CN117664128A - Scheduling path generation method, generation device, robot, and storage medium - Google Patents

Abstract

The application discloses a scheduling path generating method, a scheduling path generating device, a robot and a computer readable storage medium, wherein the scheduling path generating method comprises the following steps: performing image processing on a grid map corresponding to the robot operation area to generate a skeleton map of a passable road in the grid map; determining skeleton nodes and skeleton endpoints of a passable road from each skeleton pixel point in the skeleton map; determining a path point of a passable road based on the skeleton node and the skeleton end point; for a target point of a scheduling task, determining a target path point closest to the target point from the path points; and generating and storing a scheduling path corresponding to the target point based on the target path point. The method can efficiently generate the scheduling path at a low cost.

Description

Scheduling path generation method, generation device, robot, and storage medium

Technical Field

The application belongs to the technical field of robot navigation, and particularly relates to a scheduling path generation method, a scheduling path generation device, a robot and a computer readable storage medium.

Background

In the scene of the operation of a plurality of robots, when the robots work simultaneously, the conditions of the robots occupying the route and the target point and the like easily occur, which may cause the collision between the robots, not only reduce the distribution efficiency, but also may cause the collision risk. To cope with this problem, the related art manually draws a scheduling path in advance on a map to generate a scheduling diagram, and then coordinates jobs of a plurality of robots according to the scheduling diagram to improve safety and efficiency of robot operation. However, conventional dispatch path drawing typically requires manual operations, which are not only time consuming and labor intensive, but also make deployment of the dispatch map relatively costly.

Disclosure of Invention

Provided are a scheduling path generation method, a scheduling path generation device, a robot, and a computer-readable storage medium, which can efficiently generate a scheduling path at a low cost.

In a first aspect, the present application provides a method for generating a scheduling path, including:

performing image processing on a grid map corresponding to the robot operation area to generate a skeleton map of a passable road in the grid map;

determining skeleton nodes and skeleton endpoints of a passable road from each skeleton pixel point in the skeleton map;

determining a path point of a passable road based on the skeleton node and the skeleton end point;

for a target point of a scheduling task, determining a target path point closest to the target point from the path points;

and generating and storing a scheduling path corresponding to the target point based on the target path point.

In a second aspect, the present application provides a scheduling path generating apparatus, including:

the first generation module is used for carrying out image processing on the grid map corresponding to the robot operation area and generating a skeleton diagram of the passable road in the grid map;

the first determining module is used for determining skeleton nodes and skeleton endpoints of the passable road from each skeleton pixel point in the skeleton map;

The second determining module is used for determining the path point of the passable road based on the skeleton node and the skeleton end point;

the third determining module is used for determining a target path point closest to the target point from the path points aiming at the target point of the scheduling task;

and the second generation module is used for generating and storing a scheduling path corresponding to the target point based on the target path point.

In a third aspect, the present application provides a robot comprising a memory, a processor and a computer program stored in said memory and executable on said processor, said processor implementing the steps of the method according to the first aspect when said computer program is executed.

In a fourth aspect, the present application provides a computer readable storage medium storing a computer program which, when executed by a processor, performs the steps of the method of the first aspect described above.

In a fifth aspect, the present application provides a computer program product comprising a computer program which, when executed by one or more processors, implements the steps of the method of the first aspect described above.

Compared with the prior art, the beneficial effects that this application exists are: the scheduling task is typically triggered during the robot job, and in order to efficiently generate the scheduling path at a lower cost, the grid map of the robot job area may be optimized. The scheduling path is mainly generated based on passable roads in the grid map, so that in order to eliminate obstacles in the grid map and effectively utilize the passable paths, the shape characteristics of the passable paths can be reserved, the skeletons of the passable paths in the grid map are extracted, and corresponding skeleton diagrams are formed. In order to reduce the complexity of the generation of the dispatching path and improve the generation efficiency of the dispatching path, each skeleton pixel point can be traversed, key points of the passable road, namely skeleton nodes and skeleton endpoints, are determined, and the skeleton nodes and the skeleton endpoints are determined to be the path points of the passable road. In order to generate all available paths related to the scheduling task, the target point of the scheduling task can be bound with the nearest path point to determine the corresponding target path point, and finally, the scheduling path corresponding to the target point can be automatically and efficiently generated and stored according to the target path point and the rest path points. When the dispatching task is triggered, the robot can quickly call the stored dispatching path according to the current position of the robot and the target point corresponding to the dispatching task, so that the dispatching reliability is improved.

It will be appreciated that the advantages of the second to fifth aspects may be found in the relevant description of the first aspect, and are not described here again.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following description will briefly introduce the drawings that are needed in the embodiments or the description of the prior art, it is obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a schematic structural diagram of a robot provided in an embodiment of the present application;

fig. 2 is a flow chart of a method for generating a scheduling path according to an embodiment of the present application;

FIG. 3 is a schematic diagram of a grid map provided by an embodiment of the present application;

FIG. 4 is a schematic diagram of a local grid map labeled with target points and waypoints provided by an embodiment of the present application;

FIG. 5 is a schematic diagram of a virtual wall map provided by an embodiment of the present application;

FIG. 6 is a schematic diagram of a grid map with topological paths drawn, provided by an embodiment of the present application;

fig. 7 is a schematic structural diagram of a scheduling path generating device provided in an embodiment of the present application;

Fig. 8 is a schematic structural diagram of a robot according to an embodiment of the present application.

