WO2021104497A1 - Procédé et système de positionnement basés sur un radar laser, ainsi que support de stockage et processeur - Google Patents

Procédé et système de positionnement basés sur un radar laser, ainsi que support de stockage et processeur Download PDF

Info

Publication number
WO2021104497A1
WO2021104497A1 PCT/CN2020/132457 CN2020132457W WO2021104497A1 WO 2021104497 A1 WO2021104497 A1 WO 2021104497A1 CN 2020132457 W CN2020132457 W CN 2020132457W WO 2021104497 A1 WO2021104497 A1 WO 2021104497A1
Authority
WO
WIPO (PCT)
Prior art keywords
laser
electronic map
cluster
positioning
lidar
Prior art date
Application number
PCT/CN2020/132457
Other languages
English (en)
Chinese (zh)
Inventor
曹军
Original Assignee
广州视源电子科技股份有限公司
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 广州视源电子科技股份有限公司 filed Critical 广州视源电子科技股份有限公司
Publication of WO2021104497A1 publication Critical patent/WO2021104497A1/fr

Links

Images

Classifications

    • 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
    • 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
    • G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
    • G01S7/48—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions

Definitions

  • This application relates to the field of radar positioning, and in particular to a positioning method and system based on lidar, a storage medium and a processor.
  • the original laser point cloud data is currently used directly for processing.
  • the environment is required to be relatively static.
  • positioning and scene reconstruction often have poor results (such as moving objects such as people and cars in the environment, or changes in the position and posture of objects in the environment).
  • a mobile robot locates a target object indoors
  • the environment changes such as objects (such as chairs, tables, etc.) originally on the map are moved to other locations, Or add some new objects in the environment that are not on the map.
  • objects such as chairs, tables, etc.
  • the number of vehicles in the garage and the position of the vehicles may change at any time.
  • the environmental information obtained by the robot and the memory information (ie, the stored information), which will cause a certain positioning error, and when the information difference is too large, it may also cause positioning failure. Failure of positioning will cause accidents in robot navigation, so it is necessary to judge the positioning results in actual use.
  • the embodiments of the present application provide a positioning method and system based on lidar, a storage medium, and a processor, so as to at least solve the technical problem in related technologies that it is impossible to determine whether the positioning result is reliable in a dynamic scene.
  • a positioning method based on lidar includes: obtaining lidar data of lidar; clustering the lidar data to obtain multiple laser clusters; During the process, each laser cluster is matched with the electronic map separately; based on the matching result of each laser cluster with the electronic map, it is determined whether the positioning result of the target object is accurate.
  • matching each laser cluster with the electronic map respectively includes: obtaining the positioning result of the target object; and mapping each laser cluster to the electronic map based on the positioning result; Match each laser cluster with the corresponding area mapped on the electronic map to obtain the matching result.
  • mapping each laser cluster to an electronic map based on the positioning result includes: reading the positioning coordinates in the positioning result; projecting the laser points of each laser cluster to the electronic map based on the positioning coordinates, and determining each The position of the laser point of the laser cluster is mapped on the electronic map.
  • match each laser cluster with the corresponding area mapped on the electronic map to obtain the matching result including: obtaining multiple laser points in any laser cluster; calculating the obstacle of the closest distance to each laser point The occupation point of the object on the electronic map, and the closest distance value between each laser point and the corresponding occupied point; count the number of laser points whose closest distance value is less than the first predetermined threshold; if the statistical number exceeds the second predetermined threshold , It is determined that the laser cluster matches the electronic map.
  • the method further includes: establishing a cache map based on the occupation point corresponding to each laser point, wherein each coordinate on the map is cached The points all represent the closest distance value.
  • the matching result includes: taking each laser cluster as the center, obtaining the coordinates of the obstacle closest to each laser cluster on the electronic map; obtaining each The distance between the laser cluster and the nearest obstacle is obtained, and multiple distance values are obtained; if the distance value is less than or equal to the third predetermined threshold, it is determined that the corresponding laser cluster matches the electronic map.
  • taking each laser cluster as the center to obtain the coordinates of the obstacle closest to each laser cluster on the electronic map includes: taking the mapping point of any laser cluster on the electronic map as the center to obtain the corresponding information on the electronic map At least one obstacle in the area; calculate the distance between the laser cluster and each obstacle in the corresponding area, and obtain the obstacle closest to the laser cluster; traverse each laser cluster to get the distance of each laser cluster on the electronic map The coordinates of the nearest obstacle.
  • determining whether the positioning result of the target object is accurate including: if the number of laser clusters matching the electronic map exceeds a fourth predetermined threshold, the matching result is that the positioning result is accurate .
  • the lidar data includes: distance information and direction information of obstacles located in the sensing range.
  • another positioning method based on lidar including: obtaining lidar data collected by lidar, where the lidar data includes: relative distance information between lidar and obstacle , Relative direction information; cluster the lidar data to obtain multiple laser clusters; in the process of locating the obstacles corresponding to the moving robot, each laser cluster is matched with the electronic map; based on each laser The matching result of the cluster and the electronic map determines whether the positioning result is accurate.
  • matching each laser cluster with an electronic map includes: obtaining the positioning result of the obstacle; and mapping each laser cluster based on the positioning result To the electronic map; each laser cluster is matched with the corresponding area mapped on the electronic map, and the matching result is obtained.
  • determining whether the positioning result of the robot is accurate includes: if the number of laser clusters matching the electronic map exceeds a predetermined threshold, the positioning result of the robot is accurate.
  • a robot including: a lidar, configured to collect lidar data; a processor, connected to the lidar, and configured to determine the positioning result of obstacles based on the lidar data; Each laser cluster obtained by clustering the lidar data is matched with the electronic map, and based on the matching result, it is determined whether the positioning result is accurate.
