Charging piles, charging pile systems, charging pile return methods, devices, and lawnmower robots

Abstract

This invention provides a charging pile, a charging pile system, a method for returning to the charging pile, a device, and a lawnmower robot. In this embodiment, only feature markers need to be set on the charging pile. The lawnmower can obtain the pose information of the feature markers through its own image recognition, thereby determining the relative pose information between the charging pile and the lawnmower for path planning, enabling the lawnmower to return to the charging pile for charging. It is low-cost, and because only feature markers need to be set on the charging pile, the structure is simple and easy to install and disassemble. Furthermore, the method for returning to the charging pile in this application uses feature markers on the charging pile for positioning, ensuring positioning accuracy. Moreover, visual imaging is not affected by noise interference from other electromagnetic waves and signals in the environment, ensuring accuracy and reliability. It also eliminates the need for additional guide lines and overcomes the defects of GPS positioning errors, fully guaranteeing the accuracy of returning to the charging pile.

Description

[0001] This application claims priority to the invention application filed on March 9, 2022, with application number “202210234711.X” and patent title “Method, Device and Lawn Mowing Robot for Reclaiming Stakes”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of gardening equipment technology, and in particular to a charging pile, a charging pile system, a charging pile return method, a device, and a lawn mowing robot. Background Technology

[0003] Lawn-mowing robots can automatically trim lawns and other surfaces, reducing the time required for manual labor. They can also return to their charging docks without human intervention, further reducing the burden on users and saving time.

[0004] Existing lawnmower return technologies mostly rely on GPS positioning, guide lines, and virtual boundary patrol. For example, in indoor use, lawnmowers typically use infrared or ultrasonic sensors for return or virtual boundary guidance based on mapping. These infrared and ultrasonic methods have a higher chance of accidental triggering and are significantly affected by environmental noise.

[0005] In outdoor applications, lawnmower robots typically rely on GPS positioning and electromagnetic guide lines in front of the charging station to accurately return to the charging pile. This GPS-based method requires high signal strength and accuracy, and also necessitates the installation of auxiliary electromagnetic guide lines on the base of the charging station, incurring material and transportation costs.

[0006] In summary, existing lawnmower robots with automatic repositioning are costly, difficult to implement, and lack stability. Summary of the Invention

[0007] To address the aforementioned problems, embodiments of this application provide a charging pile, a charging pile system, a charging pile return method, an apparatus, and a lawnmower robot, which at least partially solve the problems mentioned above.

[0008] One or more embodiments of the present invention provide a charging pile, which has multiple feature identifiers pre-set on it. At least two of the feature identifiers have different distribution surfaces from the remaining feature identifiers. The feature identifiers are used for image recognition by a lawnmower robot so that the lawnmower robot can dock with the charging pile for charging.

[0009] According to another aspect of this application, a charging station is provided, which is provided with an identification component for indicating the location of the charging station so that a lawnmower robot can dock with the charging station for charging.

[0010] According to another aspect of this application, a charging pile system is provided, including a charging pile as described in any of the above aspects and an identification device, wherein a fourth identification component is provided on the identification device, the fourth identification component being used to assist in determining the location of the charging pile.

[0011] According to another aspect of this application, a method for guiding a lawnmower robot back to a charging station is provided. The lawnmower robot is equipped with an image acquisition device, and the charging station has feature markers distributed on it. The method includes: acquiring available image frames acquired by the image acquisition device of the lawnmower robot in the area in front of the charging station, wherein the number of feature markers identified in the image frames is greater than or equal to N, where N is a positive integer greater than or equal to 3; determining the relative pose of the lawnmower robot relative to the charging station based on the two-dimensional coordinates of the identified feature markers in the image frames, the three-dimensional coordinates of the feature markers in the world coordinate system, and the imaging parameters of the image acquisition device; and driving the lawnmower robot back to the charging station based on the relative pose.

[0012] According to another aspect of this application, a return-to-charging-pile device is provided. The device guides a lawnmower robot back to a charging pile. The lawnmower robot is equipped with an image acquisition device, and the charging pile has distributed feature markers. The device includes: an acquisition module for acquiring available image frames acquired by the image acquisition device of the lawnmower robot in the area in front of the charging pile, wherein the number of feature markers identified in the image frames is greater than or equal to N, where N is a positive integer greater than or equal to 3; a determination module for determining the relative pose of the lawnmower robot relative to the charging pile based on the two-dimensional coordinates of the identified feature markers in the image frames, the three-dimensional coordinates of the feature markers in the world coordinate system, and the imaging parameters of the image acquisition device; and a driving module for driving the lawnmower robot back to the charging pile based on the relative pose.

[0013] According to another aspect of this application, a lawn mowing robot is provided, which is equipped with an image acquisition device and includes a controller for performing the above-described method.

[0014] According to another aspect of this application, a computer storage medium is provided, wherein a computer program is stored within the computer storage medium, and the computer program, when executed by a processor, implements the above-described method.

[0015] This method utilizes available image frames acquired by an image acquisition device in the area in front of the charging pile to identify the two-dimensional coordinates of the feature marker. Then, by combining the three-dimensional coordinates of the feature marker in the world coordinate system and the imaging parameters of the image acquisition device, the relative pose between the lawnmower robot and the charging pile can be accurately determined. Based on the relative pose, the lawnmower robot can be guided to accurately return to the charging pile. This positioning method using feature markers ensures positioning accuracy, and the visual imaging is not affected by noise from other electromagnetic waves and signals in the environment, ensuring accuracy and reliability. It also eliminates the need for additional guide lines and overcomes the defects of GPS positioning errors, fully guaranteeing the accuracy of returning to the charging pile.

[0016] Furthermore, in this embodiment, only an identification component needs to be installed on the charging pile. The lawnmower can obtain the pose information of the feature markers distributed in the identification component through its own image recognition, thereby determining the relative pose information between the charging pile and the lawnmower for path planning, enabling the lawnmower to return to the charging pile for charging. This approach is low-cost, and because only feature markers need to be installed on the charging pile, the structure is simple and easy to install and disassemble. Attached Figure Description

[0017] The accompanying drawings are intended only to illustrate and explain this application and do not limit the scope of this application.

[0018] Figure 1A This is a schematic flowchart illustrating the steps of the backfilling method in Embodiment 1 of this application;

[0019] Figure 1B This is a schematic diagram of a feature identifier distribution according to Embodiment 1 of this application;

[0020] Figure 1C This is a schematic diagram of another feature identifier distribution according to Embodiment 1 of this application;

[0021] Figure 1D This is a schematic diagram of the area in front of the pile in Embodiment 1 of this application;

[0022] Figure 1E This is a flowchart illustrating a sub-step of step S102 in Embodiment 1 of this application;

[0023] Figure 2 A structural block diagram of the pile-back device provided in Embodiment 2 of this application;

[0024] Figure 3A This is a schematic diagram showing the distribution of marking components on a charging pile according to an exemplary embodiment of this application;

[0025] Figure 3B This is a schematic diagram showing the distribution of marking components on a charging pile according to an exemplary embodiment of this application;

[0026] Figure 4This is a schematic diagram showing the distribution of marking components on a charging pile according to an exemplary embodiment of this application;

[0027] Figure 5 This is an exploded view of the identification component of a charging pile according to an exemplary embodiment of this application;

[0028] Figure 6 This is a schematic diagram of the structure of the reflective element of the identification component of the charging pile according to an exemplary embodiment of this application;

[0029] Figure 7 This is a schematic diagram of a charging pile system according to an exemplary embodiment of this application;

[0030] Explanation of reference numerals in the attached figures:

[0031] 10. Charging pile; 11. Charging pile body; 111. Mounting slot; 12. Top cover; 13. Base plate; 2. Identification component; 20. Feature identifier; 211. First identification component; 212. Second identification component; 213. Third identification component; 214. Fourth identification component; 22. Reflective element; 221. Light-transmitting part; 222. Reflective part; 23. Mounting plate; 3. Charging structure; 41. First distribution surface; 42. Second distribution surface; 43. Third distribution surface; 44. Fourth distribution surface; 50. Identification device. Detailed Implementation

[0032] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0033] To facilitate explanation and understanding, before describing the re-staking method, the structure and working environment of the lawnmower robot will be briefly described as follows:

[0034] In this embodiment, the lawn mowing robot can be an automatic lawn mower, which includes several main parts such as a main body, an image acquisition device (also known as a vision acquisition device), a navigation position estimation device, a visual position estimation device, and an autonomous movement device.

[0035] The main body includes a shell and a lawnmower assembly, which is used to trim the grass in the lawn to an appropriate height. The autonomous movement device includes drive wheels, a drive motor, and a motor controller. The motor controller is connected to the drive motor to control the rotation of the drive motor, which in turn drives the drive wheels, thereby enabling the lawnmower robot to move forward, backward, and turn.

[0036] Image acquisition devices can include cameras, video cameras, and so on. Different image acquisition devices may have different imaging parameters, such as aperture, focal length, and optical axis.