Detailed Description

In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system configurations, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. It will be apparent, however, to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.

In the related art, the operation of a plurality of robots is coordinated by generating a scheduling diagram, so that the running safety and efficiency of the robots are improved. However, conventional dispatch path drawing typically requires manual operations, which are not only time consuming and labor intensive, but also make deployment of the dispatch map relatively costly.

In order to solve the problem, the present application proposes a scheduling path generation method capable of efficiently generating a scheduling path at a low cost. The control method proposed in the present application will be described below by way of specific examples.

The method for generating the scheduling path of the robot can be applied to the robot, can also be applied to the electronic equipment or cloud server and the like which can establish communication connection with the robot and control the robot based on the communication connection. Such as a mobile phone, a tablet computer, a vehicle-mounted device, an augmented reality (augmented reality, AR)/Virtual Reality (VR) device, a notebook computer, an ultra-mobile personal computer, a UMPC, a netbook, a personal digital assistant (personal digital assistant, PDA), and the like, the embodiments of the present application do not limit the specific type of the electronic device.

By way of example only, referring to fig. 1, the physical structure of the robot may include a housing 10 carrying an article, a mobile chassis 11, a function controller providing user operation, an underlying controller for map generation and path planning, and an element controller controlling a mobile unit and an environment detection unit; the mobile chassis is provided with at least two groups of driving wheels 110, and each group of driving wheels 110 is respectively positioned at one side of the mobile chassis 11; the element controller controls the traveling speed of the driving wheel 110.

The mobile unit is provided with at least two groups of driving wheels 110, and each group of driving wheels 110 is respectively positioned on one side of the chassis 10; the element controller controls the traveling speed of the driving wheel 110. Specifically, of the drive wheels 110 provided in the mobile unit, at least one set of drive wheels 110 serves as a left drive wheel, while at least one set of drive wheels 110 serves as a right drive wheel, the left and right drive wheels being located on opposite sides of the chassis 10. Optionally, the moving unit may further include at least two sets of driven wheels, one set of driven wheels corresponds to one set of driven wheels, where at least one set of driven wheels is used as a left driven wheel, and at least one set of driven wheels is used as a right driven wheel, where the left driven wheel and the right driven wheel are used to assist the left driving wheel and the right driving wheel to drive the housing 10 and the chassis 11 of the robot to move, so as to reduce the load pressure of the driving wheel 110.

In order to illustrate the technical solutions proposed in the present application, the following description will be given of each embodiment.

Fig. 2 shows a schematic flowchart of a method for generating a scheduling path, which is provided by the application and is executed by a robot, and may include:

step 210, the robot performs image processing on the grid map corresponding to the operation area, and generates a skeleton map of the passable road in the grid map.

Generally, the scheduling task is triggered when a collision or possible collision occurs in a scenario where multiple robots work simultaneously. In order to enable each robot to complete a scheduling task through a scheduling path and continue subsequent jobs, the robot may process a grid map corresponding to a job area thereof so as to generate the scheduling path at a low cost and efficiently.

For example, an exemplary diagram of a grid map may refer to FIG. 3, which may be generated by a robot prior to a task by scanning a work area based on its own equipped radar system. Correspondingly, the scheduling path is generated in the process of deploying the grid map after the drawing is completed, and the scheduling path is generated in advance, so that the scheduling path is beneficial to direct calling in the running process of the robot, and the scheduling efficiency can be improved. For a small service robot, such as a meal delivery robot, the calculation power is limited, the concurrency capability is not high, and the calculation power of the robot during working can be effectively saved by generating a scheduling path in advance, so that the stability of a motion sensing system of the robot during working is enhanced, and the stability of delivery is ensured.

The dispatching path is mainly generated based on passable roads in the grid map, and the robot can generate a skeleton map by extracting skeletons of the passable roads in the grid map.

And 220, the robot determines skeleton nodes and skeleton endpoints of the passable road from each skeleton pixel point in the skeleton map.

Step 230, the robot determines the path point of the passable road based on the skeleton node and the skeleton end point.

After obtaining the skeleton diagram, in order to reduce complexity of scheduling path generation and improve scheduling path generation efficiency, the robot may generate the scheduling path based on the key points of the passable road. In view of the fact that the skeleton nodes and the skeleton endpoints are located at key positions of the passable road, the skeleton nodes and the skeleton endpoints are used as key points, topology features of the passable road can be effectively reserved, and the generated scheduling path is visual and concise. In particular, the skeletal nodes may include intersections.

The skeleton node and the skeleton end point can be considered to represent main path points on the passable road, and the extension and crossing conditions of the passable road can be reflected, so that the robot can determine the skeleton node and the skeleton end point as the path points of the passable road. If adjacent path points are connected with each other, a topology path on the passable road can be obtained, so that the generation of the scheduling path is more efficient and convenient.

Step 240, the robot determines a target path point closest to the target point from the path points for the target point of the scheduling task.

In order to generate all available paths related to the scheduled task based on the topology path, the robot may bind the target point of the scheduled task with the nearest path point, thereby determining a target path point corresponding to the target point. It will be appreciated that the number of target points of the scheduling task is at least one, whereby each target point may be individually bound to the corresponding nearest waypoint to determine each target waypoint.

For example only, assuming that the robot is a meal delivery robot, the target point may be a position that may be used for avoiding and stopping when triggering scheduling in the meal delivery process, for example, a delivery point, a meal delivery point, an origin point, a fixed avoiding point, a charging pile point, and the like of a dining table corresponding to a meal delivery task.

