Wall-climbing robot positioning system and method based on spatial structure circular descriptor

Abstract

The present disclosure relates to a wall-climbing robot positioning system and method based on a spatial structure circular descriptor. A depth image sensor (RGB-D) camera outputs a depth image; collects three-dimensional point cloud data of three-dimensional space wall points; and conformally maps the outer contour of the wall-climbing robot onto a minimum bounding sphere (MBS). A spatial structure circular descriptor (SSCD) algorithm module maps the three-dimensional point cloud data to a two-dimensional positioning and navigation plane, and forms an SSCD image thereon; matches the point cloud features of the SSCD image from time k-1 to time k, and completes the positioning of the wall-climbing robot on the three-dimensional space wall at the current moment. An onboard inertial measurement unit (IMU) measurement module collects the angular velocity and acceleration of the wall-climbing robot; and obtains the posture information of the wall-climbing robot at the current moment by pre-integrating the collected angular velocity and acceleration. Thus, the spatial structure circular descriptor (SSCD) algorithm maps the three-dimensional data to a two-dimensional plane, and accurately positions and navigates the wall-climbing robot.

Description

Technical Field

[0001] The present disclosure relates to the technical field of autonomous positioning and navigation in three-dimensional space, and more specifically, to a wall-climbing robot positioning system and method based on a spatial structure circular descriptor. Background Art

[0002] This Background generally presents the context of the present disclosure. To the extent described in this Background, the work of the presently named inventors and aspects of the specification that may not constitute prior art at the time of filing are neither expressly nor impliedly admitted to be prior art to the present disclosure.

[0003] With the rapid development of automation and intelligent technology, wall-climbing robots, as special robots that can operate on vertical or unstructured surfaces, have increasingly prominent application potential in building inspection, pipeline inspection, disaster relief and other fields. However, these robots face major challenges in autonomous positioning and navigation in three-dimensional space in order to achieve efficient and accurate operation in complex or dynamically changing environments.

[0004] Existing wall-climbing robots mainly rely on traditional hardware systems such as ultrasonic, infrared or contact sensors for positioning and navigation. However, these traditional hardware systems have limitations in environmental adaptability and positioning accuracy in three-dimensional indoor spaces, especially when facing complex or irregular three-dimensional wall environments, where the challenges of navigation and positioning are more significant; in addition, existing navigation algorithms have poor adaptability in dynamic environments and are easily disturbed by the environment, resulting in navigation deviations; at the same time, the high demand for computing resources by complex algorithms also brings challenges to resource-constrained mobile platforms.

[0005] In order to enable wall-climbing robots to adapt to more complex and changeable operating environments, it is necessary to explore and develop more accurate and robust three-dimensional space autonomous positioning and navigation systems and methods for wall-climbing robots, so as to significantly improve the autonomous positioning and navigation capabilities of wall-climbing robots in complex environments.

[0006] In summary, the key technologies in the current field of wall-climbing robots are not yet mature, and the existing technologies cannot meet the actual needs. To this end, the technical problem to be solved by the present disclosure is to overcome the problems of poor autonomous positioning and navigation accuracy of wall-climbing robots in the prior art, poor adaptability to complex and changeable operating environments, and the urgent need for a wall-climbing robot positioning system and method based on a spatial structure circular descriptor, which aims to solve the problems of poor accuracy and poor adaptability in the prior art, and can quickly, effectively and accurately realize the autonomous positioning and navigation of wall-climbing robots in complex and dynamically changing three-dimensional complex wall environments. Summary of the invention

[0007] The technical problem disclosed in the present invention is: to design a new wall-climbing robot positioning system and method in view of the defects of the existing wall-climbing robot autonomous positioning method. The defects of the prior art are fully overcome, and a wall-climbing robot positioning system and method based on a spatial structure circular descriptor are proposed, aiming to solve the problems of poor accuracy and poor adaptability in the prior art, and can quickly, effectively and accurately realize the autonomous positioning and navigation of the wall-climbing robot in a complex and dynamically changing three-dimensional complex wall environment.

[0008] In order to solve at least one of the above technical problems, a specific embodiment of the present disclosure proposes a wall-climbing robot positioning system based on a spatial structure circular descriptor (SSCD), comprising:

[0009] A depth image sensor (RGB-D) camera is configured to output a depth image; and, based on the output depth image, collect three-dimensional point cloud data of wall points in a three-dimensional space; and, conformally map the outer contour of the wall-climbing robot onto a minimum bounding sphere (MBS);

[0010] The spatial structure circular descriptor (SSCD) algorithm module is configured to map the three-dimensional point cloud data of the three-dimensional space wall points collected by the depth image sensor (RGB-D) camera to a two-dimensional positioning and navigation plane, and form a spatial structure circular descriptor (SSCD) image on the two-dimensional positioning and navigation plane; and, match the point cloud features of the two spatial structure circular descriptor (SSCD) images from time k-1 to time k, and complete the positioning of the wall-climbing robot on the three-dimensional space wall at the current moment; and,

[0011] The onboard inertial measurement unit (IMU) measurement module is configured to collect the angular velocity and acceleration of the wall-climbing robot; and obtain the posture information of the wall-climbing robot at the current moment by pre-integrating the collected angular velocity and acceleration.

