们的组合实现。本领域的技术人员应当理解，可以在实践中使用微处理器或者数字信号处理器(DSP)来实现根据本公开实施例的计算处理设备中的一些或者全部部件的一些或者全部功能。本公开还可以实现为用于执行这里所描述的方法的一部分或者全部的设备或者装置程序(例如，计算机程序和计算机程序产品)。这样的实现本公开的程序可以存储在计算机可读介质上，或者可以具有一个或者多个信号的形式。这样的信号可以从因特网网站上下载得到，或者在载体信号上提供，或者以任何其他形式提供。
例如，图7示出了可以实现根据本公开的方法的计算处理设备。该计算处理设备传统上包括处理器1010和以存储器1020形式的计算机程序产品或者计算机可读介质。存储器1020可以是诸如闪存、EEPROM(电可擦除可编程只读存储器)、EPROM、硬盘或者ROM之类的电子存储器。存储器1020具有用于执行上述方法中的任何方法步骤的程序代码1031的存储空间1030。例如，用于程序代码的存储空间1030可以包括分别用于实现上面的方法中的各种步骤的各个程序代码1031。这些程序代码可以从一个或者多个计算机程序产品中读出或者写入到这一个或者多个计算机程序产品中。这些计算机程序产品包括诸如硬盘，紧致盘(CD)、存储卡或者软盘之类的程序代码载体。这样的计算机程序产品通常为如参考图8所述的便携式或者固定存储单元。该存储单元可以具有与图7的计算处理设备中的存储器1020类似布置的存储段、存储空间等。程序代码可以例如以适当形式进行压缩。通常，存储单元包括计算机可读代码1031’，即可以由例如诸如1010之类的处理器读取的代码，这些代码当由计算处理设备运行时，导致该计算处理设备执行上面所描述的方法中的各个步骤。
另外，在本公开实施例的描述中，除非另有明确的规定和限定，术语“安装”、“相连”、“连接”应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或一体地连接；可以是机械连接，也可以是电连接；可以是直接相连，也可以通过中间媒介间接相连， 可以是两个元件内部的连通。对于本领域的普通技术人员而言，可以具体情况理解上述术语在本公开中的具体含义。
在本公开的描述中，需要说明的是，术语“中心”、“上”、“下”、“左”、“右”、“竖直”、“水平”、“内”、“外”等指示的方位或位置关系为基于附图所示的方位或位置关系，仅是为了便于描述本公开和简化描述，而不是指示或暗示所指的装置或元件必须具有特定的方位、以特定的方位构造和操作，因此不能理解为对本公开的限制。此外，术语“第一”、“第二”、“第三”仅用于描述目的，而不能理解为指示或暗示相对重要性。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的系统、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在本公开所提供的几个实施例中，应该理解到，所揭露的系统、装置和方法，可以通过其它的方式实现。以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，又例如，多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些通信接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本公开各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个处理器可执行的非易失的计算机可读取存储介质中。基于这样的理解，本公开的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本公开各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-OnlyMemory)、随机存取存储器(RAM，RandomAccessMemory)、磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是：以上所述实施例，仅为本公开的具体实施方式，用以说明本公开的技术方案，而非对其限制，本公开的保护范围并不局限于此，尽管参照前述实施例对本公开进行了详细的说明，本领域的普通技术人员应当理解：任何熟悉本技术领域的技术人员在本公开揭露的技术范围内，其依然可以对前述实施例所记载的技术方案进行修改或可轻易想到变化，或者对其中部分技术特征进行等同替换；而这些修改、变化或者替换，并不使相应技术方案的本质脱离本公开实施例技术方案的精神和范围，都应涵盖 在本公开的保护范围之内。因此，本公开的保护范围应所述以权利要求的保护范围为准。
本文中所称的“一个实施例”、“实施例”或者“一个或者多个实施例”意味着，结合实施例描述的特定特征、结构或者特性包括在本公开的至少一个实施例中。此外，请注意，这里“在一个实施例中”的词语例子不一定全指同一个实施例。
在此处所提供的说明书中，说明了大量具体细节。然而，能够理解，本公开的实施例可以在没有这些具体细节的情况下被实践。在一些实例中，并未详细示出公知的方法、结构和技术，以便不模糊对本说明书的理解。
在权利要求中，不应将位于括号之间的任何参考符号构造成对权利要求的限制。单词“包含”不排除存在未列在权利要求中的元件或步骤。位于元件之前的单词“一”或“一个”不排除存在多个这样的元件。本公开可以借助于包括有若干不同元件的硬件以及借助于适当编程的计算机来实现。在列举了若干装置的单元权利要求中，这些装置中的若干个可以是通过同一个硬件项来具体体现。单词第一、第二、以及第三等的使用不表示任何顺序。可将这些单词解释为名称。
最后应说明的是：以上实施例仅用以说明本公开的技术方案，而非对其限制；尽管参照前述实施例对本公开进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本公开各实施例技术方案的精神和范围。