By way of example only, assuming that there are N possible target points for the dispatch task, for each target point the robot may match it with a target waypoint for which the distance to the target point is closest.

For example only, for each target point, the robot may determine a distance between the target point and each of the waypoints, obtain a plurality of distances, determine a minimum distance therefrom, and determine a waypoint corresponding to the minimum distance as a target waypoint.

Alternatively, referring to fig. 4, in determining the distance between the target point n and each of the waypoints, an abnormal distance determined by crossing an obstacle (e.g., a virtual wall) should be excluded. For example, although n and waypoint M ₁ The distance between them is the shortest, but N and M ₁ Spaced apart by walls, if M ₁ The target path point determined as n may cause inaccuracy of the planned scheduling path. Thus, the robot can combine n with M ₁ The distance between them is deleted as an abnormal distance. After a valid distance is maintained, M can be determined based on the minimum distance ₂ And determining a target path point as n, and ensuring that the target path point is the path point with the smallest actual passing distance with the target point.

Optionally, after the skeleton map is generated, a skeleton in a certain range near the target point can be reserved according to the coordinates of the target point, so that the target point can be ensured to be reached through the skeleton.

Step 250, the robot generates and stores a dispatch path corresponding to the target point based on the target path point.

After each target path point is obtained, the robot can generate a scheduling path corresponding to the scheduling task, namely an optional scheduling path corresponding to the target point according to the target path point. In order to facilitate the transfer of the robot, after the robot generates the dispatching path, the dispatching path can be stored and can be stored in the local or server of the robot, so that the dispatching task can be transferred quickly after being triggered, and the dispatching efficiency is improved.

Illustratively, the present embodiment can bind the nearest route point for each possible used scheduling point (origin, charging pile point, etc.) through skeleton diagram processing, and generate and store in advance all possible scheduling paths for each route point on the map that can reach the route point. When the robot runs to trigger the scheduling, the path point closest to the robot can be determined according to the self-positioning, the scheduling point to be transmitted is determined according to the scheduling rule, the stored scheduling path is directly obtained according to the information of the two points, and the path planning time is greatly reduced.

Illustratively, each scheduling path is stored at the time of storage, as well as path points (including skeleton nodes, skeleton endpoints) and corresponding type information, such as how many coordinates a certain path point is and what type is the skeleton endpoint. The method is convenient for directly using the dispatching path during dispatching and also is convenient for quickly calculating by using path point information when real-time dispatching calculation is needed.

In the embodiment, a robot firstly extracts a skeleton of a passable path from a grid map corresponding to an operation area of the robot to obtain a skeleton map; and then determining key points of the passable path, namely skeleton nodes and skeleton endpoints from the skeleton diagram, and determining the skeleton nodes and the skeleton endpoints as path points to obtain the topological path formed by connecting the path points. In order to generate all available paths related to the scheduling task based on the topology path, the robot may bind the target point of the scheduling task with the corresponding nearest path point, thereby determining a target path point, and then the robot may automatically generate and store the scheduling path corresponding to the scheduling task according to the target path point with high efficiency. In view of the fact that the scheduling paths are generated in the map deployment stage, correspondingly, if corresponding scheduling tasks are triggered in the running process of the robot, the stored scheduling paths can be directly called, and scheduling efficiency is improved.

In some embodiments, in some application scenarios, some areas are open but belong to a manually set forbidden area, for example, for a robot delivering a meal, although it may reach the kitchen, since the kitchen is generally not open to the outside, therefore belongs to a forbidden area, and the robot should be prevented from entering the area. In order to avoid that the dispatch path passes through the forbidden area, the robot may further execute the following steps before performing image processing on the grid map corresponding to the working area:

and A1, acquiring a virtual wall map of a working area by a robot.

In order to meet special requirements in the practical application process, the robot can acquire a virtual wall map corresponding to the operation area. Referring to fig. 5, the virtual wall map may include a robot self-complement boundary, and a forbidden region and forbidden lines manually set in an actual application scene. In combination with the virtual wall map, the robot is facilitated to better understand and adapt to the working environment of the robot, and the robot is limited to move in a designated area so as to prevent the robot from crossing the boundary or entering a forbidden area.

In order to be able to label forbidden areas and forbidden lines in the virtual wall map into the grid map, the two maps can be combined to generate a corresponding fusion map. Wherein, in order to achieve accurate fusion, the size of the grid map may be consistent with the size of the virtual wall map. It will be appreciated that the same map size, map scale and coordinate system facilitates accurate and rapid fusion of the two maps.

And A2, the robot calculates the union of the values of each pair of pixel points with the same coordinate information in the virtual wall map and the grid map to obtain each updated pixel point.

And A3, the robot updates the grid map based on each updated pixel point and each coordinate information.

In particular, the process of generating a fused map involves a union operation of the values of pixels in the two maps that have the same coordinate information. The robot completes this process by performing the following steps for each coordinate information: firstly, a robot determines a pair of pixel points corresponding to the coordinate information, wherein one pixel point is from a pixel point corresponding to the coordinate information in a grid map, and the other pixel point is from a pixel point corresponding to the coordinate information in a virtual wall map; and then, performing union operation on the values of the two pixels to obtain the value of the new pixel corresponding to the coordinate information. Specifically, all the pixels in the virtual wall map can be traversed according to the coordinates, and if the pixel value of the obstacle/virtual wall point is 0, the pixel value of the pixel point corresponding to the coordinates in the grid map is updated to 0 when the pixel point with the pixel value of 0 is encountered. That is, for a pixel point, as long as the pixel point is an unvented point in either the grid map or the virtual wall map, the updated grid map is an unvented point. By the method of combining the pixel values, the map, particularly the obstacle, can be combined quickly and efficiently.

