A target identification method and system suitable for military management

Abstract

The application provides a target identification method and system suitable for military management, and relates to the field of automatic target identification; the method comprises the following steps: dividing a target identification grid template according to a three-dimensional entity model of a target identification area; acquiring standard remote sensing images corresponding to a plurality of time periods in a geographic information database and environment characteristics of each standard remote sensing image; for any target identification grid template, performing image fusion on the behavior characteristics and the standard remote sensing images according to the behavior characteristics of a target of interest, so as to obtain a template image of each target of interest; constructing a convolutional neural network model, performing model training and optimization according to the template image, and obtaining a target identification model; and identifying the target of interest according to the target identification model; the application divides the target identification area into grids, and uses the behavior characteristics of the target of interest to predict the behavior of the target, so as to use the predicted behavior as training data to train the model, and then realize high-accuracy identification of the target.

Description

Technical Field

[0001] This invention relates to the field of automatic target identification technology, and specifically to a target identification method and system suitable for military management. Background Technology

[0002] Military target identification technology refers to the accurate identification and judgment of targets on land, sea, or in the air through specific technical means, providing reliable basis for command and decision-making and accurate guidance and support for military operations. For example, in war exercises, the military can identify enemy targets to determine the enemy's troop strength, equipment, and other information; in counter-terrorism operations, the police can identify suspected targets to track and combat terrorists; and in maritime patrols, the coast guard can identify suspicious vessels to combat piracy and illegal fishing boats.

[0003] In military target identification, target information can be acquired through various channels, such as acquiring images containing military target information via vehicle-mounted, airborne, satellite, and UAV platforms. When targets are concealed in complex terrain and topographical environments, a large number of images need to be acquired for comparison and calculation. This results in a large computational load and low accuracy in processing and analyzing such image data. In military operations, decision-making needs to be rapid and accurate. Therefore, a method for intelligent target identification and judgment is needed that can not only perform rapid and automatic analysis of massive amounts of data but also ensure data accuracy, thus providing important evidence for military patents. Summary of the Invention

[0004] The purpose of this invention is to provide a target identification method and system suitable for military management. It uses machine learning to match and train the environmental features of the target identification area with the behavioral features of the target to obtain a target identification model, and finally uses the model to identify the target. This solution can not only automatically analyze the data to quickly complete the target identification, but also effectively ensure the accuracy of the identification.

[0005] To achieve the above objectives, the present invention proposes the following technical solution:

[0006] In a first aspect, the present invention proposes a target identification method suitable for military management, comprising:

[0007] Based on the 3D solid model of the target recognition area, divide the target recognition mesh template;

[0008] Obtain the standard remote sensing images corresponding to each target recognition grid template in the geographic information database for multiple time periods, as well as the environmental features of each standard remote sensing image;

[0009] For any target recognition grid template, based on the behavioral characteristics of several targets of interest within the template, the behavioral characteristics are fused with the standard remote sensing image and its environmental features to obtain template images of each target of interest.

[0010] A convolutional neural network model is constructed, and the model is trained and optimized based on the template image to obtain a target recognition model;

[0011] The target of interest is identified based on the target recognition model.

[0012] Furthermore, the process of dividing the target recognition mesh template based on the three-dimensional solid model of the target recognition region includes:

[0013] Based on the terrain distribution of the three-dimensional entity model, several common recognition areas and several key recognition areas for easily hidden targets are obtained.

[0014] Any of the aforementioned key identification regions is divided into several planar regions using a plane as the dividing standard;

[0015] A spatial coordinate system is established for each of the aforementioned ordinary recognition regions and each of the aforementioned planar regions, and each of the aforementioned regions and the corresponding spatial coordinate system constitutes the target recognition grid template.

[0016] Furthermore, the process of acquiring the environmental features includes:

[0017] Based on the image content of the standard remote sensing image, determine a plurality of first reference points and at least one second reference point contained in the standard remote sensing image; wherein, the entire first reference point is displayed in the standard remote sensing image, and a portion of the second reference point is displayed in the standard remote sensing image;

[0018] Obtain the spatial coordinates of multiple first reference points and second reference points in the corresponding target recognition grid template, and associate the spatial coordinates.

