A task execution method, device, storage medium and electronic device

Abstract

The specification discloses a task execution method, device, storage medium and electronic equipment. The task execution method comprises the following steps: acquiring a map image corresponding to a target area and updating in real time, determining at least one task area according to the connection relationship between each intersection and different roads, for each task area, determining the topography type corresponding to the task area according to the altitude distribution of the task area, and determining the target distance between each first unmanned device and the center position of the task area, and then combining the area size of each task area, taking all task areas can be searched and complex topography can be finely searched as the target, allocating the task area responsible for searching to each first unmanned device, searching through each first unmanned device, and sending a task execution instruction to a second unmanned device after a target object is searched, so that each second unmanned device executes a task.

Description

TECHNICAL FIELD

[0001] The present specification relates to the technical field of computer technology, and particularly relates to a task execution method and device, a storage medium and an electronic device. BACKGROUND

[0002] In recent years, with the development of technology, a series of unmanned device technologies such as unmanned aerial vehicles, unmanned vehicles, unmanned ships and unmanned underwater vehicles have rapidly progressed, among which, the cooperative technology of unmanned clusters is widely applied to many fields such as agriculture, manufacturing, transportation and military, so that it is possible to search for targets in a large range through unmanned device clusters.

[0003] However, the current cooperative technology of unmanned device clusters cannot effectively cooperate between unmanned devices of different structures, and in the process of searching for target objects, the device cluster is usually taken as a unit, and the independent control of individual devices in the device cluster cannot be realized, which causes a large waste of search resources in the process of searching for target objects.

[0004] Therefore, how to realize effective cooperation between different types of unmanned devices in the process of searching for target objects through unmanned device clusters, fully exert the advantages of unmanned devices of different structures, and realize the independent control of individual devices to reduce the waste of resources in the search process is a problem to be solved. SUMMARY

[0005] The present specification provides a task execution method, device, storage medium and electronic device to partially solve the above problems existing in the prior art.

[0006] The present specification adopts the following technical solutions:

[0007] The present specification provides a task execution method, comprising:

[0008] obtaining a map image corresponding to a target area;

[0009] determining at least one task area according to the connection relationship between each intersection and different roads in the map image;

[0010] For each task area, according to the altitude distribution of the task area, the topography type corresponding to the task area is determined, and the distance between each first unmanned device and the center position of the task area is determined as a target distance;

[0011] According to the topography type corresponding to each task area, the target distance and the area size corresponding to each task area, all task areas can be searched as targets, and each first unmanned device is allocated a task area responsible for searching;

[0012] The first unmanned device searches the task area responsible for searching by itself, and sends a task execution instruction to the second unmanned device after searching for a target object, so that the second unmanned device executes the task according to the task execution instruction.

[0013] Optionally, at least one task area is determined according to the connection relationship between each intersection and different roads in the map image, specifically including:

[0014] A road network topology image of the target area is constructed with each intersection in the map image as a node and different roads as edges.

[0015] A grid image of the target area is constructed according to the road network topology image, and the at least one task area is determined according to the grid image.

[0016] Optionally, for each task area, the corresponding topographic type of the task area is determined according to the corresponding elevation distribution of the task area, specifically including:

[0017] The corresponding elevation map of the grid image is constructed.

[0018] For each task area, the corresponding elevation distribution of the task area is determined according to the pixel value size distribution of the corresponding elevation map of the task area.

[0019] The corresponding topographic type of the task area is determined according to the elevation distribution.

[0020] Optionally, the grid image of the target area is constructed according to the road network topology image, and the at least one task area is determined according to the grid image, specifically including:

[0021] The minimum circumscribed rectangular area of the road network topology image is determined.

[0022] The latitude and longitude coordinates corresponding to each pixel point in the grid image are determined according to the actual latitude and longitude maximum value and the latitude and longitude minimum value of the minimum circumscribed rectangular area, and the preset width value and the preset height value of the grid image.

[0023] Optionally, each first unmanned device is assigned a task area responsible for searching according to the topographic type corresponding to each task area, the target distance, and the area size corresponding to each task area, so that all task areas can be searched as targets, specifically including:

[0024] The search difficulty corresponding to each task area is determined according to the topographic type corresponding to each task area, the target distance, and the area size.

[0025] According to the order of the search difficulty from high to low, the task areas responsible for the search are allocated to each first unmanned device until all the task areas are allocated to the first unmanned devices for the search.

[0026] According to the order of the search difficulty from high to low, the task areas responsible for the search are allocated to each first unmanned device until all the task areas are allocated to the first unmanned devices for the search, specifically including:

[0027] For each first unmanned device, if the topography type of the task area allocated to the first unmanned device is a specified type, and the area of the task area is not less than the mean value of the areas of all the task areas, no other task area is allocated to the first unmanned device.

[0028] According to the order of the search difficulty from high to low, the task areas responsible for the search are allocated to each first unmanned device until all the task areas are allocated to the first unmanned devices for the search, specifically including:

[0029] If the search difficulty of the task area allocated to each unmanned device is greater than a preset difficulty, or the search difficulty of all the task areas is less than the preset difficulty, the task areas responsible for the search are continuously allocated to each first unmanned device until all the task areas are allocated to the first unmanned devices for the search.

