Three-dimensional network type target reconnaissance method and system and application

Abstract

The invention belongs to the technical field of detection, and particularly relates to a three-dimensional network type target reconnaissance method and system and application, and the reconnaissance method comprises the steps that 1, a slave requesting reconnaissance sends a reconnaissance request packet to a host; 2, after receiving the reconnaissance request packet, the host determines a reconnaissance target according to a specific mode, and sends a content response packet to the slave requesting reconnaissance; step 3, the slave requesting reconnaissance carries out coarse scanning reconnaissance on the reconnaissance target according to the content response packet, and the coarse scanning reconnaissance content is returned to the host; step 4, the host analyzes and processes the content of the coarse scanning reconnaissance to determine whether a suspicious target exists, and if the suspicious target is found, the step 5 is executed; and step 5, the host sends a fine scanning indication packet to the slave requesting reconnaissance, the slave reconnaissance is carried out on the suspicious target according to the fine scanning indication packet, and the reconnaissance content is returned to the host to obtain detection result information. The reconnaissance capability and efficiency of the reconnaissance system are enhanced by applying the three-dimensional network type target reconnaissance method.

Description

Technical Field

[0001] The present invention belongs to the field of detection technology, and in particular relates to a three-dimensional network target reconnaissance method, system and application. Background Art

[0002] Aerial reconnaissance, characterized by its high timeliness and flexibility, is primarily used for aerial photography, surveying and mapping, intelligence reconnaissance, maritime, power, and oil patrols, field search and rescue, and disaster monitoring. For example, in intelligence reconnaissance systems, aerial reconnaissance overcomes the limitations of ground-based reconnaissance due to the curvature of the earth and terrain obstructions, while also compensating for the lack of detail and timeliness in satellite reconnaissance. It is currently an effective means of conducting reconnaissance. Certain conditions for aerial reconnaissance, such as approach direction, flight altitude, search and coverage method, photographic scale, and speed ratio, are all dependent on factors such as the mission requirements, the operating mode and capabilities of the reconnaissance equipment, the reconnaissance vehicle, the time of day, and environmental conditions.

[0003] Currently, aerial reconnaissance primarily relies on drones (UAVs) equipped with various reconnaissance equipment to detect and identify targets on the ground and surface. However, the small size of individual drones makes it difficult to carry high-powered reconnaissance equipment, limiting their reconnaissance range. Furthermore, damage to a drone can render it incapable of completing its mission. Therefore, there is an urgent need for a reconnaissance method to achieve effective aerial reconnaissance.

[0004] In view of this, the present invention is proposed. Summary of the Invention

[0005] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art and to provide a three-dimensional network target reconnaissance method, system and application.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In one aspect, the present invention provides a three-dimensional network target reconnaissance method, comprising the following steps:

[0008] Step 1: The slave device requesting reconnaissance sends a reconnaissance request packet to the master device;

[0009] Step 2: After receiving the reconnaissance request packet, the host determines the reconnaissance target according to a specific method and sends a content response packet to the slave requesting reconnaissance. The content response packet includes the name, height, area, resolution and address of the reconnaissance target requested;

[0010] Step 3: The slave device requesting reconnaissance performs a rough scan of the reconnaissance target determined in step 2 according to the content of the content response packet, and returns the content of the rough scan to the master device;

[0011] Step 4: The host analyzes and processes the content of the rough scan reconnaissance to determine whether there is a suspicious target. If no suspicious target is found in the content of the rough scan reconnaissance, the host obtains the detection result information of the reconnaissance target. Otherwise, step 5 is executed.

[0012] Step 5: The host sends a fine scan instruction packet to the slave requesting reconnaissance. The slave performs fine scan reconnaissance on the suspicious target according to the content of the fine scan instruction packet and returns the content of the fine scan reconnaissance to the host. The host obtains the detection result information of the suspicious target after analysis.

[0013] Furthermore, in step 1, the reconnaissance request packet includes the name, height, and area of ​​the requested reconnaissance.