[0012] In order to solve at least one of the above technical problems, a specific embodiment of the present disclosure further proposes a wall-climbing robot positioning method based on a spatial structure circular descriptor (SSCD), comprising the following steps:

[0013] A depth image sensor (RGB-D) camera outputs a depth image; and,

[0014] Based on the output depth image, the depth image sensor (RGB-D) camera collects three-dimensional point cloud data of the wall points in the three-dimensional space; and,

[0015] The depth image sensor (RGB-D) camera conformally maps the contour of the wall-climbing robot onto the minimum bounding sphere (MBS);

[0016] The spatial structure circular descriptor (SSCD) algorithm module maps the three-dimensional point cloud data of the three-dimensional space wall points collected by the depth image sensor (RGB-D) camera to a two-dimensional positioning navigation plane, and forms an image of the spatial structure circular descriptor (SSCD) on the two-dimensional positioning navigation plane; and,

[0017] The SSCD algorithm module matches the point cloud features of two SSCD images from time k-1 to time k to complete the positioning of the wall-climbing robot on the three-dimensional wall at the current moment; and

[0018] The onboard inertial measurement unit (IMU) measurement module collects the angular velocity and acceleration of the wall-climbing robot;

[0019] The onboard inertial measurement unit (IMU) measurement module obtains the posture information of the wall-climbing robot at the current moment by pre-integrating the collected angular velocity and acceleration.

[0020] In order to solve at least one of the above technical problems, a specific embodiment of the present disclosure further proposes a wall-climbing robot positioning method based on a spatial structure circular descriptor (SSCD), comprising the following steps:

[0021] The spatial structure circular descriptor (SSCD) algorithm module initializes the three-dimensional space wall, establishes the minimum bounding sphere (MBS) of the three-dimensional space wall point cloud at this time, maps the minimum bounding sphere (MBS) of the three-dimensional space to the two-dimensional positioning navigation plane, and obtains the first spatial structure circular descriptor (SSCD1);

[0022] After the wall-climbing robot is adsorbed on the three-dimensional space wall, the spatial structure circular descriptor (SSCD) algorithm module maps the wall-climbing robot and the three-dimensional space wall together on the two-dimensional positioning and navigation plane, and obtains the second spatial structure circular descriptor (SSCD2) of the two-dimensional plane positioning and navigation map at time k-1;

[0023] Point cloud feature matching is performed between the first spatial structure circular descriptor (SSCD1) and the second spatial structure circular descriptor (SSCD2) to determine the position of the wall-climbing robot on the wall surface of the three-dimensional space.

[0024] Further preferably, the wall-climbing robot positioning method further comprises the following steps:

[0025] After the wall-climbing robot moves on the three-dimensional wall, the spatial structure circular descriptor (SSCD) algorithm module obtains the third spatial structure circular descriptor (SSCD3) of the two-dimensional positioning plane navigation map at time k.

[0026] The spatial structure circular descriptor (SSCD) algorithm module completes the real-time positioning of the wall-climbing robot by performing feature matching on the second spatial structure circular descriptor (SSCD2) at time k-1 and the third spatial structure circular descriptor (SSCD3) at time k; and, through the point cloud changes of the wall-climbing robot, the spatial structure circular descriptor (SSCD) algorithm module performs real-time correction on the posture information of the wall-climbing robot measured by the onboard inertial measurement unit (IMU).

[0027] In summary, compared with the prior art, the present disclosure provides a wall-climbing robot positioning system and method based on a spatial structure circular descriptor, which has the following beneficial technical effects:

[0028] The beneficial technical effects of the present disclosure include:

[0029] 1. The Spatial Structure Circular Descriptor (SSCD) algorithm maps three-dimensional data to a two-dimensional plane for precise positioning and navigation: Through the three-dimensional wall feature conversion technology based on the Spatial Structure Circular Descriptor (SSCD), complex three-dimensional wall features are converted into three-dimensional point cloud data carrying spatial structure information and mapped to a two-dimensional plane, thereby achieving precise positioning and navigation of complex three-dimensional environments and improving the positioning accuracy of wall-climbing robots in complex environments.

[0030] 2. The Spatial Structure Circular Descriptor (SSCD) algorithm maps three-dimensional data to a two-dimensional plane to improve data processing efficiency: The Spatial Structure Circular Descriptor (SSCD) algorithm maps three-dimensional point cloud spatial information data to a two-dimensional plane, achieves data compression, and realizes efficient data processing while retaining key spatial structure information, thereby improving the autonomous navigation and positioning efficiency of wall-climbing robots in complex environments.

