Navigation positioning method and system for distribution unmanned aerial vehicle in high-speed service area

Abstract

The invention discloses a navigation positioning method and system for a high-speed service area distribution unmanned aerial vehicle, and relates to the technical field of video measurement. The navigation and positioning method for the distribution unmanned aerial vehicle in the high-speed service area comprises the following steps: acquiring ground point cloud data of the high-speed service area; traversing the ground point cloud data to obtain a first frame of point cloud data; traversing the first frame of point cloud data to obtain a first cloud point set; obtaining an extension trend value based on the first cloud point set; and obtaining the slender obstacle probability of the first cloud point set based on the extension trend value. According to the method and the device, the point cloud sets with the slender obstacle characteristics in the single-frame point cloud data are identified, and the point cloud sets in the multi-frame point cloud data are matched, so that the slender obstacles in the delivery area of the high-speed service area can be accurately identified.

Description

Technical Field

[0001] The present application relates to the field of video measurement technology, and specifically to a navigation and positioning method and system for delivery drones in highway service areas. Background Art

[0002] Highway service area delivery drones use drones to rapidly deliver items such as express delivery, catering, and emergency supplies. Equipped with automatic navigation and landing capabilities, they significantly improve delivery efficiency in highway service areas, reduce labor costs, and meet the needs of drivers and passengers for convenient and efficient service. They are particularly suitable for delivering supplies in crowded areas, with inconvenient transportation, or in emergencies. During their missions, highway service area delivery drones face complex and changing environments, including obstacles such as parked vehicles, pedestrians, and billboards. Accurate obstacle location is crucial for flight safety. It not only effectively avoids collisions, ensures stable drone flight, and safely delivers items, but also enhances the intelligence of route planning and delivery efficiency. It is a key technical support for achieving unmanned, highly reliable delivery services.

[0003] Existing technologies often challenge drones flying near highway service areas with identifying slender obstacles like cables and flagpoles. These obstacles are small, have low reflectivity, and have low laser hit rates. They often appear sparse and fragmented in the lidar point cloud, making them more susceptible to misidentification or omission during dynamic flight. Summary of the Invention

[0004] The purpose of this application is to provide a navigation and positioning method and system for delivery drones in highway service areas, so as to solve the technical problem that drones have difficulty in identifying slender obstacles.

[0005] To achieve the above objectives, this application provides the following technical solutions: In the first aspect, the present application proposes a technical solution for a navigation and positioning method for a delivery drone in a highway service area. The navigation and positioning method for a delivery drone in a highway service area comprises: Acquire ground point cloud data of a highway service area; the ground point cloud data includes multiple frames of point cloud data; Traversing the ground point cloud data to obtain a first frame of point cloud data; the first frame of point cloud data is point cloud data in any frame of the ground point cloud data in which the probability of a slender obstacle is not obtained; Traversing the first frame of point cloud data to obtain a first cloud point set; the first cloud point set is any cloud point set in the first frame of point cloud data that has not obtained a stretching trend value; Based on the first cloud point set, a stretching trend value is obtained; the stretching trend value is at least used to characterize the symmetry of each cloud point in the first cloud point set; Based on the stretching tendency value, obtaining a slender obstacle probability of the first cloud point set; the slender obstacle probability is at least used to represent the probability that each cloud point in the first cloud point set belongs to the same slender obstacle; Constructing a three-dimensional model of the highway service area based on the probability of the slender obstacle; The delivery drone is navigated and positioned based on the three-dimensional model.

[0006] As a specific solution in the technical solution of the present application, traversing the first frame of point cloud data to obtain a first cloud point set includes: Based on the first frame of point cloud data, a second frame of point cloud data is obtained; the second frame of point cloud data is the point cloud data of the first frame of point cloud data excluding all ground cloud points; Based on the second frame of point cloud data, a third frame of point cloud data is obtained; the third frame of point cloud data is the point cloud data of the second frame of point cloud data excluding the cloud points corresponding to all non-slender obstacles; Traversing the third frame of point cloud data to obtain a first cloud point; the first cloud point is any cloud point in the third frame of point cloud data that is not included in any cloud point set; Based on the first cloud point, a plurality of near cloud points are obtained; the near cloud points are cloud points in the third frame point cloud data whose distance from the first cloud point is less than or equal to a first preset value, or cloud points in the third frame point cloud data whose distance from any near cloud point is less than or equal to a first preset value; The first cloud point and each of the near cloud points are formed into a set to obtain a cloud point set; Traverse each cloud point set to obtain the first cloud point set.

[0007] As a specific solution in the technical solution of the present application, obtaining a second frame of point cloud data based on the first frame of point cloud data includes: Traversing the first frame of point cloud data to obtain a second cloud point and a third cloud point; the second cloud point is any cloud point in the first frame of point cloud data that has not been confirmed to be a ground cloud point; the third cloud point is the cloud point with the lowest vertical height in the first frame of point cloud data; Based on the second cloud point and the third cloud point, obtaining a first distance; the first distance is the distance between the second cloud point and the third cloud point in the vertical direction; If the first distance is less than or equal to a second preset value, the second cloud point is eliminated; otherwise, the second cloud point is retained.

[0008] As a specific solution in the technical solution of the present application, obtaining a third frame of point cloud data based on the second frame of point cloud data includes: Obtaining a plurality of clusters from the second frame of point cloud data by using a density-based clustering algorithm; The cloud points belonging to each cluster are removed from the second frame of point cloud data to obtain the third frame of point cloud data.

[0009] As a specific solution in the technical solution of this application, the step of obtaining multiple near cloud points based on the first cloud point includes: Obtain a first near cloud point and a second near cloud point; the first near cloud point and the second near cloud point are obtained based on the update; the initial first near cloud point is the first cloud point, and the initial second near cloud point is any cloud point in the third frame of point cloud data whose distance from the first near cloud point is less than or equal to a first preset value; Based on the first near cloud point and the second near cloud point, acquiring an extension direction; the extension direction is from the first near cloud point to the second near cloud point; Based on the second near cloud point and the extension direction, a third near cloud point is obtained, and the first near cloud point and the second near cloud point are updated; the third near cloud point is the cloud point in the third frame point cloud data that is closest to the second near cloud point along the extension direction, and the distance between the second near cloud point and the third near cloud point is less than or equal to a first preset value; the updated first near cloud point is equal to the second near cloud point before the update, and the updated second near cloud point is equal to the third near cloud point before the update.

[0010] As a specific solution in the technical solution of the present application, obtaining the stretching trend value based on the first cloud point set includes: Based on the first cloud point set, an intermediate cloud point is obtained; the intermediate cloud point is a cloud point with the smallest extension angle in the first cloud point set; Based on the first cloud point set and the intermediate cloud point, a second cloud point set and a third cloud point set are obtained; the second cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located before the intermediate cloud point; the third cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located after the intermediate cloud point; Based on the second cloud point set, a first trend value is obtained; the first trend value is used to at least characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the second cloud point set; Based on the third cloud point set, a second trend value is obtained; the second trend value is used to at least characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the third cloud point set; The stretching trend value is obtained based on the first trend value and the second trend value.