  • the lidar data includes: relative distance information and relative direction information between the robot and the obstacle.
  • the positioning result is accurate.
  • a positioning device based on lidar including: an acquisition module configured to acquire lidar data of lidar; and a clustering module configured to cluster lidar data, Obtain multiple laser clusters; the matching module is set to match each laser cluster with the electronic map in the process of positioning the target object; the positioning module is set to match the result of each laser cluster with the electronic map, Determine whether the positioning result of the target object is accurate.
  • a storage medium includes a stored program, wherein the device where the storage medium is located is controlled to execute the above-mentioned lidar-based positioning method when the program is running.
  • a processor configured to run a program stored in a memory, wherein the above-mentioned laser radar-based positioning method is executed when the program is running.
  • the lidar data is clustered, and each cluster obtained by the clustering is matched with the electronic map, and the positioning result of the target object is determined according to the matching result.
  • This method can be based on the clustering.
  • the method of class matches each cluster with the map. Therefore, it can be matched by class as a unit. Even if the environment changes, it can be judged whether the positioning result is reliable, which solves the problem that the related technology cannot judge the positioning result in a dynamic scene. Whether it is reliable technical question.
  • Fig. 1 is a schematic structural diagram of a robot according to an embodiment of the present application.
  • FIG. 2 is a schematic flowchart of a method for positioning lidar according to an embodiment of the present application
  • Fig. 3 is a schematic diagram of the distribution of obstacle points according to an embodiment of the present application.
  • FIG. 4 is a schematic diagram of the distribution of obstacle points after clustering according to an embodiment of the present application.
  • FIG. 5 is a schematic diagram of a map after determining whether a positioning result of a robot moving area is reliable according to an embodiment of the present application
  • Fig. 6 is a schematic flowchart of an optional laser radar positioning method according to an embodiment of the present application.
  • Fig. 7 is a schematic structural diagram of a laser radar positioning device according to an embodiment of the present application.
  • Fig. 8 is a schematic flowchart of another method for positioning a lidar according to an embodiment of the present application.
  • Lidar It is a radar system that emits a laser beam to detect the position of an obstacle. Its working principle is to transmit a detection signal (laser beam), and then receive the signal reflected from the obstacle in the environment and compare it with the transmitted signal, and make appropriate processing. , The relevant information of the target can be obtained, such as the target's distance, azimuth, height, speed, attitude, and even shape parameters, so as to detect, track and recognize the target.
  • Lidar can include laser transmitters, optical receivers, turntables, and information processing systems.
  • Map A map constructed based on a 2D lidar SLAM, which may be a grid map in this embodiment of the application, and the value in each grid point represents the probability that the point is occupied by an obstacle.
  • the robot after the robot obtains the positioning result, it directly projects the lidar data onto the map, and then judges whether the current positioning is reliable based on the matching of all lidar data with the map.
  • the environment where the robot is located changes greatly, the measurement data obtained by the lidar will be quite different from the map. Even if the positioning is completely correct, the matching degree between the lidar data and the map is not high. Therefore, this method is not suitable for dynamic scenes with large environmental changes
  • the lidar data is clustered, and the lidar data falling on different objects are divided into different categories. After the positioning result is generated, the lidar data is projected onto the map according to the positioning result, and the matching of each laser cluster with the map is calculated to determine whether the positioning result is reliable. It will be described in detail below in conjunction with specific embodiments.
  • Fig. 1 is a schematic structural diagram of a robot according to an embodiment of the present application. As shown in Fig. 1, the robot 1 includes: a laser radar 10 and a processor 12, among which:
  • Lidar 10 is set to collect Lidar data.
  • the lidar 10 may be provided in the robot 10.
  • the robot 10 may be a mobile robot such as a small handling and logistics robot or a sweeping robot.
  • the lidar may be integrated in the robot body to be set as a real-time positioning robot.
  • the location on the map at work For example, when a sweeping robot is performing cleaning work, it emits light signals through lidar at regular intervals, and after receiving the echo of the light signal, it sends the lidar data to the positioning device, and the positioning device receives the robot’s After the lidar data, and then map the lidar data to the electronic map, the location of the obstacle can be judged.
  • the robot 10 can move in a target area when it is moving.
  • the target area can be the active area of the robot 10. In some application scenarios, it can also be expressed as the sensing area of lidar.
  • radar sensing is a wireless Perception technology, by analyzing the received target echo characteristics, extracts and discovers the target's position, shape, motion characteristics and motion trajectory, and can further infer the characteristics of the target and the environment.
  • the target echo feature is the process in which a beam of wireless light signal hits an object in front of it, reflected back and received again.
  • the distance between the lidar and the object can be calculated based on the launch time to when the object is turned back and the echo is received, and the size of the received echo or the strength of the signal can be judged Find out the size, shape and speed of the object.
  • the sensing area refers to the range that the lidar signal can cover. In some embodiments of the present application, the sensing area is the sensing range of the lidar, which can be a circular plane or a fan-shaped plane, for example, the central angle of the fan is 90 Degree or 50 degrees, etc., the radius is r. Generally, the value of r can be related to the measurement parameters of the lidar or be set according to the needs of the user within the range of the measurement parameters.
  • a rectangular coordinate system is established with the collection point of the lidar as the center.
  • the sensing range of the lidar can be the first quadrant, the second quadrant, or the third quadrant, or the fourth quadrant, or even adjacent Any combination between the two quadrants, such as the range formed by the first quadrant and the second quadrant, is used as the sensing range of the lidar.
  • the above-mentioned lidar data may include, but is not limited to: a collection of multiple laser points with distance information and direction information collected in real time.
  • the lidar data includes: relative distance information and relative direction information between the lidar and an obstacle.
