Foot type robot positioning method and system

Abstract

The invention discloses a foot type robot positioning method and system, and belongs to the technical field of robots. Preprocessing the data; reprocessing the data; presetting a feature map database; matching data; querying and positioning; wherein the data reprocessing comprises the following steps: acquiring invariant feature distribution according to the preprocessed data; and obtaining a distribution set of the invariant features in the same coordinate system according to the invariant feature distribution. The technical problem that the foot type robot cannot be accurately positioned due to the external environment is solved, and the technical effect of accurately positioning the foot type robot is achieved.

Description

Technical Field

[0001] The present invention relates to the field of robot technology, and in particular to a legged robot positioning method and system. Background Art

[0002] The environmental perception system is a very important component of legged robots, and positioning is the basis of the environmental perception system. Commonly used positioning technologies include satellite positioning, inertial navigation, dead reckoning, map matching, and sensor positioning. In recent years, with the expansion of application scenarios, harsh environments such as no satellite signals and no light have posed new challenges to the robot's perception system.

[0003] In the prior art, the accuracy of satellite positioning is greatly reduced due to weak or no GPS signals on mountain roads due to tree obstruction or other reasons. In addition, unstructured roads have fewer environmental features that can be used for positioning. In addition, the snowy environment will cause the information expressed by the three-dimensional point cloud data obtained by sensors such as lidar and depth cameras to be unable to be matched with the original map, making it impossible to complete accurate positioning. Summary of the invention

[0004] In order to solve the above problems, the first aspect of the present invention provides a legged robot positioning method, comprising:

[0005] Data collection; data preprocessing; data reprocessing; preset feature map database; data matching; and query positioning;

[0006] Wherein, the data reprocessing includes:

[0007] Obtain invariant feature distribution based on preprocessed data;

[0008] According to the distribution of invariant features, the distribution set of invariant features in the same coordinate system is obtained.

[0009] In some embodiments, the data reprocessing further includes:

[0010] Construct the descriptor of the invariant features according to the distribution set of the invariant features in the same coordinate system;

[0011] Perform invariant feature detection on the descriptor of the invariant feature to determine the position distribution of the invariant feature in the descriptor;

[0012] According to the position distribution of the invariant features in the descriptor, the quantity distribution of the invariant features in the descriptor is determined.

[0013] In some embodiments, obtaining a distribution set of invariant features in the same coordinate system according to the invariant feature distribution includes:

[0014] Construct a reference coordinate system;

[0015] determining coordinates of the first invariant feature in the reference coordinate system;

[0016] Using a coordinate transformation method, the coordinates of the second invariant feature are determined in the reference coordinate system;

[0017] determining a distance between the first invariant feature and the second invariant feature based on the coordinates of the first invariant feature and the coordinates of the second invariant feature;

[0018] If the distance between the first invariant feature and the second invariant feature is less than a second preset threshold, the first invariant feature and the second invariant feature are the same invariant feature, and the second invariant feature is removed from the reference coordinate system.

[0019] In some embodiments, the data matching includes:

[0020] Expand the quantity distribution of invariant features in the descriptor into the quantity distribution of invariant features in the one-dimensional descriptor;

[0021] The quantitative distribution of the invariant features in the one-dimensional descriptor is matched with the feature map database to determine the target descriptor.

[0022] In some embodiments, determining the target descriptor comprises:

[0023] Assume the size of the one-dimensional descriptor is N r ×N s , the number of invariant features in the one-dimensional descriptor is distributed as S i =[m 1 , m 2 , ..., m N ];

[0024] Assume that the number of invariant features in the matching descriptors in the feature map database is distributed as S j =[n 1 , n 2 , ..., n N ];

[0025] Then the similarity calculation function is:

[0026]

[0027] Among them, δ is the similarity, N = N r ×N s .

[0028] In some embodiments, the calculating the matching target descriptor according to the similarity further comprises:

[0029] Determine the maximum similarity between the one-dimensional descriptor and the matching descriptor;

[0030] in,

[0031] δ(S i , S j ) is the maximum similarity, S i and S j Similarity results of

[0032] If the maximum similarity is greater than a fourth preset threshold, the matching descriptor is determined to be the target descriptor.

