Lidar-based robot recharging method, device and mowing robot

Abstract

The present disclosure relates to a laser radar-based robot return charging method and device, wherein the method guides the robot to return charging by detecting the position of the charging pile through the laser radar arranged on the robot, and when the position of the charging pile is identified by using the initial point cloud data detected by the laser radar, the initial point cloud set corresponding to the charging pile in the detected initial point cloud data is subjected to a completion processing and / or a noise point cloud processing, so that the target point cloud set of the charging pile identified based on the corrected initial point cloud data is more accurate, the position of the charging pile identified by the robot is more accurate, the planned return charging path is more accurate, and the return charging efficiency is improved.

Description

Technical Field

[0001] This disclosure relates to the field of robot control technology, and in particular to a robot recharging method, device, and lawn mowing robot based on lidar. Background Technology

[0002] Currently, guiding lawnmowers back to their charging stations typically utilizes infrared sensors. Specifically, multiple infrared transmitters are placed on the charging station, for example, one on the left edge, one in the middle, and one on the right edge. These transmitters emit infrared signals, which are received by the lawnmower's infrared receiver. Based on these signals, the lawnmower determines the charging station's location and then moves accordingly. Guided by the infrared signals, the lawnmower continues to move until it is directly facing the charging station, at which point it moves forward to align with the charging port and begin charging. However, infrared sensors have the following drawbacks: 1. The infrared transmitters have relatively low power and short transmission distance, making it impossible to guide the lawnmower back to the charging station when it is far away. 2. The signals emitted by the infrared transmitters are easily interfered with by sunlight outdoors, resulting in poor stability. 3. Interference can exist between the infrared signals from multiple transmitters on the charging station, affecting the lawnmower's ability to determine the charging station's location and impacting recharging efficiency. Therefore, how to guide lawnmowers back to their charging stations quickly and efficiently is a problem that urgently needs to be solved. Summary of the Invention

[0003] To address the aforementioned technical issues, this disclosure provides a robot recharging method, apparatus, and lawnmower robot based on lidar.

[0004] In a first aspect, this disclosure provides a robot recharging method based on lidar, comprising:

[0005] Acquire the first and second initial point cloud data detected by the lidar on the robot during its movement;

[0006] Point cloud recognition is performed on the first initial point cloud data to obtain the first initial point cloud set of the charging pile. The first initial point cloud set is then filled and / or noise point cloud is removed to obtain the first target point cloud set.

[0007] Point cloud recognition is performed on the second initial point cloud data to obtain the second initial point cloud set of the charging pile. The second initial point cloud set is then filled and / or noise point cloud is removed to obtain the second target point cloud set.

[0008] Point cloud matching and point cloud identification are performed based on the first target point cloud set, the third initial point cloud set, the second target point cloud set, and the fourth initial point cloud set to obtain the third target point cloud set corresponding to the charging pile; wherein, the third initial point cloud set includes other point clouds in the first initial point cloud data other than the first initial point cloud set, and the fourth initial point cloud set includes other point clouds in the second initial point cloud data other than the second initial point cloud set;

[0009] The target pose of the charging pile is determined based on the third target point cloud set;

[0010] Based on the target pose of the charging pile, a return path is planned, and the robot is controlled to move along the return path to the location of the charging pile for charging.

[0011] In some embodiments, the completion processing and / or noise point cloud removal processing of the target initial point cloud set of the charging pile includes:

[0012] Using a preset 3D point cloud model of the charging pile, the target initial point cloud set of the charging pile is subjected to completion processing and / or noise point cloud removal processing, wherein the target initial point cloud set is either the first initial point cloud set or the second initial point cloud set.

[0013] In some embodiments, the step of using a preset 3D point cloud model of the charging pile to perform completion processing and / or noise point cloud removal processing on the target initial point cloud set of the charging pile includes:

[0014] The Iterative Closest Point (ICP) algorithm is used to perform point cloud matching between the 3D point cloud model of the charging pile and the target initial point cloud set of the charging pile. Based on the point cloud matching results, one of the overlapping point clouds is retained and combined with the non-overlapping point clouds to complete the target initial point cloud set.

[0015] In some embodiments, the step of using a preset 3D point cloud model of the charging pile to perform completion processing and / or noise point cloud removal processing on the target initial point cloud set of the charging pile includes:

[0016] The area to be completed is determined based on the initial point cloud set of the charging piles.

[0017] Determine the point cloud data corresponding to the region to be completed in the three-dimensional point cloud model of the charging pile;

[0018] The initial target point cloud set of the charging pile is completed using the point cloud data corresponding to the region to be completed in the three-dimensional point cloud model.

[0019] In some embodiments, determining the region to be completed based on the target initial point cloud set of the charging pile includes:

[0020] The first shape region of the charging pile is obtained based on the initial point cloud set of the target of the charging pile;

[0021] The first shape area of ​​the charging pile is mapped to the standard shape area of ​​the charging pile, and the area in the standard shape area that is not covered by the first shape area is determined as the area to be completed.