[0030] Optionally, the search of the task area responsible for the search by each first unmanned device specifically includes:

[0031] For each task area, according to the road direction of the task area, a search path of the unmanned device responsible for the search of the task area is determined;

[0032] According to the search path, the target object is searched on the road of the task area, and after the target object is searched, the target object is tracked.

[0033] Optionally, for each task area, according to the road direction of the task area, a search path of the unmanned device responsible for the search of the task area is determined, specifically including:

[0034] According to the road direction of the task area, the search range of each first unmanned device responsible for the search of adjacent two task areas does not overlap, and the search range can cover the own task area as a target, and the search path is determined.

[0035] Optionally, the method further includes:

[0036] If the task for the target object has been executed, other target objects in the non-road area are searched.

[0037] Optionally, each first unmanned device searches a task area responsible for by itself, and sends a task execution instruction to each second unmanned device after searching a target object, so that each second unmanned device executes a task according to the task execution instruction, and the task execution instruction specifically includes:

[0038] For each target object, a task execution difficulty corresponding to the target object is determined according to a number of connected roads at a trajectory end point of the target object.

[0039] The second unmanned devices responsible for executing tasks on the target objects are allocated in an order from small to large according to the task execution difficulties.

[0040] Optionally, the second unmanned devices responsible for executing tasks on the target objects are allocated in an order from small to large according to the task execution difficulties, and the allocation specifically includes:

[0041] For each target object, a specified number of second unmanned devices are selected as candidate second unmanned devices according to distances between the second unmanned devices and the target object.

[0042] Task paths of the candidate second unmanned devices for executing tasks on the target object are determined.

[0043] The candidate second unmanned devices execute tasks on the target object according to the respective task paths.

[0044] Optionally, the first unmanned device includes a drone, and the second unmanned device includes an unmanned vehicle.

[0045] The specification provides a task execution device, which includes:

[0046] An acquisition module acquires a map image corresponding to a target area.

[0047] A first determination module determines at least one task area according to a connection relationship between each intersection and different roads in the map image.

[0048] A second determination module determines, for each task area, a topographic type corresponding to the task area according to an altitude distribution of the task area, and determines a distance between each first unmanned device and a center position of the task area as a target distance.

[0049] A distribution module allocates, for each first unmanned device, a task area responsible for searching according to the topographic type corresponding to each task area, the target distance, and an area size corresponding to each task area, so that all the task areas can be searched as targets.

[0050] The execution module searches the task area responsible for searching by each first unmanned device, and sends a task execution instruction to the second unmanned device after searching for a target object, so that each second unmanned device executes a task according to the task execution instruction.

[0051] The present specification provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the above-mentioned task execution method.

[0052] The present specification provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the above-mentioned task execution method when executing the program.

[0053] The above-mentioned at least one technical solution adopted by the present specification can achieve the following beneficial effects:

[0054] The task execution method provided by the present specification includes: a server acquires a map image corresponding to a target area, determines at least one task area according to the connection relationship between each intersection and different roads, for each task area, determines the topographic type corresponding to the task area according to the altitude distribution of the task area, and determines the target distance between each first unmanned device and the center position of the task area, according to the topographic type corresponding to each task area, the target distance, and the area size of each task area, all task areas can be searched as targets, and each first unmanned device is allocated a task area responsible for searching, searching is performed by each first unmanned device, and a task execution instruction is sent to a second unmanned device after searching for a target object, so that each second unmanned device executes a task.

[0055] As can be seen from the above method, the present scheme can allocate corresponding first unmanned devices to search each task area according to the topographic type corresponding to each task area, the target distance, and the area size, and send a task execution instruction to each second unmanned device after searching for a target object, so as to realize independent control of each unmanned device by allocating corresponding task areas to different first unmanned devices, and also realize cooperative execution of tasks for target objects by unmanned devices with different structures, improve the accuracy of target search, and reduce the waste of resources in the task execution process. BRIEF DESCRIPTION OF DRAWINGS

[0056] The accompanying drawings used in the description of the present specification are used to provide further understanding of the present specification, form a part of the present specification, and the illustrative embodiments of the present specification and their description are used to explain the present specification, and do not constitute an improper limitation on the present specification. In the drawings:

[0057] Figure 1 It is a flowchart of a task execution method provided in the present specification;

[0058] Figure 2 A schematic diagram of a road network topology image of a target area provided in the present specification;

[0059] Figure 3 A schematic diagram of a grid image of a target area provided in the present specification;

[0060] Figure 4 A schematic diagram of a task area division provided in the present specification;

[0061] Figure 5 A schematic diagram of an updating method of a grid image provided in the present specification;

[0062] Figure 6 A schematic diagram of a process in which a first unmanned device and a second unmanned device cooperatively perform a task provided in the present specification;

[0063] Figure 7 A schematic diagram of a task path planning result on a grid image provided in the present specification;

[0064] Figure 8 A schematic diagram of an adjusting method of a task path provided in the present specification;

[0065] Figure 9 A schematic diagram of an apparatus for performing a task provided in the present specification;

[0066] Figure 10 A schematic diagram of an electronic device corresponding to Figure 1 provided in the present specification. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical solutions and advantages of the present specification clearer, the technical solutions of the present specification will be described in detail below with reference to the embodiments of the present specification and corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present specification, rather than all the embodiments. Based on the embodiments in the present specification, all other embodiments obtained by a person of ordinary skill in the art without any creative work fall within the scope of protection of the present specification.

