WO2020019221A1

WO2020019221A1 - Method, apparatus and robot for autonomous positioning and map creation

Info

Publication number: WO2020019221A1
Application number: PCT/CN2018/097134
Authority: WO
Inventors: 徐泽元
Original assignee: 深圳前海达闼云端智能科技有限公司
Priority date: 2018-07-26
Filing date: 2018-07-26
Publication date: 2020-01-30
Also published as: CN109074085B; CN109074085A

Abstract

A method for autonomous positioning and map creation, the method being applied to a robot (10) and comprising: obtaining a distance observation value of the robot (10) for a landmark point (101); obtaining the location of the robot (10) in a map (102); adding a new landmark point comprising a fixed article in the landmark point to the map, and obtaining a pose of the new landmark point according to the location and the distance observation value (103). The described method only adds a landmark of a fixed object into a map, avoids changes to a landmark when a robot refers to the landmark, and reduces the influence of the surrounding environment on robot positioning and map creation, such that the positioning of a robot and the calculation of a landmark in a map are more accurate. Further disclosed are an apparatus for autonomous positioning and map creation, a robot for implementing the described method, a non-volatile computer readable storage medium for implementing the described method, and a computer program product for implementing the described method.

Description

Method, device and robot for autonomous positioning and map establishment

Technical field

The embodiments of the present application relate to the field of artificial intelligence, for example, to an autonomous positioning and map building method, device, and robot.

Background technique

SLAM (Simultaneous Localization and Mapping), that is, simultaneous positioning and map reconstruction, refers to a robot equipped with a specific sensor. In the absence of prior information about the environment, it builds an incremental map of the environment through the robot's motion process and estimates its own Posture to realize autonomous positioning and navigation of the robot. With the development of science and technology, more and more applications based on SLAM.

In the process of researching the prior art, the inventors found that there are at least the following problems in the related art: In the prior art, when positioning a robot, it is usually based on the position of each landmark in an existing map and the distance observation of the landmark by the robot at the current moment. Value to estimate the robot's position on the map at the current moment. The pose calculation of new landmarks in the map depends on the robot's position in the map and the above-mentioned distance observations. If the landmark referenced by the robot changes, the robot positioning will be inaccurate, and the pose calculation of the new landmark in the map will also be inaccurate. As a result, robot positioning and map building are greatly affected by the environment.

Summary of the Invention

An object of the embodiments of the present application is to provide an autonomous positioning and map building method, device, and robot, which can reduce the influence of the surrounding environment on the autonomous positioning and map building of the robot.

In a first aspect, an embodiment of the present application provides a method for autonomous positioning and map establishment. The method is applied to a robot. The method includes:

Obtaining a distance observation value of the robot to a landmark point;

Obtaining a position of the robot in a map;

A new landmark point belonging to a fixed object among the landmark points is added to the map, and the pose of the new landmark point is obtained according to the position and the distance observation value.

In a second aspect, an embodiment of the present application further provides an autonomous positioning and map establishment device, where the device is applied to a robot, and the device includes:

An observation distance acquisition module, configured to acquire a distance observation value of the robot to a landmark point;

A positioning module, configured to obtain a position of the robot in a map;

A mapping module is configured to add a new landmark point belonging to a fixed object among the landmark points to the map, and obtain a pose of the new landmark point according to the position and the distance observation value.

In a third aspect, an embodiment of the present application further provides a robot, including:

At least one processor; and

A memory connected in communication with the at least one processor; wherein,

The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can execute the foregoing method.

The method, device and robot for autonomous positioning and map establishment provided by the embodiments of the present application only add road signs belonging to a fixed object to the map, so as to avoid the situation that the robot changes the road signs when referring to the road signs, and reduce the robot's positioning and map establishment in the surrounding environment. Influence, so that the positioning of the robot and the calculation of road signs in the map are more accurate.

BRIEF DESCRIPTION OF THE DRAWINGS

One or more embodiments are exemplified by the pictures in the accompanying drawings. These exemplary descriptions do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the drawings in the drawings do not constitute a limitation on scale.