[0033] In some embodiments, the query positioning includes:

[0034] Determine a sampling point for generating a target descriptor according to the target descriptor in the feature map database;

[0035] Determine the current position of the legged robot according to the sampling points of the generated target descriptor;

[0036] Determining the position of the current legged robot according to the sampling points of the generated target descriptor comprises:

[0037] Filter the sampling points with the highest occurrence frequency from the sampling points used to generate the target descriptor;

[0038] The location reliability is calculated based on the number of occurrences of the sampling point with the highest frequency;

[0039] The calculation method of the positioning reliability is as follows:

[0040]

[0041] Among them, confidence is the positioning confidence;

[0042] If the positioning confidence is greater than the fifth preset threshold, the sampling point with the highest occurrence frequency is determined as the positioning position of the legged robot, otherwise the positioning fails.

[0043] In some embodiments, the data preprocessing includes: reducing redundancy and segmenting the ground;

[0044] Wherein, reducing redundancy includes:

[0045] Sort the collected data according to time sequence;

[0046] Acquire a first data set from the collected data according to the sampling distance;

[0047] Wherein, acquiring a first data set from the collected data according to the sampling distance includes:

[0048] If the distance between the sampling point and the previous sampling point is greater than a first preset threshold, the data of the sampling point is added to the first data set.

[0049] In some embodiments, segmenting the ground comprises:

[0050] According to the first data set, obtaining a plane data set;

[0051] The plane data set is removed from the first data set to obtain a second data set.

[0052] A second aspect of the present invention provides a legged robot positioning system, comprising:

[0053] The acquisition module is used to collect point cloud data of the target scene;

[0054] A preprocessing module, used to preprocess the point cloud data of the target scene;

[0055] A reprocessing module, used for reprocessing the preprocessed data;

[0056] A preset module, used to preset a feature map database according to the reprocessed data;

[0057] A matching module, used to match the reprocessed positioning data with the feature map database;

[0058] The query module is used to determine the position of the legged robot according to the matching results.

[0059] By adopting the above technical solution, the present invention mainly has the following technical effects:

[0060] By collecting data and matching it with a preset feature map database after processing, the sampling points with the highest frequency are determined, and the position of the legged robot is determined using the positioning confidence. This solves the technical problem that the legged robot cannot be accurately positioned due to the external environment, and achieves the technical effect of accurately positioning the legged robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 is a schematic diagram of an application scenario of a legged robot positioning method according to some embodiments of this specification;

[0062] Figure 2 is an exemplary flow chart of a legged robot positioning method according to some embodiments of this specification;

[0063] Figure 3 is a schematic diagram of a descriptor of an invariant feature according to some embodiments of this specification;

[0064] Figure 4is a schematic diagram of the position distribution of invariant features in descriptors according to some embodiments of this specification;

[0065] Figure 5 is a (partial) schematic diagram of a feature map database according to some embodiments of this specification;

[0066] Figure 6 It is a module diagram of a legged robot target following system according to some embodiments of this specification. DETAILED DESCRIPTION

[0067] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of this specification. For ordinary technicians in this field, this specification can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0068] It should be understood that the "system", "device", "unit" and / or "module" used herein are a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0069] As shown in this specification, unless the context clearly indicates an exception, the words "a", "an", "a kind" and / or "the" do not refer to the singular, but also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0070] Flowcharts are used in this specification to illustrate the operations performed by the system according to the embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed precisely in order. Instead, the steps may be processed in reverse order or simultaneously. At the same time, other operations may also be added to these processes, or one or more operations may be removed from these processes.

[0071] Figure 1 is a schematic diagram of an application scenario of a legged robot positioning method according to some embodiments of this specification;

[0072] In some embodiments, the legged robot positioning system 100 can be positioned in real time by implementing the legged robot positioning method disclosed in this specification. For example, in a typical application scenario, the legged robot positioning system 100 can collect real-time information through the acquisition device 140, and determine the real-time position through the server 110 based on the real-time information, wherein the real-time information at least includes point cloud data of the target scene, and the real-time position is determined through the server 110 based on the point cloud data of the target scene.

