A method and device for detecting dust on the surface of a lidar

Abstract

The present application provides a method and device for detecting dust on the surface of a lidar. The dust detection method for the lidar surface in the present application can respectively determine whether each laser detection point of the lidar is a dust point based on point cloud data, and on this basis, evaluate the dust severity using the distribution of dust points for corresponding processing, thereby overcoming the problem that the obstacle detection accuracy of the lidar is affected by dust attached to the surface, and providing guarantee for the driving safety of a mobile robot in a complex environment. Moreover, compared with the discrimination method of directly determining whether to perform dust treatment based on the presence or absence of dust points, the method adopted in the present application of determining dust statistical information based on the dust distribution and determining the corresponding treatment plan based on the dust statistical information is based on more comprehensive judgment information, so that on the basis of ensuring the driving safety of the mobile robot, false alarm situations can be further reduced, and the influence of the dust environment on the operation efficiency of the mobile robot can be reduced.

Description

Technical Field

[0001] This application relates to sensor detection technology, and particularly to a method and device for detecting dust on the surface of a lidar. Background Art

[0002] Obstacle detection based on lidar is a key technology in mobile robots. To avoid collisions between mobile robots and people or objects, continuous obstacle detection and positioning are required during their driving, so as to perform obstacle avoidance driving according to the detected obstacle positions.

[0003] However, when a mobile robot is driving in a complex environment, dust may adhere to the surface of the lidar, affecting the accuracy of radar detection data, resulting in the mobile robot being unable to timely and accurately perceive obstacle objects in the environment and thus prone to collisions, which is not conducive to the driving safety of the mobile robot. Summary of the Invention

[0004] Embodiments of this application provide a method and device for detecting dust on the surface of a lidar, which solve the problem that dust on the surface of the lidar affects the driving safety of a mobile robot by judging the dust situation on the surface of the lidar and performing corresponding processing.

[0005] In a first aspect, embodiments of this application provide a method for detecting dust on the surface of a lidar, which is applied to a mobile robot including a lidar. The lidar includes a plurality of laser detection points, and the method includes:

[0006] Obtain the point cloud data collected by the lidar, and determine dust points from the plurality of laser detection points based on the point cloud data. The dust points are laser detection points with dust on the surface;

[0007] Determine dust statistical information based on the distribution of all dust points, and determine the dust processing method of the lidar based on the dust statistical information;

[0008] Process the dust on the surface of the lidar based on the dust processing method.

[0009] In a possible implementation manner, the determining the dust points from the plurality of laser detection points based on the point cloud data includes:

[0010] For each laser detection point, determine the laser intensity and the target object distance corresponding to the laser detection point based on the point cloud data;

[0011] Determine a first confidence level corresponding to the above laser detection point based on the above laser intensity, determine a second confidence level corresponding to the above laser detection point based on the distance of the above target object, and determine a third confidence level corresponding to the above laser detection point based on the distance of the target object corresponding to the above laser detection point and the distance of the target object corresponding to the adjacent laser detection point of the above laser detection point;

[0012] Determine a target confidence level based on the above first confidence level, the above second confidence level, and the above third confidence level;

[0013] If the above target confidence level is greater than a preset confidence level threshold, determine the above laser detection point as a dust point, otherwise, determine the above laser detection point as a non-dust point.

[0014] In a possible implementation, the above dust statistical information includes the continuous number of dust points. Determining the dust statistical information based on the distribution of all dust points includes:

[0015] Select dust points with continuous positions from all dust points based on the distribution of all dust points;

[0016] Determine the continuous number of dust points based on the number of dust points with continuous positions.

[0017] In a possible implementation, the above dust statistical information includes the angle proportion of dust points. Determining the dust statistical information based on the distribution of all dust points includes:

[0018] Normalize the distribution of all the above dust points according to the angle distribution to a preset angle value, and determine the angle value occupied by the dust points with continuous positions based on the normalized distribution of all dust points;

[0019] Determine the angle proportion of dust points based on the above preset angle value and the angle value occupied by the dust points with continuous positions.

[0020] In a possible implementation, the above dust statistical information includes the continuous number of dust points. Determining the dust processing method of the above lidar based on the above dust statistical information includes:

[0021] If the continuous number of the above dust points is greater than a preset continuous number threshold, determine the dust processing method of the above lidar as a first dust processing method, and the first dust processing method is used to instruct the above mobile robot to stop for inspection;

[0022] If the continuous number of the above dust points is not greater than a preset continuous number threshold, determine the dust processing method of the above lidar as a second dust processing method, and the second dust processing method is used to indicate that there is a dust risk on the surface of the above lidar.

[0023] In a possible implementation, the above dust statistical information includes the continuous number of dust points and the total number of dust points. Determining the dust processing method of the lidar based on the above dust statistical information includes:

[0024] If the continuous number of dust points is greater than a preset continuous number threshold, and the total number of dust points is greater than a preset total number threshold, then determine that the dust processing method of the lidar is the first dust processing method, and the first dust processing method is used to instruct the mobile robot to stop for inspection;

[0025] If the continuous number of dust points is not greater than the preset continuous number threshold, and / or the total number of dust points is not greater than the preset total number threshold, then determine that the dust processing method of the lidar is the second dust processing method, and the second dust processing method is used to indicate a warning of a dust risk on the surface of the lidar.

[0026] In a possible implementation, the above dust statistical information includes the continuous number of dust points and the total number of dust points. Determining the dust processing method of the lidar based on the above dust statistical information includes:

[0027] Determine a detection identifier corresponding to the dust statistical information at any moment within the statistical period based on the dust statistical information at that moment; wherein, if the continuous number of dust points in the dust statistical information at that moment is greater than a preset continuous number threshold, and the total number of dust points is greater than a preset total number threshold, then the detection identifier is the first value; if the continuous number of dust points is not greater than the preset continuous number threshold, and / or the total number of dust points is not greater than the preset total number threshold, then the detection identifier is the second value;

[0028] Obtain the detection identifiers corresponding to the dust statistical information at all moments within the statistical period, and count the first number of the detection identifiers that are the first value;

[0029] If the first number is greater than a preset threshold, then determine that the dust processing method of the lidar is the first dust processing method, and the first dust processing method is used to instruct the mobile robot to stop for inspection;

[0030] If the first number is not greater than the preset threshold, then determine that the dust processing method of the lidar is the second dust processing method, and the second dust processing method is used to indicate a warning of a dust risk on the surface of the lidar.