FIG. 1 is a schematic diagram of an application scenario of the method and device for autonomous positioning and map establishment of the present application;

2 is a schematic diagram of robot positioning and mapping in an embodiment of the present application;

FIG. 3 is a flowchart of an embodiment of an autonomous positioning and map establishment method of the present application; FIG.

4 is a flowchart of steps for obtaining a distance observation value of a robot to a landmark in an embodiment of an autonomous positioning and map establishment method of the present application;

FIG. 5 is a flowchart of an embodiment of an autonomous positioning and map building method of the present application;

FIG. 6 is a schematic structural diagram of an embodiment of an autonomous positioning and map building device of the present application;

FIG. 7 is a schematic structural diagram of an embodiment of an autonomous positioning and map building device of the present application;

FIG. 8 is a schematic diagram of a hardware structure of a robot according to an embodiment of the present application.

detailed description

In order to make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments These are part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

The method and device for autonomous positioning and map establishment provided in this application are applicable to the application scenario shown in FIG. 1. The application scenario includes a robot 10, where the robot 10 is a mobile robot, and the robot refers to a robot with certain artificial intelligence. Machines, for example, sweeping robots, humanoid robots, self-driving cars, etc. In order to complete the task of the user or other situations, the robot 10 needs to move in an unknown environment. In order to achieve autonomous positioning and navigation during the movement, it is necessary to build an incremental map and estimate its own position at the same time.

For convenience, please refer to FIG. 2, a segment of continuous movement of the robot 10 is divided into discrete moments t = 1,... K, and at each moment, the position of the robot 10 is represented by x _n , that is, x ₁ , x ₂ … x _k , It constitutes the trajectory of the robot 10. Assume that the map is composed of many landmarks, such as y ₁ , y _2, and y _3. At each moment, the robot 10 will observe a part of the landmarks and obtain their distance observations z (that is, the distance between x and y). . The positioning of the robot 10 is to estimate the position (x) of the robot 10 in the map. The robot 10 can estimate the position of the robot 10 in the existing map by observing the distance of the landmarks. The establishment of the map is to estimate the position (y) of the landmark in the map, and the position (y) of the landmark can be obtained according to the position (x) and the distance observation (z) of the robot 10. The positioning and mapping of the robot 10 is a continuous process. As the position of the robot 10 changes, the robot 10 will observe new landmarks and continuously add new landmarks to the map.

The road signs are, for example, walls, windows, pillars, trees, buildings, tables, cabinets, flowers, signs, people, pets, vehicles, and the like. In some embodiments, the mobile attributes of walls, windows, columns, trees, buildings, etc. may be defined as "fixed objects", and the mobile attributes of tables, cabinets, flowers, signs, etc. may be defined as "movable objects" , Define the moving attributes of people, pets, vehicles, etc. as "moving objects". The robot 10 only adds the landmarks belonging to the fixed object to the map. Since these landmarks are not easy to change, it can avoid the changes of the landmarks when the robot 10 refers to the landmarks, and can reduce the impact of the surrounding environment on the robot's positioning and map creation, so that The positioning of the robot and the calculation of road signs in the map are more accurate. It should be noted that the above-mentioned definition of the movement attribute can be defined in advance according to the application scenario of the robot 10, and is not absolute. The movement attribute of the same object in some application scenarios is "movable objects", and in other application scenarios it may be " Fixed object. "

FIG. 3 is a schematic flowchart of an autonomous positioning and map establishment method according to an embodiment of the present application. The method may be executed by the robot 10 in FIG. 1. As shown in FIG. 3, the method includes:

101: Obtain a distance observation value of a robot to a landmark point.

Among them, the distance observation of the robot 10 to the landmarks can be based on a visual method, such as obtaining an image in front of the robot vision through a binocular camera or a depth camera, and then obtaining the robot 10 and each of the pixels by obtaining depth information of each pixel in the image. Distance between road signs. In other embodiments, the robot may also measure the distance between the robot 10 and the road sign by other methods.

Taking a distance-observed value obtained by a binocular camera based on a vision method as an example, please refer to FIG. 4. The robot 10 obtains a distance-observed value on a road sign, including:

1011: Obtain a first image and a second image in the visual range through a binocular camera.