Claims (13)

  1. 一种可移动设备的重定位方法，其特征在于，所述可移动设备的重定位方法包括：
    获取传感器在工作环境中采集的传感器数据；
    在根据所述传感器数据确定的第一识别标识的数量大于或等于数量阈值的情况下，从所述工作环境的全局地图中所记录的第二识别标识中，确定与所述第一识别标识匹配的第二识别标识；所述第一识别标识为所述工作环境中设置的识别标识；
    根据所述第一识别标识与所述第二识别标识的对应关系，计算得到所述可移动设备的初始位姿信息；
    根据所述初始位姿信息、所述全局地图、所述传感器数据，确定所述可移动设备的目标位姿信息。
  2. 根据权利要求1所述的方法，其特征在于，所述从所述工作环境的全局地图中所记录的第二识别标识中，确定与所述第一识别标识匹配的第二识别标识，包括：
    分别计算两个所述第一识别标识间的第一线段，与两个所述第二识别标识间的第二线段之间的相似度；
    确定相似度大于或等于相似度阈值的目标第一线段和目标第二线段，其中，所述目标第一线段两端的第一识别标识与所述目标第二线段两端的第二识别标识一一对应。
  3. 根据权利要求2所述的方法，其特征在于，所述分别计算两个所述第一识别标识间的第一线段，与两个所述第二识别标识间的第二线段之间的相似度，包括：
    根据所述第一线段的第一长度和所述第二线段的第二长度的差值的绝对值，以及所述第一长度和所述第二长度中的较小值，分别计算所述第一线段与所述第二线段之间的相似度。
  4. 根据权利要求2或3所述的方法，其特征在于，每个所述第一线段为第一基准识别标识与任一其他第一识别标识之间的线段，所述第一基准标 识是根据所述传感器数据确定的多个第一识别标识之一；
    每个所述第二线段为第二基准识别标识与任一其他第二识别标识之间的线段，所述第二基准标识是所述全局地图中所记录的多个第二识别标识之一。
  5. 根据权利要求1至4中任一项所述的方法，其特征在于，在确定与所述第一识别标识匹配的第二识别标识的过程中，终止匹配的条件包括：完成了所有第一识别标识和第二识别标识之间的匹配、匹配时长大于或等于预设时长、得到的所述对应关系的数量大于或等于对应关系数量阈值中的任意一种。
  6. 根据权利要求1至5中任一项所述的方法，其特征在于，还包括：
    在所述第一识别标识的数量小于数量阈值的情况下，根据所述传感器数据构建局部地图；
    将所述局部地图转换为局部灰度图，以及将所述全局地图转换为全局灰度图；
    将所述灰度局部图与所述全局灰度图进行匹配计算，得到至少一个位姿信息以及每个位姿信息对应的匹配得分；
    将所述匹配得分最大的位姿信息作为所述初始位姿信息。
  7. 根据权利要求1至6中任一项所述的方法，其特征在于，所述根据所述初始位姿信息、所述全局地图、所述传感器数据，确定所述可移动设备的目标位姿信息，包括：
    通过所述初始位姿信息，在所述全局地图中确定局部搜索区域；
    采用所述传感器数据，在所述局部搜索区域中进行位姿匹配操作，并根据匹配度最大的位姿信息得到所述目标位姿信息。
  8. 根据权利要求7所述的方法，其特征在于，所述传感器包括激光雷达传感器，所述传感器数据包括激光点云数据，所述识别标识包括反光识别标识，所述全局地图包括根据激光点云数据构建的栅格地图；
    所述采用所述传感器数据，在所述局部搜索区域中进行位姿匹配操作，并根据匹配度最大的位姿信息得到所述目标位姿信息，包括：
    根据所述传感器数据、搜索窗口的大小以及搜索步长，确定针对所述局 部搜索区域的多个候选位姿信息；
    在所述局部搜索区域中，确定每个所述候选位姿信息对应的激光点云数据所击中的栅格；
    计算击中的所有的栅格的占用概率平均值，并将所述占用概率平均值作为所述候选位姿信息的匹配度；
    根据匹配度最大的候选位姿信息得到所述目标位姿信息。
  9. 根据权利要求1至7中任一项所述的方法，其特征在于，所述识别标识用于在所述工作环境中区分于其他物体，提供路标参考。
  10. 根据权利要求9所述的方法，其特征在于，所述识别标识包括反光识别标识、或标识符。
  11. 一种计算处理设备，其特征在于，包括：
    存储器，其中存储有计算机可读代码；
    一个或多个处理器，当所述计算机可读代码被所述一个或多个处理器执行时，所述计算处理设备执行如权利要求1-10中任一项所述的可移动设备的重定位方法。
  12. 一种计算机程序，包括计算机可读代码，当所述计算机可读代码在计算处理设备上运行时，所述计算处理设备执行如权利要求1-10中任一项所述的可移动设备的重定位方法。
  13. 一种计算机可读介质，其中存储了如权利要求12所述的计算机程序。
PCT/CN2022/094718 2021-07-29 2022-05-24 可移动设备的重定位方法及装置 WO2023005384A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110867220.4A CN113776533A (zh) 2021-07-29 2021-07-29 可移动设备的重定位方法及装置
CN202110867220.4 2021-07-29

