Method and device for carrying out datum point extraction and collaborative distribution on isomorphic fixed group targets

Abstract

The invention provides a reference point extraction and cooperative distribution method and device for isomorphic fixed group targets, and the method comprises the steps: obtaining a target region image of the isomorphic fixed group targets, and recognizing a plurality of targets forming the isomorphic fixed group targets from the target region image; dividing a reference group including a preset reference point from the plurality of targets; a sliding window is adopted to sequentially intercept a plurality of comparison marshals along the front boundary lines of the targets; the similarity between each comparison group and the reference group is calculated, and the comparison group with the maximum similarity with the reference group is screened out to serve as a group reference group; determining a target reference point based on the marshalling reference group, and carrying out the cooperative distribution of the flight based on the target reference point; wherein the target number in the reference group is equal to the target number in each comparison group. Through the method disclosed by the invention, the probability of overlapping strike can be reduced, and the cooperative efficiency between aircrafts is improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of aircraft manufacturing, and in particular to a method and device for extracting and coordinating reference points for a homogeneous fixed group of targets. Background Art

[0002] With the rapid development of guidance technology and target image intelligent recognition technology, collaborative operations on multiple isomorphic fixed targets through multiple aircraft have gradually become a new operation mode in the future. When performing collaborative operations on multiple isomorphic fixed targets, if the distance between the targets is small and the total number of targets is much larger than the total number of aircraft in the multi-aircraft, it is generally necessary to group the multiple targets first, and then select the collision targets corresponding to each aircraft in the group for collision operations. However, in the case where there is no data link between the aircraft and information exchange cannot be carried out, each aircraft cannot establish a coordinate system with reference to the same reference point to achieve the grouping of multiple targets, which is easy to produce overlapping strikes and reduce the collaborative efficiency between aircraft. Therefore, how to make the reference points selected by each aircraft the same or converge within a smaller range in the absence of a data link and information exchange, improve the consistency of the grouping between the aircraft, and then reduce the probability of overlapping strikes and improve the collaborative efficiency between aircraft is a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention

[0003] In view of this, the present disclosure proposes a method and device for extracting and collaboratively allocating reference points for homogeneous fixed group targets. In the absence of data links and information exchange, the reference points selected by each aircraft can be made the same or converged within a smaller range, thereby improving the consistency of grouping between aircraft, thereby reducing the probability of overlapping strikes and improving the collaborative efficiency between aircraft.

[0004] According to a first aspect of the present disclosure, a method for extracting and co-assigning reference points for a homogeneous fixed group of targets is provided, comprising: Acquire a target area image of a homogeneous fixed group of targets, and identify a plurality of targets constituting the homogeneous fixed group of targets from the target area image; Dividing a reference group including a preset reference point from the plurality of targets; Using a sliding window to sequentially extract multiple comparison groups along the front boundary lines of the multiple targets; Calculating the similarity between each of the comparison groups and the reference group, and selecting the comparison group with the greatest similarity to the reference group as the reference group; determining a target reference point based on the marshaling reference group, and co-assigning flights based on the target reference point; The number of targets included in the benchmark grouping and each of the comparison groups is equal.

[0005] In a possible implementation, when acquiring a target area image of a homogeneous fixed group of targets, the process includes: Acquire a regional image of the area where the homogeneous fixed group targets are located, and determine the center of the field of view from the regional image; Based on the field of view center and the inertial navigation error square value, the target area image is cut out from the area image.

[0006] In a possible implementation, when dividing the reference group including the preset reference point from the plurality of targets, it is implemented based on a pre-constructed reference group recognition model.

[0007] In a possible implementation, when constructing the reference grouping recognition model, it includes: Acquire a set number of sample images of the area where the homogeneous fixed group targets are located; Marking the preset reference points and reference groups including the preset reference points in each of the sample images to form a plurality of training images with markings; The preloaded machine learning model is trained using each of the training images to obtain the reference grouping recognition model.

[0008] In a possible implementation, when calculating the similarity between each of the comparison groups and the reference group, the following steps are included: Calculating the grid features corresponding to each of the comparison groups and the grid features corresponding to the reference group; Calculating the information entropy of each of the comparison groups relative to the reference group based on the grid features corresponding to each of the comparison groups and the grid features corresponding to the reference group; Based on the information entropy of each of the comparison groups relative to the reference group, the similarity between each of the comparison groups and the reference group is determined.