[0019] Furthermore, the process of obtaining the template image includes:

[0020] For any of the target recognition grid templates, several motion trajectories of each target within the template are planned based on the content distribution of the standard remote sensing image and the movement patterns of the target of interest.

[0021] For any motion trajectory, based on the first and second reference points of the target recognition grid template, select several reference trajectory points, and determine the spatial coordinates of each reference trajectory point in the target recognition grid template;

[0022] For any reference trajectory point, the target of interest and the corresponding standard remote sensing image of the target recognition grid template are fused according to the spatial coordinates of the reference trajectory point to obtain a template image.

[0023] Furthermore, the acquisition of the environmental features also includes:

[0024] The color features of the standard remote sensing image are determined based on the time period and image content corresponding to the standard remote sensing image.

[0025] Adjust the first reference point and the second reference point according to the color characteristics.

[0026] Furthermore, it also includes:

[0027] For any of the target recognition grid templates, a reference region is obtained based on the first and second reference points selected within it; wherein, the reference region is the largest closed region obtained by connecting the first and second reference points as boundary points in an unordered manner, and the area of ​​the closed region is not less than half of the area of ​​the target recognition grid template;

[0028] Real-time remote sensing images of the target recognition grid template are acquired sequentially according to a set periodicity.

[0029] When a non-interested target enters the area of ​​the target recognition grid template based on the real-time remote sensing image, and the non-interested target enters the reference area within a set time interval, and the non-interested target stays in the reference area for more than a set time threshold, then the non-interested target is changed to an interested target.

[0030] Secondly, this invention discloses a target identification system suitable for military management, comprising:

[0031] The partitioning module is used to partition the target recognition mesh template based on the 3D solid model of the target recognition region;

[0032] The acquisition module is used to acquire standard remote sensing images corresponding to each target identification grid template in multiple time periods in the geographic information database, as well as the environmental features of each standard remote sensing image.

[0033] The image fusion module is used to identify a grid template for any target, and based on the behavioral characteristics of several targets of interest within the template, fuse the behavioral characteristics with the standard remote sensing image and its environmental features to obtain template images of each target of interest.

[0034] The model training module is used to construct a convolutional neural network model and train and optimize the model based on the template image to obtain a target recognition model.

[0035] The identification module is used to identify the target of interest based on the target identification model.

[0036] Furthermore, the partitioning module, based on the three-dimensional solid model of the target recognition region, divides the execution unit of the target recognition mesh template, including:

[0037] The acquisition unit is used to acquire several common recognition areas and several key recognition areas of easily hidden targets based on the terrain distribution of the three-dimensional entity model.

[0038] A partitioning unit is used to divide any of the key identification regions into several planar regions using a plane as the partitioning standard.

[0039] The establishment unit is used to establish a spatial coordinate system for any of the general recognition regions and any of the planar regions respectively, and any of the regions and the corresponding spatial coordinate systems constitute the target recognition grid template.

[0040] Thirdly, the present invention discloses an electronic device including at least one processor coupled to a memory, the memory storing a program or instructions that run on the processor, the program or instructions being executed by the processor to implement the steps of the target identification method applicable to military management as described above.

[0041] Fourthly, the present invention proposes a readable storage medium having a program or instructions stored thereon, characterized in that, when the program or instructions are executed by a processor, they implement the steps of the target identification method applicable to military management as described above.

[0042] As can be seen from the above technical solutions, the technical solutions of the present invention have achieved the following beneficial effects:

[0043] This invention discloses a target identification method and system applicable to military management. The method includes: dividing the target identification grid template according to a three-dimensional entity model of the target identification area; acquiring standard remote sensing images corresponding to each target identification grid template at multiple time periods in a geographic information database, as well as the environmental features of each standard remote sensing image; for any target identification grid template, based on the behavioral characteristics of several targets of interest within the template, performing image fusion with the standard remote sensing image and its environmental features to obtain template images of each target of interest; constructing a convolutional neural network model, and training and optimizing the model based on the template images to obtain a target identification model; and identifying the targets of interest based on the target identification model. The technical solution of this invention involves obtaining training data for training the target identification model based on the terrain distribution, environmental features, and behavioral characteristics of the target interest area, and then performing target identification based on the model. The target identification model obtained by matching the identification area with the behavioral characteristics of the targets of interest not only enables automatic analysis of the target identification process but also has high accuracy, effectively providing decision-making reference for military management.