[0011] As a specific solution in the technical solution of the present application, obtaining the stretching trend value based on the first trend value and the second trend value includes: Based on the first cloud point set, obtaining a first stretching angle and a second stretching angle; the first stretching angle is the maximum stretching angle corresponding to each cloud point in the first cloud point set; the second stretching angle is the minimum stretching angle corresponding to each cloud point in the first cloud point set; Obtaining a stability value based on the first extension angle and the second extension angle; wherein the stability value is at least used to characterize a difference between the first extension angle and the second extension angle; The stretching trend value is obtained based on the stability value, the first trend value, and the second trend value.

[0012] As a specific solution in the technical solution of the present application, obtaining the first trend value based on the second cloud point set includes: Based on the second cloud point set, obtaining a third stretching angle and a fourth stretching angle; the third stretching angle is the stretching angle corresponding to the middle cloud point; the fourth stretching angle is the stretching angle corresponding to the cloud point in the second cloud point set that is farthest from the middle cloud point; Obtaining a first difference based on the third extension angle and the fourth extension angle; the first difference is the fourth extension angle minus the third extension angle; Based on the second cloud point set, a plurality of second difference values ​​are obtained; the second difference value is an absolute value of a difference between extension angles corresponding to two adjacent cloud points in the second cloud point set; The first trend value is obtained based on the first difference value and each second difference value.

[0013] As a specific solution in the technical solution of the present application, obtaining the probability of a slender obstacle in the first cloud point set based on the stretching trend value includes: Based on the ground point cloud data, a second frame of point cloud data is acquired; the second frame of point cloud data is point cloud data in the ground point cloud data that is temporally adjacent to the first frame of point cloud data; Based on the second frame of point cloud data, a fourth cloud point set is obtained; the fourth cloud point set is the cloud point set with the highest similarity to the first cloud point set in the second frame of point cloud data; Obtaining similarity based on the first cloud point set and the fourth cloud point set; Based on the similarity and the stretching tendency value, a slender obstacle probability of the first cloud point set is obtained.

[0014] In a second aspect, this application proposes a technical solution for a navigation and positioning system for delivery drones in high-speed service areas. The navigation and positioning system for delivery drones in high-speed service areas includes: A camera device is used to obtain ground point cloud data of a highway service area; the ground point cloud data includes multiple frames of point cloud data; A processing device, configured to traverse the ground point cloud data to obtain a first frame of point cloud data; the first frame of point cloud data is point cloud data of any frame of the ground point cloud data in which no slender obstacle probability is obtained; and traversing the first frame of point cloud data to obtain a first cloud point set; the first cloud point set is any cloud point set in the first frame of point cloud data that has not obtained a stretching trend value; And, based on the first cloud point set, obtaining a stretching trend value; the stretching trend value is at least used to characterize the symmetry of each cloud point in the first cloud point set; and, based on the stretching tendency value, obtaining a slender obstacle probability of the first cloud point set; the slender obstacle probability is at least used to represent the probability that each cloud point in the first cloud point set belongs to the same slender obstacle; and, constructing a three-dimensional model of the highway service area based on the probability of the elongated obstacle; A control device is used to navigate and locate the delivery drone based on the three-dimensional model.

[0015] As a specific solution in the technical solution of the present application, the processing device is further used to obtain a second frame of point cloud data based on the first frame of point cloud data; the second frame of point cloud data is the point cloud data of the first frame of point cloud data excluding all ground cloud points; And, based on the second frame of point cloud data, obtaining a third frame of point cloud data; the third frame of point cloud data is the point cloud data of the second frame of point cloud data excluding all cloud points corresponding to non-slender obstacles; and traversing the third frame of point cloud data to obtain a first cloud point; the first cloud point is any cloud point in the third frame of point cloud data that is not included in any cloud point set; And, based on the first cloud point, a plurality of near cloud points are obtained; the near cloud points are cloud points in the third frame point cloud data whose distance from the first cloud point is less than or equal to a first preset value, or cloud points in the third frame point cloud data whose distance from any near cloud point is less than or equal to a first preset value; and forming a set of the first cloud point and each of the near cloud points to obtain a cloud point set; And, traverse each cloud point set to obtain the first cloud point set.

[0016] As a specific solution in the technical solution of the present application, the processing device is further used to traverse the first frame of point cloud data to obtain a second cloud point and a third cloud point; the second cloud point is any cloud point in the first frame of point cloud data that has not been confirmed to be a ground cloud point; the third cloud point is the cloud point with the lowest height in the vertical direction in the first frame of point cloud data; And, based on the second cloud point and the third cloud point, obtaining a first distance; the first distance is the distance between the second cloud point and the third cloud point in the vertical direction; And, if the first distance is less than or equal to a second preset value, the second cloud point is eliminated; otherwise, the second cloud point is retained.

[0017] As a specific solution in the technical solution of the present application, the processing device is further used to obtain a plurality of clusters from the second frame point cloud data by using a density-based clustering algorithm; Furthermore, cloud points belonging to each cluster are removed from the second frame of point cloud data to obtain the third frame of point cloud data.

[0018] As a specific solution in the technical solution of the present application, the processing device is further used to obtain a first near cloud point and a second near cloud point; the first near cloud point and the second near cloud point are obtained based on the update; the initial first near cloud point is the first cloud point, and the initial second near cloud point is any cloud point in the third frame point cloud data whose distance from the first near cloud point is less than or equal to a first preset value; and, based on the first near cloud point and the second near cloud point, obtaining an extension direction; wherein the extension direction points from the first near cloud point to the second near cloud point; And, based on the second near cloud point and the extension direction, a third near cloud point is obtained, and the first near cloud point and the second near cloud point are updated; the third near cloud point is the cloud point in the third frame point cloud data that is closest to the second near cloud point along the extension direction, and the distance between the second near cloud point and the third near cloud point is less than or equal to a first preset value; the updated first near cloud point is equal to the second near cloud point before the update, and the updated second near cloud point is equal to the third near cloud point before the update.

[0019] As a specific solution in the technical solution of the present application, the processing device is further used to obtain an intermediate cloud point based on the first cloud point set; the intermediate cloud point is a cloud point with the smallest extension angle in the first cloud point set; And, based on the first cloud point set and the intermediate cloud point, a second cloud point set and a third cloud point set are obtained; the second cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located before the intermediate cloud point; the third cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located after the intermediate cloud point; and, based on the second cloud point set, obtaining a first trend value; the first trend value is at least used to characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the second cloud point set; and obtaining a second trend value based on the third cloud point set; wherein the second trend value is at least used to characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the third cloud point set; And, based on the first trend value and the second trend value, the stretching trend value is obtained.

[0020] As a specific solution of the technical solution of the present application, the processing device is further used to obtain a first stretching angle and a second stretching angle based on the first cloud point set; the first stretching angle is the maximum stretching angle corresponding to each cloud point in the first cloud point set; the second stretching angle is the minimum stretching angle corresponding to each cloud point in the first cloud point set; and obtaining a stability value based on the first extension angle and the second extension angle; wherein the stability value is at least used to characterize the difference between the first extension angle and the second extension angle; And, based on the stability value, the first trend value and the second trend value, the stretching trend value is obtained.