In general, the grid map may be updated by determining a value of a new pixel corresponding to each coordinate information and using each new pixel and each coordinate information, to obtain an updated grid map. This is equivalent to constructing a new fusion map, replacing the grid map with the fusion map, and supplementing the information on the virtual wall map, such as the forbidden region, to the new grid map. The optimization method is favorable for seamlessly fusing the information of the two maps, and provides more accurate and comprehensive environment perception for the robot so as to meet the special requirements in the practical application process.

Accordingly, in order to optimize the accuracy and reliability of the scheduling path generation. The foregoing step 210 specifically includes: and performing image processing on the updated grid map.

In this embodiment, by combining the information of the grid map and the virtual wall map, the robot can obtain a comprehensive map which is more accurate and closer to the actual demand, that is, an updated grid map, thereby improving the accuracy and reliability of the scheduling path generation. The optimization method is helpful for ensuring that the robot perceives obstacles, boundaries and other key features more accurately in the actual environment, so that a more reliable basis is provided for scheduling path generation.

In some embodiments, in order to simplify the structure of the passable path as much as possible on the premise of retaining the shape characteristics of the passable road, so as to improve the generation efficiency of the scheduling path, the robot may extract the skeleton of the passable road and generate the skeleton map by the following steps:

step 211, the robot converts the grid map into a channel map with a preset number of bits.

Step 212, the robot performs smoothing operation on the channel map to obtain a binary map.

The robot may perform a "binary image refinement" operation on the grid map in order to be able to extract skeleton contours of passable roads, whether based on the grid map or based on the updated grid map. The main purpose of the binary image refinement is to extract the main skeleton or central axis of the object or shape from the binary image. This process typically involves refining the object boundaries or contours in the binary image to eliminate unnecessary detail, preserving the primary shape features, and thus a more compact and representative central axis or skeleton.

Therefore, the robot can extract the main skeleton outline of the passable road in the grid map by performing a binary image refinement operation on the grid map. In the process, the influence of a part of obstacles can be eliminated, the main shape characteristics of the passable road are reserved, and the skeleton profile of the passable road which is more compact and has stronger identification is obtained.

In view of the fact that the basis of this operation is a binary image, before the "binary image refinement" operation is performed on the grid map, the robot converts the grid map into a binary image for subsequent extraction of the skeleton contour.

Specifically, in order to perform the subsequent "binary image refinement" operation, the robot first converts the grid map into a channel map of a preset number of bits, for example, a channel map of 8 bits. The one-channel image, namely the gray image, can be endowed with a gray value for each pixel point, which is beneficial to simplifying image information, thereby reducing the complexity of subsequent image processing. In addition, in order to prevent noise and fine structures of irregularities in the image from affecting extraction of skeleton contours, the robot may then perform a smoothing operation on a channel map. Through the smoothing operation, the generated binary image can be ensured to be clearer and more continuous, and the stability and the accuracy of the binary image refining operation are improved.

Where smoothing is typically implemented using filters or convolution products. Illustratively, gaussian filtering (Gaussian Smoothing), mean Smoothing (Mean Smoothing), bilateral filtering (Bilateral Filtering) and the like can be used to smooth a channel image, so as to obtain a more accurate binary image, so as to extract skeleton contours of passable roads with follow-up.

And 213, the robot extracts the width information of the road corresponding to each pixel point from the binary image.

And 214, the robot extracts the skeleton outline of the passable road from the binary image based on the width information to generate a skeleton image.

In order to extract the skeleton outline of the passable road more finely, the robot first extracts the width information of the road corresponding to each pixel point from the binary image. The extraction of the width information not only can more comprehensively extract the characteristics of the passable road, but also is helpful for accurately describing the actual characteristics of the road. By acquiring the width information, the robot can more accurately understand the actual width change and detail change of the road. This process helps to capture morphological features of the road, such as that some segments are wide, some segments are narrow, and some segments are rugged, which makes skeleton contours not limited to straight lines but also to diverse representations such as curves.

Such detailed width information is critical to understanding the structure and characteristics of the road network. The robot can better identify and analyze the structure and the hierarchical relationship of the road through the accurate perception of the width information. Such a process flow helps to avoid inaccurate or distorted contour extraction due to lack of width information, thereby constructing a more accurate skeleton map. By combining the width information and the skeleton outline, the robot can describe the road network more comprehensively and carefully, and provides a more powerful foundation for subsequent processing and analysis.

For example, a specific direction may be preset to extend from a pixel point to two sides of a road in a straight line, and a distance between two intersection points of the specific direction and a road boundary is width information of the road corresponding to the pixel point. For example, the width information of each pixel point may be extracted directly using the distance conversion function. Such as cv in OpenCV. This function effectively calculates the distance of each pixel point in the binary image to the nearest non-zero pixel, thereby efficiently providing characteristic information of the communicable paths in the binary image.