That is, a first image is obtained by a camera located on the left side of the robot 10 and a second image is obtained by a camera located on the right side of the robot 10, wherein the left camera and the right camera can be set on the left and right eyes of the robot 10, respectively Office.

1012: performing image recognition on the first image and the second image respectively, identifying a category of each region in the image, and determining a mobile attribute according to the category, marking the category and the mobile attribute for each region, The moving attributes include moving objects, movable objects, and fixed objects.

Specifically, for the first image and the second image to perform image recognition, for example, a neural network model based on deep learning can be used to identify, and the category of each object in the image can be identified. At the same time, the movement attributes of each object in the image can be determined according to the movement attributes defined in advance by the object category, and the category and movement attributes (such as moving objects, movable objects, and fixed objects) can be marked for the pixels of the corresponding area of the object. For example, if the object types in the first image are identified as tables, people, walls, etc. through image recognition, according to the definition of tables, people, and walls in advance, their moving attributes are respectively movable objects, moving objects, and fixed objects. The categories and moving attributes of the pixels corresponding to the areas in the table in the image are "table" and "movable objects", and the categories and moving attributes of the pixels corresponding to the areas in the image are "people" and "moving objects" ", For the pixels in the image corresponding to the area of the wall, the category and moving attributes are" wall "and" fixed object ". Those skilled in the art can understand that when marking categories and moving attributes on pixels in practical applications, computer symbols representing actual categories and actual moving attributes may be used instead of the actual categories and actual moving attributes themselves. Among them, the moving attributes of each object can be defined in advance according to the application scenario of the robot 10.

1013: Extract feature points based on the first image and the second image, and remove feature points belonging to a moving object.

Specifically, for extracting feature points from each pixel of the first image and the second image, an algorithm such as SIFT or ORB may be used. Feature points are generally some "stable points" in the image, and will not disappear due to changes in perspective, lighting changes, and noise interference, such as corner points, edge points, bright points in dark areas, and dark points in bright areas. After the feature points are extracted, the feature points marked as moving objects in the feature points can be removed. Because the moving object has a high probability of movement, if the robot 10 uses these moving objects as a reference, the position positioning of the robot 10 will be inaccurate. Therefore, in this step, landmarks whose moving attribute is a moving object can be eliminated.

In some other embodiments, after each area in the image is identified, the area where the moving attribute belongs to the moving object can also be masked directly. When performing feature point extraction in this way, feature points located in a shielded area are not extracted, so that the extracted feature points do not include feature points whose moving attribute is a moving object.

1014: Perform feature point matching based on the first image and the second image after removing feature points, and the feature point matching is performed between feature points of the same category to obtain a road in the image whose moving attribute is a non-moving object The distance measurement of the punctuation point.

Specifically, feature point matching may be performed based on, for example, a stereo matching algorithm, and matching may be performed between feature points of the same category. For example, the feature points are matched between the feature points belonging to the "table" and the feature points are matched to the "wall". This can reduce the matching range and improve the validity of the matching results. After the feature points are matched, the matching result can be used to calculate the disparity of a point on the first image and the second image according to the principle of triangulation to determine the depth of the feature point, that is, the distance of the robot 10 from the feature point.

102: Obtain a position of the robot in a map.

103: Add a new landmark point belonging to a fixed object among the landmark points to the map, and obtain a pose of the new landmark point according to the position and the distance observation value.

Please refer to FIG. 2, when the robot 10 starts a movement, the starting point (x ₁ ) of the robot 10 movement can be set as a dot, and the robot can observe the road signs y ₁ and y ₂ at this position, assuming the road signs y ₁ and y ₂ The moving attributes of are fixed objects, and road signs y ₁ and y _{2 are} added to the map. Because the robot 10 is at the dot at this time, according to the distance observation values z ₁ and z ₂ of the road signs y ₁ and y ₂ by the robot here, the postures of the road signs y ₁ and y ₂ on the map can be obtained.