[0068] The technical solutions provided by the embodiments of the present specification will be described in detail below with reference to the drawings.

[0069] Figure 1 A schematic diagram of a flow of a method for performing a task provided in the present specification, comprising the following steps:

[0070] S101: Obtain a map image corresponding to a target area.

[0071] In fields such as military, agriculture and transportation, scenarios often involve the need to surround and capture target objects in a fixed area. For example, during military exercises, vehicles, people or animals in a designated area are often searched, and tasks are performed on the target objects after the target objects are found, thereby completing the exercise training.

[0072] In actual applications, due to the large range of the area where the target objects are located, multiple unmanned devices need to be coordinated to search, and different structures of unmanned devices (such as unmanned aerial vehicles, unmanned vehicles, unmanned ships and unmanned underwater vehicles) need to be cooperated to perform tasks.

[0073] Based on this, the present specification provides a task execution method, which allocates a task area responsible for searching to different unmanned devices in a cluster of unmanned devices, and performs a task on a target object through other structures of unmanned devices after the target object is found.

[0074] In the present specification, the execution subject of the method for implementing task execution can be a designated device such as a server. For ease of description, the present specification only takes the server as the execution subject to illustrate the task execution method provided in the present specification.

[0075] In addition, the unmanned devices in the present specification can include a plurality of first unmanned devices and a plurality of second unmanned devices. The first unmanned devices can include unmanned aerial vehicles, and the second unmanned devices can include unmanned vehicles. Of course, other unmanned devices can also be included, which are not specifically limited in the present specification.

[0076] The server needs to obtain a map image corresponding to the target area. In actual applications, the map image can be collected by remote sensing devices such as satellites and sent to the server for storage.

[0077] S102: Determine at least one task area according to the connection relationship between each intersection and different roads in the map image.

[0078] The server can construct a road network topology image of the target area according to the connection relationship between each intersection and different roads in the map image, taking each intersection in the map image as a node and different roads as edges. For ease of understanding, the present specification provides a road network topology image diagram, as shown in Figure 2 .

[0079] Figure 2 The present specification provides a road network topology image diagram of the target area.

[0080] The server connects each intersection node with the connecting road between them as an edge, thereby constructing a road network topology image of the target area.

[0081] Then the server can take the minimum circumscribed rectangle region of the road network topology image, such as the dashed line region in FIG. 8, determine the extreme values of the latitude and longitude of the region (i.e. the latitude and longitude values corresponding to the four corners of the rectangular region respectively). The server can determine the latitude and longitude coordinates corresponding to each pixel point in the grid image according to the latitude and longitude extreme values and the preset width value and the preset height value of the grid image. Figure 2

[0082] Specifically, when the preset width value of the grid single-channel image is gm_wth and the preset height value is gm_hgt, the quantization compensation of the latitude and longitude can be represented by the following formula:

[0083]

[0084] Where Δlongi is the quantization step of longitude, Δlati is the quantization step of latitude, max_longi is the maximum longitude, min_longi is the minimum longitude, max_lati is the maximum latitude, and min_lati is the minimum latitude.

[0085] After determining the quantization steps of longitude and latitude, the server can determine the actual latitude and longitude coordinates corresponding to each pixel point coordinate (x, y) on the grid image, i.e. (min_longi+Δlongi*x, min_lati+Δlati*y).

[0086] In this way, the server can construct a one-to-one mapping relationship between the pixel coordinates on the grid image and the actual latitude and longitude of the map image, and map the road network topology nodes in the map image and the connected roads between them to the grid image.

[0087] Further, the server can first generate an initial single-channel grid image with pixel values all being 0 and a size of gm_wth×gm_hgt, and then map each intersection in the map image as a node to the initial grid image, and the pixel value of the corresponding pixel point can be set to 255. At the same time, the server maps the connected roads between each intersection as an edge to the initial grid image, and the pixel value of the corresponding line segment can also be set to 255. Then the server can perform a dilation operation on the initial grid image after adding nodes and edges, so that there is a connected path between any two nodes of the grid image. In order to facilitate understanding, the present specification provides a grid image schematic diagram of a target region, as shown in FIG. 9. Figure 3

[0088] Figure 3 A grid image schematic diagram of a target region provided in the present specification.

[0089] ​​Wherein, the black region corresponds to pixel value 0, the white region corresponds to pixel value 255, the region with pixel value 255 represents the passable region, and the region with pixel value 0 represents the unknown region without road.

[0090] It should be noted that the grid image generation process has quantization error, and the smaller the values of Delta longi and Delta lati, the smaller the error. However, since a certain distance threshold is used as a constraint in the process of cooperative search, tracking and cooperative task execution of the target object, the deviation of the target position in the actual longitude and latitude map caused by the quantization error is not sensitive in the process of executing the task of the target object. Proper control of the quantization step can reduce the data processing amount as much as possible while meeting the application accuracy.

[0091] After determining the grid image, the server can determine each task area according to the grid image, wherein each task area corresponds to a smallest closed area surrounded by a plurality of roads. In order to facilitate understanding, the present specification provides a task area allocation schematic diagram, as shown in Figure 4 .

