CN-116844153-B - Target identification method and system suitable for military management

CN116844153BCN 116844153 BCN116844153 BCN 116844153BCN-116844153-B

Abstract

The invention provides a target identification method and a target identification system suitable for military management, and relates to the field of automatic target identification; the method comprises the steps of dividing a target identification grid template according to a three-dimensional entity model of a target identification area, obtaining standard remote sensing images and environment features of the standard remote sensing images, corresponding to the target identification grid template in a geographic information base, of a plurality of time periods, respectively, conducting image fusion on the behavior features of any target identification grid template and the standard remote sensing images according to the behavior features of the target of interest to obtain template images of the target of interest, constructing a convolutional neural network model, conducting model training and optimization according to the template images to obtain a target identification model, and conducting identification of the target of interest according to the target identification model. According to the invention, the target recognition area is subjected to grid division, and the behavior of the target is expected to be used as training data for model training according to the behavior characteristics of the target, so that the high-accuracy recognition of the target is realized.

Inventors

XU HAN
SHEN YUE
JI WEI
CHEN QI
Zuo Xunxun

Assignees

西安速度时空大数据科技有限公司

Dates

Publication Date: 20260508
Application Date: 20230625

Claims (8)

1. A target identification method suitable for military management, comprising: dividing a target recognition grid template according to the three-dimensional entity model of the target recognition area; Obtaining standard remote sensing images corresponding to a plurality of time periods of each target identification grid template in a geographic information base respectively and environmental characteristics of each standard remote sensing image; carrying out image fusion on any target identification grid template according to the behavior characteristics of a plurality of targets of interest in the template, and respectively carrying out image fusion on the behavior characteristics, the standard remote sensing image and the environment characteristics thereof to obtain template images of the targets of interest; constructing a convolutional neural network model, and carrying out model training and optimization according to the template image to obtain a target recognition model; Identifying the target of interest according to the target identification model; The environmental characteristic acquisition process comprises the following steps: Determining a plurality of first reference points and at least one second reference point contained in the standard remote sensing image according to the image content of the standard remote sensing image, wherein the whole of the first reference points is displayed in the standard remote sensing image, and part of the second reference points is displayed in the standard remote sensing image; acquiring space coordinates of a plurality of first reference points and second reference points in the corresponding target identification grid templates, and correlating the space coordinates; the template image acquisition process comprises the following steps: Planning a plurality of motion tracks of each target of interest in any target identification grid template according to the content distribution of the standard remote sensing image and the action rule of the target of interest; selecting a plurality of reference track points for any motion track according to a first reference point and a second reference point of the target identification grid template, and determining the space coordinates of each reference track point in the target identification grid template; And carrying out image fusion on any reference track point and the standard remote sensing image of the target of interest and the corresponding target identification grid template according to the space coordinates of the reference track point, so as to obtain a template image.
2. The method for target recognition applicable to military management according to claim 1, wherein the process of dividing the target recognition grid template according to the three-dimensional solid model of the target recognition area comprises: Acquiring a plurality of common identification areas and a plurality of key identification areas of easy-to-conceal targets according to the relief distribution of the three-dimensional entity model; Dividing any key identification area into a plurality of plane areas by taking a plane as a division standard; And respectively establishing a space coordinate system for any common identification area and any plane area, wherein any area and the corresponding space coordinate system form the target identification grid template.
3. The method for target identification for military management according to claim 2, wherein the obtaining of the environmental characteristic further comprises: determining the color characteristics of the standard remote sensing image according to the time period and the image content corresponding to the standard remote sensing image; and adjusting the first reference point and the second reference point according to the color characteristics.
4. The method for target identification for military management according to claim 2, further comprising: Obtaining a reference area according to the first reference point and the second reference point selected in any target identification grid template, wherein the reference area is a closed area with the largest area, which is obtained by unordered connection of the first reference point and the second reference point serving as boundary points, and the area of the closed area is not less than half of the area of the target identification grid template; Acquiring real-time remote sensing images of the target identification grid template according to a set period sequence; and when a non-target enters the region of the target recognition grid template according to the real-time remote sensing image detection and recognition, the non-target enters the reference region within a set time interval, and the stay time of the non-target in the reference region exceeds a set time threshold, changing the non-target as the target of interest.
5. A target recognition system adapted for military management, comprising: The dividing module is used for dividing the target recognition grid template according to the three-dimensional entity model of the target recognition area; The acquisition module is used for acquiring standard remote sensing images corresponding to a plurality of time periods of each target identification grid template in a geographic information base and environment characteristics of each standard remote sensing image, wherein the acquisition process of the environment characteristics comprises the steps of determining a plurality of first reference points and at least one second reference point contained in the standard remote sensing image according to image contents of the standard remote sensing images, wherein the whole of the first reference points is displayed in the standard remote sensing images, and part of the second reference points is displayed in the standard remote sensing images; The image fusion module is used for carrying out image fusion on the behavior characteristics of a plurality of targets in any target identification grid template and the standard remote sensing images and the environment characteristics thereof respectively to obtain template images of the targets in attention, wherein the acquisition process of the template images comprises the steps of planning a plurality of motion tracks of the targets in attention according to the content distribution of the standard remote sensing images and the action rules of the targets in attention for any target identification grid template, selecting a plurality of reference track points for any motion track according to the first reference point and the second reference point of the target identification grid template, determining the space coordinates of the reference track points in the target identification grid template, carrying out image fusion on the targets in attention and the standard remote sensing images of the corresponding target identification grid template according to the space coordinates of the reference track points, and further obtaining a template image; the model training module is used for constructing a convolutional neural network model, and carrying out model training and optimization according to the template image to obtain a target recognition model; and the identification module is used for identifying the target of interest according to the target identification model.
6. The target recognition system for military management of claim 5, wherein said partitioning module partitions the execution unit of the target recognition grid template according to the three-dimensional solid model of the target recognition area, comprising: The acquisition unit is used for acquiring a plurality of common identification areas and a plurality of key identification areas of the easily hidden targets according to the geographic distribution of the three-dimensional entity model; The dividing unit is used for dividing any key identification area into a plurality of plane areas by taking a plane as a dividing standard; The establishing unit is used for respectively establishing a space coordinate system for any common identification area and any plane area, and any area and the corresponding space coordinate system form the target identification grid template.
7. An electronic device comprising at least one processor coupled to a memory storing a program or instructions that, when executed by the processor, implement the steps of the method for target identification for military management of any of claims 1 to 4.
8. A readable storage medium having stored thereon a program or instructions which when executed by a processor realizes the steps of the object identification method for military management according to any of claims 1 to 4.