[0022] In some embodiments, the step of using a preset 3D point cloud model of the charging pile to perform completion processing and / or noise point cloud removal processing on the target initial point cloud set of the charging pile includes:

[0023] The initial center position of the charging pile is determined based on the target initial point cloud set of the charging pile.

[0024] The center position of the three-dimensional point cloud model of the charging pile is set at the initial center position, and the point cloud data outside the outline of the three-dimensional point cloud model is removed.

[0025] In some embodiments, determining the initial center position corresponding to the charging pile based on the target initial point cloud set of the charging pile includes:

[0026] The initial point cloud position of the charging pile is obtained by fitting the point cloud data of each point cloud in the target initial point cloud set using the Random Sample Consensus (RANSAC) algorithm.

[0027] Secondly, this disclosure provides a robot recharging device based on lidar, comprising:

[0028] The data acquisition module is used to acquire the first and second initial point cloud data detected by the lidar on the robot during its movement.

[0029] The data processing module is further configured to perform point cloud recognition on the first initial point cloud data to obtain a first initial point cloud set of the charging pile, and to perform completion processing and / or noise point cloud removal processing on the first initial point cloud set to obtain a first target point cloud set.

[0030] The data processing module is further configured to perform point cloud recognition on the second initial point cloud data to obtain the second initial point cloud set of the charging pile, and to perform completion processing and / or noise point cloud removal processing on the second initial point cloud set to obtain the second target point cloud set.

[0031] The data processing module is further configured to perform point cloud matching and point cloud recognition based on the first target point cloud set, the third initial point cloud set, the second target point cloud set, and the fourth initial point cloud set to obtain the third target point cloud set corresponding to the charging pile; wherein, the third initial point cloud set includes other point clouds in the first initial point cloud data other than the first initial point cloud set, and the fourth initial point cloud set includes other point clouds in the second initial point cloud data other than the second initial point cloud set;

[0032] The data processing module is further configured to determine the target pose of the charging pile based on the third target point cloud set.

[0033] The path planning module is used to plan the recharge path based on the target pose of the charging pile.

[0034] The motion control module is used to control the robot to move along the recharge path to the location of the charging station for charging.

[0035] Thirdly, this disclosure provides an electronic device, including: a memory and a processor;

[0036] The memory is configured to store computer program instructions;

[0037] The processor is configured to execute the computer program instructions, causing the electronic device to implement the method as described in the first aspect and any one of the first aspects.

[0038] Fourthly, this disclosure provides a computer-readable storage medium, comprising: computer program instructions; and an electronic device executing the computer program instructions to cause the electronic device to perform the method as described in the first aspect and any one of the first aspects.

[0039] Fifthly, this disclosure provides a lawnmower robot, comprising: at least one lidar, a memory, and a processor; wherein, in a lawnmower recharging scenario, the at least one lidar is used to detect initial point cloud data during the movement of the lawnmower; the memory is used to store the initial point cloud data detected by the at least one lidar; the memory also stores computer program instructions; the processor executes the computer program instructions to perform the method described in the first aspect and any one of the first aspects based on the initial point cloud data detected by the at least one lidar.

[0040] In a sixth aspect, this disclosure provides a computer program product that an electronic device runs, causing the electronic device to perform the method as described in the first aspect and any one of the first aspects.

[0041] This embodiment provides a robot recharging method, device, and lawnmower robot based on LiDAR. The method guides the robot to recharge by detecting the location of the charging pile using a LiDAR mounted on the robot. When identifying the location of the charging pile, the initial point cloud set corresponding to the charging pile in the detected initial point cloud data is completed and / or noisy point cloud is removed. In this way, the target point cloud set of the charging pile is more accurately identified based on the corrected initial point cloud data, and the location / pose of the charging pile identified by the robot is more accurate, making the planned recharge path more accurate and improving recharge efficiency. Attached Figure Description

[0042] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0043] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram of the hardware structure of a lawnmower robot according to an embodiment of the present disclosure;

[0045] Figure 2 A flowchart illustrating a robot recharging method based on lidar provided in an embodiment of this disclosure;

[0046] Figure 3 A flowchart illustrating a robot recharging method based on lidar provided in another embodiment of this disclosure;

[0047] Figure 4 A flowchart illustrating a robot recharging method based on lidar provided in another embodiment of this disclosure;

[0048] Figure 5 A flowchart illustrating a robot recharging method based on lidar provided in another embodiment of this disclosure;

[0049] Figure 6 A schematic diagram of the structure of a robot recharging device based on lidar provided in an embodiment of this disclosure;

[0050] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0051] To better understand the above-mentioned objectives, features, and advantages of this disclosure, the solutions disclosed herein will be further described below. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0052] Numerous specific details are set forth in the following description in order to provide a full understanding of this disclosure, but this disclosure may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only some, and not all, of the embodiments of this disclosure.