[0092] Figure 4 A task area allocation schematic diagram is provided in the present specification.

[0093] As can be seen from Figure 4 , the target area has 8 smallest closed areas surrounded by different roads, and the 8 smallest closed areas correspond to 8 task areas respectively.

[0094] In the actual search process of the target object by the first unmanned device (unmanned aerial vehicle), it is often found that some temporary obstacles block the passable road. Therefore, the server can update the intersection of the area affected by the obstacle according to the obstacle information collected by the first unmanned device.

[0095] In this process, the server can sample and map the corner points on the obstacle contour in the actual map image to the grid image. In order to facilitate understanding, the present specification provides a grid image updating method schematic diagram, as shown in Figure 5 .

[0096] Figure 5 A grid image updating method schematic diagram is provided in the present specification.

[0097] Wherein, a is the grid image determined by the server, b is the mask (mask) image after a new obstacle is found and a new area is added. The mask image can be any shape of closed polygon. Then the server can combine the inverse color images of a and b to obtain the updated grid image c.

[0098] In this process, the server can detect and identify the obstacles and obstacle types in the image collected by the first unmanned device through computer vision (CV) object detection. In this process, the server can generate an initial quantization image with all pixel values being 0 and the same resolution as the grid image. Then, the server maps the current pixel coordinates on the initial quantization image to the latitude and longitude map, converts the obtained latitude and longitude coordinates to coordinates on the image collected by the first unmanned device, and then obtains the effective pixel value mean, and determines the quantization visible light image according to the effective pixel value mean.

[0099] Then, the server determines the classification results of the existing obstacles according to the quantization visible light image, including: new obstacles, reduced obstacles, and no obstacles. Then, the server updates the grid image according to the classification results.

[0100] When the server determines that no new obstacles are added, the initial grid image is not updated. When the server determines that new obstacles are added, the initial grid image can be updated according to the process shown in FIG. 8 to obtain an updated grid image. Figure 5 When the server determines that the obstacles are reduced, the server first obtains the mask graph mask_rm (the foreground pixel value is 255) corresponding to the eliminated obstacle area, and then fuses the initial grid image and mask_rm to obtain an updated grid image.

[0101] S103: For each task area, determine the topography type corresponding to the task area according to the altitude distribution of the task area, and determine the distance between each first unmanned device and the center position of the task area as the target distance.

[0102] Specifically, the server can construct an elevation map corresponding to the grid image, and for each task area, determine the altitude distribution of the task area according to the pixel value size distribution of the elevation map corresponding to the task area.

[0103] In the process of constructing the elevation map, the server can first generate a single-channel quantization elevation map with the same resolution as the grid image. For each pixel point in the elevation map, the larger the pixel value corresponding to the pixel point, the higher the actual altitude corresponding to the pixel point. The server can extract the minimum circumscribed rectangle of each task area contour, and determine the altitude distribution of the task area according to the altitude distribution in the circumscribed rectangle. Of course, the server can also directly determine the altitude distribution of each task area.

[0104] The server can then determine the terrain type of each task area based on its altitude. For example, in task areas with drastic altitude changes, the elevation map image has rich texture and large gray-scale variance, indicating that the actual terrain type is complex terrain such as mountains, ravines, and valleys. In contrast, in task areas with relatively gentle altitude changes, the elevation map image has monotonous texture and small gray-scale variance, and its terrain type is usually plains, lakes, or seas.

[0105] In addition, the server can also determine the actual physical distance cost_dist from the center of each task area to each first unmanned device (drone). ij This serves as the target distance for the task area, thus determining the target distance for each task area.

[0106] After determining the target distance for each task area, the server can normalize it to the range of 0 to 1 using the sum of the actual path lengths around all task areas as the base, thereby obtaining the first search difficulty index for each area.

[0107] S104: Based on the terrain type corresponding to each task area, the target distance, and the area size of each task area, with the goal of ensuring that all task areas can be searched, assign a task area to each first unmanned device to be responsible for searching.

[0108] After determining the terrain type corresponding to each task area, the server can determine the second search difficulty index corresponding to each task area based on the terrain type. For task areas with complex terrain such as mountains-hills, valleys-depressions, etc., a higher second search difficulty index can be set, i.e., cost_search = 1. For task areas with open terrain such as plains-sea, a lower second search difficulty index can be set, i.e., cost_search = 0.5.

[0109] Additionally, the server can also calculate the cost_area based on the area corresponding to each task region. ij The total area of ​​each task region is used as the base to normalize it to a range of 0 to 1, thus obtaining the third search difficulty index corresponding to each task region. Specifically, the server can calculate the search difficulty index corresponding to each task region using the area of ​​each task region in the raster image and the total area of ​​all task regions.

[0110] To ensure that task areas with high search difficulty (mainly referring to complex terrain) can be searched in detail, the server can determine the search difficulty of each task area based on the terrain type, target distance, and area size. Then, the server assigns task areas to each first unmanned device in descending order of search difficulty until all task areas have been assigned to the corresponding first unmanned device for search.

