An unmanned reconnaissance system for confined and narrow spaces

Abstract

The present invention belongs to the field of communications technology and specifically discloses an unmanned reconnaissance system for confined and narrow spaces. The system comprises a control terminal, a 5G base station, and an unmanned vehicle. The unmanned vehicle is equipped with a communication module, an image acquisition module, an encryption module, and a data packet construction module. The 5G base station serves as a data transfer and processing station between the control terminal and the unmanned vehicle, decrypting, identifying, and storing data packet structures acquired via a data channel before sending them to the control terminal. The control terminal then sends instructions to the unmanned vehicle based on the identified data packet structures to complete reconnaissance. By establishing dedicated communication channels and constructing different data packet structures, the present invention enhances data processing accuracy and meets operational requirements in specific environments.

Description

Technical Field

[0001] The present invention belongs to the field of communication technology, and in particular relates to an unmanned reconnaissance system for a confined and narrow space. Background Art

[0002] During field operations, it is inevitable to encounter special closed and narrow environments such as tunnels, pits, and electronic shielding areas. When it is necessary to conduct reconnaissance in these areas, due to the lack of understanding of the internal terrain of the area and the possible inability to communicate with the outside world, rash entry may pose a safety hazard.

[0003] Therefore, providing an unmanned reconnaissance system for use in confined and narrow spaces is a technical problem that needs to be urgently solved by those skilled in the art. Summary of the Invention

[0004] The purpose of the present invention is to overcome the defects of the prior art that the internal conditions of a special environment cannot be obtained and that the system cannot communicate with the outside, and to provide an unmanned reconnaissance system and method for a confined and narrow space.

[0005] In a first aspect, the present invention provides an unmanned reconnaissance system for confined and narrow spaces, comprising a control terminal, a 5G base station, and an unmanned vehicle, wherein the unmanned vehicle is provided with a communication module, an image acquisition module, an encryption module, a data packet construction module, and a 3D model construction module; the communication module is connected to the 5G base station via a leaky cable wired communication connection and a 5G wireless communication connection, forming a wired channel and a wireless channel; the image acquisition module is used to collect data of the area to be reconnaissanced; the data packet construction module is used to construct the collected data into a data packet structure; the encryption module is used to encrypt the data packet structure; and the 3D model construction module constructs a three-dimensional model based on the collected data;

[0006] The wired channel and the wireless channel respectively establish MIMO channel models, wherein the MIMO channel models include at least three channels, namely a data channel, a control channel and an RTS signal channel, and the three wireless transmission channels all use 4.9 GHz and 700 MHz dual-frequency electromagnetic waves;

[0007] The 5G base station acts as a data transfer processing station between the control end and the unmanned vehicle, decrypting, identifying, and storing the data packet structure obtained through the data channel before sending it to the control end;

[0008] The control end sends instructions to the unmanned vehicle based on the identified data packet structure to complete the reconnaissance.

[0009] A further solution is that the data packet structure includes a data number, a data type, a collection time and a priority;

[0010] The collection time is the number of seconds from 00:00:00 on January 1, 1970 to the collection time.

[0011] A further solution is that the encrypted data packet structure is decrypted through the IP of the control end; the identification of the data packet structure includes but is not limited to extracting features from the input data through a feature extraction model to obtain corresponding feature data, and then classifying the feature data through a classification model to obtain the corresponding category; the storage of the data packet structure includes sorting the identified data packet structure based on the weight of priority and acquisition time and storing it on the control end.

[0012] A further solution is that the category includes at least video data and audio data, and the unmanned vehicle further includes a data processing module, and the data module has built-in computer vision algorithms, speech recognition algorithms, and natural language processing algorithms;

[0013] Processing video information through computer vision algorithms: Identifying and tracking reconnaissance targets based on target detection and tracking algorithms. Counting and statistical analysis are performed based on the recognition results to calculate a modeling subset for 3D modeling. The modeling subset is a set of valid images in each frame of the video data.

[0014] Processing audio data through speech recognition algorithms: Converting audio to text based on speech recognition algorithms and simultaneous interpretation functions; Recognizing people's voiceprints based on voiceprint recognition algorithms and combining them with video face recognition algorithms to confirm the speaker's identity;

[0015] The text information is deduplicated, normalized, and disambiguated through natural language processing algorithms: the text data information obtained based on speech recognition is translated into a recognizable language using machine translation, and the summary information of the text is extracted based on the summary extraction algorithm; the main keywords in the text are extracted based on the named entity recognition algorithm.