[0021] As a specific solution of the technical solution of the present application, the processing device is further used to obtain a third stretching angle and a fourth stretching angle based on the second cloud point set; the third stretching angle is the stretching angle corresponding to the intermediate cloud point; the fourth stretching angle is the stretching angle corresponding to the cloud point in the second cloud point set that is farthest from the intermediate cloud point; and obtaining a first difference based on the third extension angle and the fourth extension angle; wherein the first difference is the fourth extension angle minus the third extension angle; and, based on the second cloud point set, obtaining a plurality of second difference values; the second difference value is an absolute value of a difference between extension angles corresponding to two adjacent cloud points in the second cloud point set; And, based on the first difference value and each second difference value, the first trend value is obtained.

[0022] As a specific solution in the technical solution of the present application, the processing device is further used to obtain a second frame of point cloud data based on the ground point cloud data; the second frame of point cloud data is point cloud data in the ground point cloud data that is temporally adjacent to the first frame of point cloud data; And, based on the second frame of point cloud data, obtaining a fourth cloud point set; the fourth cloud point set is the cloud point set with the highest similarity to the first cloud point set in the second frame of point cloud data; and, obtaining similarity based on the first cloud point set and the fourth cloud point set; And, based on the similarity and the stretching tendency value, a slender obstacle probability of the first cloud point set is obtained.

[0023] Compared with the prior art, the present invention has the following advantages: This application identifies point cloud sets containing slender obstacle features in single-frame point cloud data, then matches point cloud sets in multi-frame point cloud data. It then uses the slender obstacle probability of each matched point cloud set and its similarity with adjacent point cloud sets to comprehensively determine whether there are slender obstacles in the delivery area of ​​a highway service area. In other words, this application can accurately identify slender obstacles in the delivery area of ​​a highway service area. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a flowchart of a navigation and positioning method for delivery drones in high-speed service areas proposed in an embodiment of the present application; Figure 2 This is a structural diagram of a navigation and positioning system for delivery drones in highway service areas proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0026] The terms "first", "second", etc. in the description of the embodiments of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. For example, the first cloud point set and the second cloud point set proposed below belong to different sets. It should be understood that the names used in this way can be interchanged where appropriate so that the embodiments described herein can be implemented in an order other than that shown or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or modules is not necessarily limited to those steps or modules clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices. The division of modules appearing in the embodiments of the present application is only a logical division. In actual applications, there may be other division methods. For example, multiple modules can be combined or integrated into another system, or some features can be ignored or not performed. In addition, the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, and the indirect coupling or communication connection between modules can be electrical or other similar forms, which are not limited in the embodiments of the present application. Moreover, the modules or sub-modules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed into multiple circuit modules, and some or all of the modules may be selected according to actual needs to achieve the purpose of the embodiment of the present application.

[0027] In order to solve the technical problem that drones have difficulty identifying slender obstacles, this application proposes an embodiment of a navigation and positioning method for delivery drones in high-speed service areas. Figure 1 As shown, the navigation and positioning method for the delivery drone in the highway service area includes steps S100 to S700.

[0028] Step S100: Acquire ground point cloud data of the highway service area.

[0029] It's important to note that before a delivery drone begins its mission at a highway service area, it first obtains the location of the drone's takeoff and landing platform at the service area it needs to reach. After the drone begins its delivery mission, it uses its onboard Global Navigation Satellite System (GNSS) receiver to obtain real-time global positioning information for navigation. When performing a delivery mission at a highway service area, the drone's flight process involves a rapid climb from the takeoff point, reaching a stable cruising altitude, and then gradually descending upon approaching the service area before landing with precision. Throughout the flight, to ensure flight safety and mission accuracy, obstacle detection is primarily focused on the approach and landing phases. During this phase, as the drone gradually descends in altitude and enters the densely populated service area, it relies on high-precision lidar for environmental perception, enabling real-time obstacle avoidance and precise platform positioning to ensure a safe and reliable landing. When the drone is 100 meters from the takeoff and landing platform, it uses lidar to perform a 3D point cloud scan of the ground, collecting real-time spatial data of the service area. Generally, it is necessary to collect multiple frames of point cloud data, analyze the multiple frames of point cloud data, build a three-dimensional model of the ground space within the service area, and then navigate the drone based on the three-dimensional model. In this embodiment, the ground point cloud data includes multiple frames of point cloud data collected when the drone lands.

[0030] Step S200: traverse the ground point cloud data to obtain the first frame of point cloud data.

[0031] In this embodiment, the first frame of point cloud data is any frame of the ground point cloud data for which the slender obstacle probability is not obtained. That is, in this embodiment, the method for obtaining the slender obstacle probability for any frame of point cloud data in the ground point cloud data is the same as the method for obtaining the slender obstacle probability for the first frame of point cloud data.

[0032] Step S300: traverse the first frame of point cloud data to obtain a first cloud point set.

[0033] In this embodiment, the first cloud point set is any cloud point set in the first frame of point cloud data for which a stretching trend value has not been obtained. That is, in this embodiment, the stretching trend value of any cloud point set in the first frame of point cloud data is obtained in the same manner as the stretching trend value of the first cloud point set.

[0034] In this embodiment, the cloud point set refers to a set of cloud points in the first frame of point cloud data that may belong to the same elongated obstacle. In the embodiment of the present application, the first cloud point set is obtained from the first frame of point cloud data in any reasonable manner.

[0035] It should be clear that when a drone first enters a highway service area to perform a delivery mission, due to the lack of a priori maps of the area, the lidar needs to collect and process a large amount of raw point cloud data, which results in a heavy computational load. In order to ensure the adequacy of data processing and the accuracy of obstacle avoidance judgment, the drone should gradually approach the take-off and landing platform at a slower speed during this stage to avoid recognition delays or obstacle avoidance failures due to flying too fast, thereby ensuring the safety and stability of the first landing process. When scanning point cloud data, it is necessary to analyze the content of the scanned point cloud data to obtain the location of the obstacle entity. In order to reduce the amount of data processing, the point cloud in the first frame of point cloud data that does not belong to a slender obstacle can be removed. Based on this, in a specific embodiment of the present application, step S300, traverses the first frame of point cloud data to obtain a first cloud point set, including steps S310 to S360.

[0036] Step S310: Based on the first frame of point cloud data, obtain a second frame of point cloud data.

[0037] In this embodiment, the second frame of point cloud data is the point cloud data obtained by removing all ground cloud points from the first frame of point cloud data. It should be understood that any reasonable method can be used to remove ground cloud points from the first frame of point cloud data to obtain the second frame of point cloud data. For example, step S310, which obtains the second frame of point cloud data based on the first frame of point cloud data, includes steps S311 to S313.

[0038] Step S311: traverse the first frame of point cloud data to obtain the second cloud point and the third cloud point.

[0039] In this embodiment, the second cloud point is any cloud point in the first frame of point cloud data that has not been confirmed to be a ground cloud point. The third cloud point is the cloud point with the lowest vertical height in the first frame of point cloud data.

[0040] Step S312: Acquire a first distance based on the second cloud point and the third cloud point.

[0041] In this embodiment, the first distance is the distance between the second cloud point and the third cloud point along the vertical direction. It should be noted that obtaining the distance between two cloud points (i.e., the second cloud point and the third cloud point) along the vertical direction in point cloud data is a mature technology and will not be described in detail here.

[0042] Step S313: If the first distance is less than or equal to a second preset value, the second cloud point is eliminated; otherwise, the second cloud point is retained.