  • the lidar 10 may be a combination of multiple lidars, which are distributed in different positions of the robot, so that different lidars form overlapping detection ranges to cover as many obstacles as possible in the active area of the robot. Things.
  • the timer in the lidar 10 calculates the feedback wave feedback time to obtain the distance information of the lidar 10 from different obstacles.
  • the long-distance obstacle feedback wave time will inevitably be The feedback wave time of obstacles shorter than the distance is longer.
  • the lidar 10 will calculate the source position of the feedback wave according to the vertical angle of the feedback wave feedback, that is, the direction of the feedback wave source, so as to obtain the position data of the surrounding environment such as obstacles, so that the lidar
  • the information of its own position that is, the position of the robot where the lidar is located
  • the relative position information of the robot and the obstacle can be determined.
  • the processor 12 is connected to the lidar 10 and is configured to determine the positioning result of the obstacle based on the lidar data; each laser cluster obtained by clustering the lidar data is matched with the electronic map, and determined based on the matching result Whether the positioning result is accurate.
  • the position or direction information of the target object (such as an obstacle) in the lidar data matches the position or direction information of the obstacle in the map, it is considered that the current positioning result for the robot is accurate.
  • the position or direction information of the target object (such as an obstacle) in the lidar data matches the position or direction information of the obstacle in the map, it can be considered that the positioning result of the robot obtained at this time is credible, that is, the positioning The result is accurate.
  • the laser cluster may be determined in the following manner: clustering the above-mentioned lidar data, and classifying the point cloud data with the same position into one category to obtain the laser cluster.
  • each laser cluster can also be screened to eliminate the laser clusters that do not meet the requirements. For example: the laser points in each laser cluster are counted. When the distance d between the laser point and the nearest obstacle is less than the second threshold Dt, it is considered that this point matches the map. Among them, Dt is usually determined according to the inherent error of the lidar. For each laser cluster, when the number of laser points matching the map in the cluster exceeds the third threshold (for example, when the number of matching laser points exceeds 90% of the number of laser points in the cluster), the laser cluster is considered to match the map; otherwise, it is not match.
  • the third threshold for example, when the number of matching laser points exceeds 90% of the number of laser points in the cluster
  • the embodiments of the present application provide a method embodiment of a laser radar positioning method. It should be noted that the steps shown in the flowchart of the accompanying drawings can be performed on a computer such as a set of computer-executable instructions. It is executed in the system, and although the logical sequence is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than here.
  • Fig. 2 is a schematic flowchart of a method for positioning lidar according to an embodiment of the present application. As shown in Fig. 2, the method includes the following steps:
  • Step S202 Obtain lidar data collected by lidar
  • the lidar may be a combination of multiple lidars, which are distributed in different positions of the robot, so that different lidars form overlapping detection ranges to cover as many obstacles as possible.
  • the timer in the lidar calculates the feedback wave feedback time to obtain the distance information of the lidar from different obstacles.
  • the long-distance obstacle feedback wave time will inevitably be shorter than the distance. Obstacles have a long feedback wave time.
  • the lidar will calculate the source position of the feedback wave according to the vertical angle of the feedback wave feedback, that is, the direction of the feedback wave source, to obtain the position data of the surrounding environment such as obstacles, so that the lidar's own position (That is, the position of the robot where the lidar is located) information can be determined, for example, the relative position information of the robot and the obstacle can be determined.
  • Step S204 clustering the lidar data to obtain multiple laser clusters
  • clustering algorithms can be used for clustering, where clustering algorithms include but are not limited to: K-means clustering algorithm, Mean-Shift clustering algorithm, etc.
  • the positioning result of the robot is accurate.
  • Figure 3 is a visual representation of the output data of the lidar, where each point is the position of the obstacle point in the environment. It can be seen from Figure 3 that it is impossible to distinguish which objects these point cloud data belong to from the original data.
  • the laser data is clustered first.
  • clustering point clouds with the same position can be clustered into a cluster.
  • the clustering method here is implemented using the RBNN (Radial Basis Function Neural Network) algorithm.
  • the RBNN algorithm is an improved algorithm based on the KNN clustering method. The distance between each data and the neighboring point is used to determine whether to return to the neighboring point. As a class.
  • the point cloud image with cluster tags is shown in Figure 4.
  • the RBNN algorithm is a normalization algorithm, which is generally used for lattice classification calculations.
  • the RBNN algorithm is used The laser clusters with similar positions can be "rounded" to obtain a precise and single laser cluster position for subsequent matching steps.
  • the clustering process it is possible to determine which laser points belong to the same object. And through clustering, the laser data can be denoised, and some irregular environmental information can be removed.
  • the specific method of removing environmental information is to use the RBNN algorithm to remove the redundant laser cluster lattice coordinates, so as to achieve the technical effect of removing noise.
  • Step S206 in the process of locating the target object, each laser cluster is matched with the electronic map respectively;
  • the matching process can be expressed in the following two ways:
  • the laser radar integrated inside the robot body is used to locate the position of the obstacle on the map in real time when the robot is working.
  • a robot based on an ARM smart chip uses a robotic arm to carry goods, it sends lidar data to the server via lidar every 0.5s (that is, the data obtained after lidar sends laser signals to obstacles and receives echo signals)
  • the server can determine the location of the obstacle based on the map of the sensing area.
  • the position of the robot on the map can also be determined based on the position of the obstacle.
  • the robot can also be provided with a GPS positioning device, so that it can be based on the GPS positioning device and laser
  • the radar data jointly determine the positioning information of the robot.
  • GPS positioning device for coarse-grained positioning and lidar data for fine-grained positioning, specifically: GPS positioning data is obtained, and the first positioning result is obtained. Because the accuracy of satellite positioning is limited, it is not suitable for large-area Indoor scenes or outdoor scenes, therefore, fine-grained positioning of radar data is required; obtain lidar data to obtain the second positioning result; use the first positioning result to obtain the first position of the robot; use the second positioning result to determine the first position Path to make corrections.