[0073] The server 110 and the terminal device 130 can be connected through the network 120, and the server 110 can be connected to the storage device 150 through the network 120. The server 110 may include a processing device 112, and the processing device 112 may be used to execute the legged robot positioning method described in some embodiments of this specification. The network 120 can connect the various components of the application scenario 100 and / or connect the system with external resources. The storage device 150 can be used to store data and / or instructions. For example, the storage device 150 can store operation information or feature map databases. The storage device 150 can be directly connected to the server 110 or located inside the server 110. The terminal device 130 refers to one or more terminal devices or software. In some embodiments, the terminal device 130 can issue instructions to the processing device 112 or receive the positioning results of the legged robot sent by the processing device 112, and display them to the user. Exemplarily, the terminal device 130 may include one or any combination of mobile devices 130-1, tablet computers 130-2, laptop computers 130-3, etc. or other devices with input and / or output functions.

[0074] The acquisition device 140 may be used to acquire real-time information of a target scene. An exemplary acquisition device 140 may include a laser radar, which acquires point cloud data of the surface of the target object.

[0075] The legged robot positioning system 100 is provided for illustrative purposes only and is not intended to limit the scope of this specification. For those of ordinary skill in the art, various modifications or variations can be made according to the description of this specification. For example, the legged robot positioning system 100 may also include a power supply device, etc. For another example, the legged robot positioning system 100 may be implemented on other devices to achieve similar or different functions. However, the changes and modifications will not deviate from the scope of this specification.

[0076] Figure 2 is an exemplary flow chart of a legged robot positioning method according to some embodiments of this specification;

[0077] like Figure 2As shown, the process 200 includes the following steps. In some embodiments, the process 200 can be executed by the processing device 112.

[0078] Step 210, data collection;

[0079] In some embodiments, step 210 may be performed by a collection module 310 . In some embodiments, the collection module 310 may acquire point cloud data of the target scene through a laser radar.

[0080] Step 220, data preprocessing;

[0081] In some embodiments, step 220 may be performed by a preprocessing module 320, by preprocessing the data collected in step 210 to obtain preprocessed data;

[0082] In some embodiments, the data preprocessing includes: reducing redundancy and segmenting the ground;

[0083] Wherein, reducing redundancy includes:

[0084] Step 2201, sorting the collected data according to the time sequence;

[0085] Step 2202: Obtain a first data set from the collected data according to the sampling distance;

[0086] In some embodiments, acquiring a first data set from the collected data according to the sampling distance includes:

[0087] If the distance between the sampling point and the previous sampling point is greater than a first preset threshold, adding the data of the sampling point to the first data set;

[0088] As an example, the first preset threshold is 5m; when the Euclidean distance between the next sampling point and the previous sampling point is greater than 5m, the point cloud data collected by the sampling point is added to the data set.

[0089] In some embodiments, segmenting the ground comprises:

[0090] Step 2203: Obtain a plane data set according to the first data set;

[0091] Step 2204: remove the plane data set from the first data set to obtain a second data set;

[0092] The reason is that the point cloud data of the ground information is distributed on a plane, so the ground features can be segmented from the first data set by acquiring a plane data set from the first data set and removing the plane data set from the first data set.

[0093] As an example, the random sampling consistent RANSANC algorithm can be used to calculate the ground model, and then the corresponding point cloud is removed from the data set.

[0094] Step 230, data reprocessing;

[0095] In some embodiments, step 230 may be performed by a reprocessing module 330 by reprocessing the data preprocessed in step 220 to obtain reprocessed data.

[0096] In some embodiments, the data reprocessing includes:

[0097] Step 2301, obtaining invariant feature distribution according to the preprocessed data;

[0098] In some embodiments, the number of tree trunk features on both sides of the road will not change with the presence or absence of snow, so the trunks of the trees on both sides of the road can be selected as invariant features, and the BEV target detection and clustering algorithm can be used to perform trunk detection on the data collected in step 210 frame by frame to obtain the position distribution of the trunks within the current frame scanning range.

[0099] Step 2302: Obtain a distribution set of invariant features in the same coordinate system according to the invariant feature distribution;

[0100] In some embodiments, after acquiring the position distribution of the tree trunks frame by frame, the position distribution of the tree trunks can be expressed in the same coordinate system by means of coordinate transformation, thereby obtaining a distribution set of invariant features in the same coordinate system.

[0101] As an example, a coordinate system may be constructed as a reference coordinate system when collecting data at the first sampling point, and then the coordinate information of the subsequently scanned tree trunk positions may be transformed into the reference coordinate system by means of coordinate transformation.

[0102] Furthermore, considering that when the coordinate information of the trunk position is transformed into the same coordinate system, due to the existence of factors such as errors, the coordinate information of the same trunk may be different after the coordinate information of the trunk position is transformed into the reference coordinate system.