[0009] In a possible implementation, when calculating the grid features corresponding to the reference grouping, the process includes: Selecting a starting point target and an ending point target from the reference group, and determining an initial vector based on the starting point target and the ending point target; Calculating the cosine distance of each target in the reference group except the starting point target relative to the initial vector; Calculating the Euclidean distance between each target in the reference group except the starting target and the starting target, and determining the range of the Euclidean distance in the reference group; The grid feature corresponding to the reference grouping is determined based on the cosine distance of each target in the reference grouping except the starting point target relative to the initial vector and the Euclidean distance between the target and the starting point target and the range of the Euclidean distance in the reference grouping.

[0010] In a possible implementation, when determining a target reference point based on the grouping reference group, the method includes: The target located in the middle of the front boundary of the grouping in the grouping reference group is used as the target reference point.

[0011] According to a second aspect of the present disclosure, a device for extracting and co-assigning reference points for a homogeneous fixed group of targets is provided, comprising: An image processing module, used for acquiring a target area image of a homogeneous fixed group of targets, and identifying a plurality of targets constituting the homogeneous fixed group of targets from the target area image; A reference group identification module, used to divide a reference group including a preset reference point from the plurality of targets; A comparison grouping interception module, used for sequentially intercepting a plurality of comparison groups along the front boundary lines of a plurality of the targets using a sliding window; A grouping reference group screening module, used for calculating the similarity between each of the comparison groups and the reference group, and screening out the comparison group with the greatest similarity to the reference group as the grouping reference group; an allocation module, configured to determine a target reference point based on the marshaling reference group, and to coordinately allocate flights based on the target reference point; The number of targets included in the benchmark grouping and each of the comparison groups is equal.

[0012] According to a third aspect of the present disclosure, a device for extracting and co-assigning reference points for a homogeneous fixed group of targets is provided, comprising: a processor; a memory for storing processor executable instructions; wherein the processor is configured to execute the method described in the first aspect of the present disclosure.

[0013] According to a fourth aspect of the present disclosure, a non-volatile computer-readable storage medium is provided, on which computer program instructions are stored, wherein the computer program instructions, when executed by a processor, implement the method described in the first aspect of the present disclosure.

[0014] The present disclosure provides a method and device for extracting and collaboratively allocating reference points for a homogeneous fixed group of targets. The method comprises: acquiring a target area image of a homogeneous fixed group of targets, and identifying a plurality of targets constituting the homogeneous fixed group of targets from the target area image; dividing a reference group including a preset reference point from the plurality of targets; sequentially intercepting a plurality of comparison groups along the front boundary lines of the plurality of targets using a sliding window; calculating the similarity between each comparison group and the reference group, and selecting the comparison group with the greatest similarity to the reference group as the grouping reference group; determining a target reference point based on the grouping reference group, and collaboratively allocating flights based on the target reference point; wherein the number of targets included in the reference group and each comparison group is equal. In the present disclosure, each sub-aircraft executing the same collaborative allocation method mentioned above can screen out the same benchmark grouping that includes the same preset benchmark point, and then calculate the target benchmark points that are the same or converge within a smaller range based on the benchmark grouping. Furthermore, since the target benchmark points selected by each sub-aircraft are the same or converge within a smaller range, the consistency of the grouping between the sub-aircraft can be improved in the subsequent collaborative allocation process, thereby reducing the probability of overlapping strikes and improving the collaborative efficiency between aircraft.

[0015] Further features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0017] Figure 1 A flow chart showing a method for extracting and co-assigning reference points for a homogeneous fixed group of targets according to an embodiment of the present disclosure; Figure 2 A schematic diagram showing the result of performing target recognition on a target area image according to an embodiment of the present disclosure; Figure 3 A schematic diagram showing a labeled training image according to an embodiment of the present disclosure; Figure 4 A schematic diagram showing a reference grouping including preset reference points according to an embodiment of the present disclosure; Figure 5 A schematic diagram of a grid feature calculation process according to an embodiment of the present disclosure is shown; Figure 6 A schematic diagram showing a comparison grouping interception result according to an embodiment of the present disclosure; Figure 7 A schematic diagram showing a comparison grouping interception result according to another embodiment of the present disclosure; Figure 8A flow chart showing a method for collaboratively allocating aircraft based on target reference points according to an embodiment of the present disclosure; Fig. 9 A schematic block diagram showing an apparatus for extracting and co-assigning reference points for a homogeneous fixed group of targets according to an embodiment of the present disclosure is shown; Fig.10 A schematic block diagram of a device for extracting and co-assigning reference points for a homogeneous fixed group of targets according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0018] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0019] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0020] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present disclosure.