Illustratively, instead of the distance conversion function, the binary map refinement may also be implemented by some special refinement algorithm. Specifically, the Zhang-Suen refinement algorithm, guo-Hall refinement algorithm and the like can be included. The algorithms aim to extract the main skeleton contour of the object in the image on the premise of retaining the topological features of the passable road. The following is a brief description of these two algorithms:

the Zhang-Suen refinement algorithm is an iterative algorithm that gradually refines the binary image by repeatedly performing two sub-steps. These two sub-steps involve the deletion and marking of pixels, respectively. The algorithm will iterate until no more pixels can be deleted. The algorithm has the advantage of being able to extract a relatively fine skeleton contour while maintaining image connectivity.

The Guo-Hall refinement algorithm is also an iterative algorithm which thinks to gradually extract the main contour of the image through multiple sub-steps. Similar to the Zhang-Suen algorithm, the Guo-Hall algorithm deletes pixels in iterations and finally gets a refined skeleton. The algorithm may be more rapid and robust in some cases.

It can be appreciated that different refinement algorithms have different advantages in terms of extracting skeleton contours, and that which refinement algorithm is specifically adopted can be selected according to specific application scenarios and requirements.

Optionally, a Zhang-Suen refinement algorithm is adopted for frameworks in a certain range near the target point to ensure connectivity, a Guo-Hall refinement algorithm is adopted for frameworks at other positions to improve efficiency, and reliability and generation efficiency of the framework graph are balanced.

In this embodiment, the robot successfully refines the fine skeleton outline of the passable road by performing a binary image refinement operation, i.e., image skeleton extraction, on the grid map. The operation filters the influence of the obstacle, and simultaneously reserves the original shape of the passable road as much as possible, so that the robot perceives the passable area more accurately when planning the scheduling path, the precision and the efficiency of path planning are improved, and the robot can deal with the navigation challenges of the complex environment more flexibly and intelligently.

In some embodiments, to determine a pixel point at a key location of the skeleton contour as a key point of a passable road, for each skeleton pixel point, the robot may perform the steps of:

step 221, the robot determines the number of adjacent pixels.

For each skeleton pixel, if the positions of the skeleton pixels are different, the number of the adjacent skeleton pixels, i.e., the adjacent pixels, is different. In particular, skeleton pixels located at key positions of the skeleton outline generally have a certain characteristic of the number of adjacent pixels corresponding to the skeleton pixels. That is, the number of adjacent pixels of the skeleton pixel may reflect the structural characteristics of the location of the skeleton pixel. Therefore, key points can be screened out through the number of adjacent pixel points of the skeleton pixel points, namely skeleton nodes and skeleton endpoints are determined.

The adjacent pixel points refer to the pixel points adjacent to the skeleton pixel point in other skeleton pixel points except the skeleton pixel point. For example only, assume a total of K skeleton pixels, K for skeleton pixels ₁ The adjacent pixels are pixels K except for the skeleton ₁ Other K-1 skeleton pixel points and skeleton pixel point K ₁ Adjacent skeleton pixel points.

Illustratively, in determining the number of adjacent pixels of each skeleton pixel, the robot may consider using a preset field, such as a four-neighborhood, a six-neighborhood, an eight-neighborhood, and the like. The eight neighborhood is preferable, and the eight neighborhood comprises eight directions around the target pixel point instead of four directions or six directions, so that the robot can more comprehensively consider adjacent pixel points in the up-down, left-right and diagonal directions of the skeleton pixel point and provide more comprehensive information for screening skeleton nodes and skeleton endpoints.

The method for determining the number of the adjacent pixels is helpful for capturing the structural characteristics of the skeleton outline more accurately, so that adjacent pixels possibly related to the diagonal direction are also considered, and the screening result is more accurate and comprehensive.

Step 222, when the number of adjacent pixels is equal to 1, the robot determines the skeleton pixel as the skeleton endpoint.

The skeleton end point is skeleton pixel with only 1 adjacent pixel point. Thus, the robot can determine the skeleton pixel point as the skeleton end point by checking the number of adjacent pixel points of each skeleton pixel and determining that the number of adjacent pixel points is equal to 1.

Step 223, when the number of adjacent pixels is greater than 2, the robot determines the skeleton node from the skeleton pixels.

A skeleton node is typically a skeleton pixel having a plurality of adjacent pixels. Thus, the robot can determine the skeleton pixel point as a skeleton node by checking the number of adjacent pixel points of each skeleton pixel and determining that the number of adjacent pixel points is greater than 2.

In this embodiment, the robot sets the screening condition of the key points according to the number features of the adjacent pixel points of the skeleton pixel points at the key positions, so that the key points of the passable road can be accurately and comprehensively determined from the skeleton pixel points. Namely, by definitely defining the characteristic conditions of the key positions, the robot can more accurately identify and select the key points in the skeleton, so that the topological path can be effectively distinguished. This facilitates efficient and accurate generation of the dispatch path by the robot from the critical locations.

In some embodiments, invalid key points are unavoidable in the determined skeleton nodes and skeleton endpoints. For example, some skeleton points may be noise points that may lead to an invalid short path on a long or closed path, or an isolated path formed by two isolated points. Obviously, the path points formed by these keypoints may reduce the accuracy of the topological path. To further improve the accuracy of the topological path, the robot may determine the path points by:

Step 231, the robot is connected into a line segment based on two adjacent skeleton target points.

In order to determine invalid key points from each skeleton node and each skeleton endpoint more intuitively, the robot can connect two skeleton target points through a straight line to form a line segment. Each line segment can more intuitively display an initial topological path formed by each current skeleton node and each skeleton endpoint, and is beneficial to optimizing key points. The two target skeleton points may be two adjacent skeleton nodes, two adjacent skeleton end points, and two adjacent skeleton end points and skeleton nodes.