[0016] A further solution is that the 5G base station sends instructions to the unmanned vehicle based on the stored data packet structure recognition results. Every time the unmanned vehicle executes an instruction, the control end deletes the data packet structure corresponding to the instruction.

[0017] A second aspect of the present invention provides an unmanned reconnaissance method for confined and narrow spaces, which utilizes the aforementioned reconnaissance system, enters the area to be reconnaissanced via an unmanned vehicle, and establishes communication with a control terminal. The method comprises the following steps:

[0018] Step 1: Establishment of communication link;

[0019] Connect unmanned vehicles to 5G base stations using a combination of wired and wireless transmission;

[0020] When the unmanned vehicle enters the area to be reconnaissanced, it immediately exchanges signaling with the 5G base station and establishes MIMO channel models for wired and wireless transmission for communication between the unmanned vehicle and the 5G base station. The MIMO channel model includes at least three channels: a data channel, a control channel, and an RTS signal channel. The three wireless transmission channels all use dual-frequency electromagnetic waves of 4.9 GHz and 700 MHz.

[0021] The unmanned vehicle obtains data on the area to be surveyed, and transmits it to the 5G base station via wired or wireless transmission in the form of an encrypted data packet structure according to the RTS / CTS protocol;

[0022] Step 2: Obtain the data packet structure of the area to be surveyed and perform target identification in real time;

[0023] The 5G base station obtains the data packet structure through the data channel, decrypts, identifies and stores the data packet structure, and then sends it to the control end;

[0024] Step 3: The control terminal remotely controls the unmanned vehicle;

[0025] The control end controls the unmanned vehicle based on the identified data packet structure.

[0026] A further solution is that the data packet structure includes a data number, a data type, a collection time and a priority;

[0027] The collection time is the number of seconds from 00:00:00 on January 1, 1970 to the collection time.

[0028] A further solution is that in step 2, the encrypted data packet structure is decrypted through the IP of the control end; the identification of the data packet structure includes but is not limited to extracting features from the input data through a feature extraction model to obtain corresponding feature data, and then classifying the feature data through a classification model to obtain the corresponding category; the storage of the data packet structure includes sorting the identified data packet structure based on the weighting of priority and acquisition time and storing it on the control end.

[0029] A further solution is that the category includes at least video data and audio data, and the identification of the data packet structure further includes:

[0030] Processing video information through computer vision algorithms: Identifying and tracking reconnaissance targets based on target detection and tracking algorithms. Counting and statistical analysis are performed based on the recognition results to calculate a modeling subset for 3D modeling. The modeling subset is a set of valid images in each frame of the video data.

[0031] Processing audio data through speech recognition algorithms: Converting audio to text based on speech recognition algorithms and simultaneous interpretation functions; Recognizing people's voiceprints based on voiceprint recognition algorithms and combining them with video face recognition algorithms to confirm the speaker's identity;

[0032] The text information is deduplicated, normalized, and disambiguated through natural language processing algorithms: the text data information obtained based on speech recognition is translated into a recognizable language using machine translation, and the summary information of the text is extracted based on the summary extraction algorithm; the main keywords in the text are extracted based on the named entity recognition algorithm.

[0033] A further solution is that the 5G base station sends instructions to the unmanned vehicle based on the stored data packet structure recognition results. Every time the unmanned vehicle executes an instruction, the control end deletes the data packet structure corresponding to the instruction.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The present invention improves data processing accuracy by establishing dedicated communication channels and constructing different data packet structures, meets combat requirements in specific environments, ensures synchronous interaction of business, measurement and control, and command and control information, and meets message throughput requirements and anti-interference performance requirements.

[0036] This invention utilizes "leaky cable + 5G" transmission technology to enable each node in the unmanned system's data link to automatically generate, process, and exchange tactical information, following the operational control and timing requirements of the communication protocol. This establishes communication links and forms a communication network with a defined topology, meeting the real-time, reliable communication requirements of combat missions. 3D data acquisition equipment carried by unmanned vehicles can be used to model the environment and determine scene distribution. AI-powered intelligent data analysis can extract a vast amount of important information from raw data without requiring human intervention, saving significant time in information extraction. The system ultimately displays the extracted, usable information to the command center, enabling expert decision-making and facilitating combat operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The following drawings are merely provided for illustrative purposes only and are not intended to limit the scope of the present invention.