[0111] Specifically, the server can determine the search loss function corresponding to each task area according to the first search difficulty index, the second search difficulty index and the third search difficulty index. The loss function can be represented by the following formula:

[0112] cost_val ij =3-(cost_search+cost_dist ij +cost_area ij )

[0113] Wherein, cost_val ij is the function value of the loss function. As can be seen from the formula, for each task area, the greater the search difficulty of the task area, the smaller the function value of the corresponding loss function, and the smaller the search difficulty of the task area, the greater the function value of the corresponding loss function.

[0114] Since the number of task areas is likely to be more than the number of first unmanned devices in actual application, some first unmanned devices may be assigned two or more task areas. For task areas with greater search difficulty (smaller loss function value), the server can preferentially assign the first unmanned device to search the task area, and try to ensure that the unmanned device searching the task area will not be assigned to search other task areas, so as to ensure that the task area with greater search difficulty can be searched more carefully.

[0115] Further, for each first unmanned device, if the topography type of the task area assigned to the first unmanned device is a specified type (such as a complex type of gully, hill, forest, etc.), and the area of the task area is not less than the average of the areas of all task areas, or the loss function value of the task area is less than a preset threshold, it means that the search difficulty of the task area is greater, so the server can no longer assign other task areas to the first unmanned device responsible for searching the task area.

[0116] It should be noted that if the search difficulty of the task area assigned to each unmanned device for the first time is greater than a preset difficulty, or the search difficulty of all task areas is less than the preset difficulty, all first unmanned devices participate in the next round of task area allocation, and the server continues to allocate the task area responsible for searching for each first unmanned device, until all task areas are allocated to the corresponding first unmanned device for searching. The above-mentioned preset difficulty can be set according to actual conditions, and the present specification does not make specific limitations thereto.

[0117] In this process, the server can control the size of the entire loss function value by adjusting the calculation strategy of the first search difficulty index, thereby regulating the allocation strategy of each task area. For each first unmanned device, if the first unmanned device is not allocated a task area with a search difficulty greater than the preset difficulty, the calculation strategy of the corresponding first search difficulty index remains unchanged.

[0118] If it is allocated a task area with a search difficulty greater than the preset difficulty, the calculation strategy of the corresponding first search difficulty index is changed to cost_dist ij = cost_dist ij / 2, thereby increasing the function value of the loss function, so that in the case where all first unmanned devices are not allocated task areas with greater difficulty in the first round, the first unmanned device will not be allocated other task areas.

[0119] S105: Each first unmanned device searches the task area it is responsible for and sends a task execution instruction to the second unmanned device after searching for the target object, so that each second unmanned device executes the task according to the task execution instruction.

[0120] After the server allocates task areas for each first unmanned device, each first unmanned device can search the task area it is responsible for. In this process, the first unmanned device can determine the search path of the unmanned device responsible for searching the task area according to the road direction of the task area, and then search for the target object on the road of the task area according to the search path, and send a task execution instruction to the second unmanned device after searching for the target object. Before the second unmanned device completes the task of the target object, the first unmanned device can first track the target object in the air.

[0121] The first unmanned device can determine a straight line segment path with its own latitude and longitude coordinates as the starting point of the trajectory and the target latitude and longitude coordinates as the end point to efficiently track the target object. When the first unmanned device approaches the target object to a preset distance, it maintains the distance from the target object in a circling manner. The preset distance can be set according to actual conditions, which is not limited in the present specification.

[0122] If the target object on the road has been captured, the first unmanned device can randomly investigate and search the non-road area of the task area to search for other target objects in the non-road area.

[0123] Further, as the first unmanned device of the two adjacent task areas may have overlapping field of view when searching, which will cause repeated field of view to some extent, therefore, the server can determine the search path according to the road direction of the task area, so that the search range of the first unmanned device responsible for searching the adjacent two task areas does not overlap, and the search range can cover the target task area.

[0124] In the present specification, the server can first set a morphological operator with appropriate shape and size, and perform an iterative erosion operation on each task sub-area assigned to the first unmanned device on the grid image. When the minimum circumscribed rectangle side length of the erosion result area ≈ the minimum circumscribed rectangle side length of the area before erosion * ft, stop the erosion operation, where ft is a size scaling factor, which can be determined according to the coverage area of the image sensor of the first unmanned device and the actual physical size of the task area to be searched.

[0125] Then the server can take the outer contour of the erosion result area, equally spaced sampling to get a set of trajectory points, and map the trajectory points on the grid image to the actual map image to generate a closed trajectory route as the search path corresponding to the task area. In this way, it is possible to cover all task areas while avoiding overlapping search ranges of first unmanned devices in adjacent task areas.

[0126] In the process of searching and capturing the target object, it needs to be completed by the first unmanned device and the second unmanned device. First, the server can plan a reasonable cruise trajectory for each first unmanned device in the task area assigned to it, and then each first unmanned device detects the target object on the given search trajectory using its image sensing device, and shares information through situation information update.

[0127] When the first unmanned device discovers the target object, it begins to track it, and only tracks the same target object in the task area within the same time period, but can still perform search detection on other target objects in the task area while tracking. At this time, the second unmanned device receives the task execution instruction and obtains the position information and route information of the target object shared by the first unmanned device, and then executes the task for the target object according to the above position information and route information.