[0053] Figure 1 This is a schematic diagram of the hardware structure of a lawnmower robot provided according to an embodiment of this disclosure. Please refer to [link / reference]. Figure 1 As shown, the lawnmower robot 100 provided in this embodiment includes at least: a mowing component 110, a moving component 120, a processor 130, a lidar 140, a memory 150, and a bus 160. The mowing component 110, the moving component 120, the processor 130, the lidar 140, and the memory 150 can communicate with each other via the bus 160. The lawnmower robot 100 may also include other sensors, such as image sensors, infrared sensors, etc. Figure 1 Not shown in the image.

[0054] The mowing assembly 110 is used to perform mowing operations, and the mowing assembly 110 includes at least a blade disc and a blade disc drive structure.

[0055] The moving component 120 is used to enable the movement of the lawnmower robot 100, and the moving component 120 includes at least wheels and a wheel drive structure.

[0056] The LiDAR 140 is used to acquire point cloud data. This disclosure does not limit the number of lines of the LiDAR 140. It should be understood that the number of lines of the LiDAR 140 typically refers to its horizontal scanning resolution. Different models of LiDAR may have different line counts, and a higher line count generally results in stronger data acquisition and environmental perception capabilities. A suitable line count can be selected based on requirements. Furthermore, this disclosure does not limit the number of LiDARs 140. When multiple LiDARs are installed on the lawnmower robot 100, their orientations can differ to detect the environment from different perspectives. For example, one LiDAR can face forward to detect the environment in front of the lawnmower robot, while another LiDAR can face backward to detect the environment behind it.

[0057] The processor 130 can call computer program instructions in the memory 150 to execute the robot recharging method based on LiDAR in the following embodiments.

[0058] The processor 130, sensor 140, and memory 150 may each include one or multiple integrated components, and this application embodiment does not specifically limit this.

[0059] Furthermore, the logical instructions in the aforementioned memory 150 can be implemented as software functional modules and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0060] Based on the above Figure 1 The description of the embodiments shown illustrates that the LiDAR-based robot recharging method provided in this application can be executed by the software and / or hardware modules of a lawnmower robot. For simplicity, a lawnmower robot is used as an example, and the LiDAR-based robot recharging method provided in this application will be described in detail below with reference to the accompanying drawings.

[0061] It should be noted that the LiDAR-based robot recharging method provided in this disclosure can be applied to, but is not limited to, lawn mowing robots, and can also be applied to robots or devices with similar problems.

[0062] Figure 2 This is a flowchart illustrating a lidar-based robot recharging method according to an embodiment of this disclosure. Please refer to [link / reference]. Figure 2 As shown, the method in this embodiment includes:

[0063] S201. Acquire the first initial point cloud data and the second initial point cloud data detected by the lidar on the robot during its movement.

[0064] The lawnmower is triggered to return to its charging dock when low battery is detected, after the mowing task is completed, or when the user actively controls the lawnmower to return to its charging dock. This application does not limit the conditions for triggering the lawnmower to return to its charging dock.

[0065] When the lawnmower robot determines it needs to return to its charging dock, it moves and simultaneously activates its LiDAR for detection. For example, the lawnmower robot can move in place, or it can determine its initial direction of movement based on point cloud data detected during the mowing process and then move accordingly.

[0066] The first and second point cloud data can be point cloud data detected by the same lidar on the lawnmower from different perspectives, or point cloud data from the same perspective at different times, or point cloud data detected by different lidars on the lawnmower.

[0067] S202. Perform point cloud recognition on the first initial point cloud data to obtain the first initial point cloud set of the charging pile. Perform completion processing and / or noise point cloud removal processing on the first initial point cloud set to obtain the first target point cloud set.

[0068] S203. Perform point cloud recognition on the second initial point cloud data to obtain the second initial point cloud set of the charging pile. Perform completion processing and / or noise point cloud removal processing on the second initial point cloud set to obtain the second target point cloud set.

[0069] A point cloud recognition algorithm can be pre-configured in the lawnmower. This algorithm analyzes the geometric and semantic information of each point cloud in the initial target point cloud data to identify the target initial point cloud set corresponding to the charging station. The target initial point cloud data can be either first initial point cloud data or second initial point cloud data. Correspondingly, the target initial point cloud set corresponding to the charging station is either the first initial point cloud set identified from the first initial point cloud data, or the second initial point cloud set identified from the second initial point cloud data.