[0037] The navigation position estimation device is used to locate the lawnmower robot, and then drive the lawnmower robot to move based on the positioning results.

[0038] The visual position estimation device can be used to process image frames acquired by the image acquisition device, and then estimate the relative pose of the lawnmower robot relative to the charging station based on the image frames. This relative pose allows for navigation and movement of the lawnmower robot. It should be noted that the visual position estimation device can be a standalone chip with computing capabilities, or it can be a chip shared with other devices; there are no restrictions on this.

[0039] like Figure 1B and Figure 1C As shown, charging pile ( Figure 1B The number 10 has Q feature identifiers. Figure 1B (20) where Q is greater than or equal to 2 and is a positive integer. The feature identifier can be any suitable pattern, shape, or structure capable of being identified, without limitation. For example, the feature identifier can be a triangular protrusion on the charging station, or a dot, rectangle, QR code, line segment, arc, etc., and will not be listed here.

[0040] Preferably, to improve the accuracy of relative pose localization, for each coordinate axis of the world coordinate system, the maximum distance between the projections of the Q feature identifiers on each coordinate axis is greater than or equal to a distance threshold (the distance threshold can be determined according to the required localization and recognition accuracy). This makes the feature identifiers on the same distribution plane more dispersed on the distribution plane, and the distance between the feature identifiers is larger, thereby helping to improve the detection efficiency and accuracy of relative pose localization. The reason is: Figure 1B In the coordinate system shown, the Z-axis represents the forward direction of the lawnmower robot. In a typical, relatively flat ground environment, the lawnmower robot's ability to mount charging stations relies more on the change in the yaw angle (i.e., the angle of rotation around the Y-axis). That is, the greater the lateral distance between the feature markers on the charging station, the greater the absolute distance of the positional change of the feature markers on the imaging plane when the relative pose of the lawnmower robot changes relative to the charging station, thus resulting in stronger anti-interference capabilities. The pitch angle (i.e., the angle of rotation around the X-axis) works similarly.

[0041] Preferably, such as Figure 1CAs shown, at least two of the Q feature identifiers are distributed on different planes than the remaining feature identifiers. That is, the multiple feature identifiers of the charging pile are located on at least two different distribution planes, thus achieving differentiation of feature identifiers in the Z-axis direction. Comparative experiments have determined that charging pile A, where all feature identifiers are located on the same distribution plane, has weaker noise resistance than charging pile B, where at least two feature identifiers are located on different distribution planes than the remaining feature identifiers. For the same number of tests, charging pile A is worse than charging pile B in both its inability to calculate relative pose and the accuracy of its calculated relative pose.

[0042] Optimizing the distribution of feature markers on the charging pile can improve the anti-interference and anti-sensitivity capabilities for relative pose recognition of lawnmower robots. However, it should be noted that the distribution method exemplified in this application is a preferred method. In other embodiments, the feature markers can adopt other distribution methods, as long as the relative pose can be calculated.

[0043] Example 1

[0044] The implementation process of the pile-back method is explained below:

[0045] like Figure 1A The diagram illustrates a flowchart of a backfilling method. In this embodiment, the method includes the following steps:

[0046] Step S102: Obtain available image frames acquired by the image acquisition device of the lawnmower in the area in front of the charging pile.

[0047] In this embodiment, the image acquisition device captures image frames of the surrounding environment during the movement of the lawnmower robot. Multiple image frames at different times form an image frame sequence (also known as video). Since the lawnmower robot's pose is different at different times, the content contained in the image frames is also different. Some image frames may contain feature identifiers, while others may not.

[0048] A usable image frame can be an image frame acquired in the area in front of the stake that contains a sufficient number of feature identifiers. These image frames can be used to calculate the relative pose of the lawnmower robot, and therefore can be considered usable image frames. "Containing a sufficient number of feature identifiers" can be defined as: the number of feature identifiers identified in the image frame is greater than or equal to N, where N is a positive integer greater than or equal to 3.

[0049] like Figure 1DAs shown, the area in front of the charging pile includes a fan-shaped area corresponding to the charging pile. The fan-shaped area is located in front of the charging pile, and the radius of the fan-shaped area is greater than 0 and less than M, where M is a positive integer. The value of M can be determined according to the imaging parameters of the image acquisition device, the size of the feature marker, etc., and is not limited thereto.

[0050] In one example, the area in front of the pile can be a fan-shaped area with a radius of 2 meters and an included angle of 120°. Of course, in other examples, the shape and size of the area in front of the pile can be chosen appropriately, and there are no restrictions on this.

[0051] Optionally, in this embodiment, to ensure that the lawnmower robot can reliably and accurately move to the charging station from any location, such as... Figure 1E As shown, step S102 can be implemented through the following sub-steps:

[0052] Sub-step S1021: Obtain the current navigation pose of the lawnmower robot.

[0053] For example, in one scenario, when the lawnmower receives a return-to-stake command in normal operating mode, the lawnmower's navigation estimation device can determine the lawnmower's navigation pose.

[0054] Sub-step S1022: Determine whether the lawn mowing robot is in the area in front of the charging pile based on the navigation pose and the preset pose of the charging pile.

[0055] The area in front of the charging station can be determined based on its pose, and then the location information in the navigation pose can be used to determine whether the lawnmower robot is in the area in front of the charging station.

[0056] If the location is within the area in front of the charging station, it means that the distance between the charging station and the lawnmower robot is sufficient to collect clear and appropriately sized feature markers, and sub-step S1023 can be executed; otherwise, the lawnmower robot needs to be driven to move to the area in front of the charging station first, so sub-step S1025 can be executed.

[0057] Sub-step S1023: If the lawn mowing robot is in the area in front of the stake, determine whether there is a usable image frame in the image frame sequence acquired by the image acquisition device.

[0058] In one feasible approach, a trained robot learning model is used to perform image recognition on each image frame in the image frame sequence to identify feature identifiers, and the number of identified feature identifiers determines whether the image frame is a usable image frame.

[0059] Taking N=3 as an example, if the number of identified feature identifiers is greater than or equal to 3, then the image frame is a usable image frame. Execute sub-step S1024; otherwise, sub-step S1026 can be executed.

[0060] Sub-step S1024: If there is an available image frame, then the existing available image frame is used as the obtained available image frame.

[0061] After obtaining a usable image frame, step S104 can be performed.

[0062] Sub-step S1025: If the mowing robot is not in the area in front of the stake, then according to the navigation pose, drive the mowing robot to move into the area in front of the stake, and return to the step of obtaining the current navigation pose of the mowing robot to continue execution.

[0063] When the navigation pose obtained by the navigation position estimation device is not within the area in front of the stake, the navigation position estimation device can guide the mowing robot to move according to the navigation pose, and then return to sub-step S1021 to execute, repeating this process until the mowing robot moves into the area in front of the stake. In this way, the mowing robot only enters the stake return mode when it moves into the area in front of the stake and can collect a sufficient number of feature markers.

[0064] Sub-step S1026: If no usable image frame exists, adjust at least one of the position and posture of the lawnmower robot, acquire a new image frame sequence, and return to the step of determining whether the usable image frame exists in the image frame sequence acquired by the image acquisition device to continue execution.

[0065] For example, when adjusting at least one of the position and posture of the mowing robot, the mowing robot is adjusted to a new navigation pose, and the mowing robot is located outside the area in front of the stake when it is in the new navigation pose; according to the new navigation pose, the mowing robot is driven into the area in front of the stake, and during the movement of the mowing robot, the image acquisition device acquires multiple image frames to form the image frame sequence.

[0066] For example, if a lawnmower robot enters the area in front of a stake but there are no usable image frames, the robot can be driven out of the area and then re-enter the area from a different angle. Then, image frames can be acquired, and the system can return to determine whether the usable image frames exist in the sequence of image frames acquired by the image acquisition device.

[0067] For example, if the lawnmower robot does not detect enough feature markers on the charging pile within 5 seconds in the area in front of the charging pile, it will randomly choose a direction to drive out of the area in front of the charging pile, drive in again at a different angle, and detect the feature markers on the charging pile again. This process will be repeated until enough feature markers are detected, and then it will enter the return-to-charging-pile mode.

[0068] If a usable image frame is obtained, step S104 can be performed.

[0069] Step S104: Determine the relative pose of the lawnmower robot with respect to the charging pile based on the two-dimensional coordinates of the identified feature identifier in the image frame, the three-dimensional coordinates of the feature identifier in the world coordinate system, and the imaging parameters of the image acquisition device.

[0070] In the docking mode, the visual position estimation device continuously updates the relative pose of the lawnmower robot and the charging station by detecting and matching image frames, thereby guiding the lawnmower robot onto the charging station and completing the docking.

[0071] The image frame detection method can be as follows: use a trained neural network model to identify the image frame, detect the feature identifiers in the image frame, and determine the two-dimensional coordinates of the feature identifiers in the image frame.

[0072] Different neural network models can be trained for different feature identifiers to detect feature identifiers in image frames.