  • each laser cluster can be mapped to the electronic map in the following ways: read the positioning coordinates in the positioning results; The laser points are projected onto the electronic map, and the location of the laser points of each laser cluster on the electronic map is determined.
  • each laser cluster is mapped to the electronic map, which is mapped according to the positioning result, that is, first obtain the positioning coordinates in the coordinate system where the obstacle is located, for example, the positioning coordinates are (a, b), and then the positioning coordinates are converted It is the coordinate in the coordinate system corresponding to the electronic map.
  • Obtain multiple laser points in any laser cluster calculate the location point on the electronic map of the obstacle that is the closest distance to each laser point, and the closest distance value between each laser point and the corresponding location point ; Count the number of laser points whose closest distance value is less than the first predetermined threshold; if the statistical number exceeds the second predetermined threshold, it is determined that the laser cluster matches the electronic map.
  • a cache map is established based on the location point corresponding to each laser point.
  • the coordinate points all represent the closest distance value.
  • Step 1) With each laser cluster as the center, obtain the coordinates of the obstacle closest to each laser cluster on the electronic map;
  • the location of each laser cluster on the electronic map is a positioning coordinate, and there are different obstacles around the positioning coordinates, then the obstacle location information can be determined according to the coordinates of the obstacle detected by the lidar on the electronic map .
  • Step 2 Obtain the distance between each laser cluster and the nearest obstacle, and obtain multiple distance values
  • the distance between the obstacle and the laser cluster is obtained by the method of step 1), but the selection of the obstacle should be determined according to the obstacle with the closest distance between the surrounding obstacle and the laser cluster, for example, the distance between the laser cluster and the first obstacle Is a, the distance between the laser cluster and the first obstacle is b, and the distance between the laser cluster and the first obstacle is c.
  • the obstacle c is selected as the obstacle with the closest distance to the laser cluster. And get the value of c as the output value.
  • Step 3 If the distance value is less than or equal to the third predetermined threshold, it is determined that the corresponding laser cluster matches the electronic map.
  • the mapping point of any laser cluster on the electronic map is taken as the center to obtain at least one obstacle in the corresponding area on the electronic map; the difference between the laser cluster and each obstacle in the corresponding area is calculated. Obtain the obstacles closest to the laser clusters; traverse each laser cluster to obtain the coordinates of the obstacles closest to each laser cluster on the electronic map.
  • the black part represents the grid points occupied by obstacles
  • the white is the grid points occupied by obstacles
  • the robot can move freely in these areas
  • the gray parts are unknown areas; after obtaining the clustering results Then, match each cluster with the map.
  • Step S208 based on the matching result of each laser cluster and the electronic map, it is determined whether the positioning result of the target object is accurate.
  • the matching result is that the positioning result is accurate.
  • the lidar data includes: relative distance information and relative direction information between the obstacle and the lidar.
  • each point in each cluster find the closest point in the map that is occupied by obstacles, and calculate the distance d between the laser point and the point.
  • calculate the distance between each location on the map and the nearest obstacle point to create a cache map.
  • the value of each point on the map indicates the distance between the point and the nearest obstacle point.
  • the laser points in each cluster are counted. When the distance d between the laser point and the nearest obstacle is less than the preset threshold Dt, it is considered that this point matches the map.
  • the Dt setting is usually the size of the inherent error of using lidar.
  • For each laser cluster when the number of laser points in the cluster matches more than 90%, it is considered that the laser cluster matches the map; otherwise, it does not match; when there are more than N laser clusters that match the map, the positioning result is judged to be reliable.
  • the positioning result is shown in Figure 6.
  • the red, green, and purple lines in the figure are laser clusters matching the map. At this time, there are 3 laser clusters matching the map, and the positioning can be judged to be reliable.
  • the embodiment of this application directly uses the original laser data to match the map.
  • the matching degree between the laser and the map is not high, and it cannot be judged. Whether the positioning is unreliable, and the object-based laser cluster is used for matching. Even if the environment changes greatly, as long as there are some static objects that can be successfully matched, then the positioning can be judged to be reliable.
  • the embodiment of the present application provides a positioning device based on lidar, which is used to implement the method shown in FIG. 2, as shown in FIG. 7, the device includes: an acquisition module 70 configured to acquire lidar data collected by lidar
  • the clustering module 72 is configured to cluster the lidar data to obtain a plurality of laser clusters;
  • the matching module 74 is configured to separate the laser clusters with the electrons in the process of locating the target object The map is matched;
  • the positioning module 76 is configured to determine whether the positioning result of the target object is accurate based on the matching result of each laser cluster and the electronic map.
  • the clustering module 72 is configured to cluster the lidar data to obtain multiple laser clusters, generate laser clusters according to the lidar measurement data, and match the laser clusters with the positioning data in the electronic map, where:
  • the laser cluster is a kind of lidar position data. After the lidar judges the positioning of the subject, it will locate the subject equipment according to the surrounding environment.
  • the positioning result of the robot is accurate.
  • the laser data is clustered first.
  • clustering point clouds with the same position can be clustered into a cluster.
  • the clustering method here is implemented using the RBNN algorithm.
  • the RBNN algorithm is an improvement of the KNN-based clustering method. It uses the distance between each data and the neighbor point to determine whether it is classified as the same with the neighboring point.
  • the point cloud image with cluster tags is shown in Figure 4.
  • the clustering process it is possible to determine which laser points belong to the same object. And through clustering, we can denoise the laser data and remove some irregular environmental information.