Publications (1)

Publication Number Publication Date
WO2023005384A1 true WO2023005384A1 (zh) 2023-02-02

Family

ID=78836491

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/094718 WO2023005384A1 (zh) 2021-07-29 2022-05-24 可移动设备的重定位方法及装置

Country Status (2)

Country Link
CN (1) CN113776533A (zh)
WO (1) WO2023005384A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117075171A (zh) * 2023-10-18 2023-11-17 新石器慧通（北京）科技有限公司 激光雷达的位姿信息确定方法、装置、设备及存储介质

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113776533A (zh) * 2021-07-29 2021-12-10 北京旷视科技有限公司 可移动设备的重定位方法及装置
CN114543808B (zh) * 2022-02-11 2024-09-27 杭州萤石软件有限公司 室内重定位方法、装置、设备及存储介质
CN117031481B (zh) * 2023-08-14 2024-09-03 北京数字绿土科技股份有限公司 一种基于投影3d激光点云的移动机器人重定位方法及系统

Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104501811A (zh) * 2014-11-24 2015-04-08 杭州申昊科技股份有限公司 一种基于环境直线特征的地图匹配方法
CN104764457A (zh) * 2015-04-21 2015-07-08 北京理工大学 一种用于无人车的城市环境构图方法
CN109655805A (zh) * 2019-01-25 2019-04-19 南京理工大学 一种基于扫描线段重合长度估计的激光雷达定位方法
CN110456797A (zh) * 2019-08-19 2019-11-15 杭州电子科技大学 一种基于2d激光传感器的agv重定位系统及方法
CN110954112A (zh) * 2019-03-29 2020-04-03 北京初速度科技有限公司 一种导航地图与感知图像匹配关系的更新方法和装置
WO2020192105A1 (zh) * 2019-03-28 2020-10-01 魔门塔(苏州)科技有限公司 一种车辆位姿的修正方法和装置
CN111856447A (zh) * 2020-07-28 2020-10-30 三一机器人科技有限公司 一种雷达定位方法、装置及存储介质
CN112255616A (zh) * 2020-10-19 2021-01-22 三一机器人科技有限公司 一种多雷达的反光柱定位方法及反光柱定位装置
CN112433211A (zh) * 2020-11-27 2021-03-02 浙江商汤科技开发有限公司 一种位姿确定方法及装置、电子设备和存储介质
US20210164800A1 (en) * 2019-11-29 2021-06-03 Aptiv Technologies Limited Method for Determining the Position of a Vehicle
CN113776533A (zh) * 2021-07-29 2021-12-10 北京旷视科技有限公司 可移动设备的重定位方法及装置
CN114088114A (zh) * 2021-11-19 2022-02-25 智道网联科技(北京)有限公司 车辆位姿校准方法、装置和电子设备

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105913082B (zh) * 2016-04-08 2020-11-27 北京邦视科技有限公司 一种对图像中目标进行分类的方法及系统
CN108759844B (zh) * 2018-06-07 2021-11-16 科沃斯商用机器人有限公司 机器人重定位与环境地图构建方法、机器人及存储介质
CN109084732B (zh) * 2018-06-29 2021-01-12 北京旷视科技有限公司 定位与导航方法、装置及处理设备
CN111367269B (zh) * 2018-12-26 2023-08-15 武汉万集信息技术有限公司 激光雷达的导航定位方法、装置及系统
CN109932713B (zh) * 2019-03-04 2021-07-09 北京旷视科技有限公司 定位方法、装置、计算机设备、可读存储介质和机器人
CN112815923B (zh) * 2019-11-15 2022-12-30 华为技术有限公司 视觉定位方法和装置
CN111060135B (zh) * 2019-12-10 2021-12-17 亿嘉和科技股份有限公司 一种基于局部地图的地图修正方法及系统
CN111273304B (zh) * 2019-12-31 2022-05-24 芜湖哈特机器人产业技术研究院有限公司 一种融合反光柱的自然定位方法及系统
CN111596298B (zh) * 2020-05-13 2022-10-14 北京百度网讯科技有限公司 目标对象的定位方法、装置、设备及存储介质
CN113126119B (zh) * 2021-04-02 2023-01-24 上海机电工程研究所 融合点云数据与机器视觉的精定位方法、系统及介质