[0103] In some embodiments, obtaining a distribution set of invariant features in the same coordinate system according to the invariant feature distribution includes:

[0104] Step 23021, constructing a reference coordinate system;

[0105] Step 23022, determining the coordinates of the first invariant feature in the reference coordinate system;

[0106] Step 23023, using a coordinate transformation method to determine the coordinates of the second invariant feature in the reference coordinate system;

[0107] Step 23024: determine the distance between the first invariant feature and the second invariant feature according to the coordinates of the first invariant feature and the coordinates of the second invariant feature;

[0108] Step 23025: If the distance between the first invariant feature and the second invariant feature is less than the second preset threshold, the first invariant feature and the second invariant feature are the same invariant feature, and the second invariant feature is removed from the reference coordinate system.

[0109] The reason is that the feature points obtained after target detection of the same tree at different sampling points often do not overlap after the coordinate system is unified, but are distributed in a surface area of ​​δ (δ≈0). Therefore, in order to solve this problem, an allowable error needs to be set when unifying the coordinate system.

[0110] Step 2303: construct a descriptor of the invariant features according to the distribution set of the invariant features in the same coordinate system;

[0111] Figure 3 is a schematic diagram of a descriptor of an invariant feature according to some embodiments of this specification;

[0112] Exemplarily, the descriptor of the invariant feature is the Scan Context descriptor, and the number of input rings and sectors is N respectively. r , N s , the maximum description radius of the descriptor is L max , then the distance resolution of the descriptor is The angular resolution is

[0113] In some embodiments, for each invariant feature in the same coordinate system, a descriptor characterizing the invariant feature is generated.

[0114] Step 2304: perform invariant feature detection on the descriptor of the invariant feature to determine the position distribution of the invariant feature in the descriptor;

[0115] Figure 4 is a schematic diagram of the position distribution of invariant features in descriptors according to some embodiments of this specification;

[0116] As an example, N r =6, N s =16,L max =40m, and the position distribution of the invariant features in the 6-row × 16-column grid is obtained.

[0117] Step 2305: Determine the quantity distribution of the invariant features in the descriptor according to the position distribution of the invariant features in the descriptor;

[0118] As an example, after obtaining the position distribution of the invariant features in a 6-row×16-column grid, the number of invariant features falling into each grid of the 6-row×16-column grid can be expressed numerically.

[0119] Figure 5 is a (partial) schematic diagram of a feature map database according to some embodiments of this specification;

[0120] Step 240, presetting a feature map database;

[0121] In some embodiments, step 240 may be performed by the preset module 340 to generate a feature map database from the data processed in step 230 .

[0122] In some embodiments, step 240 can collect data of the target scene in advance and perform data preprocessing and data processing steps before the legged robot performs positioning work to obtain a descriptor set for each invariant feature, and then generate a feature map database, and preset the feature map database in the storage device 150.

[0123] In some embodiments, the preset feature map database includes:

[0124] Step 2401, expand the quantity distribution of the invariant features in the descriptor into the quantity distribution of the invariant features in the one-dimensional descriptor;

[0125] As an example, the quantity distribution of the invariant features in a 6-column×16-row grid is expanded to the quantity distribution of the invariant features in a 1-row×96-column grid.

[0126] Step 2402: Obtain a feature map database according to the quantity distribution of the invariant features in the one-dimensional descriptor;

[0127] In some embodiments, the feature map database is a collection of quantity distributions of each invariant feature in a one-dimensional descriptor. As an example, 10 descriptors of size 6 columns × 16 rows are merged into a quantity distribution of invariant features of 10 rows × 96 columns.

[0128] Step 250, data matching;

[0129] In some embodiments, step 250 can be performed by the matching module 350 to match the data collected by the acquisition device 140 during positioning with the feature map database; in some embodiments, the data collected by the acquisition device 140 is real-time data collected when the legged robot needs to be positioned.