[0073] Based on the detected 2D coordinates of the feature identifiers, the known 3D coordinates of the feature identifiers in the world coordinate system, and the imaging parameters of the image acquisition device, the relative pose at the current moment is determined. The calculation process for the relative pose is as follows: Based on the 2D coordinates and the imaging parameters of the image acquisition device, the 2D coordinates can be converted into 3D coordinates in the coordinate system of the image acquisition device. Then, based on the 3D coordinates of the feature identifiers in the coordinate system of the image acquisition device and their 3D coordinates in the world coordinate system, the mapping relationship from the coordinate system of the image acquisition device to the world coordinate system can be determined. This mapping relationship is the relative pose.

[0074] Step S106: Drive the lawnmower robot back to the charging station according to the relative pose.

[0075] The relative pose can be adjusted to make the lawnmower face the charging station, thus enabling the lawnmower to return to the charging station.

[0076] As can be seen from the foregoing process, the ability of the lawnmower robot to quickly and accurately return to the charging station is greatly affected by the accuracy of the two-dimensional coordinate recognition of the feature markers in the image frame. Since the image frame is acquired during the movement of the lawnmower robot, it is inevitably affected by motion blur. In order to reduce the adverse effects of motion blur on feature marker detection and two-dimensional coordinate accuracy, in this embodiment, when the image acquisition device acquires the image frame, with the ambient light intensity fixed, the exposure time of the image frame is positively correlated with the relative distance, where the relative distance is the distance between the lawnmower robot and the charging station.

[0077] For example, under normal circumstances, the exposure time of an image acquisition device is determined based on the intensity of ambient light. The brighter the ambient light, the shorter the exposure time, and the shorter the exposure time, the less motion blur is produced.

[0078] In the actual operation of the lawnmower robot, if the feature markers in the image frame taken when the distance between the lawnmower robot and the charging station is greater than X meters (X is a positive number and can be determined as needed) are missed due to blurring or other reasons, the problem can be solved by twisting the lawnmower robot or moving it closer to the charging station. Since the lawnmower robot is far from the charging station at this time, the omission of feature markers or the low accuracy of the matched two-dimensional coordinates on the image frame has little impact on the navigation accuracy and can be compensated for by adjusting subsequent image frames.

[0079] When the distance between the lawnmower and the charging station is less than or equal to Y (where Y is a positive number and less than X), the close proximity can easily lead to missed detection of feature markers or insufficient accuracy of 2D coordinates due to motion blur, resulting in excessive movement deviation of the lawnmower and inability to accurately return to the charging station. To avoid this problem and ensure that the brightness of the captured image frames meets the detection requirements, under the same light intensity, the exposure time of the image acquisition device decreases as the distance decreases until the distance reaches a threshold. This effectively reduces motion blur.

[0080] For example, in one scenario, when light intensity and distance are constant, the exposure time remains constant. When the light intensity remains constant but the distance decreases by z, the exposure time T will decrease by a*z ms. Here, 'a' can be set appropriately as needed. When the light intensity changes, the exposure time T will either increase or decrease accordingly. The effects of light intensity and distance on exposure time are independent of each other.

[0081] Furthermore, research has found that, due to the characteristics of the image acquisition device, when the ambient light intensity remains constant, the shorter the exposure time, the lower the scene reproduction of the image frame, and the more appropriate the image frame appears to be darker overall. However, experimental comparisons have shown that, compared to the adverse effects of motion blur on the accuracy of backtracking, the adverse effects of reduced image frame reproduction on backtracking accuracy are far less than those of motion blur. Therefore, appropriately reducing the exposure time at close range can help improve accuracy.

[0082] Optionally, to further improve accuracy, data augmentation techniques can be used to improve the accuracy of feature identification for image frames with low fidelity, thereby avoiding missed detections and improving robustness.

[0083] In this embodiment, the two-dimensional coordinates of the feature marker are identified by using available image frames acquired by the image acquisition device in the area in front of the charging pile. Then, by combining the three-dimensional coordinates of the feature marker in the world coordinate system and the imaging parameters of the image acquisition device, the relative pose between the lawnmower robot and the charging pile can be accurately determined. Based on the relative pose, the lawnmower robot is guided to return to the charging pile accurately. This positioning using feature markers ensures positioning accuracy, and the visual imaging is not affected by noise from other electromagnetic waves and signals in the environment, ensuring accuracy and reliability. It also eliminates the need for additional guide lines and overcomes the defects of GPS positioning errors, fully guaranteeing the accuracy of returning to the charging pile.

[0084] By combining the image frames acquired by the image acquisition device with the navigation position estimation device, which does not require high precision, the lawnmower robot can accurately return to its charging dock and recharge from any direction within the working area.

[0085] During the approach to the charging pile, the characteristic that visual information such as image frames becomes more accurate as the distance increases enables stable charging. In particular, the feature identifier distribution in this embodiment has better robustness and better resistance to detection sensitivity. The problem of motion blur, which is difficult for visual solutions to handle, is also solved by adjusting the exposure time according to the distance.

[0086] Example 2

[0087] Reference Figure 2 The diagram shows the structural block diagram of the pile-back device in this embodiment 2.

[0088] The device is used to guide the lawnmower robot back to the charging station. The lawnmower robot is equipped with an image acquisition device, and the charging station has feature markers distributed on it. The device includes:

[0089] The acquisition module 202 is used to acquire available image frames acquired by the image acquisition device of the lawnmower in the area in front of the charging pile, wherein the number of feature identifiers identified in the image frames is greater than or equal to N, and N is a positive integer greater than or equal to 3.

[0090] The determining module 204 is used to determine the relative pose of the lawnmower robot relative to the charging pile based on the two-dimensional coordinates of the identified feature identifier in the image frame, the three-dimensional coordinates of the feature identifier in the world coordinate system, and the imaging parameters of the image acquisition device.

[0091] The drive module 206 is used to drive the lawnmower robot back to the charging station according to the relative pose.

[0092] Optionally, the area in front of the charging pile includes a sector-shaped area corresponding to the charging pile, the sector-shaped area is located in front of the charging pile, and the radius of the sector-shaped area is greater than 0 and less than M, where M is a positive integer.

[0093] Optionally, the acquisition module 202 is used to acquire the current navigation pose of the lawnmower robot; determine whether the lawnmower robot is in the area in front of the charging pile based on the navigation pose and the preset pose of the charging pile; if the lawnmower robot is in the area in front of the charging pile, determine whether there is a usable image frame in the image frame sequence acquired by the image acquisition device; if there is a usable image frame, use the existing usable image frame as the acquired usable image frame.

[0094] Optionally, the acquisition module 202 is further configured to, if no available image frame exists, adjust at least one of the position and posture of the lawnmower robot, acquire a new image frame sequence, and return to the step of determining whether the available image frame exists in the image frame sequence acquired by the image acquisition device to continue execution.

[0095] Optionally, the acquisition module 202 is used to adjust the lawn mower to a new navigation pose when at least one of the position and posture of the lawn mower is adjusted, wherein the lawn mower is located outside the area in front of the stake when it is in the new navigation pose; and to drive the lawn mower into the area in front of the stake according to the new navigation pose, wherein the image acquisition device acquires multiple image frames to form the image frame sequence during the movement of the lawn mower.

[0096] Optionally, the acquisition module 202 is further configured to, if the mowing robot is not in the area in front of the stake, drive the mowing robot to move into the area in front of the stake according to the navigation pose, and return to the step of acquiring the current navigation pose of the mowing robot to continue execution.

[0097] Optionally, the charging pile is provided with Q feature identifiers, where Q is greater than or equal to 2 and is a positive integer. For each coordinate axis of the world coordinate system, the maximum distance of the projection of the Q feature identifiers on each coordinate axis is greater than or equal to a distance threshold.

[0098] Optionally, the distribution plane of at least two of the Q feature identifiers is different from the distribution plane of the remaining feature identifiers.

[0099] Optionally, when the image acquisition device acquires the image frame, with the ambient light intensity fixed, the exposure time of the image frame is positively correlated with the relative distance, where the relative distance is the distance between the lawnmower robot and the charging pile.

[0100] The device can achieve the corresponding effect of the method, so it will not be described in detail here.

[0101] Example 3

[0102] According to an embodiment of this application, a lawnmower robot is provided. The lawnmower robot is equipped with an image acquisition device, which includes a controller. The controller is used to execute the aforementioned method. The controller can realize the operation corresponding to the aforementioned method and achieve the corresponding effect, which will not be described in detail here.

[0103] Optionally, the lawnmower also includes a charging unit, which is located below the image acquisition device, closer to the ground than the image acquisition device. The charging unit is used to interface with the charging structure of a charging station to charge the lawnmower.

[0104] According to another aspect of this application, a computer storage medium is provided, wherein a computer program is stored within the computer storage medium, and the computer program, when executed by a processor, implements the above-described method. This program can perform the operations corresponding to the aforementioned method and achieve the corresponding effects, which will not be elaborated further.