[0014] Furthermore, in step 2, the host stores a correspondence table of all ranges in the reconnaissance area and their coordinate positions; the process of determining the reconnaissance target is: the host checks whether the content in the reconnaissance request packet is saved in the correspondence table; if so, all geographical ranges that meet the reconnaissance request packet are determined in the reconnaissance area, and the geographical range closest to the host is determined as the reconnaissance target; otherwise, the reconnaissance request packet is discarded.

[0015] Furthermore, in step 4, the content of the rough scan reconnaissance is analyzed and processed as follows:

[0016] Step 4.1, obtaining multiple images containing all objects within the reconnaissance target from the rough scan reconnaissance content;

[0017] Step 4.2: Use the multiple images in step 4.1 as input to the object recognition model, obtain an object recognition result through the object recognition model, and then determine whether the object is a suspicious target based on the object recognition result.

[0018] Furthermore, step 5 includes the following process:

[0019] Step 5.1: The host determines the coordinates and orientation of the suspicious target from the correspondence table and constructs a fine-scan instruction package, wherein the fine-scan instruction package includes the coordinates and orientation of the suspicious target and the reconnaissance resolution;

[0020] Step 5.2: After receiving the fine scan instruction packet sent by the host, the slave performs fine scan reconnaissance on the suspicious target according to the content of the fine scan instruction packet, obtains multiple suspicious target images, and returns the multiple suspicious target images to the host;

[0021] Step 5.3: The host uses the multiple suspicious target images as input to the object recognition model, and determines the object recognition result through the object recognition model, that is, obtains the detection result information of the suspicious target.

[0022] In another aspect, the present invention provides a system for applying the above-mentioned three-dimensional network target reconnaissance method, comprising a plurality of reconnaissance devices distributed over a reconnaissance area, wherein information is transmitted between the plurality of reconnaissance devices via radio communication;

[0023] Each of the reconnaissance devices includes a host, and any two of the hosts are connected to each other.

[0024] Furthermore, each of the reconnaissance devices further includes at least four slaves, which are circumferentially distributed around the host, and the slaves transmit information to the corresponding host via radio communication.

[0025] On the other hand, the present invention provides an application based on the three-dimensional network target reconnaissance system as described above, wherein the three-dimensional network target reconnaissance system is used for intelligence reconnaissance, field search and rescue, or disaster monitoring.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] 1. The three-dimensional networked target reconnaissance system comprises multiple reconnaissance units, distributing various reconnaissance equipment previously integrated on a single drone across multiple units, reducing the number of devices required per unit. Furthermore, once multiple units have reached a designated airspace, they can cover a large area through signal transmission. Over the designated area, these units continuously exchange information in real time, creating a "target detection network" and enhancing regional reconnaissance capabilities. This ensures that even if some units are damaged, sufficient units remain available to conduct reconnaissance missions. Furthermore, thanks to the "target detection network," damaged units can relay their information to intact units, minimizing the impact on the reconnaissance mission and ensuring continued target reconnaissance.

[0028] 2. Multiple slaves simultaneously conduct rough reconnaissance under the guidance of the master, expanding the scan area by at least (1 + the number of slaves) times, effectively increasing the reconnaissance range. Secondly, through short-term rough reconnaissance, the master can quickly obtain preliminary image information of the reconnaissance area and quickly locate the area where suspicious targets are located through processing and analysis. Subsequently, when the slaves conduct fine reconnaissance, they can confirm the detection results of suspicious targets in a relatively short time, greatly improving the timeliness of reconnaissance. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The accompanying drawings are incorporated in and constitute a part of this specification and, together with the description, serve to explain the principles of the invention.

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0031] Figure 1 This is a flow chart of the three-dimensional network target reconnaissance method of the present invention;

[0032] Figure 2 A schematic structural diagram of a three-dimensional network target reconnaissance system provided by an embodiment of the present invention;

[0033] Figure 3 A schematic structural diagram of a reconnaissance device provided in an embodiment of the present invention.