[0044] When acquiring environmental features, this invention first divides the target recognition area into grids to narrow down the specific target recognition range. Then, based on the behavioral characteristics of the target of interest, it predicts the behavior of the target and uses this as training data to train the model, thereby enabling high-accuracy target recognition.

[0045] It should be understood that all combinations of the foregoing concepts and the additional concepts described in more detail below can be considered part of the inventive subject matter of this disclosure, provided that such concepts do not contradict each other.

[0046] The foregoing and other aspects, embodiments, and features of the teachings of the present invention will be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the invention, such as features and / or beneficial effects of exemplary embodiments, will become apparent from the following description or may be learned through practice of specific embodiments according to the teachings of the present invention. Attached Figure Description

[0047] The accompanying drawings are not drawn to scale. In the drawings, each identical or nearly identical component shown in the various figures may be denoted by the same reference numeral. For clarity, not every component is labeled in each figure. Embodiments of various aspects of the invention will now be described by way of example and with reference to the accompanying drawings, wherein:

[0048] Figure 1 This is an overall flowchart of the target identification method for military management disclosed in the embodiments of this application;

[0049] Figure 2This is a flowchart of the target identification grid template disclosed in an embodiment of this application;

[0050] Figure 3 The environmental feature acquisition process disclosed in the embodiments of this application Figure 1 ;

[0051] Figure 4 This is a flowchart illustrating the template image acquisition process disclosed in an embodiment of this application.

[0052] Figure 5 The environmental feature acquisition process disclosed in the embodiments of this application Figure 2 ;

[0053] Figure 6 This is a flowchart illustrating the process of determining targets of interest as disclosed in an embodiment of this application;

[0054] Figure 7 This is a schematic diagram of an electronic device according to an embodiment of this application;

[0055] Figure 8 This is a framework diagram of a target identification system suitable for military management disclosed in an embodiment of this application. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art to which this invention pertains.

[0057] The terms "first," "second," and similar words used in the specification and claims of this patent application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, unless the context clearly indicates otherwise, the singular forms of "an," "a," or "the," etc., do not indicate a quantity limitation, but rather indicate the presence of at least one. Terms such as "comprising" or "including" mean that the element or object preceding "comprising" encompasses the features, integrals, steps, operations, elements, and / or components listed following "comprising" or "including," and do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; these relative positional relationships may change accordingly when the absolute position of the described object changes.

[0058] Military target identification technology plays a crucial role in command and decision-making in military management. However, the speed and accuracy of target identification limit its application. Target identification speed depends on the algorithm's ability to process massive amounts of data, while accuracy depends on its classification capabilities. This invention aims to propose a target identification method and system suitable for military management. First, environmental characteristics of the target identification area are fused with behavioral characteristics of the target of interest and used as training data. Then, a convolutional neural network model is used to train the target identification model, improving accuracy while maintaining processing speed.

[0059] The target identification method applicable to military management described in this embodiment will be specifically introduced below with reference to the accompanying drawings.

[0060] like Figure 1 As shown, the method includes:

[0061] Step S102: Divide the target recognition mesh template according to the three-dimensional solid model of the target recognition area;

[0062] The purpose of dividing the target recognition grid template based on the 3D solid model is to find the locations where the target of interest is likely to pass through. For example, for targets of interest that are easy to hide during operation, the 3D solid model can discover multiple areas that are easy to hide in. These areas can be identified by finely dividing the recognition area and focusing on them. In addition, for some inconveniently observed recessed areas, planar observation can also be carried out by finely dividing the recognition area. In other words, the purpose of dividing the target recognition grid template is to track the target of interest better and more accurately.