[0021] <Method Example> Figure 1 A flow chart of a method for extracting and collaboratively allocating reference points for a homogeneous fixed group of targets according to an embodiment of the present disclosure is shown. The method is independently implemented by the seeker heads of each sub-aircraft in a multi-aircraft, so that multiple sub-aircraft can locate the same target reference point or a target reference point that converges within a smaller range through the same calculation process without a data link and without the ability to interact, thereby ensuring the consistency of the grouping of each sub-aircraft based on the target reference point, reducing the overlap rate of subsequent strikes, and improving collaborative efficiency. Since the collaborative allocation process of each sub-aircraft seeker is the same, this embodiment takes a sub-aircraft seeker as an example to explain in detail the algorithm steps when performing collaborative allocation. Figure 1 As shown, the method includes steps S1100-S1500.

[0022] S1100, obtaining a target area image of a homogeneous fixed group target, and identifying multiple targets constituting the homogeneous fixed group target from the target area image. In this embodiment, the homogeneous fixed group target refers to a fixed group target in which the total number of targets is much larger than the number of sub-aircraft in the multi-aircraft for coordinated collision, the distribution of each target has a certain degree of non-uniformity, and the distance between each target is small.

[0023] In a possible implementation, when acquiring a target area image of a homogeneous fixed group of targets, the following steps may be included: First, obtain the regional image of the area where the homogeneous fixed group of targets is located, and determine the center of the field of view from the regional image. Specifically, when performing collaborative allocation calculation, it is necessary to first obtain the regional image of the area where the homogeneous fixed group of targets is located through a camera device. After obtaining the regional image, the center position of the field of view can be further determined from the regional image ( , ).

[0024] Second, based on the center of the field of view and the square value of the inertial navigation error, the target area image is cut out from the area image. Specifically, the center position of the field of view ( , ) as the reference, expand the inertial guidance error square value twice to the left in the horizontal direction of the regional image As the left border of the target area image, that is, the left border of the target area image = -2 . Expand the square value of the inertial guidance error to the right by twice the horizontal direction of the regional image As the right border of the target area image, that is, the right border of the target area image = +2 The upper and lower boundaries of the target region image are the upper and lower boundaries of the original region image. The image surrounded by the upper, lower, left, and right boundaries of the target region image is the final target region image.

[0025] After determining the target region image of the homogeneous fixed group target, the target recognition model can be used to identify multiple targets constituting the homogeneous fixed group target. In a specific example (hereinafter collectively referred to as Example 1), the result of target recognition on the target region image can be as follows: Figure 2 shown.

[0026] After the target recognition of the target area image is completed, step S1200 may be executed to divide the multiple targets in the target area image into reference groups including preset reference points.

[0027] In a possible implementation, when a reference group including a preset reference point is divided from a plurality of objects in a target area image, it is implemented based on a pre-built reference group recognition model. The reference group recognition model is a machine learning model that can accurately divide a reference group including a preset reference point from a plurality of objects in a target area image.

[0028] It should be noted here that, in this embodiment, a reference grouping identification model needs to be pre-built. In a possible implementation, the steps of building the reference grouping identification model can be as follows: First, a set number of sample images of the area where the homogeneous fixed group targets are located are obtained. Specifically, regional images of homogeneous fixed group targets under different meteorological conditions, different time periods and different viewing angles are collected to obtain a set number of regional images. The target regional image is intercepted from each regional image as the sample image. The method of capturing the target regional image is as described above and will not be repeated here.

[0029] Second, a preset reference point and a reference group including the preset reference point are marked in each sample image to form a plurality of training images with annotations. Specifically, the annotation process of each sample image is as follows: First, the number of targets in each group is determined according to the total number of sub-aircraft in the multi-aircraft and the total number of targets in the isomorphic fixed group target, wherein the number of targets in each group = the total number of targets in the isomorphic fixed group target / the total number of aircraft. Then, the isomorphic fixed group targets are evenly divided into multiple groups according to the number of targets in each group. In the process of grouping, if the number of remaining targets is less than the number of targets in the above-mentioned groups, the remaining targets are directly used as the last group. After completing the grouping of the targets, the front boundary line of the isomorphic fixed group targets (the boundary line indicated by the flight direction of the aircraft) is determined, and a specific target is identified from the multiple targets close to the midpoint of the front boundary line and marked as a preset reference point, and the group including the preset reference point is marked as a reference group. The sample image marked with the preset reference point and the reference group is used as a training image. The preset reference points marked in each sample image correspond to the same physical target in the isomorphic fixed group target. In this way, when the machine learning model is trained later, the machine learning model can accurately identify the unified preset reference points in different input images and determine the basically consistent reference grouping based on the preset reference points. In a specific example, the marked training image is as follows: Figure 3 Referring to the above method, a set number of training images can be obtained.