[0031] 3. Conformal mapping technology simplifies spatial information processing: Conformal mapping technology is used to project the normal of the wall-climbing robot's outer contour onto the three-dimensional wall surface, and the distance difference is used as the elevation information, thereby realizing the representation of the surface morphology in three-dimensional space on a two-dimensional plane, simplifying the processing of spatial information data.

[0032] 4. Onboard inertial measurement unit (IMU) and point cloud feature matching to enhance system robustness: The integrated autonomous positioning and navigation functions reduce the reliance on external reference systems. The onboard inertial measurement unit (IMU) and point cloud feature matching are used to correct the attitude and position in real time. Even in the case of noise or partial failure of sensor data, the robustness of the system can be improved to ensure stable navigation performance. The integrated autonomous positioning and navigation functions reduce the reliance on external reference systems, allowing the wall-climbing robot to achieve precise positioning and navigation even without external signals.

[0033] 5. Combined circular and sector-shaped partitioning to improve positioning accuracy: Through the mapping and area division of the two-dimensional positioning navigation plane, the use of combined circular and sector-shaped partitioning enables the wall-climbing robot to understand and process spatial information more intuitively, thereby improving positioning accuracy.

[0034] 6. Combination of the onboard inertial measurement unit (IMU) measurement module and the SSCD algorithm to enhance adaptability to dynamic environments: Combining the onboard inertial measurement unit (IMU) measurement module with the spatial structure circular descriptor (SSCD) algorithm, preliminary measurement and positioning are achieved through pre-integration of the measurement values ​​of the onboard inertial measurement unit (IMU), and then the posture and position are corrected through point cloud feature matching of the spatial structure circular descriptor (SSCD), thereby improving the stability and reliability of the wall-climbing robot in dynamic environments.

[0035] 7. Improve safety: Through the recognition and response of abnormal wall features, it can identify abnormal wall types such as inner corners, outer corners and discontinuous walls, so that the wall-climbing robot can actively avoid dangerous areas and establish reasonable control strategies to enhance safety.

[0036] Through the above technical scheme, the positioning method of the wall-climbing robot disclosed in the present invention provides an efficient, low-cost, and easy-to-operate solution for the positioning of the wall-climbing robot for autonomous positioning and navigation in three-dimensional space, which has significant technological advancement and market application potential.

[0037] Further areas of applicability will become apparent from the description provided herein.It should be understood that the description and specific examples are intended for purposes of illustration only and are not intended to limit the scope of the present disclosure.

[0038] The above features and advantages and other features and advantages of the presently disclosed systems and methods are apparent from the detailed description, including the claims and exemplary embodiments, when taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The following is a brief introduction to the drawings required for use in the specific implementation or description of the prior art.

[0040] Obviously, some drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0041] In order to more clearly illustrate the specific implementation manner of the present disclosure or the technical solutions in the prior art, more details, features and advantages of the present disclosure are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which:

[0042] Figure 1A schematic diagram of an autonomous positioning and navigation system of a wall-climbing robot on a three-dimensional space wall in a specific embodiment of the present disclosure;

[0043] Figure 2 is a schematic diagram of the principle of a spatial structure circular descriptor (SSCD) algorithm in a specific embodiment of the present disclosure;

[0044] Figure 3 It is a schematic diagram of a method for autonomous positioning and navigation of a wall-climbing robot on a wall surface in a three-dimensional space in a specific embodiment of the present disclosure;

[0045] Figure 4 The figure is a schematic diagram of the area division of the spatial structure circular descriptor (SSCD) mapping in a specific embodiment of the present disclosure. DETAILED DESCRIPTION

[0046] In order to more clearly understand the above-mentioned purposes, features and advantages of the present disclosure, the present disclosure is further described in detail below in conjunction with specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0047] In the following description, many specific details will be set forth with reference to several examples of the present disclosure illustrated in the accompanying drawings to facilitate a full understanding of the present disclosure. Whenever possible, the same or similar reference numerals are used in the drawings and the description to represent the same or similar components or steps.

[0048] However, the present disclosure may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present disclosure is not limited by the specific embodiments disclosed below.

[0049] like Figure 1 As shown, the embodiment of the present disclosure provides an autonomous positioning and navigation system for a wall-climbing robot on a three-dimensional wall surface based on a spatial structure circular descriptor (SSCD), the system comprising: a depth image sensor (RGB-D) camera, a spatial structure circular descriptor (SSCD) algorithm module and an airborne inertial measurement unit (IMU) measurement module, wherein:

[0050] Depth image sensor output: Depth image sensor (RGB-D) camera, capable of outputting multiple depth images.