  • the specific method of removing environmental information is to use the RBNN algorithm to remove the redundant laser cluster lattice coordinates, so as to achieve the technical effect of removing noise.
  • the matching module 74 is configured to match each laser cluster with the electronic map during the process of locating the target object in the sensing area.
  • the matching process can be expressed as the following two Ways:
  • each laser cluster can be mapped to the electronic map in the following ways: read the positioning coordinates in the positioning results; The laser points are projected onto the electronic map, and the location of the laser points of each laser cluster on the electronic map is determined.
  • Step 1) With each laser cluster as the center, obtain the coordinates of the obstacle closest to each laser cluster on the electronic map;
  • Step 2 Obtain the distance between each laser cluster and the nearest obstacle, and obtain multiple distance values
  • Step 3 If the distance value is less than or equal to the third predetermined threshold, it is determined that the corresponding laser cluster matches the electronic map.
  • the positioning module 76 is configured to determine whether the positioning result of the target object is accurate based on the matching result of each laser cluster and the electronic map. In some embodiments of the present application, if the number of laser clusters matching the electronic map exceeds the fourth predetermined threshold, the matching result is that the positioning result is accurate.
  • each point in each cluster find the closest point in the map that is occupied by obstacles, and calculate the distance d between the laser point and the map point.
  • calculate the distance between each location on the map and the nearest obstacle point to create a cache map.
  • the value of each point on the map indicates the distance between the point and the nearest obstacle point.
  • the laser points in each cluster are counted. When the distance d between the laser point and the nearest obstacle is less than the preset threshold Dt, it is considered that this point matches the map. Dt is usually the inherent error of lidar.
  • Dt is usually the inherent error of lidar.
  • the laser cluster when more than 90% of the laser points in the cluster match, the laser cluster is considered to match the map; otherwise, it does not match; when there are more than N laser clusters matching the map, the positioning result is judged to be reliable.
  • the positioning result is shown in Figure 6. The red, green, and purple parts of the figure are laser clusters matching the map. At this time, there are 3 laser clusters matching the map, and the positioning can be judged to be reliable.
  • the embodiment of this application directly uses the original laser data to match the map.
  • the matching degree between the laser and the map is not high, and it cannot be judged. Whether the positioning is unreliable, and the object-based laser cluster is used for matching. Even if the environment changes greatly, as long as there are some static objects that can be successfully matched, then the positioning can be judged to be reliable.
  • FIG. 8 is a schematic flowchart of another method for positioning a lidar according to an embodiment of the present application. As shown in FIG. 8, the method includes:
  • Step S802 Obtain lidar data collected by lidar, where the lidar data includes: relative distance information and relative direction information between the lidar and the obstacle;
  • Step S804 clustering the lidar data to obtain multiple laser clusters
  • Step S806 in the process of locating the obstacle corresponding to the moving robot, matching each of the laser clusters with the electronic map respectively;
  • Step S808 based on the matching result of each laser cluster and the electronic map, it is determined whether the positioning result is accurate.
  • matching each laser cluster with the electronic map respectively includes: obtaining the positioning result of the obstacle; As a result of the positioning, each laser cluster is mapped to the electronic map; each laser cluster is matched with a corresponding area respectively mapped on the electronic map to obtain the matching result.
  • An embodiment of the present application provides a storage medium, the storage medium includes a stored program, wherein the device where the storage medium is located is controlled to execute the above-mentioned lidar-based positioning method when the program is running.
  • the method may be: obtaining lidar data collected by lidar; clustering the lidar data to obtain multiple laser clusters; in the process of locating the target object, separate each of the laser clusters Matching with an electronic map; based on the matching result of each laser cluster and the electronic map, it is determined whether the positioning result of the target object is accurate.
  • a processor configured to run a program stored in a memory, wherein the above-mentioned lidar-based positioning method is executed when the program is running.
  • the method may be: acquiring lidar data collected by lidar; clustering the lidar data to obtain multiple laser clusters; in the process of locating the target object in the sensing area, the Each laser cluster is matched with an electronic map respectively; based on the matching result of each laser cluster with the electronic map, it is determined whether the positioning result of the target object is accurate.
  • the disclosed technical content can be implemented in other ways.
  • the device embodiments described above are only illustrative.
  • the division of the units may be a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components may be combined or may be Integrate into another system, or some features can be ignored or not implemented.
  • the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, units or modules, and may be in electrical or other forms.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.
  • the functional units in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.
  • the above-mentioned integrated unit can be implemented in the form of hardware or software functional unit.
  • the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium.
  • the technical solution of the present application essentially or the part that contributes to the existing technology or all or 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 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 method described in each embodiment of the present application.
  • the aforementioned storage media include: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes. .
  • the lidar data is clustered, and each cluster obtained by the clustering is matched with the electronic map, and the positioning result of the target object is determined according to the matching result.
  • This method can be based on the clustering.
  • the method of class matches each cluster with the map. Therefore, it can be matched by class as a unit. Even if the environment changes, it can be judged whether the positioning result is reliable, which solves the problem that the related technology cannot judge the positioning result in a dynamic scene. Whether it is reliable technical question.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Remote Sensing (AREA)
  • General Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Probability & Statistics with Applications (AREA)
  • Electromagnetism (AREA)
  • Traffic Control Systems (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
  • Optical Radar Systems And Details Thereof (AREA)