[0128] When the second unmanned device executes the task for the target object, it can take into account the search for the target object, but does not take responsibility for tracking the target object. In this way, even if the first unmanned device needs to return for supplies, the search for the target object will not be interrupted during this period. In order to facilitate understanding, the present specification provides a process diagram of the first unmanned device and the second unmanned device cooperating to execute the task, as shown in Figure 6 .

[0129] Figure 6 A schematic diagram of a process for a first unmanned device to perform a task in cooperation with a second unmanned device is provided in the present specification.

[0130] If a second unmanned device (an unmanned vehicle) has already performed a task for a target object, the unmanned vehicle can be responsible for searching for other target objects while performing the task for the target object. If no unmanned vehicle has performed a task for a target object, only a UAV searches for and tracks the target object, and stops searching for the target object after returning for supplies, at which time the unmanned vehicle is responsible for searching for the target object.

[0131] When a new target object is detected, the server can update the search list and determine whether the added target object is in a monitored (tracked) state. If so, the server further determines whether the target object is located in another task area. If not, the server iterates through other UAVs that do not track a target object and updates the tracking list of the UAV when the UAV is close enough to the target object to track the target object.

[0132] In the present specification, the task performed by each second unmanned device for a target object can be a task of surrounding the target object. During the performance of the task for the target object, when each direction passing path of the location of the escaped target object has an unmanned vehicle surrounding the target object, and the distance between each unmanned vehicle and the target object has reached a target distance, the task of surrounding the target object can be considered to be completed.

[0133] During the performance of the task for the target object, the movement speed of the target object can be too fast, and the target object can always reach the end intersection node before its own trajectory is updated. Therefore, the server can determine the task difficulty according to the number of connected paths corresponding to the intersection node at the end of the trajectory of the target object. The more paths connected to the intersection node at the end of the trajectory of the target object, the greater the task difficulty.

[0134] For example, if there are 3 connected paths between a node and other nodes, the task difficulty of the target object whose trajectory end is the node can be set to 3. When multiple target objects are searched at the same time, the server can assign a second unmanned device responsible for performing a task for each target object in order of task difficulty from small to large, so as to preferentially perform a task for a target object with higher task difficulty to prevent the target object from escaping. The task execution strategy for the target object can be represented by the following formula:

[0135]

[0136] wherein M is the number of searched target objects, diff is the task difficulty corresponding to the target object, and N is the number of currently idle second unmanned devices.

[0137] In the process of assigning the second unmanned device to perform the task for the target object, the server can implement iterative assignment for each required escape target execution unit until the full amount is reached, and according to the distance between each second unmanned device and the target object, the server can preferentially select a specified number of second unmanned devices closest to the target object as candidate second unmanned devices to perform the task.

[0138] In the process of assigning the second unmanned device to perform the task for the target object, the server can implement iterative assignment for each required escape target execution unit until the full amount is reached, and according to the distance between each second unmanned device and the target object, the server can preferentially select a specified number of second unmanned devices closest to the target object as candidate second unmanned devices to perform the task.

[0139] For each second unmanned device assigned to perform the task for a target object, the server can determine its corresponding optimal trajectory endpoint network node. Each second unmanned device can use the pixel point on the grid image corresponding to its own position as the trajectory starting point, and the pixel point on the grid image corresponding to the intersection node assigned by the Hungarian algorithm as the trajectory endpoint, and use the D*lite path planning algorithm to optimize the trajectory to obtain the global optimal task path on the grid image. The server can sample the optimal task path onto the road image and add the escape target trajectory endpoint intersection node as the task path endpoint to obtain the complete task path. In order to facilitate understanding, the present specification provides a task path planning result schematic diagram on a grid image, as shown in Figure 7 .

[0140] Figure 7 A task path planning result schematic diagram on a grid image is provided in the present specification.

[0141] In which, Figure 7 A part of the grid image, white pixels are passable areas, gray pixels are forbidden areas, and black pixels are determined task paths. The server can select the start and end points of the task path through the above method, and then find the optimal task path through the start and end points in the intersection nodes of the grid image.

[0142] In addition, in the process of determining the task path, part of the passable path may be blocked, so the server can adjust the task path while avoiding the blockage. In order to facilitate understanding, the present specification provides a task path adjustment method schematic diagram, as shown in Figure 8 .

[0143] Figure 8A method for adjusting a task path is provided in the specification.

[0144] The gray area is a forbidden area, and Fig. a is the optimal task path obtained between the starting point and the ending point on the original map. When (2, D) becomes an unreachable area, the server can adjust the previous trajectory in real time to achieve the obstacle avoidance strategy as shown in Fig. b.

[0145] From the above method, it can be seen that the scheme can allocate corresponding first unmanned devices to search for each task area according to the corresponding topographic type, target distance, and area size of each task area, and send execution instructions of the task to each second unmanned device after searching for the target object. By allocating corresponding task areas to different first unmanned devices, independent control of individual unmanned devices is realized, and different structures of unmanned devices are also realized to cooperatively execute tasks for target objects, improving the accuracy of target search and reducing the waste of resources in the task execution process.

[0146] In addition, the improved target hunting method of the specification adapts to the application requirements of topology dynamic adjustment and topology switching, has generalization ability, individual unmanned vehicles in the unmanned cluster have no formation behavior, are highly autonomous, and have low structural coupling.

[0147] The above is a method for executing one or more embodiments of the specification, based on the same idea, the specification also provides a corresponding task execution device, as shown in Figure 9 .