Description

Target identification method and system suitable for military management Technical Field The invention relates to the technical field of automatic target identification, in particular to a target identification method and system suitable for military management. Background The military target recognition technology is used for accurately recognizing and judging targets on the ground, on the sea or in the air through specific technical means, providing reliable basis for command decision and accurate guidance and guarantee for military operations, for example, in war exercises, army can determine information such as forces, equipment and the like of the other side by recognizing hostile targets, in anti-terrorist operations, police can track and strike terrorists by recognizing suspected targets, in marine patrol, sea police can strike pirates, illegal fishing boats and the like by recognizing suspicious ships. In the process of military target identification, the acquisition of target information can be achieved through various channels, and images containing military target information are conventionally acquired through platforms such as vehicle-mounted, airborne, satellite, unmanned aerial vehicle and the like. When the target is hidden in the environment of complex terrains and landforms, a large number of pictures need to be acquired for comparison calculation, so that the image data information is processed and analyzed, the calculated amount is large, and the recognition accuracy is low. In military operations, decision making needs to be quick and accurate, so that a method capable of intelligently identifying and judging targets is needed, mass data can be quickly and automatically analyzed, data accuracy can be guaranteed, and important basis is provided for military patents. Disclosure of Invention The invention aims to provide a target recognition method and a target recognition system suitable for military management, which are characterized in that a machine learning mode is adopted to carry out matching training on environmental characteristics of a target recognition area and interesting target behavior characteristics to obtain a target recognition model, and finally the model is adopted to carry out target recognition. In order to achieve the above purpose, the present invention proposes the following technical scheme: In a first aspect, the present invention provides a target recognition method suitable for military management, including: dividing a target recognition grid template according to the three-dimensional entity model of the target recognition area; Obtaining standard remote sensing images corresponding to a plurality of time periods of each target identification grid template in a geographic information base respectively and environmental characteristics of each standard remote sensing image; carrying out image fusion on any target identification grid template according to the behavior characteristics of a plurality of targets of interest in the template, and respectively carrying out image fusion on the behavior characteristics, the standard remote sensing image and the environment characteristics thereof to obtain template images of the targets of interest; constructing a convolutional neural network model, and carrying out model training and optimization according to the template image to obtain a target recognition model; and identifying the target of interest according to the target identification model. Further, the process of dividing the target recognition grid template according to the three-dimensional entity model of the target recognition area includes: Acquiring a plurality of common identification areas and a plurality of key identification areas of easy-to-conceal targets according to the relief distribution of the three-dimensional entity model; Dividing any key identification area into a plurality of plane areas by taking a plane as a division standard; And respectively establishing a space coordinate system for any common identification area and any plane area, wherein any area and the corresponding space coordinate system form the target identification grid template. Further, the process for obtaining the environmental characteristic includes: Determining a plurality of first reference points and at least one second reference point contained in the standard remote sensing image according to the image content of the standard remote sensing image, wherein the whole of the first reference points is displayed in the standard remote sensing image, and part of the second reference points is displayed in the standard remote sensing image; And acquiring space coordinates of the plurality of first reference points and the plurality of second reference points in the corresponding target recognition grid templates, and correlating the space coordinates. Further, the process of obtaining the template image includes: Planning a plurality of motion tracks of each target o