[0070] Point cloud recognition algorithms can be, but are not limited to, algorithms based on geometric features, machine learning, feature descriptors, and image processing. Specifically, algorithms based on geometric features include methods such as normal-based methods, curvature-based methods, and edge extraction methods, which utilize the geometric properties of point clouds for the identification and segmentation of charging stations. Machine learning algorithms include methods such as support vector machines, random forests, and deep learning, which learn features from sample data through pre-trained models for classification and recognition. Algorithms based on feature descriptors include methods such as SHOT (Signature of Histograms of Orientations), FPFH (Fast Point Feature Histograms), and 3D-SIFT (3D Scale Invariant Feature Transform), which extract local feature descriptors from point clouds for matching and recognition. Image processing algorithms, for example, convert point cloud data into depth images, project them onto an image plane for processing, and use image processing techniques for object detection and recognition.

[0071] By using the aforementioned arbitrary point cloud recognition algorithm to detect and identify charging piles from the initial point cloud data of the target, the target initial point cloud set corresponding to the charging pile can be obtained.

[0072] Next, the lawnmower robot completes and / or removes noisy point clouds from the initial target point cloud set of the charging pile, and obtains a more accurate charging pose based on the processed point cloud data.

[0073] In some embodiments, the lawnmower robot uses a pre-created 3D point cloud model of the charging pile to perform completion processing and / or noise point cloud removal processing on the target initial point cloud set of the charging pile. The 3D point cloud model of the charging pile is obtained through computer modeling based on the shape of the charging pile, accurately representing its shape, size, and other attributes. During completion processing, the 3D point cloud model of the charging pile is used as a priori value to fill in missing parts of the actual detected point cloud data, ensuring the accuracy of the filled point cloud data. During noise point cloud removal processing, the 3D point cloud model of the charging pile is used as a priori value to accurately determine the noise point cloud, thereby enabling its removal.

[0074] It should be noted that the order of performing completion processing and noise removal point cloud processing is not limited. Completion processing can be performed first, followed by noise removal, or vice versa. Specifically, if completion processing is performed first, the lawnmower robot can perform noise removal processing on the completed point cloud set; if noise removal processing is performed first, the lawnmower robot can perform completion processing on the point cloud set after noise removal.

[0075] S204. Perform point cloud matching and point cloud recognition based on the first target point cloud set, the third initial point cloud set, the second target point cloud set, and the fourth initial point cloud set to obtain the third target point cloud set corresponding to the charging pile; wherein, the third initial point cloud set includes other point clouds in the first initial point cloud data other than the first initial point cloud set, and the fourth initial point cloud set includes other point clouds in the second initial point cloud data other than the second initial point cloud set.

[0076] Combining the first target point cloud set and the third initial point cloud set yields the corrected first initial point cloud data, and combining the second target point cloud set and the fourth initial point cloud set yields the corrected second initial point cloud data.

[0077] Based on this, the lawnmower robot performs point cloud matching and point cloud recognition sequentially on the corrected first initial point cloud data and the corrected second initial point cloud data.

[0078] Point cloud matching involves transforming the corrected first and second initial point cloud data into the same coordinate system based on the relative transformation relationship between them, thus aligning them within that coordinate system. The relative transformation relationship can include translation, rotation, scaling, etc. Furthermore, the coordinate system used for alignment during point cloud matching can be a coordinate system established based on the current position and orientation of the lawnmower robot.

[0079] For example, the ICP algorithm can be used to perform point cloud matching on the corrected first initial point cloud data and the corrected second initial point cloud data. The detailed point cloud matching process is as follows:

[0080] Step a1, Initialization: Select an initial transformation matrix to align the corrected first initial point cloud data and the corrected second initial point cloud data to be matched (assuming the corrected first point cloud data is the reference point cloud and the corrected second point cloud data is the target point cloud). Empirical values, pre-estimated transformation matrices, or other initialization methods can be used.

[0081] Step a2: Finding Corresponding Point Pairs: For each point in the reference point cloud, establish a correspondence by finding the nearest point in the target point cloud. Euclidean distance or other distance metrics can be used to calculate the distance between two points, and the nearest point is selected as the corresponding point.

[0082] Step a3, Outlier Filtering: By setting a threshold, point pairs whose distance exceeds the threshold are considered outliers, thus excluding their influence on the matching. This can improve the robustness and accuracy of the matching.

[0083] Step a4: Transformation Calculation: Based on the corresponding point pairs, calculate an optimal transformation matrix to align the reference point cloud to the target point cloud. Common transformation matrices include translation, rotation, and scaling. The optimal transformation can be calculated using the least squares method or other optimization methods.

[0084] Step a5, Transformation Update: Apply the transformation matrix to the reference point cloud to obtain a new transformed point cloud. Then repeat steps 2-4 until the termination condition is met. The termination condition can be that the number of iterations reaches the upper limit, or the change in the transformation matrix is ​​less than a threshold, etc.

[0085] Step a6, Output the result: The final transformation matrix is ​​the transformation relationship that best matches the reference point cloud and the target point cloud. This transformation relationship can be used to align the corrected first point cloud data and the corrected second point cloud data to the same coordinate system.