[0031] In a possible implementation, after determining that the dust processing method of the lidar is the second dust processing method, the method further includes:

[0032] Control the above mobile robot to avoid obstacles and drive based on the point cloud data corresponding to the laser detection points among the above multiple laser detection points that are not identified as dust points.

[0033] In a second aspect, an embodiment of the present application further provides a dust detection device for the surface of a lidar, which is applied to a mobile robot including a lidar. The above lidar includes a plurality of laser detection points. The device includes:

[0034] A dust point determination unit, configured to obtain the point cloud data collected by the above lidar, and determine dust points from the above multiple laser detection points based on the above point cloud data. The above dust points are laser detection points with dust on the surface;

[0035] A processing method determination unit, configured to determine dust statistical information based on the distribution of all dust points, and determine the dust processing method of the above lidar based on the above dust statistical information;

[0036] A dust processing unit, configured to process the dust on the surface of the above lidar based on the above dust processing method.

[0037] In a third aspect, an embodiment of the present application further provides an electronic device, which includes: a processor and a machine-readable storage medium;

[0038] The above machine-readable storage medium stores machine-executable instructions that can be executed by the above processor;

[0039] The above processor is used to execute the machine-executable instructions to implement the above method for detecting dust on the surface of the lidar.

[0040] In a fourth aspect, an embodiment of the present application further provides a machine-readable storage medium. The above machine-readable storage medium stores machine-readable instructions. When the above machine-readable instructions are called and executed by a processor, the above machine-readable instructions cause the above processor to implement the above method for detecting dust on the surface of the lidar.

[0041] It can be seen from the above technical solutions that the method for detecting dust on the surface of the lidar in this embodiment can respectively determine whether each laser detection point of the lidar is a dust point based on the point cloud data, and on this basis, evaluate the severity of the dust using the distribution of the dust points for corresponding processing, thereby overcoming the problem that the obstacle detection accuracy of the lidar is affected by dust attached to the surface, and providing guarantee for the driving safety of the mobile robot in a complex environment.

[0042] Furthermore, compared with the discrimination method of directly determining whether to perform dust treatment based on the presence or absence of dust spots, the method adopted in this embodiment of determining dust statistical information based on the dust distribution and determining the corresponding treatment plan based on the dust statistical information is based on more comprehensive judgment information, so that on the basis of ensuring the driving safety of the mobile robot, false alarms can be further reduced, and the impact of the dust environment on the operation efficiency of the mobile robot can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The accompanying drawings herein are incorporated into and constitute a part of this specification, showing embodiments consistent with the present disclosure and, together with the specification, are used to explain the principles of the present disclosure.

[0044] Figure 1 It is a flowchart of the method provided by the embodiment of the present application;

[0045] Figure 2 It is a flowchart of dust spot judgment provided by the embodiment of the present application;

[0046] Figure 3 It is a flowchart of dust treatment method judgment provided by the embodiment of the present application;

[0047] Figure 4 It is a structure diagram of the device provided by the embodiment of the present application;

[0048] Figure 5 It is a structure diagram of the electronic device provided by the embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] Exemplary embodiments will be described in detail herein, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0050] The terms used in the present application are for the purpose of describing particular embodiments only and are not intended to limit the present application. The singular forms "a", "the", and "said" used in the present application and the appended claims are also intended to include the plural forms unless the context clearly dictates otherwise.

[0051] In order to enable those skilled in the art to better understand the technical solutions provided by the embodiments of the present application and to make the above-mentioned objects, features, and advantages of the embodiments of the present application more apparent and understandable, the technical solutions in the embodiments of the present application will be further described in detail below with reference to the accompanying drawings.

[0052] See Figure 1 , Figure 1This is a flowchart of the method provided by the embodiments of the present application. This process can be applied to any mobile robot equipped with a lidar, such as a logistics robot, a cleaning robot, etc. The above lidar contains multiple laser detection points; in this embodiment, the types, uses of the mobile robot, and the models of the lidar are not limited.

[0053] As Figure 1 shown, this process may include the following steps:

[0054] Step 101, obtain the point cloud data collected by the lidar, and determine the dust points from the above-mentioned multiple laser detection points based on the above point cloud data.

[0055] In this embodiment, the above point cloud data is the data collected by the lidar installed on the mobile robot. This data may include information such as the reflection intensity, distance, and angle of each laser point in the radar. The above information comes from each laser detection point in the radar respectively.

[0056] In this embodiment, dust points, that is, laser detection points with dust on the surface, can be selected from the multiple laser detection points included in the lidar based on the above point cloud data for subsequent operations such as selecting the dust treatment method. Specifically, it is possible to determine whether a point is a dust point based on the point cloud data of a certain laser detection point, or it is also possible to further comprehensively judge in combination with the point cloud data of other laser detection points adjacent to this detection point, etc. This embodiment does not limit the specific determination method of the dust points; subsequent will be combined with Figure 2 the shown process to exemplarily give one or more optional implementation manners, which will not be elaborated here for the time being.

[0057] Step 102, determine the dust statistics information based on the distribution of all dust points, and determine the dust treatment method of the above lidar based on the above dust statistics information.

[0058] In this embodiment, after identifying the dust points from all the laser detection points of the lidar, the dust statistics information that can measure the severity of the dust attached to the surface of the lidar can be determined based on the distribution of the identified dust points, such as the number, position, or adjacent relationship, etc., and the dust treatment method can be decided based on this dust statistics information, so that the mobile robot can adopt a matching treatment method for different degrees of dust attachment severity.