[0030] Third, the preloaded machine learning model is trained using each training image to obtain a reference grouping recognition model. By training the loaded machine learning model using a set number of training images, a reference grouping recognition model that can automatically divide a reference group including a preset reference point from multiple targets in the target area image can be obtained.

[0031] After the benchmark grouping recognition model is constructed, the benchmark grouping recognition model is loaded into the seeker heads of each sub-aircraft under the multi-aircraft. In this way, after the seeker heads of each aircraft complete the target recognition of the target area image, they can call the pre-loaded benchmark grouping recognition model to recognize the basic grouping. Specifically, the target area image after target recognition is input into the benchmark grouping recognition model, and the benchmark grouping recognition model can divide the multiple targets in the target area image into benchmark groups including preset benchmark points. In the above Example 1, the benchmark grouping including preset benchmark points divided by the benchmark grouping recognition model is as follows Figure 4 shown.

[0032] After dividing the reference group, the grid features of the reference group can be calculated. In a possible implementation, the following steps can be included when calculating the grid features corresponding to the reference group: First, select the starting target and the end target from the benchmark group, and determine the initial vector based on the starting target and the end target. Figure 5 As shown in the figure, the target near the upper left corner of the grid boundary in the reference group is taken as the starting target, the target with the largest distance from the starting target in the reference group is taken as the ending target, and then the vector from the ending target to the starting target is determined as the initial vector in the reference group (i.e. Figure 5 The grid margins are the boundaries between groups.

[0033] Second, calculate the cosine distance of each target in the reference group except the starting target relative to the initial vector. Specifically, for each target except the starting target, construct a vector from each target to the starting target as the target vector corresponding to each target ( Figure 5 Only the target vectors corresponding to the four targets in the group are shown, and the target vectors corresponding to other targets are not shown). The cosine distance between the target vector corresponding to each target and the initial vector is calculated, and the cosine distance between the target vector corresponding to each target and the initial vector is used as the cosine distance of each target relative to the initial vector. The calculation formula of the cosine distance is as follows: In the formula, X is the initial vector, Y i The first target in the group except the starting point itarget, D( X , Y i ) is the first target other than the starting point i The cosine distance between each target and the initial vector, where n is the total number of targets in the group.

[0034] Third, the Euclidean distance between each target other than the starting target and the starting target in the reference group is calculated, and the range of the Euclidean distance in the reference group is determined. Specifically, for each target other than the starting target, the Euclidean distance between each target and the starting target is calculated, thereby obtaining multiple calculated values ​​of the Euclidean distance, and the difference between the maximum value and the minimum value of the multiple Euclidean distances is calculated as the range of the Euclidean distance in the reference group.

[0035] Fourth, based on the cosine distance of each target in the benchmark group except the starting point target relative to the initial vector and the Euclidean distance between each target and the starting point target, as well as the range of the Euclidean distance in the benchmark group, the grid feature corresponding to the benchmark group is determined. Specifically, after the above calculation, each target in the benchmark group except the starting point target has a corresponding cosine distance value and Euclidean distance value. The cosine distance values ​​are sorted from small to large according to the Euclidean distance values ​​corresponding to each target, and the range of the Euclidean distance in the benchmark group is added to the end of the sorting, and the grid feature corresponding to the benchmark group can be obtained.

[0036] For example, when the benchmark group o includes n targets, the n-1 targets except the starting target will correspond to a cosine distance value and a Euclidean distance value, that is, n-1 cosine distance values ​​and n-1 Euclidean distance values ​​will be calculated, and a range will be calculated based on the n-1 Euclidean distance values. The cosine distance values ​​are sorted in the order of the Euclidean distance values ​​corresponding to each target from small to large, and a cosine distance value sequence consisting of n-1 cosine distance values ​​will be obtained. , add the range A of the Euclidean distance within the reference group o to the end of the cosine distance value sequence o The grid characteristics of the benchmark group o can be obtained . Where o is the number of the base group.