[0063] As a specific implementation method, such as Figure 2 As shown, the process of dividing the target recognition mesh template in step S102 specifically includes the following steps: Step S1022: Based on the terrain distribution of the three-dimensional entity model, obtain several ordinary recognition areas and several key recognition areas for easily hidden targets; Step S1024: Divide any one of the key recognition areas into several planar areas using a plane as the dividing standard; Step S1026: Establish a spatial coordinate system for any one of the ordinary recognition areas and any one of the planar areas respectively. The origin of the spatial coordinate system can be uniformly selected as the center point of the area. Any one of the areas and the corresponding spatial coordinate system constitute the target recognition mesh template.

[0064] For a target recognition grid template, the observed content is relatively independent. Because it has an independent spatial coordinate system, the coordinates and position can be independently represented and displayed when determining the target of interest.

[0065] As an optional implementation, each target identification grid template is displayed independently in a planar mode in the military management decision-making center, and the real-time status of each target identification grid template is displayed synchronously to assist decision-makers in observation.

[0066] Step S104: Obtain the standard remote sensing images corresponding to each target identification grid template in the geographic information database for multiple time periods, as well as the environmental features of each standard remote sensing image;

[0067] This solution aims to fully utilize the environmental features of the target recognition area, matching these features with the behavioral features of the target of interest, and using this data to train a neural network target recognition model, thereby reducing the error caused by the environment of the recognition area in target recognition. Specifically, the process of acquiring environmental features is implemented based on the steps S1022-S1026 described above, such as... Figure 3 As shown, the following steps are used to obtain the reference points: Step S1042, based on the image content of the standard remote sensing image, determine multiple first reference points and at least one second reference point contained in the standard remote sensing image; wherein, the entire first reference point is displayed in the standard remote sensing image, and a portion of the second reference point is displayed in the standard remote sensing image; Step S1044, obtain the spatial coordinates of the multiple first reference points and the second reference points in the corresponding target recognition grid template, and associate the spatial coordinates. The purpose of selecting reference points is to provide a reference for the trajectory of the target when the entire target recognition grid template is used as the area of ​​the target's movement, so that the trajectory can be planned according to the target's movement pattern. Optionally, reference points can be selected as prominent structures in the area, such as trees with significantly prominent shapes, rocks protruding from the ground, grassland, etc. Reference points can also be selected by performing image recognition on the standard remote sensing image and using the recognition results as reference points. Furthermore, in implementation, the spatial coordinates are associated by storing each spatial coordinate as a data group in the corresponding target recognition grid template, and highlighting each target recognition grid template independently in a planar mode in the decision center.

[0068] As another possible implementation method, such as Figure 5 As shown, the acquisition of the environmental features further includes the following steps: Step S1046: Determine the color features of the standard remote sensing image based on the time period and image content corresponding to the standard remote sensing image. Step S1048: Adjust the first reference point and the second reference point based on the color features. That is, as the season changes, the reference points are adjusted to some appearances that contradict the color features of that season. Of course, the reference points can also be set independently during exercises.

[0069] Step S106: For any target identification grid template, based on the behavioral characteristics of several targets of interest within the template, perform image fusion with the standard remote sensing image and its environmental features to obtain template images of each target of interest.

[0070] In this embodiment, image fusion involves predicting the motion trajectory of the target of interest. The predicted trajectory points are spatially hypothesized and fused with a standard remote sensing image and its environmental features to obtain a predicted image, i.e., a template image. Specifically, this fusion process is implemented based on steps S1042-S1044 described above, and the template image is as follows: Figure 4 As shown, the following steps are used to obtain the target recognition grid template: Step S1062, for any target recognition grid template, several motion trajectories of each target within the template are planned based on the content distribution of the standard remote sensing image and the movement patterns of the target of interest. Step S1064, for any motion trajectory, several reference trajectory points are selected based on the first and second reference points of the target recognition grid template, and the spatial coordinates of each reference trajectory point are determined in the target recognition grid template. Step S1066, for any reference trajectory point, the target of interest and the corresponding standard remote sensing image of the target recognition grid template are fused based on the spatial coordinates of the reference trajectory point to obtain a template image.