When the robot 10 moves to the position x ₂ , the distance observation values of the robot 10 to the road signs y ₁ and y ₂ are z ₁ ′ and z ₂ ′, according to z ₁ ′ and z ₂ ′ on the map (the map obtained at the position x ₁ ) Location search (i.e. positioning). When a certain position in the map is searched, the distance between the position and each landmark in the map is referred to as the positioning distance, and the distance observed by the robot 10 for each corresponding landmark is referred to as the observation distance. If the coincidence degree of the observation distance is greater than a preset threshold, the position is regarded as an estimated position of the x ₂ position.

Only the road signs y ₁ and y ₂ are shown in FIG. 2. Actually each road sign includes a plurality of feature points (road sign points). Calculating the above-mentioned coincidence degree is actually calculating the degree at which the positioning distance of each landmark point in the map matches the actual observation distance of each corresponding landmark point by the robot 10. For example, if the preset threshold is 70, the landmark points whose positioning distance matches the observation distance are marked as 1, and the landmark points whose positioning distance does not match the observation distance are marked as 0. If there are more than 70 landmark points and the positioning distance of the position is consistent with the actual observation distance, the position can be considered as the position of the robot 10 on the map. Otherwise, you need to search for the new location again.

Wherein, judging whether the positioning distance of the landmark points meets the observation distance needs to be based on the corresponding landmark points. Therefore, before calculating the degree of coincidence, the correspondence between the landmark points observed by the robot 10 and the landmark points in the map needs to be matched. The feature point matching between the landmark points observed by the robot 10 and the landmark points in the map may be performed to determine the landmark points corresponding to the landmark points observed by the robot in the map. Wherein, in some embodiments, when each landmark point is added to the map, the category to which each landmark point belongs may also be marked, and the feature point matching may be performed between feature points of the same category to reduce the matching range and improve the matching. The validity of the results.

In some embodiments, when each landmark point is added to the map, a moving weight value may also be marked for each landmark point. For example, although the trees and building movement properties in the map are fixed objects, relatively speaking, buildings are less prone to movement, so you can set the building's movement weight value to be greater than the tree movement weight value. . For example, set the moving weight value of the building to 3 and the moving weight value of the tree to 2. Correspondingly, the above-mentioned degree of coincidence can also be obtained by combining the moving weight values corresponding to each landmark point in the map. In some embodiments, the coincidence degree and the moving weight value are positively correlated. The above example also illustrates that it is assumed that the moving weight value of the landmark y ₁ is 3, the moving weight value of the landmark y ₂ is 1, and the preset threshold is still 70. If the signs y ₁ 20 positioning distance compliance feature point observation distance, road signs y ₂ 15 positioning distance compliance feature point observation distance, the degree of compliance of 20 * 3 + 15 * 1 = 75> 70, this Position location succeeded. If the signs y ₁ 8 positioning from compliance feature point observation distance, road signs y ₂ 45 positioning distance compliance feature point observation distance, the degree of compliance for the 8 * 3 + 45 * 1 = 69 <70, this Location location failed, you need to search for a new location on the map again.

Among them, in some embodiments, in order to reduce the search range, the calculation amount is reduced. It is also possible to first estimate the displacement of the robot 10 by a detection device such as a sensor, and then estimate the position of the robot 10, and then perform a position search within a certain range near the estimated position in the map. Specifically, by estimating the displacement of the robot 10 between the first time and the second time, it is possible to perform feature point matching on the depth maps of the landmark points obtained at the first time and the second time, based on the depth of the same feature point in different depth maps. The displacement of the robot 10 is obtained. This displacement can be further filtered and fused with the attitude information provided by the Inertial Measurement Unit (IMU).

When the robot 10 moves to the position x ₄ , the robot observes a new landmark y ₃ at this position. If the landmark y ₃ also belongs to a fixed object, the landmark y _{3 is} added to the map. The pose of the road sign y ₃ in the map is obtained according to the position x ₄ obtained by the positioning and the distance observation value of the road sign y ₃ by the robot 10 at this position.

Because each landmark is composed of multiple feature points, the feature points corresponding to the landmark are also added to the map. These feature point clouds are combined to form a map.