Step 232, if the length of the line segment meets the preset condition, the robot deletes at least one skeleton target point corresponding to the line segment.

For a line segment formed by two skeleton endpoints, if the length of the line segment is smaller than the first preset length, it is indicated that the two skeleton endpoints are most likely two noise points, in which case the line segment formed by the two skeleton endpoints is a large probability of an isolated line segment. To reduce unnecessary computation, the two skeleton endpoints may be deleted.

For a line segment formed by two skeleton nodes, if the length of the line segment is smaller than a second preset length, the distance between the two skeleton nodes is too close. That is, there are two skeleton nodes on a shorter path, which may make the local path in the topology path too complex, which is not beneficial to efficiently generating the scheduling path. Therefore, the robot can reserve one skeleton node from the two skeleton nodes, so that the structure of the topological path is more concise and clear. Optionally, the first preset length is smaller than the second preset length. Illustratively, the first preset length may be 0.1m to 0.2m and the second preset length may be 0.3m to 0.4m.

Illustratively, the reserved skeleton nodes may be random reservations, or skeleton nodes that may be optimized may be determined based on the distance between each skeleton node and surrounding skeleton nodes or skeleton endpoints, and deleted.

For a line segment formed by a skeleton node and a skeleton endpoint, if the length of the line segment is less than a third length threshold, it is indicated that the skeleton endpoint may be a noise point. I.e., skeleton nodes and skeleton endpoints are likely to derive an isolated path on the primary path. With this isolated path, the amount of computation tends to be increased. Thus, for such two skeleton target points, skeleton endpoints may be pruned to filter isolated paths that are not valid on the primary path.

Optionally, after generating the skeleton map, the skeleton in a certain range near the target point is reserved according to the coordinates of the target point, so as to ensure that the target point can be reached through the skeleton.

Step 233, the robot determines the rest skeleton nodes and skeleton end points as path points.

Through the steps, the framework nodes and the framework endpoints are optimized, so that the rest framework nodes and framework endpoints can more clearly and accurately represent the topological path. As such, the robot may determine the remaining skeleton nodes and skeleton endpoints as path points to generate a more optimal topological path based on each path point.

In this embodiment, the robot optimizes each skeleton node and each skeleton endpoint to obtain a better path point, so as to generate a simpler and more accurate topology path, and the topology path helps the robot to generate a scheduling path more accurately and efficiently.

In some embodiments, when the distance between two path points is too large, multiple robots cannot easily distinguish which path point the current position corresponding to each robot belongs to during scheduling, so that the conflict is not well resolved, and the split line segments are beneficial to binding the robots distributed in a centralized way (the principle of not penetrating through the wall nearby) on different path points respectively so as to perform conflict calculation. Furthermore, too large a distance between two waypoints may also cause the resulting topological path to ignore local structures in the environment, such as curved roads or suddenly appearing obstacles, and thus lack adaptability to environmental changes. The scheduling paths obtained based on such topology path planning may have insufficient accuracy, are difficult to provide enough details, and also easily cause robots to need to make extensive adjustments on the paths in actual execution, increase energy consumption and may cause motion delays. Therefore, the robot can filter invalid route points based on the length of the line segment, and can add route points based on the length of the line segment. The specific steps are as follows:

And B1, if the length of the line segment is greater than a length threshold, dividing the line segment into at least two sub-line segments based on a preset length by the robot.

And B2, the robot determines the end point of each sub-line segment as a path point.

If the length of the line segment exceeds a preset length threshold, the robot may adopt a segmentation strategy to divide the line segment into at least two sub-line segments. The end points of each sub-line segment may be determined as path points to ensure fine granularity of the path, preventing longer line segments from missing local structure and critical details in path planning. The strategy is beneficial to improving the flexibility and adaptability of the scheduling path generation, so that the robot is better adapted to complex environments.

In this embodiment, the robot obtains at least two sub-line segments by dividing the overlong line segments; and then determining the end point of each sub-line segment as a path point so as to ensure the fine granularity of a long path and prevent the longer line segment from missing local structures and key details in path planning. This helps the robot to accurately and efficiently generate a high quality dispatch path.

By way of example only, based on the determined route points of any one of the embodiments or combinations of the embodiments, a corresponding topology route may be obtained after the route points are connected to each other. Wherein the topological path in the grid map can be seen in fig. 6. The path formed by connecting each path point in series in the middle of the passable road is the topological path. It can be appreciated that the robot can generate a clear and easily understood path based on the topology structure of the passable road, and can provide expansibility for the dispatching task. The scheduling path is generated based on the topology path, so that the calculation complexity can be reduced, the scheduling path corresponding to the multiple target points is matched with the scheduling task, and the robot can flexibly and efficiently complete the scheduling task in a complex environment.

In some embodiments, after the target point binds the corresponding target path point, the robot may generate the scheduling path by:

and C, taking the target path point as a starting point, and searching all passable paths between the starting point and each other path point based on a depth-first traversal algorithm to obtain a scheduling path corresponding to each target point.

In order for the robot to reach the target point of the scheduled task from any position in the previous map, the robot may take the target waypoint as a starting point and then systematically explore all possible traversable paths from the starting point to each of the other waypoints using a depth-first traversal algorithm. In this process, the robot explores continuously along the path in a recursive manner and records each feasible path. Finally, the robot obtains a scheduling path corresponding to the target point. The scheduling paths are subjected to depth-first search, so that possible path combinations are covered, comprehensive and effective scheduling paths are provided for scheduling tasks of the robot, and the scheduling reliability can be improved.