[0105] According to another aspect of this application, a charging pile 10 is provided, wherein Q feature identifiers 20 are provided on the charging pile 10, and the distribution plane of at least two of the Q feature identifiers 20 is different from the distribution plane of the remaining feature identifiers 20, and Q is greater than or equal to 2 and is a positive integer.

[0106] According to the charging pile provided in the embodiment, the charging pile is provided with Q feature markers 20. The distribution plane of at least two of the Q feature markers 20 is different from the distribution plane of the remaining feature markers 20, and Q is greater than or equal to 2 and is a positive integer. In this embodiment, only feature markers 20 need to be set on the charging pile. The lawnmower can obtain the pose information of the feature markers 20 through its own image recognition, thereby determining the relative pose information between the charging pile and the lawnmower for path planning, enabling the lawnmower to return to the charging pile for charging. The cost is low, and since only feature markers 20 need to be set on the charging pile, the structure is simple and easy to install and disassemble; and by making the distribution plane of at least two of the multiple feature markers 20 different from the distribution plane of the remaining feature markers 20, the accuracy of the lawnmower in calculating the relative pose information of the charging pile through image recognition can be improved.

[0107] like Figure 1B and Figure 1C As shown, charging pile ( Figure 1B The number 10) is equipped with Q feature identifiers 20 ( Figure 1BIn the case of feature identifier 20), Q is greater than or equal to 2 and is a positive integer. Feature identifier 20 can be any suitable pattern, shape, or structure that can be used for identification, and there are no restrictions on it. For example, feature identifier 20 can be a triangular protrusion set on the charging pile, or a dot, a rectangular pattern, a QR code, or a line segment, an arc, etc., which will not be listed here.

[0108] Preferably, to improve the accuracy of relative pose localization, for each coordinate axis of the world coordinate system, the maximum distance between the projections of the Q feature identifiers 20 on each coordinate axis is greater than or equal to a distance threshold (the distance threshold can be determined according to the required localization and recognition accuracy). This makes the feature identifiers 20 on the same distribution plane more dispersed on the distribution plane, and the distance between the feature identifiers 20 is larger, thereby helping to improve the detection efficiency and accuracy of relative pose localization. The reason is: Figure 1B In the coordinate system shown, the Z-axis represents the forward direction of the lawnmower robot. In a relatively flat ground environment, the lawnmower robot's ability to mount charging stations relies more on the change of the yaw angle (i.e., the angle of rotation around the Y-axis). That is, the greater the lateral distance between the feature markers 20 on the charging station, the greater the absolute distance of the position change of the feature markers 20 on the imaging plane when the relative pose of the lawnmower robot changes relative to the charging station, thus making the anti-interference ability stronger. The pitch angle (i.e., the angle of rotation around the X-axis) is similar.

[0109] Preferably, such as Figure 1C As shown, at least two of the Q feature identifiers 20 are distributed on different planes than the remaining feature identifiers 20. That is, the multiple feature identifiers 20 of the charging pile are located on at least two different distribution planes, thus achieving differentiation of the feature identifiers 20 in the Z-axis direction. Comparative experiments have determined that charging pile A, where all feature identifiers 20 are located on the same distribution plane, has weaker noise resistance than charging pile B, where at least two feature identifiers 20 are located on different distribution planes than the remaining feature identifiers 20. For the same number of tests, charging pile A is worse than charging pile B in both its inability to calculate relative pose and the accuracy of its calculated relative pose.

[0110] Optimizing the distribution of feature markers 20 on the charging pile can improve the anti-interference and anti-sensitivity capabilities for relative pose recognition of the lawnmower robot. However, it should be noted that the distribution method exemplified in this application embodiment is a preferred method. In other embodiments, feature markers 20 can adopt other distribution methods, as long as the relative pose can be calculated.

[0111] According to another aspect of this application, a charging pile 10 is provided. For ease of explanation and understanding, before describing the charging pile of this embodiment, the overall structure of the charging pile 10 is briefly described as follows:

[0112] The charging pile 10 may include a charging pile body 1 and a charging component 3. The charging pile body 1 includes an upper cover 12, a charging pile body 11 and a bottom plate 13. The charging component 3 is installed on one side of the charging pile body 11. The upper cover 12 is connected to the upper end of the charging pile body 11 and the bottom plate 13 is connected to the lower end of the charging pile body 11.

[0113] Reference Figures 1A to 7 As shown, in this embodiment, the charging pile 10 has multiple feature identifiers 20 pre-set on it. At least two of the feature identifiers 20 have different distribution surfaces than the remaining feature identifiers 20. The feature identifiers 20 are used for image recognition by the lawnmower robot so that the lawnmower robot can dock with the charging pile 10 for charging.

[0114] For example, feature identifier 20 is an identifier point. An identifier point refers to a point with abstract meaning. It can be an explicit identifier point, such as the endpoint of a line segment, the corner of a shape, or the corner of a QR code. It can also be a semantically implicit identifier point, such as the midpoint of the intersection of different surfaces, or the semantic information of the identifier itself, such as the wrist of a human-shaped identifier or the center point of the densest grass in a grass identifier. An identifier point can be a point on a surface, or a raised or recessed point on that surface. In this embodiment, feature identifier 20 refers to an explicit identifier point on the charging pile 10. It can be an identifier point of the charging pile 10 itself, such as the endpoint of the edge of the surface of the charging pile 10, or an identifier point designed according to the image recognition algorithm requirements of the lawnmower robot. For example, different shaped identifier components 2 are set on the charging pile 10, and feature identifier 20 is the corner of the shape of the identifier component 2. This embodiment does not limit this.

[0115] The image acquisition device of the lawnmower acquires usable image frames in the area in front of the charging pile 10. According to a preset recognition algorithm, it identifies preset feature marks 20 on the surface of the charging pile 10 from the image frames. The preset feature marks 20 are marker points on the charging pile 10 that are pre-set according to the preset recognition algorithm. Based on the two-dimensional coordinates of the identified feature marks 20 in the image frame, the three-dimensional coordinates of the feature marks 20 in the world coordinate system, and the imaging parameters of the image acquisition device, the relative pose of the lawnmower relative to the charging pile 10 is determined. Based on the relative pose, the lawnmower is driven back to the charging pile 10.

[0116] Reference Figure 1CThe charging pile 10 is provided with multiple feature markers 20. The distribution plane of at least two of the feature markers 20 is different from that of the remaining feature markers 20. That is, the multiple feature markers 20 provided by the charging pile 10 are located on at least two different distribution planes. This achieves the differentiation of feature markers 20 in the Z-axis direction. Comparative experiments have determined that the charging pile 10A, in which all feature markers 20 are located on the same distribution plane, has weaker noise resistance than the charging pile 10B, in which at least two feature markers 20 are located on different distribution planes from the remaining feature markers 20. For the same number of tests, the charging pile 10A is worse than the charging pile 10B in both its inability to calculate relative pose and the accuracy of its calculated relative pose.

[0117] In this embodiment, only feature markers 20 need to be set on the charging pile 10. The lawnmower robot can obtain the pose information of the feature markers 20 through its own image recognition, thereby determining the relative pose information between the charging pile 10 and the lawnmower robot for path planning, enabling the lawnmower robot to return to the charging pile for charging. This method is low-cost, and because only feature markers 20 need to be set on the surface of the charging pile 10, the structure is simple and easy to install and disassemble. Furthermore, by ensuring that the distribution surfaces of at least two of the multiple feature markers 20 are different from the distribution surfaces of the remaining feature markers 20, the accuracy of the lawnmower robot in calculating the relative pose of the charging pile 10 through image recognition can be improved.

[0118] Referring to Figure 3, in one specific implementation, the surface of the charging pile 10 is provided with an identification component 2. The identification component 2 includes a first identification component 211 and a second identification component 212. The first identification component 211 and the second identification component 212 are distributed on different surfaces on the charging pile 10. At least two feature marks 20 are distributed on the identification component 2.

[0119] For example, the number of the first marking component 211 and the second marking component 212 can each be greater than or equal to 1, and are positive integers. The marking components 2 can have different shapes, such as rectangles, triangles, circles, or L-shapes. The first marking component 211 and the second marking component 212 can use marking components 2 of the same shape, or they can use marking components 2 of different shapes. For example, the first marking component 211 can use four L-shaped marking components 2, and the second marking component 212 can use one strip-shaped marking component 2. This embodiment does not impose any limitations on this. Each marking component 2 can be a single shape, or it can be a combination of multiple different shapes. The marking component 2 can be planar, such as a sticker, or three-dimensional, such as a plastic reflective component.

[0120] The number of feature markers 20 on each marker component 2 can be greater than or equal to 2 and be a positive integer. The distribution of feature markers 20 on different shaped marker components 2 is different and can be preset according to the robot image recognition algorithm of the lawnmower. For example, the feature markers 20 on the rectangular marker component 2 can be set at least two of the four corner points, and the feature markers 20 on the L-shaped marker component 2 can be set at least two of the four endpoints, etc.