[0086] The ICP algorithm continuously optimizes the matching results by iteratively updating the transformation matrix until the best matching effect is achieved, thus exhibiting high accuracy and robustness.

[0087] Then, all point cloud data aligned to the same coordinate system are treated as a whole and point cloud recognition is performed to obtain the target point cloud set corresponding to the charging pile. The specific implementation method of point cloud recognition can be referred to the detailed description of S203 in this embodiment. For the sake of brevity, it will not be repeated here.

[0088] S205. Determine the target pose of the charging pile based on the third target point cloud set.

[0089] In some embodiments, the lawnmower robot can use the RANSAC algorithm to fit the point cloud data in the third target point cloud set to obtain the center position of the charging pile, which is the position of the charging pile, and determine the orientation of the charging pile based on the contour of the third target point cloud set, thereby obtaining the target pose of the charging pile.

[0090] The following is a detailed fitting process of the RANSAC algorithm:

[0091] Step b1, Random Sampling: Randomly select a set of point clouds from the third target point cloud set of the charging piles as sample points. These samples should be representative enough of the entire dataset, and the number should be large enough to ensure accuracy.

[0092] Step b2, Model Fitting: Fit a model using the selected sample points. In this case, a simple method can be used, such as calculating the geometric center of the selected sample points as the center of the fit.

[0093] Step b3, Interior Point Filtering: For unselected point clouds in the third target point cloud set, classify them as interior or exterior points by calculating their distance from the fitted model (e.g., the center location). Typically, points with a distance less than a predefined threshold are considered interior points, otherwise they are considered exterior points.

[0094] Step b4: Evaluate the model: Calculate the number of interior points as a reference indicator for evaluating the fitted model. The more interior points, the better the fit.

[0095] Step b5, Iteration Process: Repeat the above steps multiple times, selecting the better model (with the most inliers) as the current fitted model each time. An upper limit on the number of iterations or a convergence condition can be set to determine the conditions for stopping the iteration.

[0096] Step b6, Optimal Model Selection: The model with the most inliers is selected as the final fit result throughout all iterations. Reusing inliers to re-estimate model parameters can further improve the fitting accuracy.

[0097] Through the above steps, the RANSAC algorithm can obtain the center position of the charging pile by fitting the point cloud in the third target point cloud set of the charging pile.

[0098] S206. Based on the target pose of the charging pile, plan the return path and control the robot to move along the return path to the location of the charging pile for charging.

[0099] The lawnmower robot determines a local map based on its own pose and the target pose of the charging station. It then uses a path planning algorithm to plan a return path on this local map, starting from its own location and ending at the charging station. The charging station's pose is considered during path planning to provide data for the algorithm, ensuring that the section of the return path closest to the charging station coincides with the extended line of the charging station. This guarantees that the lawnmower robot can move along the return path to face the charging station, thereby improving its recharging efficiency.

[0100] The process involves differentiating the recharge path to calculate the lawnmower robot's direction of movement. Using a set speed, the moving components are controlled to ensure the robot moves along the recharge path to the charging station for charging. The recharge path can be planned using any existing path planning algorithm; this disclosure does not impose any limitations on this method.

[0101] The method in this embodiment guides the robot to return to charging by detecting the location of the charging pile using a lidar mounted on the robot. When identifying the location of the charging pile, the initial point cloud sets of the charging pile in the first initial point cloud data and the second initial point cloud data are respectively processed to complete and / or remove noisy point clouds. In this way, the target point cloud set of the charging pile obtained by point cloud matching and point cloud recognition based on the corrected first initial point cloud data and the corrected second initial point cloud data is more accurate, and the location / pose of the charging pile identified by the robot is more accurate, making the planned return charging path more accurate and improving the return charging efficiency.

[0102] Next, through Figures 3 to 5 The illustrated embodiment details the implementation methods for completing the initial point cloud set of charging piles and removing noisy point clouds. Specifically, Figure 3 and Figure 4 These are two exemplary implementations of completion processing. Figure 5 This is an example of an implementation of noise removal point cloud processing.

[0103] Figure 3 This is a flowchart illustrating a lidar-based robot recharging method according to an embodiment of this disclosure. Please refer to [link / reference]. Figure 3 As shown, the method in this embodiment includes:

[0104] S301. Use the ICP algorithm to perform point cloud matching between the 3D point cloud model of the charging pile and the target initial point cloud set of the charging pile.

[0105] S302. Based on the point cloud matching results, retain one of the overlapping point clouds and combine it with the non-overlapping point clouds to complete the target initial point cloud set.

[0106] The implementation process of point cloud matching using the ICP algorithm and Figure 2 The process of matching the corrected first initial point cloud data and the corrected second initial point cloud data using the ICP algorithm described in the embodiment is similar and can be referred to accordingly. Figure 2 The detailed description in the embodiments differs in that the three-dimensional point cloud model and the target initial point cloud set are used as the target point cloud and the reference point cloud for matching, respectively.