[0037] After completing the division of the benchmark grouping, step 1300 can be executed, and a plurality of comparison groups can be sequentially intercepted along the front boundary lines of the plurality of targets using a sliding window. The size of the sliding window is determined based on the size of the benchmark grouping. Specifically, the number of target rows and the number of targets in each row in the benchmark grouping are first determined, and the number of target rows in the benchmark grouping is used as the number of target rows covered by the sliding window, and the number of targets in each row in the benchmark grouping is used as the number of targets in each row covered by the sliding window. The size of the sliding window is determined based on the number of target rows covered by the sliding window and the number of targets in each row, that is, the size of the sliding window is determined by the number of target rows covered and the number of targets in each row. For example, in a benchmark grouping such as Figure 4 In the embodiment shown, which includes 3 rows of targets and each row includes 5 targets, each sliding of the sliding window will cover 3 rows of targets and the number of targets in each row is 5.

[0038] When intercepting the comparison group, the sliding window intercepts the comparison group from left to right along the front boundary. After completing the interception of a group of comparison groups, it moves one column of targets to the left and continues to intercept the next comparison group until the group can cover a column of targets on the rightmost edge, and then stops intercepting the comparison group. For example, in the target area image, multiple targets such as Figure 6 When the sliding window is shown, it will first be cut out along the leftmost side of the front boundary line. Figure 6 After extracting the comparison group 1, move one column to the right and continue to extract the comparison group 1. Figure 7 The comparison group 2 shown in , and so on, until all the comparison groups are extracted.

[0039] After the multiple comparison groups are intercepted, step S1400 may be executed to calculate the similarity between each comparison group and the reference group, and select the comparison group with the greatest similarity to the reference group as the grouping reference group.

[0040] In a possible implementation, the similarity between each comparison grouping and the reference grouping is characterized by the information entropy between each comparison grouping and the reference grouping. In this implementation, the similarity calculation steps between each comparison grouping and the reference grouping can be as follows: First, the grid features corresponding to each comparison group and the grid features corresponding to the benchmark group are calculated. Specifically, the grid features corresponding to the benchmark group have been described above and will not be repeated here. The calculation process of the grid features corresponding to each comparison group and the grid features corresponding to the benchmark group is the same and will not be repeated here.

[0041] Second, based on the grid features corresponding to each comparison group and the grid features corresponding to the reference group, the information entropy of each comparison group relative to the reference group is calculated. Specifically, the calculation formula for the information entropy of the comparison group relative to the reference group is as follows: In the formula, The first grid feature of the base group o i elements, is the first grid feature in the jth comparison group i elements, is the information entropy of the jth comparison group relative to the benchmark group o, and n is the total number of targets in each group.

[0042] Referring to the above formula, the information entropy of each comparison group relative to the benchmark group can be calculated.

[0043] Third, based on the information entropy of each comparison group relative to the benchmark group, the similarity between each comparison group and the benchmark group is determined. Specifically, the greater the information entropy, the smaller the similarity between the comparison group and the benchmark group. Therefore, the negative number of the information entropy of each comparison group relative to the benchmark group is used as the similarity between each comparison group and the benchmark group.

[0044] After determining the similarity between each comparison grouping and the benchmark grouping, the comparison grouping with the greatest similarity to the benchmark grouping is selected as the grouping benchmark group.

[0045] It should be noted here that in another possible implementation, the comparison grouping can be intercepted while calculating the similarity between the comparison grouping and the benchmark grouping, until the comparison grouping with the smallest similarity to the benchmark grouping is found, then the interception of the comparison grouping is stopped, and the comparison grouping with the smallest similarity to the quasi-grouping is used as the grouping benchmark group. This implementation can greatly reduce the amount of calculation of the grouping benchmark group and improve the efficiency of collaborative allocation.

[0046] S1500, determining a target reference point based on the marshaling reference group, and collaboratively allocating flights based on the target reference point.

[0047] In a possible implementation, when determining the target reference point based on the marshaling reference group, a target in the marshaling reference group located in the middle of the front boundary of the marshaling is used as the target reference point.

[0048] After the target reference point is determined, the aircraft can be coordinated and allocated based on the target reference point. The specific steps are as follows: Figure 8 As shown, steps S1510-S1530 are included.

[0049] S1510, calculating the positional relationship between each target in the target area image and the target reference point, and sorting each target based on the positional relationship between each target and the target reference point.