[0051] 3D data acquisition by depth image sensor: Based on each depth image output by the depth image sensor (RGB-D) camera, the depth image sensor (RGB-D) camera can collect 3D point cloud data containing more than 100,000 3D space wall points. These 3D point cloud data represent the 3D shape, position and characteristics of the 3D space wall of the steel structure, as well as the 3D shape of the wall-climbing robot. These 3D point cloud data provide the necessary data basis for subsequent conformal mapping. The 3D point cloud data of the 3D space wall collected by the depth image sensor (RGB-D) camera includes: elevation information of the wall-climbing robot and the 3D space wall, convexities and concavities of the 3D space wall and other spatial point modal information, providing rich environmental information for the robot's positioning and navigation.

[0052] Homothetic Mapping technology simplifies spatial information processing: The conformal mapping technology is used to project the normal of the wall-climbing robot's outline onto the three-dimensional space wall, and the distance difference is used as the elevation information, which finally realizes the representation of the curved surface in the three-dimensional space on the two-dimensional plane, simplifying the processing of spatial information data. Adopting conformal mapping (homothetic mapping) for three-dimensional point cloud data to simplify spatial information processing: Adopting conformal mapping technology, the depth image sensor (RGB-D) camera projects the normal of the wall-climbing robot's outline, that is, the normal of the wall-climbing robot's outline is mapped onto the three-dimensional space wall. This step is to determine the relative position and posture of the robot when it contacts the three-dimensional space wall. Homothetic Mapping: Homothetic mapping, or similarity mapping, has the invariance of angle preservation and scaling rate. This means that after this homothetic mapping, the angle and shape of the figure remain unchanged, but the size changes. That is to say, when mapping one area to another area, not only the direction and angle are maintained, but also the proportional relationship of the enlargement or reduction of the shape is maintained. In two-dimensional space, homothetic mapping is usually expressed as a transformation multiplied by a constant scaling factor, such as scaling and translation. Homothetic Mapping, the analysis of homothetic mapping involves its angle preservation and scaling rate invariance. The Riemann mapping theorem, an important theorem in complex function theory, shows that in a simply connected region, there is a unique conformal mapping that maps it to the inside of the unit circle. To understand conformal mapping from a geometric perspective, we can consider any directed continuous curve on a plane, which can be described by a parametric equation. By studying the derivative of the curve (the angle between the tangent of the curve and the real axis), we can understand the changes in the curve, including its scaling and rotation. In conformal mapping, this transformation not only maintains the angle of the curve, but also the shape of the curve, but only changes the size of the curve. The results of conformal mapping can be used to guide the motion and control strategies of wall-climbing robots. Through conformal mapping technology, the processing of spatial information data is simplified. The wall-climbing robot can adapt to the three-dimensional shape of the wall in three-dimensional space more accurately, accurately represent the surface shape in three-dimensional space on a two-dimensional plane, improve the wall-climbing robot's wall-climbing adaptability, stability and efficiency, realize the wall-climbing robot's precise positioning and navigation on the wall in three-dimensional space, and improve the efficiency of data processing.

[0053] The distance difference between the outline of the wall-climbing robot and the wall surface of the steel structure in three-dimensional space and the minimum bounding sphere (MBS) is calculated, and the distance difference is set as the elevation information on the conformal three-dimensional surface (minimum bounding sphere, MBS), which is used to represent the surface morphology in the three-dimensional point cloud data space of the wall-climbing robot and the steel structure wall surface on the two-dimensional positioning and navigation plane. In this way, conformal mapping is achieved, that is, the surface morphology in three-dimensional space is accurately represented on the two-dimensional positioning and navigation plane.

[0054] The two-dimensional positioning and navigation plane is divided into multiple areas. Starting from the center of the circle, the images of these spatial structure circular descriptors (SSCD) on the two-dimensional positioning and navigation plane are decomposed into ring-shaped and sector-shaped combined partitions; each ring-shaped partition and sector-shaped partition correspond to the latitude and longitude of the minimum bounding sphere (MBS). The two-dimensional positioning and navigation plane can map the relative position relationship between the wall-climbing robot and the three-dimensional space wall.

[0055] Positioning of three-dimensional point cloud data: The real-time three-dimensional point cloud data perceived by the wall-climbing robot is combined with the structured circular descriptor (SSCD) algorithm to extract the features of the spatial structured circular descriptor (SSCD) image at the current viewing angle. The structured circular descriptor (SSCD) algorithm module maps the three-dimensional point cloud data of the three-dimensional wall points collected by the depth image sensor (RGB-D) camera to the two-dimensional positioning navigation plane (forming the image of the spatial structured circular descriptor (SSCD) on the two-dimensional positioning navigation plane), and then compares the point cloud feature matching of the two spatial structured circular descriptor (SSCD) images from time k-1 to time k, that is, the features of the spatial structured circular descriptor (SSCD) image extracted at time k are matched with the features of the spatial structured circular descriptor (SSCD) image in the three-dimensional wall environment collected in advance at time k-1 for point cloud feature matching, thereby identifying that only the three-dimensional points in the area where the wall-climbing robot is located have changed, and then completing the positioning of the wall-climbing robot on the three-dimensional wall at the current moment.