For example, when the target points are plural, the robot may perform step C on the path point corresponding to each target point to obtain the scheduling path corresponding to the target point. If the scheduling path corresponding to one target point is called a group of scheduling paths, the robot can determine how many groups of scheduling paths, and each target point corresponds to each group of scheduling paths one by one.

Illustratively, the algorithm for generating the scheduling paths may employ other path planning algorithms besides the depth-First Search algorithm, such as the Breadth-First Search (BFS) algorithm, the a-x algorithm, and the like. The specific algorithm to be adopted can be determined according to actual situations and requirements, and is not limited in the application.

In this embodiment, the robot systematically searches passable paths from the starting point to each other path point by using the target path point as the starting point through the depth-first search algorithm, thereby obtaining all possible scheduling paths corresponding to the target point. This depth-first strategy not only provides more alternative paths, but also enhances the reliability of scheduling tasks by covering the possible path combinations entirely.

It should be understood that the sequence number of each step in the foregoing embodiment does not mean that the execution sequence of each process should be determined by the function and the internal logic of each process, and should not limit the implementation process of the embodiment of the present application in any way.

Fig. 7 shows a block diagram of the configuration of the scheduling path generating apparatus 7 according to the embodiment of the present application, and for convenience of explanation, only the portions related to the embodiment of the present application are shown.

Referring to fig. 7, the scheduling path generating apparatus 7 includes:

a first generation module 71, configured to perform image processing on a grid map corresponding to the robot operation area, and generate a skeleton map of a passable road in the grid map;

a first determining module 72, configured to determine skeleton nodes and skeleton endpoints of the passable road from each skeleton pixel point in the skeleton map;

a second determining module 73, configured to determine a path point of the passable road based on the skeleton node and the skeleton end point;

a third determining module 74, configured to determine, for the target point of the scheduling task, a target path point closest to the target point from the path points;

the second generating module 75 is configured to generate and store a scheduling path corresponding to the target point based on the target point.

Optionally, the generating means 7 may further include:

the acquisition module is used for acquiring a virtual wall map of the operation area before performing image processing on the grid map corresponding to the robot operation area; the virtual wall map and the grid map have the same size;

the computing module is used for merging the values of each pair of pixel points with the same coordinate information in the virtual wall map and the grid map to obtain each updated pixel point;

an updating module for updating the grid map based on each updating pixel point and each coordinate information;

Accordingly, the first generating module 71 is specifically configured to perform image processing on the updated grid map.

Alternatively, the first generating module 71 may include:

the conversion unit is used for converting the grid map into a channel map with preset digits;

the smoothing unit is used for executing smoothing operation on a channel map to obtain a binary map;

the first extraction unit is used for extracting the width information of the road corresponding to each pixel point from the binary image;

and the second extraction unit is used for extracting the skeleton outline of the passable road from the binary image based on the width information to generate a skeleton image.

Alternatively, the first determining module 72 may include:

a first determining unit, configured to determine, for each skeleton pixel point, the number of adjacent pixel points; the adjacent pixel points refer to the pixel points adjacent to the skeleton pixel point in other skeleton pixel points except the skeleton pixel point;

a second determining unit configured to determine the skeleton pixel point as a skeleton endpoint when the number of adjacent pixel points is equal to 1;

and the third determining unit is used for determining the skeleton node from the skeleton pixel points when the number of the adjacent pixel points is more than 2.

Optionally, the second generating module 75 is specifically configured to: and searching all passable paths from the starting point to each other path point based on the depth-first traversal algorithm by taking the target path point as the starting point, and obtaining the scheduling path corresponding to each target point.

Alternatively, the second determining module 73 may include:

the connecting unit is used for connecting the two adjacent skeleton target points into a line segment, wherein the skeleton target points comprise skeleton nodes and/or skeleton endpoints;

the optimizing unit is used for deleting at least one skeleton target point corresponding to the line segment if the length of the line segment meets the preset condition;

and the fourth determining unit is used for determining the rest skeleton nodes and skeleton end points as path points.

Optionally, the optimizing unit may be further configured to segment the line segment into at least two sub-line segments based on a preset length if the length of the line segment is greater than the length threshold;

the fourth determining unit may be further configured to determine an end point of each sub-line segment as a path point.

It should be noted that, because the content such as the information interaction and the execution process between the above devices/units is based on the same concept as the method embodiment of the present application, specific functions and technical effects thereof may be referred to in the method embodiment section, and will not be described herein again.

Fig. 8 is a schematic structural diagram of a physical layer of a robot according to an embodiment of the present disclosure. As shown in fig. 8, the robot 8 of this embodiment includes: at least one processor 80 (only one shown in fig. 8), a memory 81, and a computer program 82 stored in the memory 81 and executable on the at least one processor 80, the processor 80 implementing steps in an embodiment of a method of generating any of the scheduling paths described above, such as steps 210-250 shown in fig. 1, when the computer program 82 is executed.

The processor 80 may be a central processing unit (Central Processing Unit, CPU), the processor 80 may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The memory 81 may in some embodiments be an internal storage unit of the robot 8, such as a hard disk or a memory of the robot 8. The memory 81 may in other embodiments also be an external storage device of the robot 8, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash Card (Flash Card) or the like, which are provided on the robot 8.