[0038] Figure 1 : Schematic diagram of the connection principle of the system of the present invention;

[0039] Figure 2 : Schematic diagram of the process of the present invention;

[0040] In the picture: 1. Unmanned vehicle; 2. 5G base station; 3. Control terminal. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution, design method and advantages of the present invention more clear, the present invention is further described in detail below through specific embodiments in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0042] Example 1

[0043] like Figure 1 As shown, the present invention provides an unmanned reconnaissance system for confined and narrow spaces, including a control terminal 3, a 5G base station 2, and an unmanned vehicle 1. Unmanned vehicle 1 can be a robot, a robot dog, or a drone. Due to its small size and agility, a robot dog is used as an example in this embodiment. The robot dog, carrying an optical fiber coil, enters special environments such as tunnels, pits, and electronically shielded areas for reconnaissance, collects video, creates real-time maps, and transmits them to the 5G base station 2 and control terminal 3. Through artificial intelligence target detection, the robot dog can detect enemies and, based on commands from the command terminal, perform tasks such as dropping bombs.

[0044] The present invention provides a robot dog with a communication module, an image acquisition module, an encryption module, a data packet construction module, and a 3D model construction module; the communication module is connected to the 5G base station 2 via a leaky cable wired communication connection and a 5G wireless communication connection, forming a wired channel and a wireless channel; the image acquisition module is used to collect data of the area to be surveyed; the data packet construction module is used to construct the collected data into a data packet structure; the encryption module is used to encrypt the data packet structure; the 3D model construction module constructs a three-dimensional model based on the collected data.

[0045] MIMO channel models are established for wired channels and wireless channels respectively. The MIMO channel model includes at least three channels, namely data channel, control channel and RTS signal channel. The three channels of wireless transmission all use 4.9GHz and 700MHz dual-frequency electromagnetic waves; the data channel is a dedicated time slot for the unmanned vehicle end to transmit the data packet structure to the 5G base station 2, the control channel is a dedicated time slot for the 5G base station 2 to forward the control terminal 3 instructions to the unmanned vehicle end, and the RTS signal channel is the remaining time slots to ensure that the unmanned vehicle end and the 5G base station 2 establish a real-time communication connection. In this embodiment In the process, wired transmission and wireless transmission communicate according to the RTS / CTS protocol, which can effectively avoid conflicts. For example, first, the robot dog sends an RTS signal to the 5G base station 2 through wireless transmission, indicating that the robot dog wants to send data to the 5G base station 2. After receiving the RTS, the 5G base station 2 sends a CTS signal to the robot dog through wireless transmission, indicating that it is ready and the robot dog can send. At the same time, the wired transmission path of the robot dog suspends sending data to the 5G base station 2. It should be noted that the premise of this process is that the robot dog maintains wireless communication with the 5G base station 2 and the communication is good. When the area to be investigated cannot communicate normally with the 5G base station 2, wired transmission is required as the first choice. At this time, the robot dog sends an RTS signal to the 5G base station 2 through wired transmission, indicating that the robot dog wants to send data to the 5G base station 2. After receiving the RTS, the 5G base station 2 sends a CTS signal to the robot dog through wired transmission, indicating that it is ready and the robot dog can send. At the same time, the wireless transmission path of the robot dog suspends sending data to the 5G base station 2, and completes the free switching of the "leaky cable + 5G" transmission technology through the communication protocol, ensuring uninterrupted communication and improving communication stability.

[0046] 5G base station 2 acts as a data transfer processing station between the control terminal 3 and the robot dog, decrypting and identifying the data packet structure obtained through the data channel and sending it to the control terminal 3;

[0047] The control terminal 3 sends instructions to the robot dog based on the identified data packet structure to complete the reconnaissance.

[0048] In the above, the data packet structure includes data number, data type, collection time and priority; as shown in the following table:

[0049] serial number Ⅰ Ⅱ Ⅲ Ⅳ Ⅴ type video Audio picture Audio video time 1651911056 1651911084 1651911096 1651911108 1651911120 Priority 1 2 3 2 1

[0050] The collection time is the number of seconds from 00:00:00 on January 1, 1970 to the collection time.