[0056] Data acquisition of the onboard inertial measurement unit (IMU): The onboard inertial measurement unit (IMU) measurement module can collect the angular velocity and acceleration of the wall-climbing robot; and obtain the posture information of the wall-climbing robot at the current moment by pre-integrating the collected angular velocity and acceleration. This information will be used for subsequent path planning to ensure the motion control of the robot. Subsequently, the robot's positioning information and the three-dimensional information of the environment will be combined, and the rapidly-extremely random tree (RRT) path planning algorithm will be applied to calculate the optimal path from the current position to the target position on the two-dimensional positioning navigation plane to generate navigation information, and inversely map it on the three-dimensional space wall.

[0057] In a specific embodiment, the onboard inertial measurement unit (IMU) and point cloud feature matching enhance the robustness of the system: the autonomous positioning and navigation functions are integrated, and the dependence on the external reference system is reduced, so that the wall-climbing robot can achieve accurate positioning and navigation even without external signals. The posture and position are corrected in real time by matching the onboard inertial measurement unit (IMU) and point cloud features. Even in the case of noise or partial failure of sensor data, the robustness of the system can be improved, the dependence on the external reference system is reduced, the autonomy and reliability of the wall-climbing robot are improved, and stable navigation performance is ensured. When positioning and navigating the wall-climbing robot, by using the spatial structure circular descriptor (SSCD) algorithm, accurate positioning and navigation of complex three-dimensional environments can be achieved, with high robustness and adaptability.

[0058] In a specific embodiment, the onboard inertial measurement unit (IMU) measurement module is combined with the spatial structure circular descriptor (SSCD) algorithm to enhance the adaptability to dynamic environments: the onboard inertial measurement unit (IMU) measurement module is combined with the spatial structure circular descriptor (SSCD) algorithm. On the one hand, the posture and position are corrected by feature matching the three-dimensional point cloud data of the spatial structure circular descriptor (SSCD); on the other hand, preliminary measurement and positioning are achieved by pre-integration of the measurement values ​​of the onboard inertial measurement unit (IMU). The stability and reliability of the wall-climbing robot in a dynamic environment are improved.

[0059] Figure 2 It is a schematic diagram of the principle of the spatial structure circular descriptor (SSCD) algorithm in a specific embodiment of the present disclosure.

[0060] First, the radius r of the minimum bounding sphere (MBS) of the three-dimensional space wall is calculated; wherein, according to the spatial structure circular descriptor (SSCD) algorithm, the three-dimensional space wall features are converted into three-dimensional point cloud data carrying spatial structure information, and the minimum bounding sphere (MBS) of the three-dimensional point cloud is established, that is, all point cloud points are surrounded by a sphere, the center and radius of the sphere are determined by the distribution of the point cloud, and the radius r of the sphere is as small as possible.

[0061] Secondly, the spatial structure information of all points on the three-dimensional space wall is projected onto the minimum bounding sphere (MBS), and one or more attribute values ​​are assigned to each point on the minimum bounding sphere (MBS) to represent the three-dimensional space information of the three-dimensional space wall and the wall-climbing robot;

[0062] Finally, the south pole O of the minimum bounding sphere (MBS) remains stationary, and each latitude line on the sphere is mapped vertically to the surroundings. At this time, the minimum bounding sphere (MBS) is mapped to a plane, thus forming a circular area, and the overall spatial structure information of the three-dimensional space wall is maintained, and the projection area remains limited. The projection image is a circular area with a radius of πr. The projection method is conformal mapping, which can retain the characteristic structural information of the original three-dimensional space wall.

[0063] The minimum bounding sphere (MBS) is projected onto the two-dimensional positioning navigation plane, and the projection area is a limited circular area, and the spatial structure of the original three-dimensional space wall can be retained.

[0064] In the three-dimensional coordinate system, for the convenience of description and calculation, such as Figure 2 As shown, the origin is moved to the center point O' of the minimum bounding sphere (MBS). For a point M(x M ,y M ,z M ), create a line starting from the center point O' and passing through the point M(x) on the wall of the three-dimensional space M ,y M ,z M ) has an intersection point S(x S ,y S ,z S ), where the intersection point S(x S ,y S ,z S )satisfy:

[0065]

[0066] The depth information is the intersection point S(x S ,y S ,z S ) provides the attribute value α S , indicating that M(x M ,y M ,z M )’s spatial information, attribute value α S The calculation formula is as follows:

[0067]

[0068] Where, d M It is the Euclidean distance between the surface point M and the center point O' in three-dimensional space.

[0069] Then, the origin is moved to the south pole O of the minimum bounding sphere (MBS), and the minimum bounding sphere (MBS) is projected onto the XOY plane, where: and

[0070]

[0071] Where r is the radius of the minimum bounding sphere (MBS), l is the arc between points O and S on the sphere The length of OQ is the distance between points O and Q on the two-dimensional plane.