[0121] In this specific implementation, the identification component 2 allows for more selectable distribution positions of the feature identifiers 20 on the charging pile 10, and the number of feature identifiers 20 that can be set is increased, thereby improving the accuracy of image recognition calculation by the lawnmower robot. In addition, by setting the identification component 2, the position and number of feature distributions can be set according to the image recognition algorithm or hardware specifications used by the lawnmower robot, which can adapt to more robot products.

[0122] In one specific implementation, the first marking component 211 is located on the first distribution surface 41 of the charging pile 10, and the second marking component 212 is located on the second distribution surface 42 of the charging pile 10. An angle is formed between the normal of the first distribution surface 41 and the normal of the second distribution surface 42, and the angle is not equal to 0. Alternatively, the normal of the first distribution surface 41 is parallel to the normal of the second distribution surface 42, and the first distribution surface 41 and the second distribution surface 42 are spaced apart in the normal direction. For example, referring to… Figure 3A and Figure 3B The first distribution surface 41 can be the first side of the charging structure 3 of the charging pile 10, and the second distribution surface 42 can be the side of the charging pile 10 that is connected to the first side, or the side, upper surface, or lower surface of the charging structure 3; the normals of the two (the dashed lines shown in the figure) form an angle A, and the degree of the angle A is not equal to 0, then the first distribution surface 41 and the second distribution surface 42 are different distribution surfaces. Alternatively, the first distribution surface 41 can be the first side of the charging structure 3 of the charging pile 10, and the second distribution surface 42 can be the side of the charging pile 10 opposite to the first side or the front end face of the charging structure 3, and the normals of the two (the dashed lines shown in the figure) are parallel, but they are spaced apart in their normal directions. Then the first distribution surface 41 and the second distribution surface 42 are different distribution surfaces. In this specific implementation, by determining the positional relationship between the distribution surfaces where the first marking component 211 and the second marking component 212 are located, it can be determined that the first marking component 211 and the second marking component 212 are located on different distribution surfaces.

[0123] In one specific implementation, the first distribution surface 41 and the second distribution surface 42 may include a plane and a curved surface. For example, the first distribution surface 41 and the second distribution surface 42 may be the side surface of the charging pile 10, or the edge where two side surfaces intersect. It should be noted that when the marking component 2 is placed on a curved surface, the normal to the location of the marking component 2 is used as the normal to the distribution surface on which the marking component 2 is located. When multiple feature markings 20 on distribution surfaces with large curvature differences are simultaneously collected, the accuracy of the final calculated relative pose information of the charging pile 10 and the lawnmower robot is higher.

[0124] In one specific implementation, the charging pile 10 includes a charging pile body 11 and a charging structure 3. The charging structure 3 is installed on a first side of the charging pile body 11, a second identification component 212 is disposed on the charging structure 3, and a first identification component 211 is disposed on the first side of the charging pile body 11 on which the charging structure 3 is installed.

[0125] For example, one end of the charging structure 3 is installed on the first side of the charging pile body 11, and the other end extends out of the charging pile body 11. The charging structure 3 is used to connect to the charging port of the lawnmower robot to charge the lawnmower robot. The second marking component 212 is disposed on the charging structure 3, and the first marking component 211 is disposed on the first side of the charging pile body 1 on which the charging structure 3 is mounted. For example, the first marking component 211 can use four L-shaped marking components, and the second marking component 212 can use one strip marking component. The strip marking component is installed on the end of the charging structure 3 that extends out of the charging pile body 11, and the four L-shaped marking components are installed on the first side of the charging pile body 11 on which the charging structure 3 is mounted. According to the projection position of the marking component 2 on the first side of the charging pile body 11 on which the charging structure 3 is mounted, the four L-shaped marking components 2 surround the strip marking component 2, which has both aesthetic appeal and the functionality of the feature marking 20.

[0126] In this specific implementation, the second marking component 212 is set on the charging structure 3, and the first marking component 211 is set on the side of the charging pile body 11 where the charging structure 3 is installed. This utilizes the structural characteristics of the charging pile 10 itself, so that the first marking component 211 and the second marking component 212 are located on different distribution surfaces. There is no need to set a separate protruding plane or curved surface. The structure is simple and can save costs.

[0127] In one specific implementation, the charging pile 10 includes a charging pile body 11 and a charging structure 3. The charging structure 3 is installed on a first side of the charging pile body 11. The marking component 2 also includes a third marking component 213. The first marking component 211 is disposed on the first side, and the third marking component 213 is disposed on at least one of the other sides or edges of the charging pile body 11 other than the first side.

[0128] For example, the first marking component 211 and the third marking component 212 may be respectively distributed on the first side of the charging pile 10 and any side adjacent to the first side, or the first side and the side opposite to the first side, or the first side and any edge (the edge refers to the intersection of two adjacent sides in the outer peripheral surface), or the first side, any edge and any side adjacent to the first side, or any side of the outer peripheral surface and the edge, etc. It should be noted that the edge here can refer to a facet, which can be a plane or a curved surface.

[0129] It should be noted that in this implementation, the surface where the first marking component 211 is located is the aforementioned first distribution surface 41, and the surface where the third marking component 213 is located is the aforementioned second distribution surface 42. The normal of the first distribution surface 41 where the first marking component 211 is located and the normal of the second distribution surface 42 where the third marking component 213 is located form an angle, and the angle is not equal to 0.

[0130] This specific implementation method ensures that the lawnmower robot can collect and recognize a greater number of feature markers 20 in all directions, thereby producing a more robust image recognition algorithm effect. For example, setting the feature marker 20 on any side adjacent to the first side allows the lawnmower robot to recognize the feature marker 20 even when the offset angle between the lawnmower robot and the first side is large, thereby adjusting the lawnmower robot's posture and achieving recharging.

[0131] Reference Figure 4 In one specific implementation, the marking component 2 is projected onto the surface of the charging pile body 11 where the marking component 2 is located to form a marking pattern. The minimum inner diameter of the marking pattern is greater than or equal to a first preset threshold, which is 3 centimeters. For example, the marking component 2 is projected onto the side of the charging pile body 11 where the marking component 2 is located to form a marking pattern, which is the shape of the marking component 2. The minimum inner diameter of the marking pattern is greater than or equal to the first preset threshold, meaning that the width of any part of the marking pattern is greater than or equal to the first preset threshold. For example, the width of the rectangle in the rectangular marking component 2 is the minimum inner diameter, the height of the triangle in the triangular marking component 2 is the minimum inner diameter, and the width of the L-shaped part in the L-shaped marking component 2 is (…). Figure 4 The width a) is the minimum inner diameter. The first preset threshold can be determined based on the size of the working area of ​​the lawnmower robot and the size of the charging pile range, so that the lawnmower robot can capture the charging pile 11 within the working area. The first preset threshold can be determined to be 3 centimeters.

[0132] In this specific embodiment, by setting the minimum inner diameter of the shape of the marking component 2 to be greater than or equal to a first preset threshold, the size of the marking component 2 is controlled to meet the image recognition requirements of the lawnmower robot, thereby avoiding the inability to accurately identify the feature marking 20 due to the size of the marking component 2 being too small, or the increased cost and the impact on the aesthetics of the charging pile 10 due to the size of the marking component 2 being too large.

[0133] In one specific implementation, the minimum spacing between the marking components 2 located on the same distribution surface of the charging pile body 11 is greater than or equal to a second preset threshold, where the second threshold is 3 centimeters. For example, referring to... Figure 4 The minimum spacing between the marking components 2 on the same side of the charging pile 11 can be represented as b. The second preset threshold can be determined based on the size of the working area of ​​the lawnmower robot and the size of the pile mounting range, so that the lawnmower robot can capture the charging pile 11 within the working area and the feature marking 20 within the pile mounting range. For example, the second preset threshold is set to 3 centimeters. In this specific embodiment, by setting the minimum spacing between adjacent marking components 2 on the same distribution surface to be greater than or equal to the second preset threshold, the spacing between the marking components 2 is controlled to meet the image recognition requirements of the lawnmower robot, thereby improving the accuracy of image recognition by the lawnmower robot.

[0134] It should be noted that this specific implementation method controls the minimum spacing between adjacent marker components 2 on the same distribution surface to be greater than or equal to the second preset threshold, rather than the maximum spacing. This can eliminate the influence of the size of the marker component 2 on the spacing between adjacent marker components 2, thus ensuring that both the size of the marker component 2 and the spacing between adjacent marker components 2 meet the image recognition requirements of the lawnmower robot.

[0135] In one specific implementation, the marking component 2 is a reflective component or a light-emitting component. The lawnmower and charging station 10 are typically used outdoors. When weather conditions are poor or there is insufficient light at night, the lawnmower's image acquisition device cannot accurately capture the feature marking 20 on the charging station 10. This specific implementation, by setting the marking component 2 as a reflective component or a light-emitting component, makes the feature marking 20 more prominent on the charging station 10, easier to identify, and improves the lawnmower's recognition accuracy.

[0136] In one specific implementation, the reflective assembly includes a reflector 22 and a mounting plate 23. The reflector 22 includes a light-transmitting part 221 and a reflective part 222. The reflective part 222 is composed of multiple cubes. The reflective part 222 of the reflector 22 is fixedly connected to the mounting plate 23. The reflective assembly is connected to one side of the charging pile body 11.