[0071] Step S108: Construct a convolutional neural network model and train and optimize the model based on the template images to obtain a target recognition model. During training, a number of template images are obtained based on the different targets of interest, the different target recognition grid templates they are in, and the different planned trajectories. These form a template image set, which is then divided into a training set and a validation set in a 7:3 ratio for model training. During model optimization, the recognition images corresponding to the real-time recognition results of the targets of interest are used as new training data to expand the template image set. The training set is updated in real time to optimize the target recognition model and improve the classification accuracy of the model.

[0072] Step S110: Identify the target of interest according to the target recognition model.

[0073] This invention applies a trained neural network model to a monitoring device, enabling the monitoring of targets of interest in each target identification grid template. Decision-makers can then make intelligent assisted decisions based on the identification results of the targets of interest.

[0074] As another possible implementation method, to ensure the accuracy of target identification, although targets of interest are given in advance, there are still some targets that are not in the target of interest list but exhibit behaviors that match those of the targets of interest. These targets also need to be given special attention; therefore, Figure 6As shown, the target recognition method disclosed in this embodiment of the invention further includes: step S1062', for any target recognition grid template, obtaining a reference region based on the selected first reference point and second reference point within it; wherein, the reference region is the largest closed region obtained by randomly connecting the first reference point and the second reference point as boundary points, and the area of ​​the closed region is not less than half of the area of ​​the target recognition grid template; step S1064', acquiring real-time remote sensing images of the target recognition grid template according to a set periodic sequence; step S1066', when a non-interested target is detected and identified entering the region of the target recognition grid template based on the real-time remote sensing image, and the non-interested target enters the reference region within a set time interval, and the non-interested target stays in the reference region for a time exceeding a set time threshold, then the non-interested target is changed to an interested target.

[0075] The target identification method proposed in this embodiment, applicable to military management, aims to address the shortcomings of existing military target identification methods, which are affected by the complex terrain and geomorphological environment where targets are hidden. These methods require the acquisition of a large number of images for comparison and calculation, resulting in high computational load and low identification accuracy when processing and analyzing such image data. The proposed method aims to meet the needs of military target identification while assisting military decision-making.

[0076] In embodiments of this application, an electronic device is also provided. This device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the target identification method for military management disclosed in the above embodiments. Taking an electronic device running on a computer as an example, such as... Figure 7 As shown, the electronic device may include one or more (only one is shown in the figure) processors (processors may include, but are not limited to, microprocessors such as MCUs or programmable logic devices such as FPGAs), a memory for storing data, and a transmission device for communication functions. Those skilled in the art will understand that... Figure 7 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device.

[0077] The aforementioned program can run in a processor or be stored in memory, i.e., in a computer-readable medium. Computer-readable media includes both permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media, such as modulated data signals and carrier waves.

[0078] These computer programs may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes can be implemented through different modules, corresponding to different method steps.

[0079] In this embodiment, such an apparatus or system is provided, which can be referred to as a target identification system suitable for military management. The system is as follows: Figure 8 As shown, the system includes: a segmentation module for segmenting target recognition grid templates based on a 3D entity model of the target recognition region; an acquisition module for acquiring standard remote sensing images corresponding to each target recognition grid template at multiple time periods in a geographic information database, as well as the environmental features of each standard remote sensing image; an image fusion module for fusing the behavioral features of several targets of interest within any target recognition grid template with the standard remote sensing images and their environmental features to obtain template images of each target of interest; a model training module for constructing a convolutional neural network model and training and optimizing the model based on the template images to obtain a target recognition model; and a recognition module for recognizing the targets of interest based on the target recognition model.

[0080] Since the system is used to implement the above method, the above-described features will not be repeated here.

[0081] For example, the execution unit of the division module for dividing the target recognition grid template according to the three-dimensional entity model of the target recognition area includes: an acquisition unit, used to acquire several ordinary recognition areas and several key recognition areas of easily hidden targets according to the terrain distribution of the three-dimensional entity model; a division unit, used to divide any one of the key recognition areas into several planar areas using a plane as the division standard; and an establishment unit, used to establish a spatial coordinate system for any one of the ordinary recognition areas and any one of the planar areas respectively, wherein any one of the areas and the corresponding spatial coordinate system constitute the target recognition grid template.