[0051] The encryption module encrypts the data collected by the image acquisition module. In order to decrypt quickly and accurately, in this embodiment, the decryption key is the IP of the control terminal 3. When the encrypted data is transmitted to the 5G base station 2, the 5G base station 2 obtains the IP of the command terminal and can quickly and accurately decrypt the encrypted file; the recognition of the data packet structure includes extracting features of the input data through a feature extraction model to obtain corresponding feature data, and then classifying the feature data through a classification model to obtain the corresponding category. The storage of the data packet structure includes sorting the identified data packet structure based on the weight of priority and acquisition time and storing it in the control terminal 3. The category includes at least video data and audio data. In response to the problems of unreliable network transmission in a strong confrontation environment, the current situation of multi-link and incomplete information situation of the unmanned reconnaissance and attack system, and precise control under restricted conditions, we study the mapping of tactical and other messages to grammatical and semantic information, so that unmanned combat units can form a unified situation understanding, decision-making control and tactical coordination, thereby realizing "machine mutual recognition", forming a unified situation understanding and combat intention, and realizing consistent intelligent interaction between unmanned system units, platforms and people, accelerating the link from terminal perception to strike action, which is also the key to realizing manned / unmanned collaboration. Therefore, the robot dog must also include a data processing module, which is composed of a core algorithm server and an edge computing device. The core algorithm server provides powerful computing power, and the edge computing provides a convenient data acquisition function, which is then sent to the server for data integration. The operation of the entire system is completed through the data interaction between them. The data processing module has built-in computer vision algorithms, speech recognition algorithms, and natural language processing algorithms.

[0052] The video information is processed using computer vision algorithms: Reconnaissance targets (including people and equipment) are identified and tracked using the target detection algorithm (YOLOv5) and the target tracking algorithm (Deep Sort). Counting and statistical analysis are performed based on the recognition results to calculate the number of enemy troops and enemy firepower. The 3D model construction module uses the data collected by the image acquisition module to obtain a modeling subset for 3D modeling. The modeling subset consists of the valid image set for each frame in the video data.

[0053] Processing audio data through speech recognition algorithms: Converting audio to text based on speech recognition algorithms and simultaneous interpretation functions; Recognizing people's voiceprints based on voiceprint recognition algorithms and combining them with video face recognition algorithms to confirm the speaker's identity;

[0054] The text information is deduplicated, normalized, and disambiguated through natural language processing algorithms: the text data information obtained based on speech recognition is translated into a recognizable language using machine translation, and the summary information of the text is extracted based on the summary extraction algorithm; the main keywords in the text are extracted based on the named entity recognition algorithm.

[0055] Since the unmanned vehicle end will send data to the control end 3 in real time, in order to avoid storing too much data, the control end 3 will delete the data packet structure according to a certain strategy. For example, the control end 3 sends an instruction to the robot dog based on the stored data packet structure recognition result. Every time the robot dog executes an instruction, the control end 3 deletes the data packet structure corresponding to the instruction.

[0056] Example 2

[0057] This application provides an unmanned reconnaissance method for a confined and narrow space, using the above-mentioned reconnaissance system, an unmanned vehicle 1 enters the area to be reconnaissanced, and establishes communication with a control terminal 3; Figure 2 As shown, the following steps are included:

[0058] Step 1: Establishment of communication link;

[0059] Connect the unmanned vehicle 1 to the 5G base station 2 using a fusion of wired and wireless transmission;

[0060] When the unmanned vehicle 1 enters the area to be surveyed, the unmanned vehicle 1 immediately exchanges signals with the 5G base station 2, and establishes MIMO channel models for wired transmission and wireless transmission respectively for communication between the unmanned vehicle 1 and the 5G base station 2. The MIMO channel model includes at least three channels, namely a data channel, a control channel, and an RTS signal channel. The three channels of the wireless transmission all use 4.9 GHz and 700 MHz dual-frequency electromagnetic waves;

[0061] The unmanned vehicle 1 obtains data of the area to be surveyed, and transmits it to the 5G base station 2 via wired or wireless transmission in the form of an encrypted data packet structure according to the RTS / CTS protocol;