[0137] For example, refer to Figure 5 and Figure 6 , Figure 6 Figure I shows an enlarged view of the cube in the reflective part 222. The reflective component can be connected to the charging pile body 11 by ultrasonic welding, snap-fit, or adhesive. The reflective part 222 of the reflective component 22 is connected to the mounting plate 23 by ultrasonic welding. The reflective component 22 can be made of transparent plastic material and can be integrally molded by injection molding. The surface of the charging pile body 11 is provided with a mounting groove 111. When the reflective component is installed in the mounting groove 111 of the charging pile body 1, the side of the reflective component 22 facing away from the charging pile body 11 is a light-transmitting part 221. The light-transmitting part 221 can be a transparent light surface. Light can pass through the light-transmitting part 221 and enter the reflective part 222 for light reflection. The side of the reflective component 22 facing the charging pile body 11 is the reflective part 222. The reflective part 222 can be composed of multiple cubes. The multiple cubes are evenly and parallelly distributed. Two adjacent cubes share one side. The edge length of the cubes can be set in the range of 0.5 mm to 1 mm, so that the cubes are smaller in size and more densely arranged. When illuminated, the reflected light is more uniform and less prone to glare. The outline of the marking component 2 is clearer and easier to identify.

[0138] In this specific implementation, by setting the marking component 2 as a reflective assembly, the structure is simple, easy to install, and makes the marking component 2 easier to be identified on the charging pile body 11.

[0139] In one specific implementation, at least one of the colors and textures of the surface of the marking component 2 differs from that of the surface of the charging pile 11. Specifically, by setting at least one of the textures and colors of the surface of the marking component 2 to be different from those of the surface of the charging pile 11, the image acquisition device of the lawnmower robot can capture the feature markings 20 on the charging pile 10 more quickly and accurately. Preferably, the greater the difference in texture and color between the surface of the marking component 2 and the surface of the charging pile 10, the easier it is to identify.

[0140] According to another aspect of this application, a charging pile 10 is provided, wherein an identification component 2 is provided on the charging pile 10, the identification component 2 being used to indicate the position of the charging pile 10 so that the lawnmower robot can dock with the charging pile 10 for charging.

[0141] For example, the marking component 2 can be one or more, and at least one feature mark 20 can be distributed on the marking component 2. The feature mark 20 can be a corner point of the shape of the marking component 2. The marking component 2 can have different shapes, such as rectangles, triangles, circles, etc. The first marking component 211 and the second marking component 212 can use the same shape of marking component 2, or they can use different shapes of marking component 2. For example, the first marking component 211 can use four L-shaped marking components 2, and the second marking component 212 can use one strip-shaped marking component 2. This embodiment does not limit this. Each marking component 2 can be a single shape or a combination of multiple different shapes. The marking component 2 can be flat, such as a sticker, or three-dimensional, such as plastic. The marking component 2 can be a light-emitting component or a reflective component, so that the feature mark 20 can be more prominent on the charging pile 10 and easier to identify. When the weather conditions are bad or there is insufficient light at night, the image acquisition device of the lawnmower can also accurately acquire the feature mark 20 on the charging pile 10, improving the recognition accuracy of the lawnmower.

[0142] The image acquisition device of the lawnmower acquires usable image frames in the area in front of the charging pile 10. According to a preset recognition algorithm, it identifies preset feature marks 20 on the surface of the charging pile 10 from the image frames. The preset feature marks 20 are marker points on the charging pile 10 that are pre-set according to the preset recognition algorithm. Based on the two-dimensional coordinates of the identified feature marks 20 in the image frame, the three-dimensional coordinates of the feature marks 20 in the world coordinate system, and the imaging parameters of the image acquisition device, the relative pose of the lawnmower relative to the charging pile 10 is determined. Based on the relative pose, the lawnmower is driven back to the charging pile 10.

[0143] In this embodiment, only the marking component 2 needs to be set on the charging pile 10. The lawnmower robot can obtain the pose information of the feature markings 20 distributed on the marking component 2 through its own image recognition, thereby determining the relative pose information between the charging pile 10 and the lawnmower robot for path planning, enabling the lawnmower robot to return to the charging pile for charging. The cost is low, and since only the feature markings 20 need to be set on the surface of the charging pile 10, the structure is simple and easy to install and disassemble.

[0144] In one specific implementation, the charging pile 10 includes a charging structure 3, and an identification component 2 is disposed on the charging structure 3. Exemplarily, the charging structure 3 is disposed on one side of the charging pile 10, and the identification component 2 is disposed on the charging structure 3. The connection method can be a detachable connection, such as a snap-fit, adhesive, or ultrasonic welding. This embodiment does not limit this.

[0145] In this specific implementation, the marking component 2 can be set only on the charging structure 3. The setting position of the marking component 2 does not need to be considered in the size design of the charging pile body 11, which can reduce the size of the charging pile body 11 and reduce costs. Secondly, the design of the texture and color of the charging pile body 11 does not need to consider the impact on the lawnmower robot's collection of information from the marking component 2.

[0146] In one specific implementation, the marking component 2 includes a first marking component 211 and a second marking component 212, and the first marking component 211 and the second marking component 212 have different distribution surfaces on the charging pile 10. For example, the distribution surface of the marking component 2 on the charging pile 10 may include side surfaces and edges, which can be planar or curved surfaces. (Refer to...) Figure 3A and Figure 3B The first marking component 211 is located on the first distribution surface 41 of the charging pile 10, and the second marking component 212 is located on the second distribution surface 42 of the charging pile 10. An angle A is formed between the normal of the first distribution surface 41 and the normal of the second distribution surface 42, and the angle A is not equal to 0. Alternatively, the normal of the first distribution surface 41 is parallel to the normal of the second distribution surface 42, and the first distribution surface 41 and the second distribution surface 42 are spaced apart in the normal direction. It should be noted that when the marking component 2 is placed on a curved surface, the normal of the location of the marking component 2 is taken as the normal of the distribution surface on which the marking component 2 is located.

[0147] In this specific implementation, by setting the first marking component 211 and the second marking component 212 on different distribution surfaces of the charging pile 10, the distribution surfaces of the multiple feature markings 20 are different. Compared with feature markings distributed on the same surface, this implementation can improve the robustness of the image recognition algorithm of the image acquisition device, reduce the probability of misidentification, and improve the accuracy of the lawnmower robot in calculating the relative pose information of the charging pile 10 through image recognition.

[0148] The image acquisition device acquires an image and identifies the first identifier component 211 and the second identifier component 212 in the image, and then extracts the feature identifiers 20 from the first identifier component 211 and the second identifier component 212 (usually the endpoints and inflection points of the first identifier component 211 and the second identifier component 212 are used as feature identifiers 20).

[0149] The specific algorithm process is as follows: Set n feature identifiers 20. First, detect the position coordinates of m (m <= n) feature identifiers 20 on the 2D image. Then, based on the parameters of the image acquisition device and the 3D coordinate positions of the feature identifiers 20, match p (p <= m <= n) valid feature identifiers 20, and further calculate the relative position between the vehicle body and the camera. When p = m = n, it is considered that the image recognition is successful. If these n feature identifiers 20 are all set on the same plane, assume that the first feature identifier is set to 4, and each first feature identifier includes 1 feature identifier 20 (identification point). The 4 (n = 4) feature identifiers 20 (identification points) form a rectangle. When problems such as sunlight irradiation, camera surface occlusion, and blurring cause detection on the 2D image, the two feature identifiers 20 on the left side of the rectangle shift in the same direction (but the overall 4 feature identifiers after the shift still satisfy the geometric features of affine transformation). Therefore, there is a probability that the algorithm can still output the pose that should have an error based on the n = p = 4 coordinates after the shift, and this pose will usually deviate severely. When there are feature identifiers 20 on non - same planes, assume that 4 feature identifiers 20 are located at the rectangle endpoints of the same distribution plane, and the 2 feature identifiers 20 included in the second identification component 212 are on another distribution plane (the distribution plane protruding 4 points), that is, n = 6; when the feature identifiers on one side of the rectangle shift and the 4 feature identifiers 20 on the same plane satisfy the geometric features after the rectangle affine transformation, but due to the existence of the 2 protruding feature identifiers 20, the algorithm cannot calculate the pose with an error based on the coordinates after the shift (because at this time p < n due to the existence of the protruding points), so it will not provide a severely deviated pose. The first identification component 211 and the second identification component 212 are set on different distribution planes of the charging pile 10, which can improve the robustness of the image recognition algorithm of the image acquisition device, reduce the probability of mis - recognition, and improve the accuracy of the relative pose information of the charging pile 10 calculated by the image recognition of the lawn mowing robot.

[0150] In a specific implementation, the charging pile 10 includes a charging pile body 11 and a charging structure 3. The first side of the charging pile body 11 is provided with a first identification component 211 and the charging structure 3, that is, the first identification component 211 is set on the side where the charging structure 3 is located.