[0130] In some embodiments, the data matching includes:

[0131] Step 2501, expand the quantity distribution of the invariant features in the descriptor into the quantity distribution of the invariant features in the one-dimensional descriptor;

[0132] Step 2502, matching the quantity distribution of the invariant features in the one-dimensional descriptor with the feature map database to determine the target descriptor;

[0133] In some embodiments, determining the target descriptor comprises:

[0134] Assume the size of the one-dimensional descriptor is N r ×N s , the number of invariant features in the one-dimensional descriptor is distributed as S i =[m 1 , m 2 , ..., m N ];

[0135] Assume that the number of invariant features in the matching descriptors in the feature map database is distributed as S j =[n 1 , n 2 , ..., n N ];

[0136] Then the similarity calculation function is:

[0137]

[0138] Among them, δ is the similarity, N = N r ×N s ;

[0139] The reason is that, due to S i and S j They are all quite sparse. Experiments have shown that using the information of zero elements for similarity calculation has very poor results, so similarity calculation can only use the information of non-zero elements.

[0140] Furthermore, considering that the descriptors generated by collecting data at different angles at the same location will be different, the "rotation invariance" property of the Scan Context descriptor needs to be used during matching. During matching, the descriptor needs to be translated in the sector direction, and then the similarity between the translated descriptor and the target descriptor is calculated. i and S j Two descriptors will generate N s Similarity results.

[0141] In some embodiments, the calculating the matching target descriptor according to the similarity further comprises:

[0142] Determine the maximum similarity between the one-dimensional descriptor and the matching descriptor;

[0143] in,

[0144] δ(S i , S j ) is the maximum similarity, S i and S j Similarity results of

[0145] If the maximum similarity is greater than a fourth preset threshold, the matching descriptor is determined to be the target descriptor.

[0146] Step 260, query location;

[0147] In some embodiments, step 260 may be performed by query module 360, which determines the sampling points for generating the target descriptor when querying the preset feature map database in step 240 according to the target descriptor, thereby completing the positioning process of the legged robot.

[0148] In some embodiments, the query positioning includes:

[0149] Step 2601, determining sampling points for generating target descriptors according to target descriptors in a feature map database;

[0150] Furthermore, after matching with the feature map database and determining the target descriptor, the sampling points for generating the target descriptor may be determined based on the target descriptor. However, the number of target descriptors is usually multiple, so the number of sampling points for generating the target descriptor may be one or more.

[0151] In some embodiments, the sampling points are sampling points at which data of the target scene is collected in advance when a feature map database is preset.

[0152] Step 2602: determine the current position of the legged robot according to the sampling points of the generated target descriptor;

[0153] In some embodiments, determining the position of the current legged robot according to the sampling points of the generated target descriptor includes:

[0154] Filter the sampling points with the highest occurrence frequency from the sampling points used to generate the target descriptor;

[0155] The location reliability is calculated based on the number of occurrences of the sampling point with the highest frequency;

[0156] The calculation method of the positioning reliability is as follows:

[0157]

[0158] Among them, confidence is the positioning confidence;

[0159] If the positioning confidence is greater than the fifth preset threshold, the sampling point with the highest occurrence frequency is determined as the positioning position of the legged robot, otherwise the positioning fails.

[0160] The reason is that each descriptor in the feature map database corresponds to a tree at that moment. However, when the data is taken to generate the map, the data is redundant and each tree may be sampled by multiple sampling points.

[0161] Figure 6 It is a module diagram of a legged robot target following system according to some embodiments of this specification.

[0162] In some embodiments, the legged robot positioning system 300 may include a collection module 310, a pre-processing module 320, a re-processing module 330, a preset module 340, a matching module 350, and a query module 360;

[0163] In some embodiments, the acquisition module 310 is used to acquire point cloud data of the target scene;

[0164] In some embodiments, the preprocessing module 320 is used to preprocess the point cloud data of the target scene;

[0165] In some embodiments, the reprocessing module 330 is used to reprocess the preprocessed data;

[0166] In some embodiments, the preset module 340 is used to preset the feature map database according to the reprocessed data;

[0167] In some embodiments, the matching module 350 is used to match the reprocessed positioning data with the feature map database;

[0168] In some embodiments, the query module 360 ​​is used to determine the position of the legged robot according to the matching results.

[0169] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of this specification. Although not explicitly stated here, those skilled in the art may make various modifications, improvements and corrections to this specification. Such modifications, improvements and corrections are suggested in this specification, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of this specification.

[0170] At the same time, this specification uses specific words to describe the embodiments of this specification. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of this specification. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures or characteristics in one or more embodiments of this specification can be appropriately combined.