[0034] Among them: 1 is the reconnaissance device; 11 is the host; 12 is the slave. DETAILED DESCRIPTION

[0035] Here, exemplary embodiments will be described in detail, and the embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are only examples consistent with some aspects of the present invention described in detail in the appended claims.

[0036] On the one hand, see Figures 2 and 3 This embodiment provides a three-dimensional network target reconnaissance system, including multiple reconnaissance devices 1 distributed in the reconnaissance area, the multiple reconnaissance devices 1 can completely cover the reconnaissance area, and the multiple reconnaissance devices 1 transmit information through radio communication to form a "target detection network"; each of the reconnaissance devices 1 includes a host 11, and any two hosts 11 are connected to each other.

[0037] Furthermore, each of the reconnaissance devices 1 further includes at least four slave machines 12 , which are circumferentially distributed around the host machine 11 , and the slave machines 12 transmit information to the corresponding host machine 11 via radio communication.

[0038] In this embodiment, the slave device 12 has reconnaissance detection and image / video acquisition capabilities, and the master device 11 has image processing capabilities. The master device 11 stores a correspondence table of all detectable ranges and their coordinate positions within the reconnaissance area. This correspondence table is regularly updated and maintained to ensure that the information is up to date.

[0039] Specifically, in this embodiment, the slaves 12 are drones, and a drone swarm is formed by four slaves 12 to perform detection, locking, and tracking tasks. The host 11 is a manned aircraft, and the display panel of the host 11 is manually operated to perform tasks such as image processing and information instruction transmission.

[0040] On the other hand, see Figures 1 to 3 , this embodiment provides a three-dimensional network target reconnaissance method, such as Figure 3 As shown, the reconnaissance device 1 includes drones T1, T2, T3, and T4. It is assumed that drone T1 needs to request a reconnaissance named Search1.

[0041] Step 1: The drone T1 requesting reconnaissance sends a reconnaissance request packet to the host 11, wherein the reconnaissance request packet includes at least the following information: the name of the reconnaissance request, the altitude of the reconnaissance request, and the area of ​​the reconnaissance request.

[0042] Step 2: After receiving the reconnaissance request packet, the host 11 determines the reconnaissance target according to a specific method and sends a content response packet to the drone T1 requesting reconnaissance. The content response packet includes the name of the reconnaissance request, the altitude requested for reconnaissance, the area requested for reconnaissance, the resolution requested for reconnaissance, and the address of the reconnaissance target;

[0043] Among them, the host 11 stores a correspondence table of all ranges and their coordinate positions in the reconnaissance area, and the correspondence table is regularly updated and maintained; the host 11 checks whether the content in the reconnaissance request packet (the reconnaissance height and the reconnaissance area) is saved in the correspondence table. If saved, all geographical ranges that meet the reconnaissance request package are determined in the reconnaissance area, and the geographical range closest to the host 11 is defined as the reconnaissance target (the address of the reconnaissance target is the coordinate position address of the reconnaissance target). The name of the reconnaissance request, the height of the reconnaissance request, the area of ​​the reconnaissance request, the resolution of the reconnaissance request, the address of the reconnaissance target and other information are further used to construct a content response packet, and finally the content response packet is returned to the drone T1 requesting reconnaissance, and the operation ends.

[0044] Step 3: The drone T1 requesting reconnaissance performs a rough scan of the reconnaissance target determined in step 2 according to the content of the content response packet, and returns the content of the rough scan to the host 11;

[0045] Drone T1 has both reconnaissance and detection, and image / video acquisition capabilities. Upon receiving the content response packet from host 11, drone T1 simultaneously activates both the reconnaissance and detection functions and the image / video acquisition functions. Drone T1 then locks onto its detection range and acquires images within the range according to a set sampling period, capturing multiple images of the reconnaissance target. Finally, drone T1 returns these images to host 11.

[0046] Specifically, in step 3, when performing image acquisition, the ground resolution of the scan is 1 to 2 meters. If the drone T1 captures a video stream, the multiple images of the reconnaissance target can refer to multiple frames of images contained in the video stream, which is not specifically limited here.