[0043] In the embodiments of the present application, the second preset value can be set as needed. For example, the second preset value can be 10 cm or 20 cm. Of course, if the height of the take-off and landing platform from the horizontal ground is 50 cm, it can be assumed that objects 50 cm away from the ground will not affect the drone. Therefore, the second preset value can also be equal to the vertical height of the take-off and landing platform (i.e., 50 cm).

[0044] Step S320: Based on the second frame of point cloud data, obtain a third frame of point cloud data.

[0045] In this embodiment, the third frame of point cloud data is the point cloud data obtained by removing all cloud points corresponding to non-slender obstacles from the second frame of point cloud data.

[0046] It's important to understand that after removing the ground point cloud, non-slender obstacle entities need to be identified. When the LiDAR scans the surface of a non-slender obstacle entity, the reflected point cloud data can characterize the location of the non-slender obstacle entity, thereby forming point cloud clusters that reflect the contour features of the non-slender obstacle entity. Based on this, step S320, based on the second frame of point cloud data, acquires a third frame of point cloud data, including steps S321 and S322.

[0047] Step S321: obtaining a plurality of clusters from the second frame of point cloud data by using a density-based clustering algorithm.

[0048] It should be noted that since non-slender obstacle entities will form multiple cloud points, the clustering algorithm can be used to cluster the cloud points belonging to the same non-slender obstacle entity into the same cluster. In other words, the cloud points in the same cluster must not belong to the cloud points formed by slender obstacles.

[0049] In the embodiment of the present application, any reasonable density-based clustering algorithm can be used to cluster the cloud points in the second frame of point cloud data. For example, the density-based clustering algorithm can be the DBSCAN algorithm, the OPTICS algorithm, or the DENCLUE algorithm.

[0050] Step S322: removing cloud points belonging to each cluster from the second frame of point cloud data to obtain the third frame of point cloud data.

[0051] It should be noted that removing some cloud points from the second frame of point cloud data is a mature technology and will not be described in detail here.

[0052] Step S330: traverse the third frame of point cloud data to obtain the first cloud point.

[0053] In this embodiment, the first cloud point is any cloud point in the third frame of point cloud data that is not included in any cloud point set. That is, in this embodiment, the manner in which any cloud point in the third frame of point cloud data is included in a certain cloud point set is the same as the manner in which the first cloud point is included in the cloud point set.

[0054] Step S340: Based on the first cloud point, obtain multiple near cloud points.

[0055] In this embodiment, the near cloud point is a cloud point in the third frame point cloud data whose distance to the first cloud point is less than or equal to a first preset value, or a cloud point in the third frame point cloud data whose distance to any near cloud point is less than or equal to a first preset value.

[0056] In the embodiment of the present application, the first preset value can be set according to demand. For example, the first preset value can be 10 cm or 20 cm.

[0057] In the embodiments of the present application, there is no limitation on the method for obtaining multiple near-cloud points. It should be understood that elongated obstacles must have an extension direction, for example, a flagpole extends vertically, a power line extends horizontally, etc. In other words, if the extension directions of the cloud points in the third frame of point cloud data are roughly the same, then these cloud points are likely to belong to the same elongated obstacle. Based on this, step S340, based on the first cloud point, obtains multiple near-cloud points, including steps S341 to S343.

[0058] Step S341: Acquire the first near-cloud point and the second near-cloud point.

[0059] In this embodiment, the first near cloud point and the second near cloud point are obtained based on an update. The initial first near cloud point is the first cloud point, and the initial second near cloud point is any cloud point in the third frame of point cloud data whose distance from the first near cloud point is less than or equal to a first preset value.

[0060] Step S342: Acquire an extension direction based on the first near cloud point and the second near cloud point.

[0061] In this embodiment, the extending direction is from the first near-cloud point to the second near-cloud point.

[0062] Step S343: Based on the second near cloud point and the extension direction, obtain a third near cloud point, and update the first near cloud point and the second near cloud point.

[0063] In this embodiment, the third near cloud point is the cloud point in the third frame of point cloud data that is closest to the second near cloud point along the extension direction, and the distance between the second near cloud point and the third near cloud point is less than or equal to a first preset value; the updated first near cloud point is equal to the second near cloud point before the update, and the updated second near cloud point is equal to the third near cloud point before the update. In other words, in this embodiment, by continuously updating the first and second near cloud points to obtain the third near cloud point along the extension direction, the obtained first cloud point and multiple near cloud points are highly likely to belong to the same elongated obstacle.

[0064] Step S350: forming a set of the first cloud points and each near cloud point to obtain a cloud point set.

[0065] It should be noted that forming a collection of cloud points is a mature technology and will not be described in detail here. In this embodiment, the number of cloud point collections that can be obtained is generally determined by the number of slender obstacles in the third frame of point cloud data.

[0066] Step S360: traverse each cloud point set to obtain the first cloud point set.

[0067] In this embodiment, the first cloud point set is any cloud point set in the first frame point cloud data for which no stretching trend value is obtained.

[0068] Step S400: obtaining a stretching trend value based on the first cloud point set.

[0069] In this embodiment, the stretching trend value is at least used to characterize the symmetry of each cloud point in the first cloud point set.

[0070] It should be noted that when there are long, slender obstacles at high altitudes, such as power lines, even if they have a certain tendency to bend, they are generally symmetrical along their midpoints. Flagpoles are also symmetrical along their midpoints. Therefore, whether each point in the first point cloud set belongs to the same long, slender obstacle can be determined based on whether the points in the first point cloud set are symmetrical. Based on this, step S400, which obtains an extension tendency value based on the first point cloud set, includes steps S410 to S450.

[0071] Step S410: Based on the first cloud point set, obtain intermediate cloud points.

[0072] In this embodiment, the intermediate cloud point is the cloud point with the smallest extension angle in the first cloud point set. As previously mentioned, in this embodiment of the present application, the cloud points in the first cloud point set are arranged approximately linearly along the extension direction. In this embodiment, obtaining the extension angle of a specific cloud point (hereinafter referred to as the target cloud point) includes steps S411 to S414.

[0073] Step S411: Based on the first cloud point set, obtain target cloud points.

[0074] In this embodiment, the target cloud point is any cloud point in the first cloud point set.

[0075] Step S412: Based on the target cloud point, obtain a fourth cloud point and a fifth cloud point.

[0076] In this embodiment, the fourth cloud point and the fifth cloud point are cloud points in the first cloud point set that are spatially adjacent to the target cloud point. In other words, the fourth cloud point is a cloud point in the first cloud point set that is located on one side of the target cloud point, and the fifth cloud point is a cloud point in the first cloud point set that is located on the other side of the target cloud point.

[0077] Step S413: Acquire a first line segment and a second line segment based on the target cloud point, the fourth cloud point, and the fifth cloud point.

[0078] In this embodiment, the first line segment is a line segment formed by a line connecting the target cloud point and the fourth cloud point; the second line segment is a line segment formed by a line connecting the target cloud point and the fifth cloud point.

[0079] Step S414: obtaining an extension angle based on the first line segment and the second line segment.

[0080] In this embodiment, the extension angle is the angle formed by the first line segment and the second line segment.