[0107] The ICP algorithm can align the 3D point cloud model and the initial target point cloud set in the same coordinate system. After alignment, some point clouds will overlap, meaning there may be two or more point clouds at a given location. Since the information of the two or more overlapping point clouds is the same, retaining any one of the overlapping point clouds can represent the information of the charging pile at that location. The point clouds to be retained can be pre-set from the 3D point cloud model, pre-set from the target point cloud set, or randomly selected; this disclosure does not impose any limitations on this. Furthermore, retaining one overlapping point cloud can reduce the computational load of subsequent point cloud matching and recognition.

[0108] After performing the above stitching operation, the aligned coordinate system contains the retained point cloud and the non-overlapping point cloud, which means that the initial point cloud set of the target has been completed.

[0109] The method in this embodiment can be understood as completing the charging piles by adding the point cloud from a pre-created, highly accurate 3D point cloud model to the point cloud set of charging piles identified from the first initial point cloud data / second initial point cloud data. The completion process enables the corrected first initial point cloud data to contain richer charging pile information, which helps to improve the accuracy of identifying the location of charging piles.

[0110] Figure 4 A flowchart illustrating a LiDAR-based robot recharging method according to another embodiment of this disclosure. Please refer to [link / reference]. Figure 4 As shown, the method in this embodiment includes:

[0111] S401. Determine the region to be completed based on the initial point cloud set of the charging pile.

[0112] In some embodiments, the lawnmower robot can determine a first shape region of the charging station based on an initial point cloud set of the target. Then, by comparing the standard shape region of the charging station with the first shape region, the area to be completed is determined.

[0113] Among them, the lawnmower robot can identify the point cloud on the edge of the charging pile from the initial point cloud set of the target, and connect these point clouds representing the edge to obtain a closed space, which is the first outer shape area of ​​the charging pile.

[0114] By comparing the standard shape area and the first shape area of ​​the charging pile, the first enclosed space can be rotated, scaled, or otherwise processed to cover the standard shape area of ​​the charging pile. The area not covered in the standard shape area is the area to be filled in.

[0115] S402. Determine the point cloud data corresponding to the area to be completed in the three-dimensional point cloud model of the charging pile.

[0116] Whether a region belongs to the area to be completed can be determined based on the position of each point cloud in the 3D point cloud model of the charging pile, thereby obtaining the point cloud data corresponding to the area to be completed.

[0117] S403. Use the point cloud data corresponding to the area to be completed in the 3D point cloud model to complete the target initial point cloud set of the charging pile.

[0118] The lawnmower adds the point cloud data of the area to be completed to the target initial point cloud set of the charging pile, thereby completing the completion. Through the completion process, the corrected first initial point cloud data can contain richer charging pile information, which helps to improve the accuracy of identifying the location of the charging pile.

[0119] Figure 5 A flowchart illustrating a LiDAR-based robot recharging method according to another embodiment of this disclosure. Please refer to [link / reference]. Figure 5 As shown, the method in this embodiment includes:

[0120] S501. Determine the initial center position of the charging pile based on the initial point cloud set of the target charging pile.

[0121] In some embodiments, the lawnmower robot uses the RANSAC algorithm to fit the point cloud in the initial target point cloud set to obtain the initial center position. The detailed fitting process is similar to the process described above for fitting the target position of the charging station using the RANSAC algorithm, and can be referred to... Figure 2 The detailed description of the illustrated embodiments will not be repeated here for the sake of brevity.

[0122] S502. Set the center position of the three-dimensional point cloud model of the charging pile to the initial center position, and remove the point cloud data outside the outline of the three-dimensional point cloud model.

[0123] The center position of the 3D point cloud model is set at the initial center position obtained by fitting in S501, and the orientation is kept consistent. Next, the outline of the 3D point cloud model is used as the dividing line. Point clouds outside the outline of the 3D point cloud model are considered noise point clouds, while point clouds inside the outline of the 3D point cloud model are considered valid point clouds. In this way, all noise point cloud data is deleted to complete the noise point cloud removal.

[0124] The method in this embodiment reduces the impact of noise point clouds by removing them, thereby improving the accuracy of the corrected initial point cloud and improving the accuracy of identifying the location of charging piles.

[0125] Figure 6 This is a schematic diagram of a LiDAR-based robot recharging device according to an embodiment of this application. Please refer to... Figure 6 As shown, the robot recharging device 600 based on LiDAR provided in this embodiment includes: a data acquisition module 601, a path planning module 602, and a motion control module 603.

[0126] The data acquisition module 601 is used to acquire the first initial point cloud data and the second initial point cloud data detected by the lidar on the robot during its movement.