[0062] Step 2: Obtain the data packet structure of the area to be surveyed and perform target identification in real time; the data packet structure includes data number, data type, collection time and priority;

[0063] The 5G base station 2 obtains a data packet structure through a data channel, decrypts, identifies, and stores the data packet structure, and then sends it to the control terminal 3. Specifically, the encrypted data packet structure is decrypted through the IP of the control terminal 3. The identification of the data packet structure includes, but is not limited to, extracting features from the input data using a feature extraction model to obtain corresponding feature data, and then classifying the feature data using a classification model to obtain the corresponding category. The storage of the data packet structure includes sorting the identified data packet structures based on a weighted order of priority and acquisition time and storing them on the control terminal 3.

[0064] The categories include at least video data and audio data, and the identification of the data packet structure also includes:

[0065] Processing video information through computer vision algorithms: Identifying and tracking reconnaissance targets based on target detection and tracking algorithms. Counting and statistical analysis are performed based on the recognition results to calculate a modeling subset for 3D modeling. The modeling subset is a set of valid images in each frame of the video data.

[0066] Processing audio data through speech recognition algorithms: Converting audio to text based on speech recognition algorithms and simultaneous interpretation functions; Recognizing people's voiceprints based on voiceprint recognition algorithms and combining them with video face recognition algorithms to confirm the speaker's identity;

[0067] The text information is deduplicated, normalized, and disambiguated through natural language processing algorithms: the text data information obtained based on speech recognition is translated into a recognizable language using machine translation, and the summary information of the text is extracted based on the summary extraction algorithm; the main keywords in the text are extracted based on the named entity recognition algorithm.

[0068] Step 3: The control terminal 3 remotely controls the unmanned vehicle 1;

[0069] The control terminal 3 sends instructions to the unmanned vehicle 1 based on the identified data packet structure. Every time the unmanned vehicle 1 completes executing an instruction, the control terminal 3 deletes the data packet structure corresponding to the instruction.

[0070] While various embodiments of the present invention have been described above, the above descriptions are intended to be illustrative, non-exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is selected to best explain the principles of the embodiments, their practical applications, or technological improvements in the marketplace, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. An unmanned reconnaissance system for confined and narrow spaces, comprising a control terminal, a 5G base station, and an unmanned vehicle, characterized in that: The unmanned vehicle is provided with a communication module, an image acquisition module, an encryption module, a data packet construction module, and a 3D model construction module; the communication module is connected to the 5G base station via a leaky cable wired communication connection and a 5G wireless communication connection, forming a wired channel and a wireless channel; the image acquisition module is used to collect data of the area to be surveyed; the data packet construction module is used to construct the collected data into a data packet structure; the encryption module is used to encrypt the data packet structure; the 3D model construction module constructs a three-dimensional model based on the collected data; The wired channel and the wireless channel respectively establish MIMO channel models, wherein the MIMO channel models include at least three channels, namely a data channel, a control channel and an RTS signal channel, and the three wireless transmission channels all use 4.9 GHz and 700 MHz dual-frequency electromagnetic waves; The 5G base station acts as a data transfer processing station between the control end and the unmanned vehicle, decrypting, identifying, and storing the data packet structure obtained through the data channel before sending it to the control end; The control end sends instructions to the unmanned vehicle based on the identified data packet structure to complete the reconnaissance; The encrypted data packet structure is decrypted by the IP of the control terminal; the identification of the data packet structure includes but is not limited to extracting features from the input data using a feature extraction model to obtain corresponding feature data, and then classifying the feature data using a classification model to obtain the corresponding category; the storage of the data packet structure includes sorting the identified data packet structures based on the weight of priority and acquisition time and storing them on the control terminal; Process video information through computer vision algorithms; The 5G base station sends instructions to the unmanned vehicle based on the stored data packet structure recognition results. Every time the unmanned vehicle executes an instruction, the control end deletes the data packet structure corresponding to the instruction.

2. The unmanned reconnaissance system for confined and narrow spaces according to claim 1, characterized in that: The data packet structure includes data number, data type, collection time and priority; The collection time is the number of seconds from 00:00:00 on January 1, 1970 to the collection time.