[0072] The projected image is a circular area with a radius of πr. In the projected two-dimensional XOY plane image, each pixel has all the attribute values ​​of the corresponding point on the minimum bounding sphere (MBS) surface.

[0073] Now, every point M(x M ,y M ,z M ) has a corresponding point Q in the spatial structured circular descriptor (SSCD).

[0074] However, since the wall may have convexities, concavities, and discontinuous walls, a point in the spatial structure circular descriptor (SSCD) may have multiple corresponding points on the three-dimensional steel structure wall, and there will be multiple attribute values ​​α S , different attribute values ​​α S Reflects different three-dimensional wall characteristics.

[0075] The Spatial Structure Circular Descriptor (SSCD) algorithm maps three-dimensional data to a two-dimensional plane for precise positioning and navigation: Through the three-dimensional wall feature conversion technology based on the Spatial Structure Circular Descriptor (SSCD), complex three-dimensional wall features are converted into three-dimensional point cloud data carrying spatial structure information and mapped to a two-dimensional plane, thereby achieving precise positioning and navigation of complex three-dimensional environments and improving the positioning accuracy of wall-climbing robots in complex environments.

[0076] The Spatial Structure Circular Descriptor (SSCD) algorithm maps three-dimensional data to a two-dimensional plane to improve data processing efficiency: The Spatial Structure Circular Descriptor (SSCD) algorithm maps three-dimensional point cloud spatial information data to a two-dimensional plane, achieves data compression, and realizes efficient data processing while retaining key spatial structure information, thereby improving the autonomous navigation and positioning efficiency of wall-climbing robots in complex environments.

[0077] Figure 3 It is a schematic diagram of a method for autonomous positioning and navigation of a wall-climbing robot on a wall surface in three-dimensional space in a specific embodiment of the present invention.

[0078] like Figure 3As shown, the spatial structure circular descriptor (SSCD) algorithm module removes abnormal points from the depth image sensor (RGB-D) camera point cloud data at time k-1 through point cloud preprocessing to obtain the point cloud preprocessing data at time k-1.

[0079] Initialize the three-dimensional space wall, establish the minimum bounding sphere (MBS) of the three-dimensional space wall point cloud at this time, map the minimum bounding sphere (MBS) of the three-dimensional space to the two-dimensional positioning navigation plane, extract the features of the three-dimensional space wall on the minimum bounding sphere (MBS), and at this time the three-dimensional space wall is a known environment, and obtain the first spatial structure circular descriptor (SSCD1);

[0080] After the wall-climbing robot is adsorbed on the three-dimensional space wall, the above steps are repeated to map the wall-climbing robot and the three-dimensional space wall together on the two-dimensional positioning and navigation plane, and the second spatial structure circular descriptor (SSCD2) of the two-dimensional plane positioning and navigation map at time k-1 is obtained;

[0081] The point cloud feature matching of the first spatial structure circular descriptor (SSCD1) and the second spatial structure circular descriptor (SSCD2) is performed. At this time, only the three-dimensional space points in the area where the wall-climbing robot is located are changed, so as to determine the position of the wall-climbing robot on the three-dimensional space wall.

[0082] After the wall-climbing robot moves on the three-dimensional wall surface, the above steps are repeated, and the spatial structure circular descriptor (SSCD) algorithm module obtains the third spatial structure circular descriptor (SSCD3) of the two-dimensional positioning plane navigation map at time k.

[0083] The spatial structure circular descriptor (SSCD) algorithm module completes the real-time positioning of the wall-climbing robot by performing feature matching on the second spatial structure circular descriptor (SSCD2) at time k-1 and the third spatial structure circular descriptor (SSCD3) at time k; and, through the point cloud changes of the wall-climbing robot, the spatial structure circular descriptor (SSCD) algorithm module performs real-time correction on the posture information of the wall-climbing robot measured by the onboard inertial measurement unit (IMU).

[0084] The onboard inertial measurement unit (IMU) measures three-axis acceleration and angular velocity. According to the gravity vector constraint and the estimated time division of the angular velocity, the roll angle θ, the pitch angle Φ and the yaw angle Ψ can be obtained respectively. An extended Kalman filter (EKF) is designed for filtering calculation to reduce the characteristics of the onboard inertial measurement unit (IMU) integral drift.

[0085] First, feature matching is performed on the spatial structured circular descriptor (SSCD) graph of the two-dimensional positioning and navigation plane to complete the posture mapping of the wall-climbing robot. Then, the extended Kalman filter (EKF) is used to compensate and calibrate the onboard inertial measurement unit (IMU) to determine the posture of the wall-climbing robot at time k.

[0086] According to the principle of spatial structure circular descriptor (SSCD) mapping, the position of the projection point Q reflects the orientation and elevation information of the spatial point S, and thus roughly reflects the approximate area of ​​the point M on the three-dimensional wall in the minimum bounding sphere (MBS). By dividing the spatial structure circular descriptor (SSCD) graph into regions, the mapping relationship between the points on the spatial structure circular descriptor (SSCD) graph and the points on the three-dimensional wall is clarified.