The method for autonomous positioning and map establishment provided in the embodiments of the present application only adds landmark points that belong to a fixed object to the map to avoid the situation that the robot changes the landmark when referring to the landmark, and reduces the influence of the surrounding environment on the robot positioning and map establishment. This makes the robot's positioning and the calculation of road signs in the map more accurate.

In other embodiments, please refer to FIG. 5. In addition to steps 101-103, the method for autonomous positioning and map establishment further includes step 104:

If the position coincides with the historical position, the position is corrected according to the landmark position pose obtained at the historical position, and the landmark position pose obtained between the historical position and the position is corrected.

In order to correct errors due to drift, a map with consistent information is obtained. The robot 10 periodically detects whether the current position is a historical position that has been visited before, that is, performs loopback detection. If the current position coincides with the historical position, the current position is corrected based on the map obtained at the historical position, and the map obtained from the historical position to the current position is corrected. Because the robot 10 removes the influence of road sign changes during positioning, it is easier for the robot 10 to find the historical position during loop detection.

Correspondingly, an embodiment of the present application further provides an autonomous positioning and map establishment device, which is used for the robot 10 shown in FIG. 1, and as shown in FIG. 6, the autonomous positioning and map establishment The apparatus 600 includes:

An observation distance acquisition module 601, configured to acquire a distance observation value of the robot to a landmark point;

A positioning module 602, configured to obtain a position of the robot in a map;

A mapping module 603 is configured to add a new landmark point belonging to a fixed object among the landmark points to the map, and obtain a pose of the new landmark point according to the position and the distance observation value.

The autonomous positioning and map building device provided in the embodiment of the present application only adds the landmarks belonging to a fixed object to the map, so as to avoid the situation that the robot changes the landmark when referring to the landmark, and reduces the influence of the surrounding environment on the robot positioning and map establishment, thereby Make the robot's positioning and the calculation of road signs in the map more accurate.

In some embodiments of the autonomous positioning and map building apparatus 600, the observation distance acquisition module 601 is specifically configured to:

Acquiring a first image and a second image in the visual range through a binocular camera device;

Performing image recognition on the first image and the second image respectively, identifying a category of each region in the image, and determining a moving attribute according to the category, marking the category and the moving attribute for each region, the Movement attributes include moving objects, movable objects and fixed objects;

Extracting feature points based on the first image and the second image, and removing feature points belonging to a moving object;

Feature point matching is performed based on the first image and the second image after the feature points are removed, and the feature point matching is performed between feature points of the same category to obtain a landmark point in the image whose moving attribute is a non-moving object. Distance observations.

Specifically, in some of these embodiments, the mapping module 603 is specifically configured to:

A new landmark point belonging to a fixed object among the landmark points is added to the map, and a category and a moving weight value are marked for the new landmark point.

Specifically, in some of these embodiments, the positioning module 602 is specifically configured to:

Match the landmark points observed by the robot at the second moment with the landmark points in the map at the first moment, and determine the corresponding landmark points in the map at the first moment by the landmark points observed by the robot at the second moment. The feature points match the category Between the same feature points;

Performing a position search in the map at the first time according to the distance observation value of the robot to a landmark point at the second time;

Combine the moving weight value of the landmark points in the map to obtain the degree of coincidence between the positioning distance and the observation distance when the robot is at a certain position in the map. If the degree of coincidence exceeds a preset threshold, locate the position as The position of the robot in the map, the positioning distance is the distance between the position in the map and each landmark point, and the observation distance is the distance observation value of the robot to each corresponding landmark point.

In other embodiments of the autonomous positioning and map establishing device 600, the device described with reference to FIG. 7 further includes:

A loopback detection module 604 is configured to, if the position coincides with a historical position, correct the position according to the landmark position pose obtained at the historical position, and the landmark position obtained between the historical position and the position posture.

It should be noted that the above-mentioned autonomous positioning and map building device can execute the autonomous positioning and map building method provided in the embodiments of the present application, and has corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in the embodiment of the autonomous positioning and map establishment device, reference may be made to the autonomous positioning and map establishment method provided in the embodiment of the present application.

FIG. 8 is a schematic diagram of a hardware structure of a robot 10 according to an embodiment of the present application. As shown in FIG. 8, the robot 10 includes:

One or more processors 11 and a memory 12. One processor 11 is taken as an example in FIG. 8.