Abstract

L'invention concerne un procédé de positionnement basé sur un radar laser, ainsi qu'un robot, un support de stockage et un processeur. Le procédé consiste : à acquérir des données de radar laser d'un radar laser (S202) ; à regrouper les données de radar laser pour obtenir plusieurs groupes de lasers (S204) ; durant le processus de positionnement d'un objet cible, à mettre en correspondance respectivement chaque groupe de lasers avec une carte électronique (S206) ; et à déterminer, sur la base d'un résultat de mise en correspondance entre chaque groupe de lasers et la carte électronique, si un résultat de positionnement de l'objet cible est exact (S208).
PCT/CN2020/132457 2019-11-29 2020-11-27 Procédé et système de positionnement basés sur un radar laser, ainsi que support de stockage et processeur WO2021104497A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201911206878.XA CN110865393A (zh) 2019-11-29 2019-11-29 基于激光雷达的定位方法及系统、存储介质和处理器
CN201911206878.X 2019-11-29

Publications (1)

Publication Number Publication Date
WO2021104497A1 true WO2021104497A1 (fr) 2021-06-03

Family

ID=69657832

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/132457 WO2021104497A1 (fr) 2019-11-29 2020-11-27 Procédé et système de positionnement basés sur un radar laser, ainsi que support de stockage et processeur

Country Status (2)

Country Link
CN (1) CN110865393A (fr)
WO (1) WO2021104497A1 (fr)