[0059] It should be noted that there are multiple optional implementation manners for the specific content, determination method of the dust statistics information, and the corresponding relationship between the dust statistics information and the dust treatment method. This embodiment does not limit this; subsequent will be combined with Figure 3 the shown process to exemplarily give one or more specific solutions, which will not be elaborated here for the time being.

[0060] Step 103: Process the dust on the surface of the lidar based on the above dust processing method.

[0061] In this embodiment, the dust adhering to the surface of the lidar carried by the mobile robot can be processed accordingly based on the dust processing method determined in step 102, thereby solving the problems that dust affects the accuracy of radar obstacle detection and the driving safety of the mobile robot, and providing a safety guarantee for the mobile robot to perform driving tasks in a complex environment.

[0062] As an optional embodiment, when it is determined that the dust on the radar surface is relatively serious, resulting in the data of certain angles of the lidar being blocked and the environmental information of the corresponding positions cannot be detected, it is considered that the current dust condition seriously affects the driving safety of the mobile robot. The first dust processing method for instructing the mobile robot to stop for inspection can be selected, so that the mobile robot stops to inspect the radar and clean the dust, etc.; when it is determined that there is a certain amount of dust on the radar surface, resulting in inaccurate ranging of some laser detection points but not enough to affect the driving of the mobile robot, the second dust processing method for indicating the risk of dust on the warning lidar surface can be selected, so that the mobile robot issues corresponding warning information to prompt the user or equipment maintenance personnel that there is dust on the radar surface and needs to be cleaned in time, etc.

[0063] As a preferred embodiment, after it is determined that the dust processing method is the above-mentioned second dust processing method, the mobile robot can be further controlled to avoid obstacles based on the point cloud data corresponding to the laser detection points not marked as dust points, that is, the data collected by the laser detection points that have been confirmed to be covered by dust is excluded, and obstacle avoidance driving is performed based on the data collected by the remaining laser detection points not covered by dust. This enables the mobile robot to exclude the interference of inaccurate data and maintain the driving state when the dust severity is relatively low, avoiding the mobile robot from frequently stopping due to a small amount of dust while ensuring safety, effectively reducing the impact of the dust environment on the operation efficiency of the mobile robot, and also avoiding the situation where the mobile robot stops due to false alarms of a small number of dust points.

[0064] So far, the Figure 1 shown process is completed.

[0065] Through Figure 1 As can be seen from the shown process, the dust detection method for the surface of the lidar in this embodiment can respectively judge whether each laser detection point of the lidar is a dust point based on the point cloud data, and on this basis, evaluate the dust severity using the distribution of the dust points for corresponding processing, thereby overcoming the problem that the accuracy of obstacle detection of the lidar is affected by the dust adhering to the surface, and providing a guarantee for the driving safety of the mobile robot in a complex environment.

[0066] Furthermore, compared with the discrimination method of directly determining whether to perform dust treatment based on the presence or absence of dust points, the method adopted in this embodiment of determining dust statistical information based on dust distribution and determining the corresponding treatment plan based on the dust statistical information is based on more comprehensive judgment information, so that on the basis of ensuring the driving safety of the mobile robot, false alarms can be further reduced, and the influence of the dust environment on the operation efficiency of the mobile robot can be reduced.

[0067] To enable those skilled in the art to better understand the technical solutions provided in the embodiments of the present application, the embodiments of the present application also provide a Figure 2 , Figure 3 shown flowchart to implement the above methods disclosed in the embodiments of the present application.

[0068] As an optional embodiment, the method of determining dust points from multiple laser detection points based on point cloud data in step 101 above can be implemented with reference to the Figure 2 shown process; specifically, the process may include the following steps:

[0069] Step 201, obtain lidar point cloud data.

[0070] In this embodiment, it is necessary to obtain the point cloud data collected by the lidar carried by the mobile robot for subsequent judgment of the dust condition on the lidar surface based on this data.

[0071] Step 202, object edge extraction.

[0072] As an optional embodiment, before judging dust points, edge extraction can be performed on the point cloud data first, and operations such as segmenting and filtering the point cloud on the surface of the object detected by the lidar from the background can be performed to obtain the object edge contour, so as to improve the subsequent recognition efficiency and accuracy.

[0073] Step 203, determine the first confidence level based on the laser intensity.

[0074] In this embodiment, since dust will interfere with the detection results of the lidar and cause "flying points" in the point cloud data that are significantly different from normal detection points in terms of reflection intensity or distance, etc., the possibility that a point is a dust point can be comprehensively judged based on the data corresponding to each laser detection point in the point cloud.

[0075] As an optional embodiment, a first confidence level can be determined based on the intensity of the laser points at the edge of the object. The above first confidence level is used to represent the likelihood that the point is a dust point when judged by the laser intensity. Optionally, since different models of lidar have differences in the detection or value range of the reflection intensity, the laser intensity values in the point cloud data can be pre-equalized to the range of 0 to 255. Based on this, the first confidence level is determined according to the preset correspondence between the laser intensity and the confidence level. Among them, under the same other conditions, the smaller the laser intensity corresponding to a certain laser detection point, the greater the likelihood that the point is a dust point, and the relatively higher the corresponding first confidence level.

[0076] Step 204: Determine a second confidence level based on the point cloud distance.

[0077] As an optional embodiment, a second confidence level can be determined based on the distance information in the point cloud data corresponding to the above laser detection point. The above second confidence level is used to represent the likelihood that the point is a dust point when judged by the point cloud distance. Among them, the point cloud distance is the distance between the point detected by the above laser detection point and the target object; under the same other conditions, the smaller the point cloud distance corresponding to a certain laser detection point, the closer the reflection position of the laser point is to the lidar, the greater the likelihood that the point is a dust point on the lidar surface, and the relatively higher the corresponding second confidence level.