The processor 11 and the memory 12 may be connected through a bus or other manners. In FIG. 8, the connection through the bus is taken as an example.

The memory 12 is a non-volatile computer-readable storage medium, and may be used to store non-volatile software programs, non-volatile computer executable programs, and modules, as corresponding to the methods for autonomous positioning and map establishment in the embodiments of the present application. Program instructions / modules (for example, the observation distance acquisition module 601, the positioning module 602, and the mapping module 603 shown in FIG. 6). The processor 11 executes various functional applications and data processing of the robot by running the non-volatile software programs, instructions, and modules stored in the memory 12, that is, the autonomous positioning and map establishment method of the above method embodiment.

The memory 12 may include a storage program area and a storage data area, where the storage program area may store an operating system and application programs required for at least one function; the storage data area may store data created according to autonomous positioning and use of a map building device, and the like . In addition, the memory 12 may include a high-speed random access memory, and may further include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 12 may optionally include a memory remotely disposed with respect to the processor 11, and these remote memories may be connected to the autonomous positioning and mapping device through a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

The one or more modules are stored in the memory 12, and when executed by the one or more processors 11, execute the autonomous positioning and map establishment method in any of the above method embodiments, for example, execute the above-described The method steps 101 to 103 in FIG. 3, the method steps 1011 to 1014 in FIG. 4, and the method steps 101 to 104 in FIG. 5; implement modules 601-603 in FIG. 6 and modules 601- in FIG. 7. 604 features.

The above product can execute the method provided in the embodiment of the present application, and has the corresponding functional modules and beneficial effects of executing the method. For technical details not described in detail in this embodiment, reference may be made to the method provided in the embodiment of the present application.

An embodiment of the present application provides a non-volatile computer-readable storage medium, where the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are executed by one or more processors, for example, as shown in FIG. 8 A processor 11 may enable the one or more processors to execute the autonomous positioning and map building method in any of the foregoing method embodiments, for example, execute the method steps 101 to 103 in FIG. 3 described above, and FIG. 4 The method includes steps 1011 to 1014 in the method, and steps 101 to 104 in the method in FIG. 5. The functions of modules 601 to 603 in FIG. 6 and modules 601 to 604 in FIG. 7 are implemented.

The device embodiments described above are only schematic, wherein 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, may be located One place, or it can be distributed across multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the objective of the solution of this embodiment.

Through the description of the above embodiments, a person of ordinary skill in the art can clearly understand that the embodiments can be implemented by means of software plus a general hardware platform, and of course, also by hardware. A person of ordinary skill in the art can understand that all or part of the processes in the method of the foregoing embodiment can be completed by using a computer program to instruct related hardware. The program can be stored in a computer-readable storage medium. When executed, the processes of the embodiments of the methods described above may be included. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RandomAccess Memory, RAM).

Finally, it should be noted that the above embodiments are only used to describe the technical solution of the present application, but not limited thereto. Under the idea of the present application, the technical features in the above embodiments or different embodiments may also be combined. The steps can be implemented in any order, and there are many other variations of different aspects of the present application as described above, for the sake of brevity, they are not provided in the details; although the present application is described in detail with reference to the foregoing embodiments, it is common in the art The skilled person should understand that they can still modify the technical solutions described in the foregoing embodiments, or equivalently replace some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions separate from the implementation of this application. Examples of technical solutions.

Claims

An autonomous positioning and map building method, which is applied to a robot, is characterized in that the method includes:

Obtaining a distance observation value of the robot to a landmark point;

Obtaining a position of the robot in a map;

A new landmark point belonging to a fixed object among the landmark points is added to the map, and the pose of the new landmark point is obtained according to the position and the distance observation value.
The method according to claim 1, wherein the acquiring the distance observation value of the robot to a landmark point comprises:

Acquiring a first image and a second image in the visual range through a binocular camera device;

Performing image recognition on the first image and the second image respectively, identifying a category of each region in the image, and determining a moving attribute according to the category, marking the category and the moving attribute for each region, the Movement attributes include moving objects, movable objects and fixed objects;