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113343061A (zh) * 2021-06-24 2021-09-03 广州高新兴机器人有限公司 一种gps激光融合slam中坐标系动态对齐方法
CN113432533A (zh) * 2021-06-18 2021-09-24 北京盈迪曼德科技有限公司 一种机器人定位方法、装置、机器人及存储介质
CN114440858A (zh) * 2022-01-25 2022-05-06 中国人民解放军总医院第一医学中心 移动机器人定位丢失检测方法、系统、设备及存储介质
CN114445572A (zh) * 2021-12-29 2022-05-06 航天时代(青岛)海洋装备科技发展有限公司 一种基于DeeplabV3+的陌生海域中障碍物即时定位与地图构建方法
CN114594437A (zh) * 2022-01-25 2022-06-07 航天南湖电子信息技术股份有限公司 雷达点迹数据自动标注方法、系统和存储介质
CN115267796A (zh) * 2022-08-17 2022-11-01 深圳市普渡科技有限公司 定位方法、装置、机器人和存储介质
CN115685990A (zh) * 2022-09-22 2023-02-03 深圳市智绘科技有限公司 自动充电方法、装置、电子设备及可读存储介质
CN115840205A (zh) * 2023-02-16 2023-03-24 济宁市质量计量检验检测研究院(济宁半导体及显示产品质量监督检验中心、济宁市纤维质量监测中心) 一种基于激光雷达技术的地形面积计量方法和系统
CN116077186A (zh) * 2023-04-07 2023-05-09 青岛大学附属医院 基于激光干涉的手术定位系统
CN116840820A (zh) * 2023-08-29 2023-10-03 上海仙工智能科技有限公司 一种检测2d激光定位丢失的方法及系统、存储介质

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110865393A (zh) * 2019-11-29 2020-03-06 广州视源电子科技股份有限公司 基于激光雷达的定位方法及系统、存储介质和处理器
CN114252883B (zh) * 2020-09-24 2022-08-23 北京万集科技股份有限公司 目标检测方法、装置、计算机设备和介质
CN112213738B (zh) * 2020-09-28 2023-02-28 劢微机器人科技(深圳)有限公司 无人搬运车地图构建方法、装置、设备及存储介质
CN112630787B (zh) * 2020-12-03 2022-05-17 深圳市优必选科技股份有限公司 定位方法、装置、电子设备及可读存储介质
CN113390415B (zh) * 2021-06-04 2024-03-01 深圳优艾智合机器人科技有限公司 机器人的定位方法、定位装置、管理系统和存储介质
CN115616578A (zh) * 2022-12-05 2023-01-17 成都航空职业技术学院 一种用于无人飞行器的雷达探测方法及装置

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005181059A (ja) * 2003-12-18 2005-07-07 Toshiba Corp 目標位置標定方法及び目標位置標定装置
CN104931977A (zh) * 2015-06-11 2015-09-23 同济大学 一种用于智能车辆的障碍物识别方法
CN105866790A (zh) * 2016-04-07 2016-08-17 重庆大学 一种考虑激光发射强度的激光雷达障碍物识别方法及系统
CN109509256A (zh) * 2018-06-21 2019-03-22 华南理工大学 基于激光雷达的建筑结构自动测量及3d模型生成方法
CN109829032A (zh) * 2019-03-14 2019-05-31 广州蓝胖子机器人有限公司 一种物品识别的方法、设备及存储介质
CN109840454A (zh) * 2017-11-28 2019-06-04 华为技术有限公司 目标物定位方法、装置、存储介质以及设备
CN110865393A (zh) * 2019-11-29 2020-03-06 广州视源电子科技股份有限公司 基于激光雷达的定位方法及系统、存储介质和处理器

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9984494B2 (en) * 2015-01-26 2018-05-29 Uber Technologies, Inc. Map-like summary visualization of street-level distance data and panorama data
CN106323273B (zh) * 2016-08-26 2019-05-21 深圳微服机器人科技有限公司 一种机器人重定位方法及装置
CN107064955A (zh) * 2017-04-19 2017-08-18 北京汽车集团有限公司 障碍物聚类方法及装置
CN107422730A (zh) * 2017-06-09 2017-12-01 武汉市众向科技有限公司 基于视觉导引的agv运输系统及其驾驶控制方法
CN108507579B (zh) * 2018-04-08 2020-04-21 浙江大承机器人科技有限公司 一种基于局部粒子滤波的重定位方法
CN109284348B (zh) * 2018-10-30 2021-02-05 百度在线网络技术（北京）有限公司 一种电子地图的更新方法、装置、设备和存储介质
CN110084272B (zh) * 2019-03-26 2021-01-08 哈尔滨工业大学（深圳） 一种聚类地图创建方法及基于聚类地图和位置描述子匹配的重定位方法
CN109917792B (zh) * 2019-04-10 2023-10-13 湖南汽车工程职业学院 一种基于无人驾驶观光电动车的自主防碰撞系统

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005181059A (ja) * 2003-12-18 2005-07-07 Toshiba Corp 目標位置標定方法及び目標位置標定装置
CN104931977A (zh) * 2015-06-11 2015-09-23 同济大学 一种用于智能车辆的障碍物识别方法
CN105866790A (zh) * 2016-04-07 2016-08-17 重庆大学 一种考虑激光发射强度的激光雷达障碍物识别方法及系统
CN109840454A (zh) * 2017-11-28 2019-06-04 华为技术有限公司 目标物定位方法、装置、存储介质以及设备
CN109509256A (zh) * 2018-06-21 2019-03-22 华南理工大学 基于激光雷达的建筑结构自动测量及3d模型生成方法
CN109829032A (zh) * 2019-03-14 2019-05-31 广州蓝胖子机器人有限公司 一种物品识别的方法、设备及存储介质
CN110865393A (zh) * 2019-11-29 2020-03-06 广州视源电子科技股份有限公司 基于激光雷达的定位方法及系统、存储介质和处理器