[0081] It should be understood that if each point in the first point cloud set belongs to a slender obstacle, and the slender obstacle exhibits a certain curvature, then the point located in the middle of the slender obstacle (i.e., the middle point in the preceding text) will have the smallest extension angle. Furthermore, the closer the point in the first point cloud set is to the ends of the slender obstacle, the larger the extension angle corresponding to that point. In other words, if each point in the first point cloud set exhibits the aforementioned trend (i.e., the extension angle of each point increases from the middle point toward the end points), then the first point cloud set is more likely to represent a point cloud set of a slender obstacle entity with a curvature.

[0082] It should be noted that if each point in the first point cloud set belongs to a slender obstacle, and the slender obstacle is straight, then the extension angle corresponding to each point in the first point cloud set is substantially equal to 180°. In other words, if the extension angles of each point in the first point cloud set are close to 180°, then the first point cloud set is more likely to represent a point cloud set of a slender obstacle entity with a straight extension trend.

[0083] In other words, if the points in the first point cloud set do not exhibit the aforementioned curved or straight trends, then the first point cloud set may not be a point cloud set formed by a slender obstacle (hereinafter referred to as an "incorrect point cloud set"). It is easy to understand that the more points in the third frame of point cloud data, the greater the probability of obtaining an incorrect point cloud set.

[0084] Step S420: Based on the first cloud point set and the intermediate cloud point, obtain a second cloud point set and a third cloud point set.

[0085] In this embodiment, the second cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located before the intermediate cloud point. The third cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located after the intermediate cloud point.

[0086] Step S430: Obtain a first trend value based on the second cloud point set.

[0087] In this embodiment, the first trend value is at least used to characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the second cloud point set.

[0088] As can be seen from the foregoing, regardless of whether the first cloud point set represents a slender obstacle with a bending extension tendency or a slender obstacle extending straight, the changes in the extension angles of each cloud point pointing from the middle cloud point to the two ends are regular. Based on this, step S430, based on the second cloud point set, obtains the first trend value, including steps S431 to S434.

[0089] Step S431: Based on the second cloud point set, obtain a third stretching angle and a fourth stretching angle.

[0090] In this embodiment, the third stretching angle is the stretching angle corresponding to the middle cloud point, and the fourth stretching angle is the stretching angle corresponding to the cloud point in the second cloud point set that is farthest from the middle cloud point.

[0091] Step S432: Obtain a first difference based on the third extension angle and the fourth extension angle.

[0092] In this embodiment, the first difference is the fourth extension angle minus the third extension angle. Based on two values ​​(i.e., the third extension angle and the fourth extension angle), obtaining the difference between the two values ​​(i.e., the first difference) is a mature technology and is not described in detail here.

[0093] Step S433: Based on the second cloud point set, obtain multiple second difference values.

[0094] In this embodiment, the second difference is the absolute value of the difference between the extension angles corresponding to two adjacent cloud points in the second cloud point set. Obtaining the absolute value of the difference between two values ​​(i.e., the extension angles corresponding to two adjacent cloud points) (i.e., the second difference) based on these two values ​​is a mature technique and will not be further described here.

[0095] Step S434: Obtain the first trend value based on the first difference and each second difference.

[0096] In the embodiments of the present application, the first trend value can be obtained based on the first difference and each second difference in any reasonable manner. For example, the first trend value can be the ratio of the first difference to the average of each second difference; or the first trend value can be the ratio of the first difference to the sum of each second difference.

[0097] In a specific embodiment of the present application, in step S434, a calculation formula for obtaining the first trend value based on the first difference and each second difference is as follows: in, Indicates the first trend value; Indicates the stretching angle corresponding to the middle cloud point (also known as the third stretching angle); represents the stretching angle corresponding to the cloud point farthest from the middle cloud point in the second cloud point set (i.e., the fourth stretching angle); b represents the number of cloud points other than the middle cloud point in the second cloud point set; represents the stretching angle corresponding to the a-th cloud point in the second cloud point set; represents the stretching angle corresponding to the a-1th cloud point in the second cloud point set; represents the absolute value; λ represents the anti-zero coefficient, which can be any positive number close to 0, for example, λ can be 0.01 or 0.001.

[0098] In this embodiment, the larger the first trend value A is, the stronger the trend change of the extension angle corresponding to each cloud point in the second cloud point set is, that is, the more likely each cloud point in the second cloud point set is a cloud point formed by a slender obstacle with a bending and extending tendency in space.

[0099] Step S440: Obtain a second trend value based on the third cloud point set.

[0100] In this embodiment, the second trend value is used to at least characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the third cloud point set. It should be noted that the method for obtaining the second trend value can refer to the first trend value and will not be repeated here.

[0101] Step S450: Acquire the stretching trend value based on the first trend value and the second trend value.

[0102] In an embodiment of the present application, the stretching trend value may be obtained based on the first trend value and the second trend value in any reasonable manner. For example, the stretching trend value may be the sum or product of the first trend value and the second trend value.

[0103] In this embodiment, if the first cloud point set represents a slender obstacle with a straight extension trend, the first trend value and the second trend value obtained may both be small. In order to be able to identify whether each cloud point in the first cloud point set is formed by a slender obstacle with a straight extension trend, in one embodiment of the present application, step S450, based on the first trend value and the second trend value, obtains the extension trend value, including steps S460 to S480.

[0104] Step S460: Based on the first cloud point set, obtain a first stretching angle and a second stretching angle.

[0105] In this embodiment, the first extension angle is the maximum extension angle corresponding to each cloud point in the first cloud point set. The second extension angle is the minimum extension angle corresponding to each cloud point in the first cloud point set. It should be noted that obtaining the maximum value (i.e., the maximum extension angle) and the minimum value (i.e., the minimum extension angle) from each value (i.e., each extension angle) is a mature technology and will not be described in detail here.

[0106] Step S470: Acquire a stability value based on the first extension angle and the second extension angle.

[0107] In this embodiment, the stability value is used to at least characterize the difference between the first extension angle and the second extension angle. It should be understood that in the embodiment of the present application, the stability value may be the difference or ratio of the first extension angle and the second extension angle.

[0108] Step S480: Acquire the stretching trend value based on the stability value, the first trend value, and the second trend value.

[0109] In an embodiment of the present application, the stretching trend value can be obtained based on the stability value, the first trend value, and the second trend value in any reasonable manner. For example, in a specific embodiment of the present application, in step S480, the stretching trend value is obtained based on the stability value, the first trend value, and the second trend value using the following calculation formula: in, Indicates the stretch trend value; A indicates the first trend value; B indicates the second trend value; Indicates the first extension angle; Indicates the second extension angle; Indicates the anti-zero coefficient, It can be any positive number close to 0, for example, It can be 0.01 or 0.001, etc.

[0110] In the embodiment of the present application, the larger the first trend value or the second trend value, the larger the stretch trend value; and the smaller the difference between the first stretch angle and the second stretch angle, the larger the stretch trend value. In other words, in this embodiment, the larger the stretch trend value, the greater the probability that each cloud point in the corresponding first cloud point set belongs to the same elongated obstacle.

[0111] In another specific embodiment of the present application, in step S480, a calculation formula for obtaining the stretching trend value based on the stability value, the first trend value, and the second trend value is as follows: in, Indicates the stretch trend value; A indicates the first trend value; B indicates the second trend value; Indicates the first extension angle; Indicates the second extension angle; Indicates the standard deviation of the stretching angle corresponding to each cloud point in the first cloud point set; Indicates the anti-zero coefficient, It can be any positive number close to 0, for example, It can be 0.01 or 0.001, etc.