[0151] The lens acquisition range of the image acquisition device of the lawn mowing robot is limited. When the lawn mowing robot moves in front of the charging pile 10 and starts to dock, the closer the lawn mowing robot is to the charging structure of the charging pile, the larger the acquisition range of the first side occupying the lens of the image acquisition device. By setting the first identification component 211 on the first side, the closer the lawn mowing robot is to the charging pile 10, the more accurate the position of the first identification component 211 is recognized, the more accurate the pose of the charging pile 10 is determined, the more accurate the pose adjustment of the lawn mowing robot is, and the more accurate the docking between the charging structure 3 and the charging port of the lawn mowing robot is.

[0152] Furthermore, the charging structure 3 protrudes from the first side. During the process from initial contact to full contact between the charging part of the lawnmower robot and the charging structure 3, the first marking component 211 located on the first side ensures that the lawnmower robot can still detect the first marking component 211 and continuously adjust the lawnmower robot's posture according to the position of the first marking component 211, ensuring that the charging part of the lawnmower robot makes full contact with the charging structure 3 for effective charging.

[0153] In one specific implementation, a second marking component 212 is provided on the charging structure 3. The second marking component 212 can be provided on the front face or the side face of the charging structure 3.

[0154] In this implementation, by setting the second marking component 212 on the charging structure 3 and setting the first marking component 211 on the first side of the charging pile body 11 where the charging structure 3 is installed, the structural characteristics of the charging pile 10 itself are utilized, so that the first marking component 211 and the second marking component 212 are located on different distribution surfaces, without the need to set a separate protruding plane or curved surface, the structure is simple and can save costs.

[0155] In one specific implementation, a second marking component 212 is provided at the suspended end of the charging structure 3, and the second marking component 212 is located on the front face. When the lawnmower robot moves to the first side of the charging pile 10 but has not yet reached the top position, the image acquisition device is directly facing the front face of the suspended end of the charging structure 3, and can completely acquire the features of the front face of the suspended end of the charging structure 3. Compared to placing the marking component 2 on the side of the charging structure 3, placing the second marking component 212 on the front face of the suspended end of the charging structure 3 ensures that the lawnmower robot can obtain all the feature markings of the second marking component 212, improving the accuracy of the pose information of the charging structure 3 and realizing recharging.

[0156] In one specific implementation, the marking component 2 further includes a third marking component 213. The third marking component 213 is disposed on the second side of the charging pile body 11, and the second side is connected to the first side. Exemplarily, the first marking component 211 and the third marking component 213 can be respectively distributed on the first side of the charging pile 10 and any side adjacent to the first side, or on the first side and any edge (an edge refers to the intersection of two adjacent sides on the outer peripheral surface), etc. It should be noted that the edge here can refer to a facet, which can be a plane or a curved surface.

[0157] It should be noted that in this implementation, the surface where the first marking component 211 is located is the aforementioned first distribution surface 41, and the surface where the third marking component 213 is located is the aforementioned second distribution surface 42. The normal of the first distribution surface 41 where the first marking component 211 is located and the normal of the second distribution surface 42 where the third marking component 213 is located form an angle, and the angle is not equal to 0.

[0158] This specific implementation ensures that the lawnmower robot can collect and recognize a greater number of feature markers 20 in all directions of the charging station 10, thereby producing a more robust image recognition algorithm effect. For example, setting the third marker component 213 on any side adjacent to the first side allows the lawnmower robot to recognize the feature markers 20 even when the offset angle between the lawnmower robot and the first side is large, thereby adjusting the lawnmower robot's posture and achieving return to the charging station.

[0159] In one specific implementation, there are multiple first identification components 211, distributed around the charging structure 3. The lawnmower robot may return to the vicinity of the charging station from different locations and then recharge. If the first identification component 211 is only placed in one position on the charging structure 3, it may be impossible to correctly identify the first identification component 211 due to obstruction by the charging structure 3, resulting in pose error.

[0160] For example, if only two first identification components 211 are set on the first side, and both are below the charging structure 3 (the first identification components 211 are closer to the base plate 13 than the charging structure 3), during the process from initial contact to full contact between the charging part of the lawnmower robot and the charging structure, the lens of the image acquisition device is above the charging structure 3 (the lens of the image acquisition device is farther from the base plate 13 than the charging structure 3), and is blocked by the charging structure 3, making it impossible to correctly identify the first identification components 211 located below the charging structure 3. By distributing multiple first identification components 211 around the charging structure 3, during the process from initial contact to full contact between the charging part of the lawnmower robot and the charging structure 3, it can be ensured that the lawnmower robot can still acquire the first identification components 211 located above the charging structure 3, reducing the positional error caused by the inability to correctly identify the first identification components 211 due to the obstruction of the charging structure 3.

[0161] In another embodiment, a plurality of first identification components 211 may be spaced apart around the charging structure 3, as shown in the figure. Figure 4 There are four first identification components 211, spaced apart around the charging structure 3. The feature markers 20 for image acquisition device lens recognition are mainly distributed at the endpoints, midpoints, or inflection points of the identification components 2. By arranging multiple first identification components 211 at intervals, the number of feature markers at the endpoints can be increased.

[0162] In one specific implementation, multiple first identification components 211 are distributed non-centrally symmetrically. When the image acquisition device is offset from the setting position of the lawnmower robot, the non-centrally symmetrical distribution of multiple first identification components 211 can make the distribution of the first identification components 211 correspond to the offset direction of the image acquisition device, thereby enabling the image acquisition device to obtain more information from the identification components 2.

[0163] In one specific implementation, the first identification component 211 includes two mutually perpendicular and intersecting strip-shaped structures. For example, referring to... Figure 4 The first marking component 211 can be L-shaped. By setting the shape of the first marking component 211 to two mutually perpendicular and intersecting strip structures, multiple dispersed feature markings can be distributed, and less material is required. Secondly, compared with multiple dot-shaped marking structures, the present implementation has strong overall structural integrity, simple structure, and is easy to install.

[0164] According to another aspect of this application, a charging pile 10 system is provided, including a charging pile 10 as described in any of the above embodiments and an identification device 50. The identification device 50 is provided with a fourth identification component 214, which is used to assist in determining the location of the charging pile 10.

[0165] Optionally, the third distribution surface 43 on the charging pile 10 is provided with an identification component 2. The third distribution surface 43 may include a first distribution surface 41 and a second distribution surface 42. The fourth distribution surface 44 on the identification device 50 is provided with a fourth identification component 214. The normal of the third distribution surface 43 and the normal of the fourth distribution surface 44 form an angle, and the angle is not equal to 0. Alternatively, the normal of the third distribution surface 43 is parallel to the normal of the fourth distribution surface 44, and the third distribution surface 43 and the fourth distribution surface 44 are spaced apart in the normal direction.

[0166] For example, refer to Figure 7 The marking device 50 can be a specially designed sign or the like. The marking device 50 and the charging pile 10 are separately set in the working area of ​​the lawnmower robot. The charging pile 10 is equipped with marking components 2, such as a first marking component 211, a second marking component 212, and a third marking component 213. The surface on which the marking component 2 is located is a third distribution surface 43. The third distribution surface 43 can include the first distribution surface 41 and the second distribution surface 42 in the above embodiments. The marking device 50 is equipped with a fourth marking component 214. The surface on which the fourth marking component 214 is located is a fourth distribution surface 44. An angle B is formed between the normal of the third distribution surface 43 (the dashed line shown in the figure) and the normal of the fourth distribution surface 44 (the dashed line shown in the figure), and the angle B is not equal to 0; or, the normal of the third distribution surface 43 is parallel to the normal of the fourth distribution surface 44, and the third distribution surface 43 and the fourth distribution surface 44 are spaced apart in the normal direction.

[0167] In this embodiment, by separately arranging the identification component 2 on the charging pile 10 and the identification device 50, the distribution surfaces of multiple feature identifications 20 are different, thereby improving the accuracy of the image recognition calculation of the relative pose information of the charging pile 10 by the lawn mowing robot.

[0168] In a specific implementation manner, the identification device 50 is a positioning signal receiving device. Exemplarily, the positioning signal receiving device can be a GPS signal receiver. The fourth identification component 214 can be arranged on the surface of the positioning signal receiving device, and the connection manner can be a detachable connection or the like. In this implementation manner, by arranging the identification component 2 on the positioning signal receiving device supporting the charging pile 10, the cost of separately arranging the identification device 50 can be reduced.

[0169] It should be noted that in the description of the present invention, the terms "first" and "second" are only used for conveniently describing different components or names, and cannot be understood as indicating or implying an order relationship, relative importance, or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" can explicitly or implicitly include at least one of these features.

[0170] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this invention belongs. The terms used in the description of the present invention herein are only for the purpose of describing specific embodiments, and are not intended to limit the present invention.

[0171] It should be noted that although the specific embodiments of the present invention have been described in detail with reference to the accompanying drawings, it should not be construed as a limitation on the protection scope of the present invention. Within the scope described in the claims, various modifications and deformations that can be made by those skilled in the art without creative labor still belong to the protection scope of the present invention.