[0047] Step 4: The host 11 analyzes and processes the content of the rough scan reconnaissance to determine whether there is a suspicious target. If no suspicious target is found in the content of the rough scan reconnaissance, the host 11 obtains the detection result information of the reconnaissance target. Otherwise, step 5 is executed.

[0048] In step 4, the analysis and processing of the rough scan reconnaissance content is as follows:

[0049] Step 4.1, obtaining multiple images containing all objects within the reconnaissance target from the rough scan reconnaissance content;

[0050] Step 4.2: Use the multiple images in step 4.1 as input to the object recognition model, obtain an object recognition result through the object recognition model, and then determine whether the object is a suspicious target based on the object recognition result.

[0051] Specifically, the object recognition model is obtained by pre-training with a large number of training samples. The training samples may be object images, and the object images may be obtained by collecting images of various objects that may appear within the reconnaissance range. For each object image, the object information contained in the object image may be calibrated. Afterwards, the object image and the calibration information corresponding to the object image may be used as inputs to the initial model, and the initial model may be trained to obtain a trained object recognition model. When performing object recognition, the host 11 may use multiple images as inputs to the object recognition model, and obtain object recognition results through the object recognition model. It should be noted that the object recognition model and the process of obtaining object recognition results through the model are prior art (such as an example provided in the Chinese patent publication number CN114429558A), which will not be described in detail here.

[0052] In addition, based on the object identification results, whether the object is a suspicious target is determined. The specific suspicious target is determined based on the actual task to be performed and is not limited here. For example, in disaster monitoring, the suspicious target can be located as a flame or a fire.

[0053] Step 5: The host 11 sends a fine-scan instruction packet to the drone T1 requesting reconnaissance. The drone T1 scouts the suspicious target according to the content of the fine-scan instruction packet and returns the reconnaissance content to the host 11. The host 11 obtains the detection result information of the suspicious target after analysis.

[0054] Furthermore, step 5 includes the following process:

[0055] Step 5.1: The host 11 determines the coordinates and orientation of the suspicious target from the corresponding relationship table, and constructs a precise scanning instruction packet using the coordinates and orientation of the suspicious target and the reconnaissance resolution;

[0056] Specifically, the ground resolution of reconnaissance can be 0.6m;

[0057] Step 5.2: After receiving the fine-scan instruction packet sent by the host 11, the drone T1 performs fine-scan reconnaissance on the suspicious target according to the content of the fine-scan instruction packet, obtains multiple suspicious target images, and returns the multiple suspicious target images to the host 11;

[0058] In step 5.3, the host 11 uses the multiple suspicious target images as input to the object recognition model and determines the detection result information of the suspicious target through the object recognition model. The implementation process here is the same as that of step 4.

[0059] In an embodiment of the present invention, the host 11 can instruct the drone T1 to perform a rough scan reconnaissance by sending the content of the content response packet to obtain multiple images of the reconnaissance target, and determine the object recognition results based on the multiple images and the object recognition model, thereby judging whether there is a suspicious target in the reconnaissance target. Once a suspicious target is found in the reconnaissance target, the host 11 will send a fine scan instruction packet to instruct the drone T1 to perform a fine scan reconnaissance to determine the accurate recognition result of the suspicious target. Since the resolution of the rough scan reconnaissance is low, it takes less time, so the host 11 can quickly receive the content of the rough scan reconnaissance. The host 11 identifies suspicious targets based on the content of the rough scan reconnaissance, improves the recognition efficiency, and narrows the existence range of suspicious targets, thereby narrowing the scope of subsequent fine scan reconnaissance, thereby shortening the time of fine scan reconnaissance, and greatly improving the timeliness of reconnaissance and identification.

[0060] On the other hand, this embodiment provides an application of the three-dimensional network target reconnaissance system as described above, and the three-dimensional network target reconnaissance system is used in the fields of intelligence reconnaissance, field search and rescue, or disaster monitoring.

[0061] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention.