[0078] Step 205: Determine a third confidence level based on the continuous angle of the point cloud.

[0079] As an optional embodiment, the third confidence level can also be determined by combining the point cloud data corresponding to other laser detection points adjacent to the above laser detection point. The above third confidence level can be used to represent the likelihood that the point is a dust point when judged by the continuous angle of the point cloud. Among them, the continuous angle of the point cloud is the included angle range corresponding to the continuous laser points in the point cloud within the detection range of the lidar, which is used to judge whether a certain laser point and other adjacent laser points are on the same surface, and to represent the size of the surface where the point is located, etc.; under the same other conditions, the smaller the continuous angle of the point cloud corresponding to a certain laser detection point, the smaller the plane where the laser point is located, the more obvious the position mutation with other surrounding laser points, the greater the likelihood that the point is a dust point on the lidar surface, and the relatively higher the corresponding third confidence level.

[0080] Among them, there are various ways to determine the continuous angle of the above-mentioned point cloud. For example, the relationship between the distance of the target object corresponding to a certain laser detection point and the distances of the target objects corresponding to other laser detection points adjacent to this point can be compared to determine whether this point and its adjacent points are on the same continuous plane. That is, if the distances of the target objects at a certain point and its adjacent points are basically equal, it indicates that this point and its adjacent points are on the same continuous plane; the angle at which the plane where this point is located is mapped in the above-mentioned lidar can be used as the continuous angle of the above-mentioned point cloud. Optionally, the above-mentioned third confidence level can also be directly determined based on the distances of the target objects of the above-mentioned detection point and its adjacent detection points, etc. This embodiment does not limit this.

[0081] As a preferred embodiment, in combination with the above steps 203 to 205, for each laser detection point, the laser intensity and the distance of the target object corresponding to this laser detection point can be determined based on the acquired point cloud data. On this basis, the first confidence level corresponding to this laser detection point is determined based on the laser intensity, the second confidence level corresponding to this laser detection point is determined based on the distance of the target object, and the third confidence level corresponding to this laser detection point is determined based on the distance of the target object corresponding to this laser detection point and the distances of the target objects corresponding to other laser detection points adjacent to this laser detection point.

[0082] Step 206, comprehensively judge whether it is a dust point based on each confidence level; if so, execute step 207, otherwise execute step 208;

[0083] Step 207, set the attribute of this point to the dust point attribute;

[0084] Step 208, set the attribute of this point to the non-dust point attribute.

[0085] In this embodiment, based on the first confidence level, the second confidence level, and the third confidence level determined in the above steps 203 to 205, the target confidence level can be obtained by means of accumulation or weighted calculation, etc., and the target confidence level is compared with a preset confidence level threshold. If the target confidence level is greater than the preset confidence level threshold, the laser detection point corresponding to this target confidence level can be determined as a dust point, and its attribute is set to the dust point attribute; otherwise, this point cannot be determined as a dust point, and its attribute can be set to the non-dust point attribute.

[0086] It should be noted that the above content such as the basis for determining the confidence level and the number of confidence level items in this embodiment are only optional embodiments. In actual use, the number of confidence level items can be increased or decreased, and the calculation basis of each confidence level can be changed according to factors such as the lidar parameter acquisition ability or actual requirements. This embodiment does not limit this.

[0087] So far, the description of Figure 2 the shown process example is completed.

[0088] As an optional embodiment, the method for determining dust statistical information based on the distribution of all dust points in step 101 above may be implemented with reference to the process shown in Figure 3 The specific process may include the following steps:

[0089] Step 301: Obtain dust point mask information.

[0090] In this embodiment, with reference to step 101 and the process shown in Figure 2 The dust points are determined from multiple laser detection points of the lidar, and the dust point mask information is obtained. The dust point mask information is information used to represent the distribution of the dust points of the lidar, which may include, for example, the positions, adjacency relationships, angular ratios, and total number of dust points of each detection point determined as a dust point among the multiple laser detection points.

[0091] Step 302: Normalize the angular distribution to 360°.

[0092] As an optional embodiment, when the dust statistical information used to determine the dust processing method in subsequent steps includes angular-related content, the obtained mask information may be normalized to a preset angular value, such as 360°, according to the angular distribution in advance to unify the detection angle range differences between different models of lidar.

[0093] Step 303A: Determine whether the angular ratio of the dust points is greater than a certain threshold. If not, execute step 304; if so, execute step 305; or make a comprehensive judgment in combination with other optional determination conditions in the subsequent steps.

[0094] As an optional embodiment, the dust statistical information may include the angular ratio of the dust points. Among them, there is a certain included angle between two adjacent laser detection points in the same detection plane of the lidar. The included angle corresponding to a plurality of continuously distributed laser detection points determined as dust points may be called the dust point angle, and the ratio of the dust point angle in the preset angular value after the above normalization is called the angular ratio of the dust points.

[0095] For example, after the angular distribution of a certain type of lidar is normalized to 360°, a 5° angle is formed between two adjacent laser detection points in the same plane. Then, when 5 adjacent laser detection points horizontally or vertically are all dust points, the corresponding dust point angle is 20°, and the proportion of the dust point angle is 20° / 360° = 5.56%. It should be noted that the above values are only provided exemplarily. In actual applications, the angles between the laser detection points of the lidar may not be evenly distributed. For example, the detection points in the direction directly in front of the lidar are relatively dense, and the corresponding angle between adjacent points is relatively smaller. The detection points on both sides of the lidar are relatively sparse, and the corresponding angle between adjacent points is relatively larger, and so on.

[0096] On this basis, the maximum value of the proportion of the dust point angle in the lidar can be determined based on the dust point distribution of the lidar, and this value is compared with a preset angle proportion threshold. If it is less than the preset angle proportion threshold, it indicates that the dust adhesion on the lidar surface is not serious, and the dust treatment method in step 304 can be executed; otherwise, it indicates that the dust adhesion is serious, and the dust treatment method in step 305 needs to be executed.