[0050] In a possible implementation, when calculating the positional relationship between each target and the target reference point, the following steps may be included: First, select the target closest to the target reference point from among all the targets as the reference target.

[0051] It should be noted here that in order to ensure that the reference targets selected by each sub-aircraft after independent calculation are the same, it is necessary to ensure that there is only one target closest to the target reference point. Therefore, after obtaining the reference target, it is necessary to further determine whether the reference target closest to the target reference point is unique: if the target reference point is unique, then perform the operation of constructing the initial vector based on the target reference point and the reference target; if the target reference point is not unique, it is necessary to select a unique reference target as the final reference target based on the preset selection strategy, and then construct the initial vector based on the final reference target and the target reference point. Among them, the preset selection strategy in each sub-aircraft is consistent to ensure that each sub-aircraft can select the same reference target when calculating independently. In a possible implementation, the preset selection strategy can be to select the reference target located on the far left from multiple reference targets with equal distances from the target reference point as the final reference target. In another possible implementation, the preset selection strategy can be to select the reference target located on the far right from multiple reference targets with equal distances from the target reference point as the final reference target.

[0052] Second, based on the reference target and the target reference point, an initial vector is constructed. Specifically, with the target reference point as the starting point and the reference target as the end point, an initial vector is constructed from the target reference point to the reference target.

[0053] Third, the line vector between each target and the target reference point is calculated. Specifically, with the target reference point as the starting point and each target as the end point, a line vector from the target reference point to each target is constructed.

[0054] Fourth, based on the line vector and the initial vector between each target and the target reference point, the positional relationship between each target and the target reference point is calculated. Specifically, the cosine distance between the line vector and the initial vector between each target and the target reference point is calculated, and the positional relationship between each target and the target reference point is characterized by the calculated cosine distance.

[0055] After calculating the positional relationship between each target and the target reference point, each target will be sorted according to the positional relationship between each target and the target reference point. In an embodiment where the positional relationship is characterized by the cosine distance between the line vector between each target and the target reference point and the initial vector, each target will be sorted in descending order of the cosine distance.

[0056] S1520: Divide the sorted targets into a number of groups equal to the number of sub-aircraft, and assign a corresponding group to each sub-aircraft.

[0057] In a possible implementation, when the sorted targets are divided into a plurality of groups equal to the number of sub-aircraft, it includes: according to the number of targets in each group determined above, the sorted targets are divided into a plurality of groups equal to the total number of sub-aircraft. For example, if the total number of targets is 30, the total number of sub-aircraft is 6, and the number of targets in each group is 30 / 6=5, then the targets sorted 1-5 are divided into the first group; the targets sorted 6-10 are divided into the second group; the targets sorted 11-15 are divided into the third group, the targets sorted 16-20 are divided into the fourth group, the targets sorted 11-25 are divided into the fifth group, and the targets sorted 26-30 are divided into the sixth group.

[0058] After completing the target grouping, a corresponding grouping will be assigned to the sub-aircraft. Specifically, each sub-aircraft in the multi-aircraft has its own number, and the numbering of each sub-aircraft is consistent with the grouping numbering rule. Continuing with the above embodiment, among the 6 groups after division, the first group is numbered 1, the second group is numbered 2, the third group is numbered 3, the fourth group is numbered 4, the fifth group is numbered 5, and the sixth group is numbered 6, then the numbers of the 6 sub-aircraft will be 1, 2, 3, 4, 5, and 6 in sequence. When assigning a corresponding group to a sub-aircraft, the group that is consistent with the sub-aircraft number is assigned to it as the corresponding group. Continuing with the above embodiment, the first group is assigned to the sub-aircraft numbered 1, the second group is assigned to the sub-aircraft numbered 2, the third group is assigned to the sub-aircraft numbered 3, the fourth group is assigned to the sub-aircraft numbered 4, and the fifth group is assigned to the sub-aircraft numbered 5.

[0059] It should be noted here that after performing independent calculations according to the above scheme, the grouping results of each sub-aircraft for the target are the same. The allocation of groupings according to the sub-aircraft numbers can achieve unique and non-overlapping groupings of sub-aircraft, thereby effectively avoiding overlapping collisions of targets by each sub-aircraft.

[0060] S1530, the sub-aircraft selects a target from the corresponding group as a cooperative collision target.

[0061] In a possible implementation, when a target in a corresponding group is selected as a cooperative collision target, it is implemented based on the distance between each target in the group and the target reference point.