[0087] Figure 4 FIG. 1 is a schematic diagram of the region division of the spatial structure circular descriptor (SSCD) mapping in a specific embodiment of the present disclosure. Figure 4 As shown, a combination of ring-shaped and fan-shaped partitions is adopted.

[0088] Combined circular and sector-shaped partitioning improves positioning accuracy: Through the mapping and area division of the two-dimensional positioning navigation plane, the combined circular and sector-shaped partitioning is adopted, so that the wall-climbing robot can understand and process spatial information more intuitively, thereby improving positioning accuracy.

[0089] Regarding the concept of latitude: different annular areas correspond to a certain latitude range on the minimum bounding sphere (MBS), and at the same time have an invariant shape for the rotation of the entire three-dimensional model, dividing the minimum bounding sphere (MBS) into different numbers of rings.

[0090] About the concept of longitude: different sector-shaped areas correspond to different longitude ranges on the minimum enclosing sphere (MBS), dividing the minimum enclosing sphere (MBS) into different quadrants.

[0091] First, through the minimum bounding sphere (MBS) point S (x S ,y S ,z S ) longitude and latitude, that is, S(x S ,y S ,z S ) can determine the number of rings and quadrants in which the point S(x S ,y S ,z S ) on the spherical surface of the minimum bounding sphere (MBS); then, combined with each point S(x S ,y S ,z S )’s attribute value α SAs well as the elevation information it carries, it is possible to accurately connect the projection point Q on the two-dimensional positioning navigation plane with the point M (x M ,y M ,z M ) one by one, thereby completing the data compression and feature extraction from 3D to 2D, as well as the inverse mapping from 2D to 3D.

[0092] The three-dimensional wall feature recognition is based on the point S (x S ,y S ,z S )'s latitude and longitude, attribute value α S And the elevation information it carries is used to determine abnormal features of the wall.

[0093] The point S(x S ,y S ,z S ), when a significant sudden change in height (elevation) occurs on the spatial structure circular descriptor (SSCD) map with the same latitude or longitude as the wall, it can be directly identified as an abnormal wall.

[0094] By performing feature matching on a database or set of abnormal wall surfaces, the abnormal wall surface types can be identified, such as inner corners, outer corners, and discontinuous walls.

[0095] Therefore, when planning the path of the wall-climbing robot on the two-dimensional positioning and navigation plane, the types of abnormal wall surfaces such as inner corners, outer corners and discontinuous walls can be identified by recognizing and responding to abnormal wall features, so that the wall-climbing robot can actively avoid dangerous areas and establish reasonable control strategies to enhance safety.

[0096] Through the above technical scheme, the positioning method of the wall-climbing robot disclosed in the present invention provides an efficient, low-cost, and easy-to-operate solution for three-dimensional space autonomous positioning and navigation technology, which has significant technological advancement and market application potential.

[0097] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0098] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present disclosure. Various modifications to the above embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to the above embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0099] The above is only a specific embodiment of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present disclosure, which should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.

[0100] The above description is merely illustrative and is in no way intended to limit the present disclosure, its application or use. The broad teachings of the present disclosure can be implemented in many forms. Therefore, although the present disclosure includes specific examples, the scope of the present disclosure should not be limited thereto, as other modifications will become apparent after studying the drawings, the specification and the appended claims.

Claims

1. A wall-climbing robot positioning system based on spatial structure circular descriptor SSCD, characterized in that: include: A depth image sensor RGB-D camera configured to output a depth image; and, based on the output depth image, collecting three-dimensional point cloud data of the three-dimensional space wall points; And, conformally mapping the outer contour of the wall-climbing robot onto the minimum bounding sphere MBS; The spatial structure circular descriptor SSCD algorithm module is configured to map the three-dimensional point cloud data of the three-dimensional space wall points collected by the depth image sensor RGB-D camera to a two-dimensional positioning navigation plane, and form an image of the spatial structure circular descriptor SSCD on the two-dimensional positioning navigation plane; and, match the point cloud features of the two spatial structure circular descriptor SSCD images from time k-1 to time k, and complete the positioning of the wall-climbing robot on the three-dimensional space wall at the current moment; and, The onboard inertial measurement unit (IMU) measurement module is configured to collect the angular velocity and acceleration of the wall-climbing robot; and obtain the posture information of the wall-climbing robot at the current moment by pre-integrating the collected angular velocity and acceleration.

2. The wall-climbing robot positioning system according to claim 1, characterized in that: The three-dimensional point cloud data includes: elevation information of the wall-climbing robot and the three-dimensional space wall, and convex and concave information of the three-dimensional space wall.