[0097] As an optional embodiment, the above dust statistical information can also be the continuous number of dust points. Based on the distribution of all dust points, the dust points with continuous positions are selected from all dust points, and the continuous number of dust points is determined based on the number of dust points with continuous positions. For example, when it is determined that in the lidar, at most 10 adjacent laser detection points horizontally or vertically are dust points, the continuous number of dust points is determined to be 10.

[0098] On this basis, the determined continuous number of dust points can be compared with a preset continuous number threshold. If it is less than the preset continuous number threshold, it indicates that the dust adhesion on the lidar surface is not serious, and the dust treatment method in step 304 can be executed; otherwise, it indicates that the dust adhesion on the radar surface is serious, and the dust treatment method in step 305 can be executed.

[0099] It should be noted that, under the condition that the total number of dust points is equal, the denser the dust points are distributed and the more adjacent dust points there are, the easier it is for the data at certain angles of the laser radar to be completely blocked by dust, making it impossible to detect environmental information, making the dust risk more serious than when the dust points are scattered; the dust point angle ratio or the continuous number of dust points included in the above-mentioned dust statistical information are used to judge the severity of dust from the perspective of whether the dust points are distributed in a clustered manner when other conditions such as the total number of dust points are the same. In subsequent examples and actual use, they can be selected or replaced according to needs, such as when the angles between all detection points of the laser radar are the same or similar, the continuous number of dust points is used for judgment, and when the density of detection points directly in front of the laser radar and on both sides is quite different, the dust point angle ratio is used for judgment, and so on. This embodiment does not limit this.

[0100] Step 304, issue a warning message, remove the data corresponding to the dust point, and perform obstacle avoidance driving.

[0101] In this embodiment, when it is determined that the dust adhesion on the surface of the laser radar is not serious and the mobile robot can continue to travel, a second dust handling method can be used to indicate and warn that there is a dust risk on the surface of the laser radar, so that the mobile robot sends an alarm message to remind the user or equipment maintenance personnel that there is dust on the radar surface and that it needs to be cleaned in time.

[0102] As a preferred embodiment, since some of the radar's laser detection points have been determined to be dust points and the mobile robot can still maintain a driving state, the mobile robot can be controlled to perform obstacle avoidance driving based on the point cloud data corresponding to the laser detection points that are not identified as dust points among the above-mentioned multiple laser detection points, and the data corresponding to the dust points are eliminated to reduce dust interference, ensuring that the data used for obstacle avoidance are accurate data that are not affected by dust.

[0103] Step 305: the mobile robot stops moving.

[0104] In this embodiment, when it is judged that the dust adhesion on the surface of the laser radar is relatively serious and the safety hazard of the mobile robot continuing to drive is relatively large, the first dust processing method for instructing the mobile robot to stop for inspection can be adopted, that is, the mobile robot stops driving and waits for the user or maintenance personnel to check or wipe the dust on the radar surface to avoid the collision risk caused by continuing to drive when the reliability of environmental data detection is low.

[0105] It should be noted that the above two dust handling methods are only provided by way of example. In actual use, other dust handling methods can be replaced or supplemented according to requirements. For example, a cleaning device can be installed beside the mobile robot radar to automatically wipe or spray the dust when severe dust is detected, or the speed of the mobile robot can be limited according to the severity of the dust, etc. This embodiment does not limit this.

[0106] As a preferred embodiment, in addition to the dust point angle ratio or the continuous number of dust points in step 303A, the dust statistics information can also include other contents for a more comprehensive determination of the severity of the dust. For example, the following steps can also be included in the process:

[0107] Step 303B, determine whether the total quantity is greater than a certain threshold.

[0108] As an alternative embodiment, the dust statistics information can also include the total number of dust points, that is, the total number of laser detection points determined as dust points in the radar.

[0109] For example, a certain type of lidar has a total of 4000 laser detection points. After being determined by the process shown in step 101 or Figure 2 1000 of them are determined as dust points, then the total number of dust points is 1000.

[0110] On this basis, the total number of dust points can be compared with a preset total quantity threshold. If it is less than the preset total quantity threshold, it indicates that the dust adhesion on the lidar surface is not serious, and the dust handling method in step 304 can be executed; otherwise, it indicates that the dust adhesion is serious, and the dust handling method in step 305 needs to be executed.

[0111] As a preferred embodiment, since the number of laser detection points of different types of lidar may vary, the dust statistics information can include the ratio of the total number of dust points, that is, the ratio of the total number of laser detection points determined as dust points in the radar to all laser detection points; and the ratio of the total number of dust points is compared with a preset total ratio threshold to determine whether the dust adhesion is serious, so that the above determination conditions can be adapted to different types of lidar.

[0112] It should be noted that step 303B can be executed after determining that the angle ratio is greater than a certain threshold in step 303A above, so that step 305 is executed only when both of the above two judgment results are "yes", and step 304 is executed if any one of the two judgment results is "no"; in addition, step 303B can replace step 303A, or change the execution order of step 303A and step 303B, etc. This embodiment does not limit the number of determination conditions or the execution order between multiple determination steps.

[0113] Step 303C, store the total number of dust points and the proportion of dust point angles in the masked information of continuous N-frame dust points.

[0114] Step 303D, determine whether the proportion of frames meeting the conditions in multiple frames is greater than a certain threshold.

[0115] As a preferred embodiment, considering that due to false alarms or external environments, etc., the distribution of dust points may change over time. Therefore, when determining the dust treatment method, the dust statistical information at each moment within a certain statistical period can be obtained, and the severity of dust corresponding to the dust statistical information at any moment can be judged respectively. Then, the dust treatment method can be comprehensively determined based on the severity of dust at all moments within this statistical period.