[0062] Specifically, for the assigned group, the Euclidean distance between each target in the group and the target reference point is calculated. The calculation formula of the Euclidean distance is as follows: Where, T c (x c, z c ) is the position coordinate of the target reference point, T i (x i , z i ) is the position coordinate of the i-th target in the group, Si (T i , T c ) is the Euclidean distance between the i-th target in the group and the target reference point.

[0063] After calculating the Euclidean distance between each target in the group and the target reference point, the target in the group with the smallest Euclidean distance is selected as the cooperative collision target. When there are more than two targets in the group with the same Euclidean distance to the target reference point, the target in the group closest to the center of the field of view can be selected as the final cooperative collision target.

[0064] It should be noted here that the method and strategy for independent calculation of each sub-aircraft is completely consistent, thereby ensuring that each sub-aircraft can achieve the purpose of non-overlapping collaborative collision target selection when calculating completely independently.

[0065] The present disclosure provides a method for extracting and collaboratively allocating reference points for a homogeneous fixed group of targets, comprising: acquiring a target area image of a homogeneous fixed group of targets, and identifying a plurality of targets constituting the homogeneous fixed group of targets from the target area image; dividing a reference group including a preset reference point from the plurality of targets; sequentially intercepting a plurality of comparison groups along the front boundary lines of the plurality of targets using a sliding window; calculating the similarity between each comparison group and the reference group, and selecting the comparison group with the greatest similarity to the reference group as the grouping reference group; determining a target reference point based on the grouping reference group, and collaboratively allocating flights based on the target reference point; wherein the number of targets included in the reference group and each comparison group is equal. In the present disclosure, each sub-aircraft executing the same collaborative allocation method mentioned above can screen out the same benchmark grouping that includes the same preset benchmark point, and then calculate the target benchmark points that are the same or converge within a smaller range based on the benchmark grouping. Furthermore, since the target benchmark points selected by each sub-aircraft are the same or converge within a smaller range, the consistency of the grouping between the sub-aircraft can be improved in the subsequent collaborative allocation process, thereby reducing the probability of overlapping strikes and improving the collaborative efficiency between the sub-aircraft.

[0066] <Device Example> Fig. 9 FIG. 1 is a schematic block diagram of an apparatus for extracting and co-assigning reference points for a homogeneous fixed group of targets according to an embodiment of the present disclosure. Fig. 9 As shown, the device 100 includes: The image processing module 110 is used to obtain a target area image of a homogeneous fixed group of targets, and identify multiple targets constituting the homogeneous fixed group of targets from the target area image; A reference group identification module 120, for dividing a reference group including a preset reference point from a plurality of targets; A comparison grouping interception module 130 is used to sequentially intercept multiple comparison groups along the front boundary lines of multiple targets using a sliding window; A grouping reference group screening module 140 is used to calculate the similarity between each comparison group and the reference group, and screen out the comparison group with the greatest similarity to the reference group as the grouping reference group; An allocation module 150, for determining a target reference point based on the marshaling reference group, and collaboratively allocating flights based on the target reference point; Among them, the number of targets included in the benchmark grouping and each comparison grouping is equal.

[0067] <Equipment Embodiment> Fig.10 A schematic block diagram of a device for extracting and co-assigning reference points to a homogeneous fixed group of targets according to an embodiment of the present disclosure is shown. Fig.10 As shown, the device 200 for extracting reference points and co-assigning a homogeneous fixed group of targets includes: a processor 210 and a memory 220 for storing executable instructions of the processor 210. The processor 210 is configured to implement any of the above-mentioned methods for extracting reference points and co-assigning a homogeneous fixed group of targets when executing the executable instructions.

[0068] Here, it should be noted that the number of processors 210 may be one or more. Meanwhile, in the device 200 for extracting reference points and co-assigning homogeneous fixed group targets in the embodiment of the present disclosure, an input device 230 and an output device 240 may also be included. The processor 210, the memory 220, the input device 230 and the output device 240 may be connected via a bus or in other ways, which are not specifically limited here.

[0069] The memory 220, as a computer-readable storage medium, can be used to store software programs, computer executable programs, and various modules, such as: programs or modules corresponding to the method for extracting reference points and co-assigning reference points to a homogeneous fixed group of targets in an embodiment of the present disclosure. The processor 210 executes various functional applications and data processing of the device 200 for extracting reference points and co-assigning reference points to a homogeneous fixed group of targets by running the software programs or modules stored in the memory 220.