[0112] In the embodiment of the present application, the larger the first trend value or the second trend value, the larger the stretch trend value. The smaller the difference between the first stretch angle and the second stretch angle, or the standard deviation of the stretch angles corresponding to each cloud point in the first cloud point set, the larger the stretch trend value. In other words, in this embodiment, the larger the stretch trend value, the greater the probability that each cloud point in the corresponding first cloud point set belongs to the same elongated obstacle.

[0113] Step S500: Based on the stretching trend value, obtain the slender obstacle probability of the first cloud point set.

[0114] In this embodiment, the slender obstacle probability is at least used to represent the probability that each cloud point in the first cloud point set belongs to the same slender obstacle.

[0115] In the embodiments of the present application, any reasonable method can be used to obtain the slender obstacle probability of the first cloud point set based on the stretching tendency value. For example, the stretching tendency value can be directly used as the slender obstacle probability of the first cloud point set. As mentioned above, the larger the stretching tendency value, the greater the probability that each cloud point in the first cloud point set belongs to the same slender obstacle.

[0116] It should be clear that if the various cloud points in the first cloud point set in the first frame of point cloud data belong to the same slender obstacle, then the slender obstacle will most likely also appear in the point cloud data of the previous frame or the next frame in the first frame of point cloud data. In other words, if a cloud point set with a high similarity to the first cloud point set appears in the point cloud data adjacent to the first frame of point cloud data, it can better reflect that the various cloud points in the first cloud point set belong to the same slender obstacle. Based on this, in order to obtain a more accurate probability of a slender obstacle in the first cloud point set, in one embodiment of the present application, step S500, based on the stretching trend value, obtains the probability of a slender obstacle in the first cloud point set, including steps S510 to S540.

[0117] Step S510: Based on the ground point cloud data, obtain a second frame of point cloud data.

[0118] In this embodiment, the second frame of point cloud data is point cloud data in the ground point cloud data that is temporally adjacent to the first frame of point cloud data.

[0119] Step S520: Acquire a fourth cloud point set based on the second frame of point cloud data.

[0120] In this embodiment, the fourth cloud point set is the cloud point set with the highest similarity to the first cloud point set in the second frame point cloud data.

[0121] Step S530: Obtain similarity based on the first cloud point set and the fourth cloud point set.

[0122] It should be noted that obtaining the similarity between two cloud point sets is a mature technology in the field of artificial intelligence, such as the chamfer distance algorithm, the bulldozer distance algorithm, or the inlier ratio algorithm.

[0123] Step S540: Based on the similarity and the stretching trend value, obtain the slender obstacle probability of the first cloud point set.

[0124] It is easy to understand that the greater the similarity between the first cloud point set and the fourth cloud point set, the greater the probability that the cloud points in the first cloud point set belong to the same slender obstacle.

[0125] In this embodiment, the slender obstacle probability of the first cloud point set can be obtained based on the similarity and the stretching tendency value in any reasonable manner. For example, the slender obstacle probability can be the sum or product of the similarity and the stretching tendency value.

[0126] In a specific embodiment of the present application, in step S540, based on the similarity and the stretching tendency value, a calculation formula for obtaining the probability of a slender obstacle in the first cloud point set is as follows: in, represents the probability of a slender obstacle in the first cloud point set; P represents the number of frames of point cloud data in the ground point cloud data; Represents a weighting function, which is used to convert the values ​​in the brackets into weight coefficients so that the sum of the weight coefficients is equal to 1; represents the maximum similarity between the fourth cloud point set in the g-th frame point cloud data and each cloud point set in its adjacent frame point cloud data; Indicates the stretching trend value of the fourth cloud point set in the g-th frame point cloud data.

[0127] In this embodiment, if the probability of a slender obstacle is The larger the value, the more similar point cloud sets (i.e., the fourth point cloud set) to the first point cloud set appear continuously in the multi-frame point cloud data. In other words, the greater the probability that each point cloud set in the multi-frame point cloud data represents the same slender obstacle. In other words, the greater the probability that each point cloud set in the first point cloud set represents the same slender obstacle.

[0128] Step S600: constructing a three-dimensional model of the highway service area based on the probability of the slender obstacle.

[0129] In an embodiment of the present application, the area formed by each cloud point in the first cloud point set corresponding to the probability of the slender obstacle being greater than or equal to the third preset value can be used as the area occupied by the slender obstacle in the real space.

[0130] In this embodiment, the third preset value may be selected according to requirements. For example, the third preset value may be equal to 0.7 or 0.78.

[0131] It should be understood that constructing a corresponding three-dimensional model of the elongated obstacle based on each cloud point in the first cloud point set is a mature technology and will not be described in detail here.

[0132] Step S700: Navigate and locate the delivery drone based on the three-dimensional model.

[0133] It should be clear that navigation and positioning of delivery drones based on three-dimensional models of highway service areas is also a mature technology and will not be discussed in detail here.

[0134] The embodiment of the navigation and positioning method for delivery drones in highway service areas proposed in this application identifies point cloud sets containing slender obstacle features in single-frame point cloud data, then matches point cloud sets in multi-frame point cloud data. The method then comprehensively determines whether there are slender obstacles in the delivery area of ​​the highway service area by combining the slender obstacle probabilities of each matched point cloud set and their similarities with adjacent point cloud sets. In other words, the embodiment of the navigation and positioning method for delivery drones in highway service areas proposed in this application can accurately identify slender obstacles in the delivery area of ​​a highway service area.

[0135] After introducing the navigation and positioning method for high-speed service area delivery drones proposed in the embodiment of this application, the following introduces an embodiment of the navigation and positioning system for high-speed service area delivery drones proposed in this application. Figure 2 As shown, the navigation and positioning system 10 for high-speed service area delivery drones includes: The camera device 11 is used to obtain ground point cloud data of the highway service area; the ground point cloud data includes multiple frames of point cloud data; The processing device 12 is configured to traverse the ground point cloud data to obtain a first frame of point cloud data; the first frame of point cloud data is point cloud data in any frame of the ground point cloud data in which the probability of a slender obstacle is not obtained; and traversing the first frame of point cloud data to obtain a first cloud point set; the first cloud point set is any cloud point set in the first frame of point cloud data that has not obtained a stretching trend value; And, based on the first cloud point set, obtaining a stretching trend value; the stretching trend value is at least used to characterize the symmetry of each cloud point in the first cloud point set; and, based on the stretching tendency value, obtaining a slender obstacle probability of the first cloud point set; the slender obstacle probability is at least used to represent the probability that each cloud point in the first cloud point set belongs to the same slender obstacle; and, constructing a three-dimensional model of the highway service area based on the probability of the elongated obstacle; The control device 13 performs navigation and positioning on the delivery drone based on the three-dimensional model.