[0148] Figure 9 A task execution device provided in the specification is shown in the figure, which includes:

[0149] The acquisition module 901 acquires the map image corresponding to the target area;

[0150] The first determination module 902 determines at least one task area according to the connection relationship between each intersection and different roads in the map image;

[0151] The second determination module 903 determines the topographic type corresponding to each task area according to the altitude distribution of the task area, and determines the distance between each first unmanned device and the center position of the task area as the target distance;

[0152] The allocation module 904 allocates the task area responsible for searching for each first unmanned device according to the topographic type corresponding to each task area, the target distance, and the area size corresponding to each task area, so that all task areas can be searched as targets.

[0153] The execution module 905 searches the task area responsible for by each first unmanned device, and sends a task execution instruction to the second unmanned device after searching the target object, so that each second unmanned device executes the task according to the task execution instruction.

[0154] Optionally, the first determination module 902 is specifically configured to construct a road network topology image of the target area by taking each intersection in the map image as a node and taking different roads as edges; construct a grid image of the target area according to the road network topology image, and determine the at least one task area according to the grid image.

[0155] Optionally, the first determination module 902 is specifically configured to construct an elevation map corresponding to the grid image; for each task area, determine the altitude distribution of the task area according to the pixel value size distribution of the elevation map corresponding to the task area; and determine the landform type of the task area according to the altitude distribution.

[0156] Optionally, the first determination module 902 is specifically configured to determine a minimum circumscribed rectangular region of the road network topology image; and determine the longitude and latitude coordinates of each pixel point in the grid image according to the actual longitude and latitude maximum value and the longitude and latitude minimum value of the minimum circumscribed rectangular region, and a preset width value and a preset height value of the grid image.

[0157] Optionally, the second determination module 903 is specifically configured to determine the search difficulty of each task area according to the landform type, the target distance and the area size corresponding to each task area; and allocate the task areas responsible for searching to each first unmanned device in descending order of the search difficulty, until all the task areas are allocated to the first unmanned devices for searching.

[0158] Optionally, the second determination module 903 is specifically configured to, for each first unmanned device, if the landform type of the task area allocated to the first unmanned device is a specified type, and the area of the task area is not less than the average of the areas of all the task areas, then no other task area is allocated to the first unmanned device.

[0159] Optionally, the second determination module 903 is specifically configured to, if the search difficulty of the task area allocated to each unmanned device is greater than a preset difficulty, or the search difficulty of all the task areas is less than the preset difficulty, then continue to allocate the task areas responsible for searching to each first unmanned device, until all the task areas are allocated to the first unmanned devices for searching.

[0160] Optionally, the execution module 905 is specifically used to: for each task area, determine the search path of the unmanned equipment responsible for searching the task area based on the road direction of the task area; search for the target object on the road in the task area according to the search path; and track the target object after it is found.

[0161] Optionally, the execution module 905 is specifically used to determine the search path based on the road direction of the task area, with the goal of ensuring that the search ranges of each first unmanned device responsible for searching two adjacent task areas do not overlap and that each search range can cover its own task area.

[0162] Optionally, the execution module 905 is further configured to search for other target objects in the non-road area if the task for the target object has been completed.

[0163] Optionally, the execution module 905 is specifically used to determine the task execution difficulty corresponding to each target object based on the number of roads connecting to the endpoint of the target object's trajectory; and to assign a second unmanned device responsible for performing tasks on each target object in order of increasing task execution difficulty.

[0164] Optionally, the execution module 905 is specifically configured to: for each target object, select a specified number of second unmanned devices as candidate second unmanned devices based on the distance between each second unmanned device and the target object; determine the task path for each candidate second unmanned device to perform the task on the target object; and execute the task for the target object through each candidate second unmanned device according to its corresponding task path.

[0165] Optionally, the first unmanned device includes a drone, and the second unmanned device includes an unmanned vehicle.

[0166] This specification also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 This provides a method for task execution.

[0167] This instruction manual also provides Figure 10 The one shown corresponds to Figure 1 A schematic diagram of the structure of an electronic device. (e.g.) Figure 10 At the hardware level, the electronic device includes a processor, internal bus, network interface, memory, and non-volatile memory, and may also include other hardware required for the business operations. The processor reads the corresponding computer program from the non-volatile memory into memory and then runs it to achieve the above-mentioned functions. Figure 1The task execution method. Of course, in addition to the software implementation, the present specification does not exclude other implementations, such as a logic device or a combination of software and hardware, and the like, that is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or a logic device.

[0168] In the present specification, the terms "comprise", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such a process, method, article or device. Without more limitations, the element defined by the statement "comprises a" does not exclude the presence of other identical elements in the process, method, article or device comprising the element.

[0169] The above is only an embodiment of the present specification and is not intended to limit the present specification. Various changes and modifications can be made to the present specification by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present specification shall be included in the scope of claims of the present specification.