[0070] The input device 230 may be used to receive input numbers or signals. The signals may be key signals related to user settings and function control of the device / terminal / server. The output device 240 may include display devices such as display screens.

[0071] <Storage Medium Embodiment> According to a fourth aspect of the present disclosure, a non-volatile computer-readable storage medium is also provided, on which computer program instructions are stored. When the computer program instructions are executed by the processor 210, any of the above-mentioned methods for extracting and collaboratively allocating reference points for homogeneous fixed group targets is implemented.

[0072] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for extracting and co-assigning reference points for a homogeneous fixed group of targets, characterized in that: include: Acquire a target area image of a homogeneous fixed group of targets, and identify a plurality of targets constituting the homogeneous fixed group of targets from the target area image; Dividing a reference group including a preset reference point from the plurality of targets; Using a sliding window to sequentially extract multiple comparison groups along the front boundary lines of the multiple targets; Calculating the similarity between each of the comparison groups and the reference group, and selecting the comparison group with the greatest similarity to the reference group as the reference group; determining a target reference point based on the marshaling reference group, and co-assigning flights based on the target reference point; The number of targets included in the benchmark grouping and each of the comparison groups is equal.

2. The method according to claim 1, characterized in that When acquiring the target area image of a homogeneous fixed group target, it includes: Acquire a regional image of the area where the homogeneous fixed group targets are located, and determine the center of the field of view from the regional image; Based on the field of view center and the inertial navigation error square value, the target area image is cut out from the area image.

3. The method according to claim 1, characterized in that When a reference group including a preset reference point is divided from the plurality of targets, it is implemented based on a pre-constructed reference group recognition model.

4. The method according to claim 1, characterized in that: When constructing the benchmark grouping identification model, it includes: Acquire a set number of sample images of the area where the homogeneous fixed group targets are located; Marking the preset reference points and reference groups including the preset reference points in each of the sample images to form a plurality of training images with markings; The preloaded machine learning model is trained using each of the training images to obtain the reference grouping recognition model.

5. The method according to claim 1, characterized in that When calculating the similarity between each of the comparison groups and the reference group, it includes: Calculating the grid features corresponding to each of the comparison groups and the grid features corresponding to the reference group; Calculating the information entropy of each of the comparison groups relative to the reference group based on the grid features corresponding to each of the comparison groups and the grid features corresponding to the reference group; Based on the information entropy of each of the comparison groups relative to the reference group, the similarity between each of the comparison groups and the reference group is determined.

6. The method according to claim 1, characterized in that When calculating the grid features corresponding to the reference grouping, it includes: Selecting a starting point target and an ending point target from the reference group, and determining an initial vector based on the starting point target and the ending point target; Calculating the cosine distance of each target in the reference group except the starting point target relative to the initial vector; Calculating the Euclidean distance between each target in the reference group except the starting target and the starting target, and determining the range of the Euclidean distance in the reference group; The grid feature corresponding to the reference grouping is determined based on the cosine distance of each target in the reference grouping except the starting point target relative to the initial vector and the Euclidean distance between the target and the starting point target and the range of the Euclidean distance in the reference grouping.

7. The method according to claim 1, characterized in that When determining a target reference point based on the grouping reference group, the method includes: The target located in the middle of the front boundary of the grouping in the grouping reference group is used as the target reference point.

8. A device for extracting reference points and co-assigning reference points to a homogeneous fixed group of targets, characterized in that: include: An image processing module, used for acquiring a target area image of a homogeneous fixed group of targets, and identifying a plurality of targets constituting the homogeneous fixed group of targets from the target area image; A reference group identification module, used to divide a reference group including a preset reference point from the plurality of targets; A comparison grouping interception module is used to sequentially intercept multiple comparison groups along the front boundary lines of multiple targets using a sliding window; A grouping reference group screening module, used for calculating the similarity between each of the comparison groups and the reference group, and screening out the comparison group with the greatest similarity to the reference group as the grouping reference group; an allocation module, configured to determine a target reference point based on the marshaling reference group, and to coordinately allocate flights based on the target reference point; The number of targets included in the benchmark grouping and each of the comparison groups is equal.

9. A device for extracting and co-assigning reference points for a homogeneous fixed group of targets, characterized in that: include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to implement the method described in any one of claims 1 to 7 when executing the executable instructions.

10. A non-volatile computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images