Further, the memory 81 may also include both an internal memory unit and an external memory device of the robot 8. The memory 81 is used for storing an operating device, an application program, a boot loader (BootLoader), data, and other programs and the like, such as program codes of computer programs and the like. The memory 81 may also be used to temporarily store data that has been output or is to be output.

It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.

The present application also provides a computer readable storage medium storing a computer program which, when executed by a processor, implements steps for implementing the various method embodiments described above.

Embodiments of the present application provide a computer program product enabling a robot to carry out the steps of the various method embodiments described above when the computer program product is run on the robot.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the present application implements all or part of the flow of the method of the above embodiments, and may be implemented by a computer program to instruct related hardware, where the above computer program may be stored in a computer readable storage medium, where the computer program, when executed by a processor, may implement the steps of each of the method embodiments described above. The computer program comprises computer program code, and the computer program code can be in a source code form, an object code form, an executable file or some intermediate form and the like. The computer readable medium may include at least: any entity or device capable of carrying computer program code to a camera device/robot, a recording medium, a computer Memory, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), an electrical carrier signal, a telecommunications signal, and a software distribution medium. Such as a U-disk, removable hard disk, magnetic or optical disk, etc.

In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.

Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus/network device and method may be implemented in other manners. For example, the apparatus/network device embodiments described above are merely illustrative, e.g., the division of modules or elements described above is merely a logical functional division, and there may be additional divisions in actual implementation, e.g., multiple elements or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection via interfaces, devices or units, which may be in electrical, mechanical or other forms.

The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

The above embodiments are only for illustrating the technical solution of the present application, and are not limiting thereof; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application, and are intended to be included in the scope of the present application.

Claims (10)

1. A method for generating a scheduling path, comprising:

performing image processing on a grid map corresponding to a robot operation area to generate a skeleton map of a passable road in the grid map;

Determining skeleton nodes and skeleton endpoints of the passable road from each skeleton pixel point in the skeleton map;

determining a path point of the passable road based on the skeleton node and the skeleton end point;

for a target point of a scheduling task, determining a target path point closest to the target point from the path points;

and generating and storing a scheduling path corresponding to the target point based on the target path point.

2. The method of generating a scheduling path according to claim 1, further comprising, before the image processing of the grid map corresponding to the robot work area:

obtaining a virtual wall map of the operation area; the virtual wall map and the grid map have the same size;

obtaining a union set of values of each pair of pixel points with the same coordinate information in the virtual wall map and the grid map to obtain each updated pixel point;

updating the grid map based on each of the updated pixel points and each of the coordinate information;

correspondingly, the image processing for the grid map corresponding to the robot working area comprises the following steps:

and carrying out image processing on the updated grid map.

3. The method for generating a scheduling path according to claim 1, wherein the generating a skeleton map of a passable road in a grid map by performing image processing on the grid map corresponding to a robot work area comprises:

converting the grid map into a channel map with preset digits;

performing smoothing operation on the one channel map to obtain a binary map;

extracting the width information of the road corresponding to each pixel point from the binary image;

and extracting the skeleton outline of the passable road from the binary image based on the width information, and generating the skeleton image.

4. The method for generating a scheduling path according to claim 1, wherein determining skeleton nodes and skeleton endpoints of the passable road from each skeleton pixel point in the skeleton map comprises:

for each of the skeleton pixel points:

determining the number of adjacent pixel points; the adjacent pixel points refer to pixel points adjacent to the skeleton pixel point in other skeleton pixel points except the skeleton pixel point;

when the number of the adjacent pixel points is equal to 1, determining the skeleton pixel points as skeleton end points;

And when the number of the adjacent pixel points is greater than 2, determining the skeleton node by the skeleton pixel points.

5. The method for generating a scheduling path according to claim 1, wherein the generating a scheduling path between the target points based on the target path points comprises:

and searching all passable paths between the starting point and each other path point based on a depth-first traversal algorithm by taking the target path point as the starting point, and obtaining a scheduling path corresponding to each target point.

6. The method of generating a scheduling path according to any one of claims 1 to 5, wherein the determining a path point of the traversable road based on the skeleton node and the skeleton end point includes:

connecting line segments based on two adjacent skeleton target points, wherein the skeleton target points comprise the skeleton nodes and/or the skeleton endpoints;

if the length of the line segment meets the preset condition, deleting at least one skeleton target point corresponding to the line segment;

and determining the rest skeleton nodes and the skeleton end points as the path points.

7. The scheduling path generating method of claim 6, wherein the generating method further comprises:

If the length of the line segment is greater than the length threshold, dividing the line segment into at least two sub-line segments based on a preset length;

and determining the end point of each sub-line segment as the path point.

8. A scheduling path generating apparatus, comprising:

the first generation module is used for carrying out image processing on a grid map corresponding to the robot operation area and generating a skeleton diagram of a passable road in the grid map;

the first determining module is used for determining skeleton nodes and skeleton endpoints of the passable road from each skeleton pixel point in the skeleton map;

the second determining module is used for determining the path point of the passable road based on the skeleton node and the skeleton end point;

a third determining module, configured to determine, for a target point of a scheduling task, a target path point closest to the target point from the path points;

and the second generation module is used for generating and storing the scheduling path corresponding to the target point based on the target path point.

9. Robot comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the method of generating a scheduling path according to any of claims 1 to 7 when executing the computer program.

10. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the scheduling path generating method according to any one of claims 1 to 7.