[0136] As a specific embodiment of the present application, the processing device 12 is further configured to obtain a second frame of point cloud data based on the first frame of point cloud data; the second frame of point cloud data is point cloud data obtained by removing all ground cloud points from the first frame of point cloud data; And, based on the second frame of point cloud data, obtaining a third frame of point cloud data; the third frame of point cloud data is the point cloud data of the second frame of point cloud data excluding all cloud points corresponding to non-slender obstacles; and traversing the third frame of point cloud data to obtain a first cloud point; the first cloud point is any cloud point in the third frame of point cloud data that is not included in any cloud point set; And, based on the first cloud point, a plurality of near cloud points are obtained; the near cloud points are cloud points in the third frame point cloud data whose distance from the first cloud point is less than or equal to a first preset value, or cloud points in the third frame point cloud data whose distance from any near cloud point is less than or equal to a first preset value; and forming a set of the first cloud point and each of the near cloud points to obtain a cloud point set; And, traverse each cloud point set to obtain the first cloud point set.

[0137] As a specific embodiment of the present application, the processing device 12 is further configured to traverse the first frame of point cloud data to obtain a second cloud point and a third cloud point; the second cloud point is any cloud point in the first frame of point cloud data that has not been confirmed to be a ground cloud point; the third cloud point is the cloud point with the lowest vertical height in the first frame of point cloud data; And, based on the second cloud point and the third cloud point, obtaining a first distance; the first distance is the distance between the second cloud point and the third cloud point in the vertical direction; And, if the first distance is less than or equal to a second preset value, the second cloud point is eliminated; otherwise, the second cloud point is retained.

[0138] As a specific embodiment of the present application, the processing device 12 is further configured to obtain a plurality of clusters from the second frame of point cloud data by using a density-based clustering algorithm; Furthermore, cloud points belonging to each cluster are removed from the second frame of point cloud data to obtain the third frame of point cloud data.

[0139] As a specific embodiment of the present application, the processing device 12 is further configured to obtain a first near cloud point and a second near cloud point; the first near cloud point and the second near cloud point are obtained based on an update; the initial first near cloud point is the first cloud point, and the initial second near cloud point is any cloud point in the third frame of point cloud data whose distance from the first near cloud point is less than or equal to a first preset value; and, based on the first near cloud point and the second near cloud point, obtaining an extension direction; wherein the extension direction points from the first near cloud point to the second near cloud point; And, based on the second near cloud point and the extension direction, a third near cloud point is obtained, and the first near cloud point and the second near cloud point are updated; the third near cloud point is the cloud point in the third frame point cloud data that is closest to the second near cloud point along the extension direction, and the distance between the second near cloud point and the third near cloud point is less than or equal to a first preset value; the updated first near cloud point is equal to the second near cloud point before the update, and the updated second near cloud point is equal to the third near cloud point before the update.

[0140] As a specific embodiment of the present application, the processing device 12 is further configured to obtain an intermediate cloud point based on the first cloud point set; the intermediate cloud point is a cloud point with the smallest extension angle in the first cloud point set; And, based on the first cloud point set and the intermediate cloud point, a second cloud point set and a third cloud point set are obtained; the second cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located before the intermediate cloud point; the third cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located after the intermediate cloud point; and, based on the second cloud point set, obtaining a first trend value; the first trend value is at least used to characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the second cloud point set; and obtaining a second trend value based on the third cloud point set; wherein the second trend value is at least used to characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the third cloud point set; And, based on the first trend value and the second trend value, the stretching trend value is obtained.

[0141] As a specific embodiment of the present application, the processing device 12 is further configured to obtain a first stretching angle and a second stretching angle based on the first cloud point set; the first stretching angle is the maximum stretching angle corresponding to each cloud point in the first cloud point set; and the second stretching angle is the minimum stretching angle corresponding to each cloud point in the first cloud point set. and obtaining a stability value based on the first extension angle and the second extension angle; wherein the stability value is at least used to characterize the difference between the first extension angle and the second extension angle; And, based on the stability value, the first trend value and the second trend value, the stretching trend value is obtained.

[0142] As a specific embodiment of the present application, the processing device 12 is further configured to obtain a third stretching angle and a fourth stretching angle based on the second cloud point set; the third stretching angle is the stretching angle corresponding to the intermediate cloud point; and the fourth stretching angle is the stretching angle corresponding to the cloud point in the second cloud point set that is farthest from the intermediate cloud point. and obtaining a first difference based on the third extension angle and the fourth extension angle; wherein the first difference is the fourth extension angle minus the third extension angle; and, based on the second cloud point set, obtaining a plurality of second difference values; the second difference value is an absolute value of a difference between extension angles corresponding to two adjacent cloud points in the second cloud point set; And, based on the first difference value and each second difference value, the first trend value is obtained.

[0143] As a specific embodiment of the present application, the processing device 12 is further configured to obtain a second frame of point cloud data based on the ground point cloud data; the second frame of point cloud data is point cloud data in the ground point cloud data that is temporally adjacent to the first frame of point cloud data; And, based on the second frame of point cloud data, obtaining a fourth cloud point set; the fourth cloud point set is the cloud point set with the highest similarity to the first cloud point set in the second frame of point cloud data; and, obtaining similarity based on the first cloud point set and the fourth cloud point set; And, based on the similarity and the stretching tendency value, a slender obstacle probability of the first cloud point set is obtained.

[0144] The embodiment of the navigation and positioning system for delivery drones in highway service areas proposed in this application identifies point cloud sets containing features of slender obstacles in single-frame point cloud data, then matches point cloud sets in multi-frame point cloud data. The system then comprehensively determines whether there are slender obstacles in the delivery area of ​​the highway service area by combining the slender obstacle probabilities of each matched point cloud set and their similarities with adjacent point cloud sets. In other words, the embodiment of the navigation and positioning system for delivery drones in highway service areas proposed in this application can accurately identify slender obstacles in the delivery area of ​​a highway service area.

[0145] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0146] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the methods, devices and equipment described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0147] In the several embodiments provided in the embodiments of the present application, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is merely a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0148] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of the present embodiment according to actual needs.

[0149] In addition, the functional modules in each embodiment of the present application can be integrated into a processing module, or each module can exist physically separately, or two or more modules can be integrated into a module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software functional modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium.

[0150] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware, or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0151] The computer program product includes one or more computer instructions. When the computer program is loaded and executed on a computer, the processes or functions described in accordance with the embodiments of the present application are fully or partially generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium may be any available medium that can be stored on a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., a floppy disk, hard disk, or magnetic tape), an optical medium (e.g., a digital versatile disk), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0152] Although the embodiments of the present application have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions, and alterations may be made to these embodiments without departing from the principles of the present application.

Claims

1. A navigation and positioning method for delivery drones in high-speed service areas, characterized in that: include: Obtain ground point cloud data of highway service areas; The ground point cloud data includes multiple frames of point cloud data; Traversing the ground point cloud data to obtain a first frame of point cloud data; The first frame of point cloud data is point cloud data of any frame of the ground point cloud data in which the probability of a slender obstacle is not obtained; Traversing the first frame of point cloud data to obtain a first cloud point set; The first cloud point set is any cloud point set in the first frame of point cloud data for which a stretching trend value has not been obtained; Based on the first cloud point set, obtaining a stretching trend value; The stretching trend value is at least used to characterize the symmetry of each cloud point in the first cloud point set; Based on the stretching tendency value, obtaining a slender obstacle probability of the first cloud point set; The slender obstacle probability is at least used to represent the probability that each cloud point in the first cloud point set belongs to the same slender obstacle; Constructing a three-dimensional model of the highway service area based on the probability of the slender obstacle; The delivery drone is navigated and positioned based on the three-dimensional model.