[0171] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, as an example and not a limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A method for positioning a legged robot, characterized in that: include: Data collection; Data preprocessing; Data reprocessing; Preset feature map database; Data matching; and query positioning; Wherein, the data reprocessing includes: Obtain invariant feature distribution based on preprocessed data; According to the distribution of invariant features, the distribution set of invariant features in the same coordinate system is obtained.

2. A legged robot positioning method according to claim 1, characterized in that: The data reprocessing also includes: Construct the descriptor of the invariant features according to the distribution set of the invariant features in the same coordinate system; Perform invariant feature detection on the descriptor of the invariant feature to determine the position distribution of the invariant feature in the descriptor; According to the position distribution of the invariant features in the descriptor, the quantity distribution of the invariant features in the descriptor is determined.

3. A method for positioning a legged robot according to claim 1, characterized in that: The step of obtaining a distribution set of invariant features in the same coordinate system according to the invariant feature distribution includes: Construct a reference coordinate system; determining coordinates of the first invariant feature in the reference coordinate system; Using a coordinate transformation method, the coordinates of the second invariant feature are determined in the reference coordinate system; determining a distance between the first invariant feature and the second invariant feature based on the coordinates of the first invariant feature and the coordinates of the second invariant feature; If the distance between the first invariant feature and the second invariant feature is less than a second preset threshold, the first invariant feature and the second invariant feature are the same invariant feature, and the second invariant feature is removed from the reference coordinate system.

4. A method for positioning a legged robot according to claim 2, characterized in that: The data matching includes: Expand the quantity distribution of invariant features in the descriptor into the quantity distribution of invariant features in the one-dimensional descriptor; The quantitative distribution of the invariant features in the one-dimensional descriptor is matched with the feature map database to determine the target descriptor.

5. A method for positioning a legged robot according to claim 4, characterized in that: Determining the target descriptor comprises: Assume the size of the one-dimensional descriptor is N r ×N s , the number of invariant features in the one-dimensional descriptor is distributed as S i =[m1,m2,...,m N ]; Assume that the number of invariant features in the matching descriptors in the feature map database is distributed as S j =[n1, n2, ..., n N ]; Then the similarity calculation function is: Among them, δ is the similarity, N = N r ×N s .

6. A method for positioning a legged robot according to claim 5, characterized in that: The calculation of the matching target descriptor according to the similarity also includes: Determine the maximum similarity between the one-dimensional descriptor and the matching descriptor; in, δ(S i , S j ) is the maximum similarity, S i and S j Similarity results of If the maximum similarity is greater than a fourth preset threshold, the matching descriptor is determined to be the target descriptor.

7. A method for positioning a legged robot according to claim 6, characterized in that: The query positioning includes: Determine a sampling point for generating a target descriptor according to the target descriptor in the feature map database; Determine the current position of the legged robot according to the sampling points of the generated target descriptor; Determining the position of the current legged robot according to the sampling points of the generated target descriptor comprises: Filter the sampling points with the highest occurrence frequency from the sampling points used to generate the target descriptor; The location reliability is calculated based on the number of occurrences of the sampling point with the highest frequency; The calculation method of the positioning reliability is as follows: Among them, confidence is the positioning confidence; If the positioning confidence is greater than the fifth preset threshold, the sampling point with the highest occurrence frequency is determined as the positioning position of the legged robot, otherwise the positioning fails.

8. A method for positioning a legged robot according to claim 1, characterized in that: The data preprocessing includes: reducing redundancy and segmenting the ground; Wherein, reducing redundancy includes: Sort the collected data according to time sequence; Acquire a first data set from the collected data according to the sampling distance; Wherein, acquiring a first data set from the collected data according to the sampling distance includes: If the distance between the sampling point and the previous sampling point is greater than a first preset threshold, the data of the sampling point is added to the first data set.

9. A method for positioning a legged robot according to claim 8, characterized in that: The segmented ground includes: According to the first data set, obtaining a plane data set; The plane data set is removed from the first data set to obtain a second data set.

10. A legged robot positioning system, characterized in that: include: The acquisition module is used to collect point cloud data of the target scene; A preprocessing module, used to preprocess the point cloud data of the target scene; A reprocessing module, used for reprocessing the preprocessed data; A preset module, used to preset a feature map database according to the reprocessed data; A matching module, used to match the reprocessed positioning data with the feature map database; The query module is used to determine the position of the legged robot according to the matching results.

Images