[0116] Specifically, for example, a detection identifier corresponding to the dust statistical information at any moment within the statistical period can be determined based on the dust statistical information at that moment; optionally, if the proportion of dust point angles in the dust statistical information at that moment is greater than the preset angle proportion threshold, and the total number of dust points is greater than the preset total number threshold, then the detection identifier is determined to be the first value. If the proportion of dust point angles is not greater than the preset angle proportion threshold, and / or the total number of dust points is not greater than the preset total number threshold, then the detection identifier is determined to be the second value.

[0117] On this basis, obtain the detection identifiers corresponding to the dust statistical information at all moments within this statistical period, count the first number of detection identifiers with the first value. If this first number is greater than the preset threshold, or the proportion of the first number in the total number of detection identifiers is greater than a certain threshold, then determine the dust treatment method of the lidar as the first dust treatment method and execute Step 305; otherwise, determine the dust treatment method of the lidar as the second dust treatment method and execute Step 304.

[0118] It should be noted that the judgment basis in the process of determining the detection identifier as the first value or the second value above can be replaced according to requirements. For example, the proportion of dust point angles can be replaced by the continuous number of dust points, the total number of dust points can be replaced by the proportion of the total number of dust points, etc. This embodiment does not limit this.

[0119] So far, the description of the Figure 3 shown process example is completed.

[0120] The method provided in the embodiments of the present application has been described above. Next, the device provided in the embodiments of the present application will be described:

[0121] See Figure 4 , Figure 4 is the structural diagram of the device provided in the embodiments of the present application. This device corresponds to Figure 1The method flow shown is applied to a mobile robot including a lidar, and the above lidar includes a plurality of laser detection points.

[0122] As Figure 4 shown, the device may include:

[0123] A dust point determination unit 401, configured to obtain the point cloud data collected by the above lidar, and determine dust points from the above plurality of laser detection points based on the above point cloud data, where the dust points are laser detection points with dust on the surface;

[0124] A processing method determination unit 402, configured to determine dust statistical information based on the distribution of all dust points, and determine the dust processing method of the above lidar based on the above dust statistical information;

[0125] A dust processing unit 403, configured to process the dust on the surface of the above lidar based on the above dust processing method.

[0126] In a possible implementation manner, when the above dust point determination unit 401 determines dust points from the above plurality of laser detection points based on the above point cloud data, it is specifically configured to: for each laser detection point, determine the laser intensity and the target object distance corresponding to the laser detection point based on the above point cloud data; determine the first confidence corresponding to the laser detection point based on the above laser intensity, determine the second confidence corresponding to the laser detection point based on the above target object distance, and determine the third confidence corresponding to the laser detection point based on the target object distance corresponding to the laser detection point and the target object distance corresponding to the adjacent laser detection point of the laser detection point; determine the target confidence based on the above first confidence, the above second confidence, and the above third confidence; if the above target confidence is greater than the preset confidence threshold, determine the laser detection point as a dust point, otherwise, determine the laser detection point as a non-dust point;

[0127] In a possible implementation manner, the above dust statistical information includes the continuous number of dust points. When the processing method determination unit 402 determines the dust statistical information based on the distribution of all dust points, it is specifically configured to: select the dust points with continuous positions from all dust points based on the distribution of all dust points; determine the continuous number of dust points based on the number of dust points with continuous positions;

[0128] In a possible implementation, the above dust statistical information includes the proportion of dust point angles. When the processing method determination unit 402 determines the dust statistical information based on the distribution of all dust points, it specifically is used for: normalizing the distribution of all the above dust points according to the angular distribution to a preset angle value, and determining the angle value occupied by the dust points with continuous positions based on the distribution of all the dust points after the normalization process; determining the proportion of dust point angles based on the above preset angle value and the angle value occupied by the dust points with continuous positions;

[0129] In a possible implementation, the above dust statistical information includes the continuous number of dust points. When the processing method determination unit 402 determines the dust statistical information based on the distribution of all dust points, it specifically is used for: if the continuous number of the above dust points is greater than a preset continuous number threshold, determining the dust processing method of the above lidar as a first dust processing method, where the first dust processing method is used to instruct the above mobile robot to stop for inspection; if the continuous number of the above dust points is not greater than the preset continuous number threshold, determining the dust processing method of the above lidar as a second dust processing method, where the second dust processing method is used to indicate a warning of a dust risk on the surface of the above lidar;

[0130] In a possible implementation, the above dust statistical information includes the continuous number of dust points and the total number of dust points. When the processing method determination unit 402 determines the dust statistical information based on the distribution of all dust points, it specifically is used for: if the continuous number of the above dust points is greater than a preset continuous number threshold, and the total number of the above dust points is greater than a preset total number threshold, determining the dust processing method of the above lidar as a first dust processing method, where the first dust processing method is used to instruct the above mobile robot to stop for inspection; if the continuous number of the above dust points is not greater than the above preset continuous number threshold, and / or the total number of the above dust points is not greater than the above preset total number threshold, determining the dust processing method of the above lidar as a second dust processing method, where the second dust processing method is used to indicate a warning of a dust risk on the surface of the above lidar;

[0131] In a possible implementation, the above dust statistics information includes the continuous number of dust points and the total number of dust points. When the processing method determination unit 402 determines the dust statistics information based on the distribution of all dust points, it is specifically used to: determine a detection identifier corresponding to the dust statistics information at any moment within the statistical period based on the dust statistics information at that moment; wherein, if the continuous number of dust points in the dust statistics information at that moment is greater than a preset continuous number threshold, and the total number of dust points is greater than a preset total number threshold, then the detection identifier is a first value; if the continuous number of dust points is not greater than the preset continuous number threshold, and / or the total number of dust points is not greater than the preset total number threshold, then the detection identifier is a second value; obtain the detection identifiers corresponding to the dust statistics information at all moments within the statistical period, and count the first number of times the detection identifier is the first value; if the first number is greater than a preset threshold, determine that the dust processing method of the lidar is a first dust processing method, and the first dust processing method is used to instruct the mobile robot to stop for inspection; if the first number is not greater than the preset threshold, determine that the dust processing method of the lidar is a second dust processing method, and the second dust processing method is used to indicate that there is a dust risk on the surface of the lidar and give an alarm.