[0127] The data processing module 601 is further configured to perform point cloud recognition on the first initial point cloud data to obtain a first initial point cloud set of the charging pile, and to perform completion processing and / or noise point cloud removal processing on the first initial point cloud set to obtain a first target point cloud set.

[0128] The data processing module 601 is further configured to perform point cloud recognition on the second initial point cloud data to obtain the second initial point cloud set of the charging pile, and to perform completion processing and / or noise point cloud removal processing on the second initial point cloud set to obtain the second target point cloud set.

[0129] The data processing module 601 is further configured to perform point cloud matching and point cloud recognition based on the first target point cloud set, the third initial point cloud set, the second target point cloud set, and the fourth initial point cloud set to obtain the third target point cloud set corresponding to the charging pile; wherein, the third initial point cloud set includes other point clouds in the first initial point cloud data other than the first initial point cloud set, and the fourth initial point cloud set includes other point clouds in the second initial point cloud data other than the second initial point cloud set.

[0130] The data processing module 601 is further configured to determine the target pose of the charging pile based on the third target point cloud set.

[0131] The path planning module 602 is used to plan the recharge path according to the target pose of the charging pile.

[0132] The motion control module 603 is used to control the robot to move along the recharge path to the location of the charging pile for charging.

[0133] In some embodiments, the data processing module 601 is specifically used to perform completion processing and / or noise point cloud removal processing on the target initial point cloud set of the charging pile using a preset three-dimensional point cloud model of the charging pile, wherein the target initial point cloud set is the first initial point cloud set or the second initial point cloud set.

[0134] In some embodiments, the data processing module 601 is specifically used to perform point cloud matching on the three-dimensional point cloud model of the charging pile and the target initial point cloud set of the charging pile using the ICP algorithm. Based on the point cloud matching result, one of the overlapping point clouds is retained and combined with the non-overlapping point cloud to complete the target initial point cloud set.

[0135] In some embodiments, the data processing module 601 is specifically used to determine the region to be completed based on the target initial point cloud set of the charging pile; determine the point cloud data corresponding to the region to be completed in the three-dimensional point cloud model of the charging pile; and use the point cloud data corresponding to the region to be completed in the three-dimensional point cloud model to complete the target initial point cloud set of the charging pile.

[0136] In some embodiments, the data processing module 601 is specifically used to obtain a first shape region of the charging pile based on the target initial point cloud set of the charging pile; map the first shape region of the charging pile to a standard shape region of the charging pile; and determine the region in the standard shape region that is not covered by the first shape region as the region to be completed.

[0137] In some embodiments, the data processing module 601 is specifically used to determine the initial center position corresponding to the charging pile based on the target initial point cloud set of the charging pile; set the center position of the three-dimensional point cloud model of the charging pile at the initial center position; and remove point cloud data outside the contour of the three-dimensional point cloud model.

[0138] In some embodiments, the data processing module 601 is specifically used to fit the point cloud data in the target initial point cloud set of the charging pile using the RANSAC algorithm to obtain the initial point cloud position of the charging pile.

[0139] The apparatus provided in this embodiment can be used to implement the technical solutions of any of the foregoing method embodiments. Its implementation principle and technical effects are similar and can be referred to in the detailed description above. For the sake of brevity, it will not be repeated here.

[0140] Figure 7 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this disclosure. Please refer to [link / reference]. Figure 7 As shown, the electronic device 700 provided in this embodiment includes a memory 701 and a processor 702.

[0141] The memory 701 can be a separate physical unit, connected to the processor 702 via a bus 703. Alternatively, the memory 701 and processor 702 can be integrated and implemented in hardware. The memory 701 stores program instructions, which the processor 702 calls to execute the LiDAR-based robot recharging method provided in any of the above method embodiments.

[0142] Optionally, when some or all of the methods in the above embodiments are implemented by software, the vehicle terminal 700 may also include only the processor 702. A memory 701 for storing programs is located outside the vehicle terminal 700, and the processor 702 is connected to the memory via circuits / wires to read and execute the programs stored in the memory. The processor 702 may be a central processing unit (CPU), a network processor (NP), or a combination of a CPU and an NP. The processor 702 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0143] The memory 701 may include volatile memory, such as random-access memory (RAM); the memory may also include non-volatile memory, such as flash memory, hard disk drive (HDD) or solid-state drive (SSD); the memory may also include a combination of the above types of memory.

[0144] For example, this disclosure provides a chip system applied to a lawnmower robot including a memory and at least one lidar; the chip system includes a processor; the processor is used to execute the lidar-based robot recharging method of the preceding embodiments.

[0145] For example, this disclosure provides a lawnmower robot, including: a memory, a processor, and at least one lidar; the lidar is used to scan and obtain point cloud data, the memory is used to store the point cloud data and computer program instructions, and the processor is used to run the computer program instructions to implement the lidar-based robot recharging method of the previous embodiment based on the point cloud data obtained by the lidar scanning.