[0172] The examples of the embodiments of the present invention are intended to briefly illustrate the technical features of the embodiments of the present invention, so that those skilled in the art can intuitively understand the technical features of the embodiments of the present invention, and are not used as an improper limitation on the embodiments of the present invention.

[0173] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A charging pile, characterized in that, The charging pile (10) is provided with an identification component (2), which is used to indicate the position of the charging pile (10) so that the lawn mower robot can be connected to the charging pile (10) for charging. The identification component (2) includes a first identification component (211) and a second identification component (212), and the first identification component (211) and the second identification component (212) are distributed on different surfaces of the charging pile (10). The charging pile (10) includes a charging pile body (11) and a charging structure (3), the first side of the charging pile body (11) is provided with the first identification component (211) and the charging structure (3), the first identification component (211) is a plurality of, and a plurality of the first identification component (211) is distributed around the charging structure (3). The second identification component (212) is arranged on the charging structure (3), one end of the charging structure (3) is provided with the second identification component (212), and the second identification component (212) is arranged on the front end surface.

2. The charging post of claim 1, wherein, The charging pile (10) includes a charging structure (3), and the identification component (2) is arranged on the charging structure (3).

3. The charging post of claim 1, wherein, The identification component (2) further includes a third identification component (213), and the second side of the charging pile body (11) is provided with the third identification component (213), and the second side is connected with the first side.

4. The charging post of claim 1, wherein, A plurality of the first identification component (211) is distributed in a non-central symmetry.

5. The charging post of claim 4, wherein, The first identification component (211) includes two strip structures which are perpendicular to each other and intersect with each other.

6. The charging station of claim 1, wherein, The first identification component (211) is located on a first distribution surface (41) of the charging pile (10), the second identification component (212) is located on a second distribution surface (42) of the charging pile (10), a normal line of the first distribution surface (41) and a normal line of the second distribution surface (42) form an included angle, and the angle of the included angle is not equal to 0; or, The normal line of the first distribution surface (41) is parallel to the normal line of the second distribution surface (42), and the first distribution surface (41) and the second distribution surface (42) are spaced in the normal direction.

7. The charging station of claim 1, wherein, The identification component (2) projects on the surface of the charging pile body (11) to form an identification pattern, and the minimum inner diameter in the identification pattern is greater than or equal to a first preset threshold, and the first preset threshold is 3 cm.

8. The charging post of claim 1 or 7, wherein, The minimum spacing between the identification components (2) on the same distribution surface of the charging pile body (11) is greater than or equal to a second preset threshold, and the second preset threshold is 3 cm.

9. The charging station of claim 1, wherein, The identification component (2) is a light reflecting component or a light emitting component.

10. The charging station of claim 9, wherein, The light reflecting component includes a light reflecting piece (22) and a mounting plate (23), the light reflecting piece (22) includes a light transmitting part (221) and a light reflecting part (222), the light reflecting part (222) is composed of a plurality of regular cubes, the light reflecting part (222) of the light reflecting piece (22) is fixedly connected with the mounting plate (23), and the light reflecting component is connected on one side of the charging pile body (11).

11. A charging station system, characterized by The charging pile (10) and the identification device (50) as claimed in any one of claims 1-10, wherein the identification device (50) is provided with a fourth identification component (214) for assisting in determining the position of the charging pile (10).

12. The charging station system of claim 11, wherein, The identification device (50) is a positioning signal receiving device.

13. A charging station, characterized in that The charging pile (10) is provided with a plurality of feature identifiers (20), at least two of the feature identifiers (20) have different distribution surfaces from the remaining feature identifiers (20), and the feature identifiers (20) are used for image recognition by a mowing robot to enable the mowing robot to dock with the charging pile (10) for charging. The surface of the charging pile (10) is provided with an identification component (2), which includes a first identification component (211) and a second identification component (212), the first identification component (211) and the second identification component (212) have different distribution surfaces on the charging pile (10), and at least two feature identifiers (20) are distributed on the identification component (2). The charging pile (10) includes a charging pile body (11) and a charging structure (3), the first side of the charging pile body (11) is provided with the first identification component (211) and the charging structure (3), the first identification component (211) is a plurality of components, and the plurality of first identification components (211) are distributed around the charging structure (3). The second identification component (212) is arranged on the charging structure (3), the second identification component (212) is arranged on the free end of the charging structure (3), and the second identification component (212) is arranged on the front end surface. The method is used for guiding a mowing robot to return to the charging pile as claimed in any one of claims 1-10, the mowing robot is provided with an image acquisition device, the charging pile is provided with feature identifiers, and the method includes:

14. A method of back-pedaling, characterized by Obtaining an available image frame acquired by the image acquisition device of the mowing robot in a pile front area of the charging pile, the number of feature identifiers identified in the image frame is greater than or equal to N, and N is a positive integer greater than or equal to 3; Determining the relative pose of the mowing robot relative to the charging pile according to the two-dimensional coordinates of the identified feature identifiers in the image frame, the three-dimensional coordinates of the feature identifiers in the world coordinate system, and the imaging parameters of the image acquisition device; Driving the mowing robot to return to the charging pile according to the relative pose. The pile front area includes a sector area corresponding to the charging pile, the sector area is located in front of the charging pile, and the radius of the sector area is greater than 0 and less than M, and M is a positive integer.

15. The method of claim 14, wherein, The method includes:

16. The method according to claim 14 or 15, characterized in that Obtaining an available image frame acquired by the image acquisition device of the mowing robot in a pile front area of the charging pile, the number of feature identifiers identified in the image frame is greater than or equal to N, and N is a positive integer greater than or equal to 3; Obtaining the navigation pose of the mowing robot at the current time; Determining whether the mowing robot is in the pile front area according to the navigation pose and the preset pose of the charging pile; If the mowing robot is in the pre-pile area, it is determined whether there is a usable image frame in the image frame sequence collected by the image collection device; If there is a usable image frame, the existing usable image frame is taken as the obtained usable image frame.

17. The method of claim 16, wherein, The method further comprises: If there is no usable image frame, at least one of the position and attitude of the mowing robot is adjusted, a new image frame sequence is obtained, and the step of determining whether there is a usable image frame in the image frame sequence collected by the image collection device is continued.

18. The method of claim 17, wherein, The step of adjusting at least one of the position and attitude of the mowing robot comprises: The mowing robot is adjusted to a new navigation attitude, and the mowing robot is located outside the pre-pile area when it is in the new navigation attitude; According to the new navigation attitude, the mowing robot is driven into the pre-pile area, and the image collection device collects multiple image frames to form the image frame sequence during the movement of the mowing robot.

19. The method of claim 16, wherein, The method further comprises: If the mowing robot is not in the pre-pile area, the mowing robot is driven to move into the pre-pile area according to the navigation attitude, and the step of obtaining the navigation attitude of the mowing robot at the current time is continued.

20. The method of claim 14, wherein, The charging pile is provided with Q feature marks, Q is greater than or equal to 2 and is a positive integer, and for each coordinate axis of the world coordinate system, the maximum distance of the projection of the Q feature marks on each coordinate axis is greater than or equal to a distance threshold.

21. The method of claim 20, wherein, The distribution planes of at least two of the Q feature marks are different from those of the remaining feature marks.

22. The method of claim 14, wherein, When the image collection device collects the image frame, the exposure time of the image frame is positively correlated with the relative distance when the light intensity of the ambient light is fixed, and the relative distance is the distance between the mowing robot and the charging pile.

23. A pile backer device, characterized by The device is used to guide the mowing robot to return to the charging pile, the mowing robot is provided with an image collection device, the charging pile is distributed with feature marks, the surface of the charging pile (10) is provided with an identification component (2), the identification component includes a first identification component (211) and a second identification component (212), the distribution planes of the first identification component (211) and the second identification component (212) on the charging pile (10) are different, and at least two feature marks (20) are distributed on the identification component (2). The charging pile (10) comprises a charging pile body (11) and a charging structure (3), a first side surface of the charging pile body (11) is provided with the first identification component (211) and the charging structure (3), the first identification component (211) is a plurality of, and a plurality of the first identification component (211) is distributed around the charging structure (3). The second identification component (212) is arranged on the charging structure (3), and the second identification component (212) is arranged on the front end face of the suspended end of the charging structure (3). An acquisition module is configured to acquire an available image frame collected by an image collection device of the mowing robot in a pre-pile area of the charging pile, and a number of feature marks identified in the image frame is greater than or equal to N, where N is a positive integer greater than or equal to 3. A determination module is configured to determine a relative pose of the mowing robot relative to the charging pile according to two-dimensional coordinates of the identified feature marks in the image frame, three-dimensional coordinates of the feature marks in a world coordinate system, and imaging parameters of the image collection device. A driving module is configured to drive the mowing robot to return to the charging pile according to the relative pose.

24. A mowing robot, on which an image capturing device is arranged, characterized in that The mowing robot comprises a controller configured to perform the method of any one of claims 14-22.

Images