[0132] In a possible implementation, after the processing method determination unit 402 determines that the dust processing method of the lidar is the second dust processing method, it is specifically further used to: control the mobile robot to perform obstacle avoidance driving based on the point cloud data corresponding to the laser detection points among the multiple laser detection points that are not marked as dust points.

[0133] Thus, the Figure 4 structural description of the shown device is completed.

[0134] The embodiments of the present application also provide Figure 4 the hardware structure of the shown device. Refer to Figure 5 , Figure 5 which is the structural diagram of the electronic device provided by the embodiments of the present application. As Figure 5 shown, the hardware structure may include: a processor and a machine-readable storage medium, and the machine-readable storage medium stores machine-executable instructions that can be executed by the above processor; the above processor is used to execute the machine-executable instructions to implement the method disclosed in the above examples of the present application.

[0135] Based on the same application concept as the above method, the embodiments of the present application also provide a machine-readable storage medium, and a number of computer instructions are stored on the machine-readable storage medium. When the computer instructions are executed by a processor, the method disclosed in the above examples of the present application can be implemented.

[0136] Exemplarily, the above machine-readable storage medium can be any electronic, magnetic, optical, or other physical storage device that can contain or store information such as executable instructions, data, and so on. For example, the machine-readable storage medium can be: RAM (Random Access Memory), volatile memory, non-volatile memory, flash memory, storage drives (such as hard disk drives), solid-state drives, any type of storage disk (such as optical discs, DVDs, etc.), or similar storage media, or a combination thereof.

[0137] The systems, devices, modules, or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer, and the specific form of the computer can be a personal computer, laptop computer, cellular phone, camera phone, smart phone, personal digital assistant, media player, navigation device, email transceiver, game console, tablet computer, wearable device, or a combination of any several of these devices.

[0138] For the convenience of description, when describing the above devices, they are described separately as various units according to their functions. Of course, when implementing the present application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0139] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the embodiments of the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program code.

[0140] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, so that the instructions executed by the processors of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0141] Moreover, these computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one process Figure 1 or more processes and / or blocks Figure 1 or the functions specified in a block or more blocks.

[0142] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, such that a series of operation steps are executed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one process Figure 1 or more processes and / or blocks Figure 1 or the functions specified in a block or more blocks.

[0143] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A method for detecting dust on the surface of a lidar, characterized in that, it is applied to a mobile robot including a lidar, the lidar includes a plurality of laser detection points, and the method includes: Obtain the point cloud data collected by the lidar, and determine dust points from the plurality of laser detection points based on the point cloud data, where the dust points are laser detection points with dust on the surface; Determine dust statistical information based on the distribution of all dust points, and determine the dust processing method of the lidar based on the dust statistical information; Process the dust on the surface of the lidar based on the dust processing method; Wherein, the dust statistical information includes the continuous number of dust points, and determining the dust statistical information based on the distribution of all dust points includes: selecting continuously located dust points from all dust points based on the distribution of all dust points; determining the continuous number of dust points based on the number of continuously located dust points.

2. The method according to claim 1, characterized in that, the determining the dust points from the plurality of laser detection points based on the point cloud data includes: For each laser detection point, determine the laser intensity and the target object distance corresponding to the laser detection point based on the point cloud data; Determine the first confidence level corresponding to the laser detection point based on the laser intensity, determine the second confidence level corresponding to the laser detection point based on the target object distance, and determine the third confidence level corresponding to the laser detection point based on the target object distance corresponding to the laser detection point and the target object distance corresponding to the adjacent laser detection point of the laser detection point; Determine the target confidence level based on the first confidence level, the second confidence level and the third confidence level; If the target confidence level is greater than the preset confidence threshold, determine the laser detection point as a dust point, otherwise, determine the laser detection point as a non-dust point.

3. The method according to claim 1, characterized in that, the dust statistical information includes the angle ratio of dust points, and determining the dust statistical information based on the distribution of all dust points includes: Normalize the distribution of all dust points according to the angle distribution to a preset angle value, and determine the angle value occupied by continuously located dust points based on the normalized distribution of all dust points; Determine the angle ratio of dust points based on the preset angle value and the angle value occupied by continuously located dust points.

4. The method according to claim 1, characterized in that, the dust statistical information includes the continuous number of dust points, and determining the dust processing method of the lidar based on the dust statistical information includes: If the continuous number of dust points is greater than the preset continuous number threshold, determine the dust processing method of the lidar as the first dust processing method, and the first dust processing method is used to instruct the mobile robot to stop for inspection; If the continuous number of dust points is not greater than the preset continuous number threshold, determine the dust processing method of the lidar as the second dust processing method, and the second dust processing method is used to indicate a warning of the risk of dust on the surface of the lidar.

5. The method according to claim 1, It is characterized in that the dust statistical information includes the continuous number of dust points and the total number of dust points, and determining the dust processing method of the lidar based on the dust statistical information includes: if the continuous number of dust points is greater than a preset continuous number threshold, and the total number of dust points is greater than a preset total number threshold, then determining the dust processing method of the lidar as a first dust processing method, where the first dust processing method is used to instruct the mobile robot to stop for inspection; if the continuous number of dust points is not greater than the preset continuous number threshold, and / or the total number of dust points is not greater than the preset total number threshold, then determining the dust processing method of the lidar as a second dust processing method, where the second dust processing method is used to indicate a warning that there is a dust risk on the surface of the lidar.