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104501811A (zh) * 2014-11-24 2015-04-08 杭州申昊科技股份有限公司 一种基于环境直线特征的地图匹配方法
CN104764457A (zh) * 2015-04-21 2015-07-08 北京理工大学 一种用于无人车的城市环境构图方法
CN109655805A (zh) * 2019-01-25 2019-04-19 南京理工大学 一种基于扫描线段重合长度估计的激光雷达定位方法
WO2020192105A1 (zh) * 2019-03-28 2020-10-01 魔门塔(苏州)科技有限公司 一种车辆位姿的修正方法和装置
CN110954112A (zh) * 2019-03-29 2020-04-03 北京初速度科技有限公司 一种导航地图与感知图像匹配关系的更新方法和装置
CN110456797A (zh) * 2019-08-19 2019-11-15 杭州电子科技大学 一种基于2d激光传感器的agv重定位系统及方法
US20210164800A1 (en) * 2019-11-29 2021-06-03 Aptiv Technologies Limited Method for Determining the Position of a Vehicle
CN111856447A (zh) * 2020-07-28 2020-10-30 三一机器人科技有限公司 一种雷达定位方法、装置及存储介质
CN112255616A (zh) * 2020-10-19 2021-01-22 三一机器人科技有限公司 一种多雷达的反光柱定位方法及反光柱定位装置
CN112433211A (zh) * 2020-11-27 2021-03-02 浙江商汤科技开发有限公司 一种位姿确定方法及装置、电子设备和存储介质
CN113776533A (zh) * 2021-07-29 2021-12-10 北京旷视科技有限公司 可移动设备的重定位方法及装置
CN114088114A (zh) * 2021-11-19 2022-02-25 智道网联科技(北京)有限公司 车辆位姿校准方法、装置和电子设备

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117075171A (zh) * 2023-10-18 2023-11-17 新石器慧通（北京）科技有限公司 激光雷达的位姿信息确定方法、装置、设备及存储介质
CN117075171B (zh) * 2023-10-18 2024-01-16 新石器慧通（北京）科技有限公司 激光雷达的位姿信息确定方法、装置、设备及存储介质

Also Published As

Publication number Publication date
CN113776533A (zh) 2021-12-10

Similar Documents

Publication Publication Date Title
WO2023005384A1 (zh) 可移动设备的重定位方法及装置
CN108827249B (zh) 一种地图构建方法和装置
KR20210020945A (ko) 창고 환경에서의 차량 추적
CN110793512A (zh) 位姿识别方法、装置、电子设备和存储介质
CN110807350A (zh) 用于面向扫描匹配的视觉slam的系统和方法
Mohamed Detection and tracking of pallets using a laser rangefinder and machine learning techniques
CN108388244A (zh) 基于人工路标的移动机器人系统、停靠方法及存储介质
WO2019205978A1 (zh) 一种图书馆机器人定位导航方法
CN111964680B (zh) 一种巡检机器人的实时定位方法
CN113253737B (zh) 货架检测方法及装置、电子设备、存储介质
Ji et al. RGB-D SLAM using vanishing point and door plate information in corridor environment
CN113485350A (zh) 一种机器人移动控制方法、装置、设备及存储介质
CN111198496A (zh) 一种目标跟随机器人及跟随方法
Skrzypczyński Mobile robot localization: Where we are and what are the challenges?
Wang et al. Autonomous landing of multi-rotors UAV with monocular gimbaled camera on moving vehicle
US20210304411A1 (en) Map construction method, apparatus, storage medium and electronic device
CN114545426A (zh) 定位方法、装置、移动机器人和计算机可读介质
CN111964681B (zh) 一种巡检机器人的实时定位系统
CN116520829A (zh) 自移动设备的设备控制方法及装置
CN111367269B (zh) 激光雷达的导航定位方法、装置及系统
CN115446846A (zh) 一种基于条码识别的图书盘点机器人
CN113960999A (zh) 移动机器人重定位方法、系统及芯片
US11506483B2 (en) Method, system and apparatus for support structure depth determination
Fu et al. Costmap construction and pseudo-lidar conversion method of mobile robot based on monocular camera
CN114019963B (zh) 一种面向桌面集群机器人的外部定位系统

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: 22847994

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 15/05/2024)