2. The navigation and positioning method for high-speed service area delivery drones according to claim 1 is characterized in that: The traversing the first frame of point cloud data to obtain a first cloud point set includes: Based on the first frame of point cloud data, a second frame of point cloud data is obtained; the second frame of point cloud data is the point cloud data of the first frame of point cloud data excluding all ground cloud points; Based on the second frame of point cloud data, a third frame of point cloud data is obtained; the third frame of point cloud data is the point cloud data of the second frame of point cloud data excluding the cloud points corresponding to all non-slender obstacles; Traversing the third frame of point cloud data to obtain a first cloud point; the first cloud point is any cloud point in the third frame of point cloud data that is not included in any cloud point set; Based on the first cloud point, a plurality of near cloud points are obtained; the near cloud points are cloud points in the third frame point cloud data whose distance from the first cloud point is less than or equal to a first preset value, or cloud points in the third frame point cloud data whose distance from any near cloud point is less than or equal to a first preset value; The first cloud point and each of the near cloud points are formed into a set to obtain a cloud point set; Traverse each cloud point set to obtain the first cloud point set.

3. The navigation and positioning method for high-speed service area delivery drones according to claim 2, characterized in that: The step of acquiring a second frame of point cloud data based on the first frame of point cloud data includes: Traversing the first frame of point cloud data to obtain a second cloud point and a third cloud point; the second cloud point is any cloud point in the first frame of point cloud data that has not been confirmed to be a ground cloud point; the third cloud point is the cloud point with the lowest vertical height in the first frame of point cloud data; Based on the second cloud point and the third cloud point, obtaining a first distance; the first distance is the distance between the second cloud point and the third cloud point in the vertical direction; If the first distance is less than or equal to a second preset value, the second cloud point is eliminated; otherwise, the second cloud point is retained.

4. The navigation and positioning method for high-speed service area delivery drones according to claim 2, characterized in that: The step of acquiring a third frame of point cloud data based on the second frame of point cloud data includes: Obtaining a plurality of clusters from the second frame of point cloud data by using a density-based clustering algorithm; The cloud points belonging to each cluster are removed from the second frame of point cloud data to obtain the third frame of point cloud data.

5. The navigation and positioning method for high-speed service area delivery drones according to claim 2, characterized in that: The acquiring a plurality of near cloud points based on the first cloud point includes: Obtain a first near cloud point and a second near cloud point; the first near cloud point and the second near cloud point are obtained based on the update; the initial first near cloud point is the first cloud point, and the initial second near cloud point is any cloud point in the third frame of point cloud data whose distance from the first near cloud point is less than or equal to a first preset value; Based on the first near cloud point and the second near cloud point, acquiring an extension direction; the extension direction is from the first near cloud point to the second near cloud point; Based on the second near cloud point and the extension direction, a third near cloud point is obtained, and the first near cloud point and the second near cloud point are updated; the third near cloud point is the cloud point in the third frame point cloud data that is closest to the second near cloud point along the extension direction, and the distance between the second near cloud point and the third near cloud point is less than or equal to a first preset value; the updated first near cloud point is equal to the second near cloud point before the update, and the updated second near cloud point is equal to the third near cloud point before the update.

6. The navigation and positioning method for delivery drones in high-speed service areas according to any one of claims 1 to 5, characterized in that: The obtaining of the stretching trend value based on the first cloud point set includes: Based on the first cloud point set, an intermediate cloud point is obtained; the intermediate cloud point is a cloud point with the smallest extension angle in the first cloud point set; Based on the first cloud point set and the intermediate cloud point, a second cloud point set and a third cloud point set are obtained; the second cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located before the intermediate cloud point; the third cloud point set includes the intermediate cloud point and all cloud points in the first cloud point set that are located after the intermediate cloud point; Based on the second cloud point set, a first trend value is obtained; the first trend value is used to at least characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the second cloud point set; Based on the third cloud point set, a second trend value is obtained; the second trend value is used to at least characterize the uniformity of the change in the extension angle between the intermediate cloud point and each cloud point in the third cloud point set; The stretching trend value is obtained based on the first trend value and the second trend value.

7. The navigation and positioning method for high-speed service area delivery drones according to claim 6, characterized in that: The acquiring the stretching trend value based on the first trend value and the second trend value includes: Based on the first cloud point set, obtaining a first stretching angle and a second stretching angle; the first stretching angle is the maximum stretching angle corresponding to each cloud point in the first cloud point set; the second stretching angle is the minimum stretching angle corresponding to each cloud point in the first cloud point set; Obtaining a stability value based on the first extension angle and the second extension angle; wherein the stability value is at least used to characterize a difference between the first extension angle and the second extension angle; The stretching trend value is obtained based on the stability value, the first trend value, and the second trend value.

8. The navigation and positioning method for high-speed service area delivery drones according to claim 7, characterized in that: The obtaining of a first trend value based on the second cloud point set includes: Based on the second cloud point set, obtaining a third stretching angle and a fourth stretching angle; the third stretching angle is the stretching angle corresponding to the middle cloud point; the fourth stretching angle is the stretching angle corresponding to the cloud point in the second cloud point set that is farthest from the middle cloud point; Obtaining a first difference based on the third extension angle and the fourth extension angle; the first difference is the fourth extension angle minus the third extension angle; Based on the second cloud point set, a plurality of second difference values ​​are obtained; the second difference value is an absolute value of a difference between extension angles corresponding to two adjacent cloud points in the second cloud point set; The first trend value is obtained based on the first difference value and each second difference value.

9. The navigation and positioning method for high-speed service area delivery drones according to claim 8, characterized in that: The obtaining, based on the stretching tendency value, a probability of a slender obstacle in the first cloud point set includes: Based on the ground point cloud data, a second frame of point cloud data is acquired; the second frame of point cloud data is point cloud data in the ground point cloud data that is temporally adjacent to the first frame of point cloud data; Based on the second frame of point cloud data, a fourth cloud point set is obtained; the fourth cloud point set is the cloud point set with the highest similarity to the first cloud point set in the second frame of point cloud data; Obtaining similarity based on the first cloud point set and the fourth cloud point set; Based on the similarity and the stretching tendency value, a slender obstacle probability of the first cloud point set is obtained.

10. A navigation and positioning system for delivery drones in high-speed service areas, characterized in that: include: A camera device is used to obtain ground point cloud data of a highway service area; the ground point cloud data includes multiple frames of point cloud data; A processing device, configured to traverse the ground point cloud data and obtain a first frame of point cloud data; The first frame of point cloud data is point cloud data of any frame of the ground point cloud data in which the probability of a slender obstacle is not obtained; And, traverse the first frame of point cloud data to obtain a first cloud point set; The first cloud point set is any cloud point set in the first frame of point cloud data for which a stretching trend value has not been obtained; and, based on the first cloud point set, obtaining a stretching trend value; The stretching trend value is at least used to characterize the symmetry of each cloud point in the first cloud point set; and, based on the stretching tendency value, obtaining a slender obstacle probability of the first cloud point set; The slender obstacle probability is at least used to represent the probability that each cloud point in the first cloud point set belongs to the same slender obstacle; and, constructing a three-dimensional model of the highway service area based on the probability of the elongated obstacle; A control device is used to navigate and locate the delivery drone based on the three-dimensional model.

Images