[0082] For example, in order to acquire the environmental features, the acquisition module includes: a first determining unit, configured to determine, based on the image content of the standard remote sensing image, a plurality of first reference points and at least one second reference point contained in the standard remote sensing image; wherein, the entire first reference point is displayed in the standard remote sensing image, and a portion of the second reference point is displayed in the standard remote sensing image; and an association module, configured to acquire the spatial coordinates of the plurality of first reference points and the second reference point in the corresponding target recognition grid template, and associate the spatial coordinates.

[0083] For example, to obtain the template image, the image fusion module includes: a planning unit, used to plan several motion trajectories of each target of interest within any target recognition grid template based on the content distribution of its standard remote sensing image and the movement patterns of the target of interest; a selection unit, used to select several reference trajectory points for any motion trajectory based on a first reference point and a second reference point of the target recognition grid template, and to determine the spatial coordinates of each reference trajectory point in the target recognition grid template; and a fusion unit, used to perform image fusion of the target of interest and the corresponding standard remote sensing image of the target recognition grid template based on the spatial coordinates of any reference trajectory point, thereby obtaining the template image.

[0084] For example, based on the specific settings of the image fusion module described above, the acquisition module further includes: a second determining unit, used to determine the color features of the standard remote sensing image according to the time period and image content corresponding to the standard remote sensing image; and an adjusting unit, used to adjust the first reference point and the second reference point according to the color features.

[0085] For example, the image fusion module further includes: a reference determination unit, used to obtain a reference region for any target recognition grid template based on the first and second reference points selected therein; wherein the reference region is the largest closed region obtained by randomly connecting the first and second reference points as boundary points, and the area of ​​the closed region is not less than half of the area of ​​the target recognition grid template; an image acquisition unit, used to acquire real-time remote sensing images of the target recognition grid template sequentially according to a set period; and a target change unit, used to change the non-interested target to an interested target when a non-interested target is detected and identified by the real-time remote sensing image entering the region of the target recognition grid template, the non-interested target enters the reference region within a set time interval, and the non-interested target stays in the reference region for more than a set time threshold.

[0086] The present invention discloses a target identification method and system applicable to military management. Its technical solution obtains training data for training a target identification model based on the terrain distribution, environmental characteristics, and behavioral characteristics of the target's area of ​​interest. Then, the target is identified based on the model. The training data is obtained by matching the environmental characteristics of the target identification area with the behavioral characteristics of the target. The trained target identification model can not only automatically analyze the target identification process, but also has high accuracy, and can effectively provide decision-making reference for military management.

[0087] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A target identification method suitable for military management, characterized in that, include: Based on the 3D solid model of the target recognition area, divide the target recognition mesh template; Obtain the standard remote sensing images corresponding to each target recognition grid template in the geographic information database for multiple time periods, as well as the environmental features of each standard remote sensing image; For any target recognition grid template, based on the behavioral characteristics of several targets of interest within the template, the behavioral characteristics are fused with the standard remote sensing image and its environmental features to obtain template images of each target of interest. A convolutional neural network model is constructed, and the model is trained and optimized based on the template image to obtain a target recognition model; The target of interest is identified based on the target recognition model. The process of acquiring the environmental features includes: Based on the image content of the standard remote sensing image, determine a plurality of first reference points and at least one second reference point contained in the standard remote sensing image; wherein, the entire first reference point is displayed in the standard remote sensing image, and a portion of the second reference point is displayed in the standard remote sensing image; Obtain the spatial coordinates of multiple first reference points and second reference points in the corresponding target recognition grid template, and associate the spatial coordinates; The process of obtaining the template image includes: For any of the target recognition grid templates, several motion trajectories of each target within the template are planned based on the content distribution of the standard remote sensing image and the movement patterns of the target of interest. For any motion trajectory, based on the first and second reference points of the target recognition grid template, select several reference trajectory points, and determine the spatial coordinates of each reference trajectory point in the target recognition grid template; For any reference trajectory point, the target of interest and the corresponding standard remote sensing image of the target recognition grid template are fused according to the spatial coordinates of the reference trajectory point to obtain a template image.