6. The method according to claim 1, It is characterized in that the dust statistical information includes the continuous number of dust points and the total number of dust points, and determining the dust processing method of the lidar based on the dust statistical information includes: determining a detection identifier corresponding to the dust statistical information at any moment within a statistical period based on the dust statistical information at that moment; wherein, if the continuous number of dust points in the dust statistical information at that moment is greater than a preset continuous number threshold, and the total number of dust points is greater than a preset total number threshold, then the detection identifier is a first value; if the continuous number of dust points is not greater than the preset continuous number threshold, and / or the total number of dust points is not greater than the preset total number threshold, then the detection identifier is a second value; obtaining the detection identifiers corresponding to the dust statistical information at all moments within the statistical period, and counting a first number of the detection identifiers with the first value; if the first number is greater than a preset threshold, then determining the dust processing method of the lidar as a first dust processing method, where the first dust processing method is used to instruct the mobile robot to stop for inspection; if the first number is not greater than the preset threshold, then determining the dust processing method of the lidar as a second dust processing method, where the second dust processing method is used to indicate a warning that there is a dust risk on the surface of the lidar.

7. The method according to any one of claims 4-6, It is characterized in that after determining that the dust processing method of the lidar is a second dust processing method, the method further includes: controlling the mobile robot to perform obstacle avoidance driving based on the point cloud data corresponding to the laser detection points that are not marked as dust points among the multiple laser detection points.

8. A dust detection device for the surface of a lidar, It is characterized in that applied to a mobile robot including a lidar, the lidar includes multiple laser detection points, and the device includes: a dust point determination unit, configured to obtain the point cloud data collected by the lidar, and determine dust points from the multiple laser detection points based on the point cloud data, where the dust points are laser detection points with dust on the surface; a processing method determination unit, configured to determine dust statistical information based on the distribution of all dust points, and determine the dust processing method of the lidar based on the dust statistical information; A dust processing unit for processing the dust on the surface of the lidar based on the dust processing method; Wherein, the dust statistical information includes the continuous number of dust points. When the processing method determination unit determines the dust statistical information based on the distribution of all dust points, it specifically is used for: selecting the dust points with continuous positions from all dust points based on the distribution of all dust points; determining the continuous number of dust points based on the number of dust points with continuous positions.

9. The device according to claim 8 ， Characterized in that Wherein When the dust point determination unit determines the dust points from the multiple laser detection points based on the point cloud data, it specifically is used for: for each laser detection point, determining the laser intensity and the target object distance corresponding to the laser detection point based on the point cloud data; determining the first confidence level corresponding to the laser detection point based on the laser intensity, determining the second confidence level corresponding to the laser detection point based on the target object distance, and determining the third confidence level corresponding to the laser detection point based on the target object distance corresponding to the laser detection point and the target object distance corresponding to the adjacent laser detection point of the laser detection point; determining the target confidence level based on the first confidence level, the second confidence level, and the third confidence level; if the target confidence level is greater than the preset confidence threshold, determining the laser detection point as a dust point, otherwise, determining the laser detection point as a non-dust point; Wherein, the dust statistical information includes the angle proportion of dust points. When the processing method determination unit determines the dust statistical information based on the distribution of all dust points, it specifically is used for: normalizing the distribution of all dust points according to the angle distribution to a preset angle value, and determining the angle value occupied by the dust points with continuous positions based on the normalized distribution of all dust points; determining the angle proportion of dust points based on the preset angle value and the angle value occupied by the dust points with continuous positions; Wherein, the dust statistical information includes the continuous number of dust points. When the processing method determination unit determines the dust statistical information based on the distribution of all dust points, it specifically is used for: if the continuous number of dust points is greater than the preset continuous number threshold, determining the dust processing method of the lidar as the first dust processing method, and the first dust processing method is used to instruct the mobile robot to stop for inspection; if the continuous number of dust points is not greater than the preset continuous number threshold, determining the dust processing method of the lidar as the second dust processing method, and the second dust processing method is used to indicate a warning of the dust risk on the surface of the lidar; Among them, the dust statistical information includes the continuous number of dust points and the total number of dust points. When the processing method determination unit determines the dust statistical information based on the distribution of all dust points, it specifically is used for: if the continuous number of dust points is greater than a preset continuous number threshold, and the total number of dust points is greater than a preset total number threshold, then determining the dust processing method of the lidar as the first dust processing method, where the first dust processing method is used to instruct the mobile robot to stop for inspection; if the continuous number of dust points is not greater than the preset continuous number threshold, and / or the total number of dust points is not greater than the preset total number threshold, then determining the dust processing method of the lidar as the second dust processing method, where the second dust processing method is used to indicate a warning of a dust risk on the surface of the lidar; Among them, the dust statistical information includes the continuous number of dust points and the total number of dust points. When the processing method determination unit determines the dust statistical information based on the distribution of all dust points, it specifically is used for: determining a detection identifier corresponding to the dust statistical information at any moment within a statistical period based on the dust statistical information at that moment; among them, if the continuous number of dust points in the dust statistical information at that moment is greater than a preset continuous number threshold, and the total number of dust points is greater than a preset total number threshold, then the detection identifier is the first value; if the continuous number of dust points is not greater than the preset continuous number threshold, and / or the total number of dust points is not greater than the preset total number threshold, then the detection identifier is the second value; obtaining the detection identifiers corresponding to the dust statistical information at all moments within the statistical period, and counting the first number of times the detection identifier is the first value; if the first number is greater than a preset threshold, then determining the dust processing method of the lidar as the first dust processing method, where the first dust processing method is used to instruct the mobile robot to stop for inspection; if the first number is not greater than the preset threshold, then determining the dust processing method of the lidar as the second dust processing method, where the second dust processing method is used to indicate a warning of a dust risk on the surface of the lidar; Among them, after the processing method determination unit determines that the dust processing method of the lidar is the second dust processing method, it specifically is further used for: controlling the mobile robot to perform obstacle avoidance driving based on the point cloud data corresponding to the laser detection points that are not marked as dust points among the multiple laser detection points.

Images