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113432533A (zh) * 2021-06-18 2021-09-24 北京盈迪曼德科技有限公司 一种机器人定位方法、装置、机器人及存储介质
CN113432533B (zh) * 2021-06-18 2023-08-15 北京盈迪曼德科技有限公司 一种机器人定位方法、装置、机器人及存储介质
CN113343061A (zh) * 2021-06-24 2021-09-03 广州高新兴机器人有限公司 一种gps激光融合slam中坐标系动态对齐方法
CN114445572A (zh) * 2021-12-29 2022-05-06 航天时代(青岛)海洋装备科技发展有限公司 一种基于DeeplabV3+的陌生海域中障碍物即时定位与地图构建方法
CN114445572B (zh) * 2021-12-29 2024-05-14 航天时代(青岛)海洋装备科技发展有限公司 一种基于DeeplabV3+的陌生海域中障碍物即时定位与地图构建方法
CN114440858B (zh) * 2022-01-25 2023-12-19 中国人民解放军总医院第一医学中心 移动机器人定位丢失检测方法、系统、设备及存储介质
CN114440858A (zh) * 2022-01-25 2022-05-06 中国人民解放军总医院第一医学中心 移动机器人定位丢失检测方法、系统、设备及存储介质
CN114594437A (zh) * 2022-01-25 2022-06-07 航天南湖电子信息技术股份有限公司 雷达点迹数据自动标注方法、系统和存储介质
CN115267796A (zh) * 2022-08-17 2022-11-01 深圳市普渡科技有限公司 定位方法、装置、机器人和存储介质
CN115267796B (zh) * 2022-08-17 2024-04-09 深圳市普渡科技有限公司 定位方法、装置、机器人和存储介质
CN115685990A (zh) * 2022-09-22 2023-02-03 深圳市智绘科技有限公司 自动充电方法、装置、电子设备及可读存储介质
CN115840205A (zh) * 2023-02-16 2023-03-24 济宁市质量计量检验检测研究院(济宁半导体及显示产品质量监督检验中心、济宁市纤维质量监测中心) 一种基于激光雷达技术的地形面积计量方法和系统
CN116077186A (zh) * 2023-04-07 2023-05-09 青岛大学附属医院 基于激光干涉的手术定位系统
CN116077186B (zh) * 2023-04-07 2023-06-30 青岛大学附属医院 基于激光干涉的手术定位系统
CN116840820B (zh) * 2023-08-29 2023-11-24 上海仙工智能科技有限公司 一种检测2d激光定位丢失的方法及系统、存储介质
CN116840820A (zh) * 2023-08-29 2023-10-03 上海仙工智能科技有限公司 一种检测2d激光定位丢失的方法及系统、存储介质

Also Published As

Publication number Publication date
CN110865393A (zh) 2020-03-06

Similar Documents

Publication Publication Date Title
WO2021104497A1 (fr) Procédé et système de positionnement basés sur un radar laser, ainsi que support de stockage et processeur
US11885872B2 (en) System and method for camera radar fusion
WO2021253430A1 (fr) Procédé de détermination de pose absolue, dispositif électronique et plateforme mobile
US11960028B2 (en) Determining specular reflectivity characteristics using LiDAR
US10223793B1 (en) Laser distance measuring method and system
CN112513679B (zh) 一种目标识别的方法和装置
US20170368686A1 (en) Method and device for automatic obstacle avoidance of robot
CN110794392B (zh) 车辆定位方法、装置、车辆及存储介质
Muresan et al. Multi-object tracking of 3D cuboids using aggregated features
CN112462368B (zh) 一种障碍物检测方法、装置、车辆及存储介质
WO2022198637A1 (fr) Procédé et système de filtrage de bruit en nuage de points et plate-forme mobile
WO2022104096A1 (fr) Détection d'objet multi-radar
Li et al. Pedestrian liveness detection based on mmwave radar and camera fusion
CN114488026A (zh) 基于4d毫米波雷达的地下停车库可通行空间检测方法
US11668799B2 (en) Histogram based L-shape detection of target objects
Liu et al. Vehicle detection and tracking with 2d laser range finders
CN117169848A (zh) 过滤玻璃噪点的方法、激光雷达及机器人
WO2021087751A1 (fr) Procédé de mesure de distance, dispositif de mesure de distance, plateforme mobile autonome et support de stockage
CN116699596A (zh) 一种车辆的毫米波雷达的速度修正方法及装置
US10012729B2 (en) Tracking subjects using ranging sensors
CN113627569B (zh) 一种用于交通大场景的雷达视频一体机的数据融合方法及装置、存储介质
Meissner et al. Combining the 2d and 3d world: A new approach for point cloud based object detection
CN114839615B (zh) 一种4d毫米波雷达目标航向角拟合方法及存储介质
CN117872310B (zh) 基于雷达的水面目标跟踪方法、装置、设备及介质
US20230375694A1 (en) Dual sensing method of object and computing apparatus for object sensing

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: 20894445

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 03/11/2022)

122 Ep: pct application non-entry in european phase

Ref document number: 20894445

Country of ref document: EP

Kind code of ref document: A1