2. The target identification method for military management according to claim 1, characterized in that, The process of dividing the target recognition mesh template according to the three-dimensional solid model of the target recognition region includes: Based on the terrain distribution of the three-dimensional entity model, several common recognition areas and several key recognition areas for easily hidden targets are obtained. Any of the aforementioned key identification regions is divided into several planar regions using a plane as the dividing standard; A spatial coordinate system is established for each of the aforementioned ordinary recognition regions and each of the aforementioned planar regions, and each of the aforementioned regions and the corresponding spatial coordinate system constitutes the target recognition grid template.

3. The target identification method for military management according to claim 2, characterized in that, The acquisition of the environmental features also includes: The color features of the standard remote sensing image are determined based on the time period and image content corresponding to the standard remote sensing image. Adjust the first reference point and the second reference point according to the color characteristics.

4. The target identification method for military management according to claim 2, characterized in that, Also includes: For any of the target recognition grid templates, a reference region is obtained based on the first and second reference points selected within it; wherein, the reference region is the largest closed region obtained by connecting the first and second reference points as boundary points in an unordered manner, and the area of ​​the closed region is not less than half of the area of ​​the target recognition grid template; Real-time remote sensing images of the target recognition grid template are acquired sequentially according to a set periodicity. When a non-interested target enters the area of ​​the target recognition grid template based on the real-time remote sensing image, and the non-interested target enters the reference area within a set time interval, and the non-interested target stays in the reference area for more than a set time threshold, then the non-interested target is changed to an interested target.

5. A target identification system suitable for military management, characterized in that, include: The partitioning module is used to partition the target recognition mesh template based on the 3D solid model of the target recognition region; The acquisition module is used to acquire standard remote sensing images corresponding to each target recognition grid template in a geographic information database for multiple time periods, as well as the environmental features of each standard remote sensing image. The acquisition process of the environmental features includes: determining multiple first reference points and at least one second reference point contained in the standard remote sensing image based on the image content; wherein the entire first reference point is displayed in the standard remote sensing image, and a portion of the second reference point is displayed in the standard remote sensing image; acquiring the spatial coordinates of the multiple first reference points and the second reference points in the corresponding target recognition grid template, and associating the spatial coordinates. An image fusion module is used to perform image fusion on any target recognition grid template, based on the behavioral characteristics of several targets of interest within the template, by fusing the behavioral characteristics with the standard remote sensing image and its environmental features to obtain template images for each target of interest. The process of obtaining the template images includes: for any target recognition grid template, planning several motion trajectories for each target of interest within it based on the content distribution of its standard remote sensing image and the movement patterns of the targets of interest; for any motion trajectory, selecting several reference trajectory points based on the first and second reference points of the target recognition grid template, and determining the spatial coordinates of each reference trajectory point within the target recognition grid template; for any reference trajectory point, fusing the target of interest with the corresponding standard remote sensing image of the target recognition grid template based on the spatial coordinates of the reference trajectory point to obtain a template image. The model training module is used to construct a convolutional neural network model and train and optimize the model based on the template image to obtain a target recognition model. The identification module is used to identify the target of interest based on the target identification model.

6. The target identification system for military management according to claim 5, characterized in that, The partitioning module divides the target recognition mesh template into execution units based on the three-dimensional solid model of the target recognition region, including: The acquisition unit is used to acquire several common recognition areas and several key recognition areas of easily hidden targets based on the terrain distribution of the three-dimensional entity model. A partitioning unit is used to divide any of the key identification regions into several planar regions using a plane as the partitioning standard. The establishment unit is used to establish a spatial coordinate system for any of the general recognition regions and any of the planar regions respectively, and any of the regions and the corresponding spatial coordinate systems constitute the target recognition grid template.

7. An electronic device, characterized in that, It includes at least one processor coupled to a memory, the memory storing a program or instructions that run on the processor, the program or instructions which, when executed by the processor, implement the steps of the target identification method for military management as described in any one of claims 1 to 4.

8. A readable storage medium having a program or instructions stored thereon, characterized in that, When the program or instructions are executed by the processor, they implement the steps of the target identification method applicable to military management as described in any one of claims 1 to 4.

Images