Claims

1. A method for task execution, characterized in that, include: Obtain the map image corresponding to the target area; Based on the connection relationships between each intersection and different roads in the map image, at least one task area is determined; For each mission area, the terrain type corresponding to the mission area is determined based on the altitude distribution of the mission area, and the distance between each first unmanned device and the center of the mission area is determined as the target distance; Based on the terrain type, target distance, and area size of each task area, and with the goal of ensuring that all task areas can be searched, each first unmanned device is assigned a task area to search. The search difficulty for each task area is determined based on its terrain type, target distance, and area size. Task areas are assigned to each first unmanned device in descending order of search difficulty until all task areas have been assigned to a first unmanned device for searching. Each first unmanned device searches its assigned task area and, upon finding a target object, sends a task execution command to a second unmanned device, enabling the second unmanned device to execute the task according to the command.

2. The method as described in claim 1, characterized in that, Based on the connection relationships between intersections and different roads in the map image, at least one task area is determined, specifically including: Using each intersection in the map image as a node and different roads as edges, a road network topology image of the target area is constructed; Based on the road network topology image, a raster image of the target area is constructed, and the at least one task area is determined based on the raster image.

3. The method as described in claim 2, characterized in that, For each mission area, the corresponding terrain type is determined based on the elevation distribution of that area, specifically including: Construct an elevation map corresponding to the raster image; For each task area, the altitude distribution of that task area is determined based on the pixel value distribution of the corresponding elevation map. Based on the elevation distribution, determine the corresponding landform type for the task area.

4. The method as described in claim 2, characterized in that, Based on the road network topology image, a raster image of the target region is constructed, and the at least one task region is determined based on the raster image, specifically including: Determine the minimum bounding rectangle region of the road network topology image; Based on the actual maximum and minimum latitude and longitude values ​​of the minimum bounding rectangle region, as well as the preset width and preset height values ​​of the raster image, the latitude and longitude coordinates corresponding to each pixel in the raster image are determined.

5. The method as described in claim 1, characterized in that, In accordance with the order of search difficulty from highest to lowest, each first unmanned device is assigned a task area to be searched, until all task areas have been assigned to the first unmanned devices for searching, specifically including: For each first unmanned device, if the terrain type of the task area assigned to the first unmanned device is a specified type, and the area of ​​the task area is not less than the average area of ​​all task areas, then no other task areas will be assigned to the first unmanned device.

6. The method as described in claim 1, characterized in that, In accordance with the order of search difficulty from highest to lowest, each first unmanned device is assigned a task area to be searched, until all task areas have been assigned to the first unmanned devices for searching, specifically including: If the search difficulty of the task area assigned to each unmanned device is greater than the preset difficulty, or the search difficulty of all task areas is less than the preset difficulty, then continue to assign task areas to each first unmanned device for searching, until all task areas are assigned to the first unmanned devices for searching.

7. The method as described in claim 1, characterized in that, Each unmanned aerial vehicle (UAV) searches its assigned task area, specifically including: For each task area, the search path for the unmanned equipment responsible for searching that task area is determined based on the road layout of that task area. Based on the search path, the target object is searched for on the roads in the task area, and after the target object is found, it is tracked.

8. The method as described in claim 7, characterized in that, For each mission area, the search path for the unmanned equipment responsible for searching that area is determined based on the road layout within that area. Specifically, this includes: Based on the road layout of the task area, the search path is determined with the goal of ensuring that the search ranges of each first unmanned device responsible for searching two adjacent task areas do not overlap and that each search range can cover its own task area.

9. The method as described in claim 7, characterized in that, The method further includes: If the task for the target object has been completed, then search for other target objects in the non-road area.

10. The method as described in claim 1, characterized in that, Each first unmanned device searches its assigned task area, and upon finding the target object, sends a task execution command to the second unmanned devices, enabling each second unmanned device to execute the task according to the command. Specifically, this includes: For each target object, the task execution difficulty corresponding to that target object is determined based on the number of roads connecting it to the endpoint of its trajectory. According to the order of task execution difficulty from low to high, the second unmanned equipment is assigned to perform the task on each target object.

11. The method as described in claim 10, characterized in that, According to the order of task execution difficulty from lowest to highest, second unmanned devices are assigned to perform tasks on each target object, specifically including: For each target object, a specified number of second unmanned devices are selected as candidate second unmanned devices based on the distance between each second unmanned device and the target object; Determine the task path for each candidate second unmanned device to perform the task on the target object; Each candidate second unmanned device executes a task targeting the object according to its corresponding task path.

12. The method as described in claim 1, characterized in that, The first unmanned device includes a drone, and the second unmanned device includes an unmanned vehicle.

13. A device for performing a task, characterized in that, include: The acquisition module retrieves the map image corresponding to the target area; The first determining module determines at least one task area based on the connection relationship between each intersection and different roads in the map image; The second determination module determines the terrain type of each task area based on the altitude distribution of that task area, and determines the distance between each first unmanned device and the center of the task area as the target distance. The allocation module assigns task areas to each first unmanned device based on the terrain type, target distance, and area size of each task area, with the goal of ensuring that all task areas can be searched. Specifically, it determines the search difficulty of each task area based on its terrain type, target distance, and area size. Task areas are then assigned to each first unmanned device in descending order of search difficulty until all task areas have been assigned to a first unmanned device for searching. The execution module searches the task area it is responsible for using each first unmanned device, and sends a task execution command to the second unmanned devices after finding the target object, so that each second unmanned device can execute the task according to the task execution command.

14. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method described in any one of claims 1 to 12.

15. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method described in any one of claims 1 to 12.

Images