3. The wall-climbing robot positioning system according to claim 1, characterized in that: The conformal mapping of the outer contour of the wall-climbing robot onto the minimum bounding sphere MBS comprises: The distance difference between the outer contour of the wall-climbing robot and the three-dimensional space wall and the minimum bounding sphere MBS is calculated, and the distance difference is set as the elevation information of the wall-climbing robot and the steel structure wall on the minimum bounding sphere MBS.

4. The wall-climbing robot positioning system according to claim 1, characterized in that: It is characterized in that The two-dimensional positioning and navigation plane is divided into multiple regions, and the images of these spatial structure circular descriptors SSCD on the two-dimensional positioning and navigation plane are decomposed into ring-shaped and fan-shaped combined partitions starting from the center of the circle; wherein, Each annular partition and sector partition corresponds to the latitude and longitude of the minimum enclosing sphere MBS respectively.

5. A wall-climbing robot positioning method based on spatial structure circular descriptor SSCD, characterized in that: The following steps are involved: The depth image sensor RGB-D camera outputs a depth image; as well as, Based on the output depth image, the depth image sensor RGB-D camera collects three-dimensional point cloud data of the three-dimensional space wall points; and, The depth image sensor RGB-D camera conformally maps the outline of the wall-climbing robot onto the minimum bounding sphere MBS; The spatial structure circular descriptor SSCD algorithm module maps the three-dimensional point cloud data of the three-dimensional space wall points collected by the depth image sensor RGB-D camera to a two-dimensional positioning navigation plane, and forms an image of the spatial structure circular descriptor SSCD on the two-dimensional positioning navigation plane; as well as, The spatial structure circular descriptor SSCD algorithm module matches the point cloud features of two spatial structure circular descriptor SSCD images from time k-1 to time k to complete the positioning of the wall-climbing robot on the three-dimensional space wall at the current moment; and, The onboard inertial measurement unit (IMU) measurement module collects the angular velocity and acceleration of the wall-climbing robot; The onboard inertial measurement unit (IMU) measurement module obtains the posture information of the wall-climbing robot at the current moment by pre-integrating the collected angular velocity and acceleration.

6. The wall-climbing robot positioning method according to claim 5, characterized in that: The three-dimensional point cloud data includes: elevation information of the wall-climbing robot and the three-dimensional space wall, and convex and concave information of the three-dimensional space wall.

7. The wall-climbing robot positioning method according to claim 5, characterized in that: The conformal mapping of the outer contour of the wall-climbing robot onto the minimum bounding sphere MBS comprises: The distance difference between the outer contour of the wall-climbing robot and the three-dimensional space wall and the minimum bounding sphere MBS is calculated, and the distance difference is set as the elevation information of the wall-climbing robot and the steel structure wall on the minimum bounding sphere MBS.

8. The wall-climbing robot positioning method according to claim 5, characterized in that: It is characterized in that The two-dimensional positioning and navigation plane is divided into multiple regions, and the images of these spatial structure circular descriptors SSCD on the two-dimensional positioning and navigation plane are decomposed into ring-shaped and fan-shaped combined partitions starting from the center of the circle; wherein, Each annular partition and sector partition corresponds to the latitude and longitude of the minimum enclosing sphere MBS respectively.

9. A wall-climbing robot positioning method based on spatial structure circular descriptor SSCD, characterized in that: The following steps are involved: The spatial structure circular descriptor SSCD algorithm module initializes the three-dimensional space wall, establishes the minimum bounding sphere MBS of the three-dimensional space wall point cloud at this time, maps the minimum bounding sphere MBS of the three-dimensional space to the two-dimensional positioning navigation plane, and obtains the first spatial structure circular descriptor SSCD1; After the wall-climbing robot is adsorbed on the three-dimensional space wall, the spatial structure circular descriptor SSCD algorithm module maps the wall-climbing robot and the three-dimensional space wall together on the two-dimensional positioning and navigation plane, and obtains the second spatial structure circular descriptor SSCD2 of the two-dimensional plane positioning and navigation map at time k-1; Point cloud feature matching is performed between the first spatial structure circular descriptor SSCD1 and the second spatial structure circular descriptor SSCD2 to determine the position of the wall-climbing robot on the wall surface of the three-dimensional space.

10. The wall-climbing robot positioning method according to claim 9, characterized in that: The following steps are also included: After the wall-climbing robot moves on the three-dimensional wall, the spatial structure circular descriptor SSCD algorithm module obtains the third spatial structure circular descriptor SSCD3 of the two-dimensional positioning plane navigation map at time k. By performing feature matching on the second spatial structure circular descriptor SSCD2 at time k-1 and the third spatial structure circular descriptor SSCD3 at time k, the spatial structure circular descriptor SSCD algorithm module completes the real-time positioning of the wall-climbing robot; and, through the point cloud changes of the wall-climbing robot, the spatial structure circular descriptor SSCD algorithm module performs real-time correction on the posture information of the wall-climbing robot measured by the onboard inertial measurement unit IMU.

Images