Extracting feature points based on the first image and the second image, and removing feature points belonging to a moving object;

Feature point matching is performed based on the first image and the second image after the feature points are removed, and the feature point matching is performed between feature points of the same category to obtain a landmark point in the image whose moving attribute is a non-moving object. Distance observations.
The method according to claim 1 or 2, wherein the adding a new landmark point belonging to a fixed object among the landmark points to the map comprises:

A new landmark point belonging to a fixed object among the landmark points is added to the map, and a category and a moving weight value are marked for the new landmark point.
The method according to claim 3, wherein the acquiring the position of the robot in a map comprises:

Match the landmark points observed by the robot at the second moment with the landmark points in the map at the first moment, and determine the corresponding landmark points in the map at the first moment by the landmark points observed by the robot at the second moment. The feature points match the category Between the same feature points;

Performing a position search in the map at the first time according to the distance observation value of the robot to a landmark point at the second time;

Combine the moving weight value of the landmark points in the map to obtain the degree of coincidence between the positioning distance and the observation distance when the robot is at a certain position in the map. If the degree of coincidence exceeds a preset threshold, locate the position as The position of the robot in the map, the positioning distance is the distance between the position in the map and each landmark point, and the observation distance is the distance observation value of the robot to each corresponding landmark point.
The method according to any one of claims 1-4, further comprising:

If the position coincides with the historical position, the position is corrected according to the landmark position pose obtained at the historical position, and the landmark position pose obtained between the historical position and the position is corrected.
An autonomous positioning and map building device, which is applied to a robot, is characterized in that the device includes:

An observation distance acquisition module, configured to acquire a distance observation value of the robot to a landmark point;

A positioning module, configured to obtain a position of the robot in a map;

A mapping module is configured to add a new landmark point belonging to a fixed object among the landmark points to the map, and obtain a pose of the new landmark point according to the position and the distance observation value.
The apparatus according to claim 6, wherein the observation distance obtaining module is specifically configured to:

Acquiring a first image and a second image in the visual range through a binocular camera device;

Performing image recognition on the first image and the second image respectively, identifying a category of each region in the image, and determining a moving attribute according to the category, marking the category and the moving attribute for each region, the Movement attributes include moving objects, movable objects and fixed objects;

Extracting feature points based on the first image and the second image, and removing feature points belonging to a moving object;

Feature point matching is performed based on the first image and the second image after the feature points are removed, and the feature point matching is performed between feature points of the same category to obtain the Distance observations.
The device according to claim 6 or 7, wherein the mapping module is specifically configured to:

A new landmark point belonging to a fixed object among the landmark points is added to the map, and a category and a moving weight value are marked for the new landmark point.
The device according to claim 8, wherein the positioning module is specifically configured to:

Match the landmark points observed by the robot at the second moment with the landmark points in the map at the first moment, and determine the corresponding landmark points in the map at the first moment by the landmark points observed by the robot at the second moment. The feature points match the category Between the same feature points;

Performing a position search in the map at the first time according to the distance observation value of the robot to a landmark point at the second time;

Combine the moving weight value of the landmark points in the map to obtain the degree of coincidence between the positioning distance and the observation distance when the robot is at a certain position in the map. If the degree of coincidence exceeds a preset threshold, locate the position as The position of the robot in the map, the positioning distance is the distance between the position in the map and each landmark point, and the observation distance is the distance observation value of the robot to each corresponding landmark point.
The device according to any one of claims 6-9, wherein the device further comprises:

A loop detection module configured to correct the position according to the landmark position pose obtained from the historical position, and the landmark position pose obtained between the historical position and the position if the position coincides with the historical position .
A robot is characterized by comprising:

At least one processor; and

A memory connected in communication with the at least one processor; wherein,

The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can execute the method according to any one of claims 1-5. method.
A non-volatile computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a robot, the robot executes claim 1 -The method according to any one of -5.
A computer program product, characterized in that the computer program product includes a computer program stored on a non-volatile computer-readable storage medium, the computer program includes program instructions, and when the program instructions are executed by a robot, The robot is caused to perform the method according to any one of claims 1-5.