[0062] It should be understood that the present invention is not limited to the above description and that various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A three-dimensional network target reconnaissance method, characterized in that: The following steps are involved: Step 1: The slave (12) requesting reconnaissance sends a reconnaissance request packet to the master (11); Step 2: After receiving the reconnaissance request packet, the host (11) determines the reconnaissance target according to a specific method and sends a content response packet to the slave (12) requesting reconnaissance, wherein the content response packet includes the name, height, area, resolution and address of the reconnaissance target requested; Step 3: The slave (12) requesting reconnaissance performs a rough reconnaissance of the reconnaissance target determined in step 2 according to the content of the content response packet, and returns the content of the rough reconnaissance to the master (11); Step 4: The host (11) analyzes and processes the content of the rough scan reconnaissance to determine whether there is a suspicious target. If no suspicious target is found in the content of the rough scan reconnaissance, the host (11) obtains the detection result information of the reconnaissance target, otherwise, step 5 is executed; Step 5: The host (11) sends a fine scan indication packet to the slave (12) requesting reconnaissance. The slave (12) performs fine scan reconnaissance on the suspicious target according to the content of the fine scan indication packet and returns the content of the fine scan reconnaissance to the host (11). The host (11) obtains the detection result information of the suspicious target after analysis.

2. The three-dimensional network target reconnaissance method according to claim 1, characterized in that: In step 1, the reconnaissance request packet includes the name, height, and area of ​​the requested reconnaissance.

3. The three-dimensional network target reconnaissance method according to claim 1, characterized in that: In step 2, the host (11) stores a correspondence table of all ranges in the reconnaissance area and their coordinate positions; the process of determining the reconnaissance target is as follows: the host (11) checks whether the content in the reconnaissance request packet is saved in the correspondence table; if so, all geographical ranges that meet the reconnaissance request packet are determined in the reconnaissance area, and the geographical range closest to the host (11) is determined as the reconnaissance target; otherwise, the reconnaissance request packet is discarded.

4. The three-dimensional network target reconnaissance method according to claim 1, characterized in that: In step 4, the analysis and processing of the rough scan reconnaissance content is as follows: Step 4.1, obtaining multiple images containing all objects within the reconnaissance target from the rough scan reconnaissance content; Step 4.2: Use the multiple images in step 4.1 as input to the object recognition model, obtain an object recognition result through the object recognition model, and then determine whether the object is a suspicious target based on the object recognition result.

5. The three-dimensional network target reconnaissance method according to claim 1, characterized in that: The step 5 includes the following process: Step 5.1, the host (11) determines the coordinate position of the suspicious target from the correspondence table, and constructs a fine scanning instruction package, wherein the fine scanning instruction package includes the coordinate position and reconnaissance resolution of the suspicious target; Step 5.2: After receiving the fine scan instruction packet sent by the host (11), the slave (12) performs fine scan reconnaissance on the suspicious target according to the content of the fine scan instruction packet, obtains multiple suspicious target images, and returns the multiple suspicious target images to the host (11); Step 5.3: The host (11) uses the multiple suspicious target images as input to the object recognition model, and determines the object recognition result through the object recognition model, that is, obtains the detection result information of the suspicious target.

6. A system using the three-dimensional network target reconnaissance method according to any one of claims 1 to 5, characterized in that: The invention comprises a plurality of reconnaissance devices (1) distributed in a reconnaissance area, wherein information is transmitted between the plurality of reconnaissance devices (1) via radio communication; Each of the reconnaissance devices (1) comprises a host computer (11), and any two of the host computers (11) are interconnected.

7. The three-dimensional network target reconnaissance system according to claim 6, characterized in that: Each of the reconnaissance devices (1) further comprises at least four slave machines (12), wherein the at least four slave machines (12) are circumferentially distributed around the host machine (11), and the slave machines (12) transmit information to the corresponding host machine (11) via radio communication.

8. An application of the three-dimensional network target reconnaissance system according to claim 6 or 7, characterized in that: The three-dimensional network target reconnaissance system is used for intelligence reconnaissance, field search and rescue or disaster monitoring.

Images