[0146] For example, this disclosure provides a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor of an electronic device so that the electronic device, when executed, implements the lidar-based robot recharging method of the preceding embodiments.

[0147] For example, this disclosure provides a computer program product that, when run on an electronic device, causes the electronic device to execute the LiDAR-based robot recharging method described in the preceding embodiments.

[0148] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0149] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A robot recharging method based on lidar, characterized in that, include: Acquire the first and second initial point cloud data detected by the lidar on the robot during its movement; Point cloud recognition is performed on the first initial point cloud data to obtain the first initial point cloud set of the charging pile. The first initial point cloud set is then filled and / or noise point cloud is removed to obtain the first target point cloud set. Point cloud recognition is performed on the second initial point cloud data to obtain the second initial point cloud set of the charging pile. The second initial point cloud set is then filled and / or noise point cloud is removed to obtain the second target point cloud set. Point cloud matching and point cloud identification are performed based on the first target point cloud set, the third initial point cloud set, the second target point cloud set, and the fourth initial point cloud set to obtain the third target point cloud set corresponding to the charging pile; wherein, the third initial point cloud set includes other point clouds in the first initial point cloud data other than the first initial point cloud set, and the fourth initial point cloud set includes other point clouds in the second initial point cloud data other than the second initial point cloud set; The target pose of the charging pile is determined based on the third target point cloud set; Based on the target pose of the charging pile, a return path is planned, and the robot is controlled to move along the return path to the location of the charging pile for charging.

2. The method according to claim 1, characterized in that, The process of completing and / or removing noisy point clouds from the initial target point cloud set of the charging pile includes: Using a preset 3D point cloud model of the charging pile, the target initial point cloud set of the charging pile is subjected to completion processing and / or noise point cloud removal processing, wherein the target initial point cloud set is either the first initial point cloud set or the second initial point cloud set.

3. The method according to claim 2, characterized in that, The step of using a preset 3D point cloud model of the charging pile to perform completion processing and / or noise point cloud removal processing on the target initial point cloud set of the charging pile includes: The iterative nearest point (ICP) algorithm is used to perform point cloud matching between the 3D point cloud model of the charging pile and the target initial point cloud set of the charging pile. Based on the point cloud matching results, one of the overlapping point clouds is retained and combined with the non-overlapping point clouds to complete the target initial point cloud set.

4. The method according to claim 2, characterized in that, The step of using a preset 3D point cloud model of the charging pile to perform completion processing and / or noise point cloud removal processing on the target initial point cloud set of the charging pile includes: The area to be completed is determined based on the initial point cloud set of the charging piles. Determine the point cloud data corresponding to the region to be completed in the three-dimensional point cloud model of the charging pile; The initial target point cloud set of the charging pile is completed using the point cloud data corresponding to the region to be completed in the three-dimensional point cloud model.

5. The method according to claim 4, characterized in that, The determination of the region to be completed based on the initial point cloud set of the charging pile includes: The first shape region of the charging pile is obtained based on the initial point cloud set of the target of the charging pile; The first shape area of ​​the charging pile is mapped to the standard shape area of ​​the charging pile, and the area in the standard shape area that is not covered by the first shape area is determined as the area to be completed.

6. The method according to claim 2, characterized in that, The step of using a preset 3D point cloud model of the charging pile to perform completion processing and / or noise point cloud removal processing on the target initial point cloud set of the charging pile includes: The initial center position of the charging pile is determined based on the target initial point cloud set of the charging pile. The center position of the three-dimensional point cloud model of the charging pile is set at the initial center position, and the point cloud data outside the outline of the three-dimensional point cloud model is removed.

7. The method according to claim 6, characterized in that, The process of determining the initial center position of the charging pile based on the target initial point cloud set of the charging pile includes: The initial point cloud position of the charging pile is obtained by fitting the point cloud data of each point cloud in the target initial point cloud set using the Random Sampling Consensus (RANSAC) algorithm.

8. An electronic device, characterized in that, include: Memory and processor; The memory is configured to store computer program instructions; The processor is configured to execute the computer program instructions, causing the electronic device to implement the method as described in any one of claims 1 to 7.

9. A readable storage medium, characterized in that, include: Computer program instructions; The electronic devices on the robot execute the computer program instructions, causing the electronic devices to perform the method as described in any one of claims 1 to 7.

10. A lawnmower robot, characterized in that, include: At least one lidar, a memory, and a processor; wherein, in the lawnmower recharging scenario, the at least one lidar is used to detect and obtain initial point cloud data during the movement of the lawnmower; The memory is used to store the initial point cloud data detected by the at least one lidar; the memory also stores computer program instructions; the processor executes the computer program instructions to perform the method as described in any one of claims 1 to 7 based on the initial point cloud data detected by the at least one lidar.

Images