3. The unmanned reconnaissance system for confined and narrow spaces according to claim 2, characterized in that: The categories include at least video data and audio data, and the unmanned vehicle further includes a data processing module, wherein the data processing module has built-in computer vision algorithms, speech recognition algorithms, and natural language processing algorithms; Reconnaissance targets are identified and tracked based on target detection and tracking algorithms. Counting and statistical analysis are performed based on the identification results to calculate a modeling subset for 3D modeling. The modeling subset is a valid image set for each frame in the video data. Processing audio data through speech recognition algorithms: Converting audio to text based on speech recognition algorithms and simultaneous interpretation functions; Based on the voiceprint recognition algorithm, the voiceprint of the person is recognized and combined with the video face recognition algorithm to confirm the identity of the speaker; The text information is deduplicated, normalized, and disambiguated through natural language processing algorithms: the text data information obtained based on speech recognition is translated into a recognizable language using machine translation, the summary information of the text is extracted based on the summary extraction algorithm; and the keywords in the text are extracted based on the named entity recognition algorithm.

4. A method for unmanned reconnaissance in a confined and narrow space, characterized in that: Using the reconnaissance system according to any one of claims 1 to 3, an unmanned vehicle enters the area to be reconnaissanced and establishes communication with a control terminal; the steps include: Step 1: Establishment of communication link; Connect unmanned vehicles to 5G base stations using a combination of wired and wireless transmission; When the unmanned vehicle enters the area to be reconnaissanced, it immediately exchanges signaling with the 5G base station and establishes MIMO channel models for wired and wireless transmission for communication between the unmanned vehicle and the 5G base station. The MIMO channel model includes at least three channels: a data channel, a control channel, and an RTS signal channel. The three wireless transmission channels all use dual-frequency electromagnetic waves of 4.9 GHz and 700 MHz. The unmanned vehicle obtains data about the area to be surveyed and transmits it to the 5G base station via wired or wireless transmission in the form of an encrypted data packet structure according to the RTS / CTS protocol; Step 2: Obtain the data packet structure of the area to be surveyed and perform target identification in real time; The 5G base station obtains the data packet structure through the data channel, decrypts, identifies and stores the data packet structure, and then sends it to the control end; Step 3: The control terminal remotely controls the unmanned vehicle; The control end sends instructions to the unmanned vehicle based on the identified data packet structure to complete the reconnaissance.

5. The unmanned reconnaissance method for a confined and narrow space according to claim 4, characterized in that: The data packet structure includes data number, data type, collection time and priority; The collection time is the number of seconds from 00:00:00 on January 1, 1970 to the collection time.

6. The unmanned reconnaissance method for a confined and narrow space according to claim 5, characterized in that: In step 2, the encrypted data packet structure is decrypted through the IP of the control end; the identification of the data packet structure includes but is not limited to extracting features from the input data through a feature extraction model to obtain corresponding feature data, and then classifying the feature data through a classification model to obtain the corresponding category; the storage of the data packet structure includes sorting the identified data packet structure based on the weighting of priority and acquisition time and storing it on the control end.

7. The unmanned reconnaissance method for a confined and narrow space according to claim 6, characterized in that: The categories include at least video data and audio data, and the identification of the data packet structure further includes: Processing video information through computer vision algorithms: Identifying and tracking reconnaissance targets based on target detection and tracking algorithms. Counting and statistical analysis are performed based on the recognition results to calculate a modeling subset for 3D modeling. The modeling subset is a set of valid images in each frame of the video data. Processing audio data through speech recognition algorithms: Converting audio to text based on speech recognition algorithms and simultaneous interpretation functions; Recognizing people's voiceprints based on voiceprint recognition algorithms and combining them with video face recognition algorithms to confirm the speaker's identity; The text information is deduplicated, normalized, and disambiguated through natural language processing algorithms: the text data information obtained based on speech recognition is translated into a recognizable language using machine translation, the summary information of the text is extracted based on the summary extraction algorithm; and the keywords in the text are extracted based on the named entity recognition algorithm.

8. The unmanned reconnaissance method for a confined and narrow space according to claim 7, characterized in that: The 5G base station sends instructions to the unmanned vehicle based on the stored data packet structure recognition results. Every time the unmanned vehicle executes an instruction, the control end deletes the data packet structure corresponding to the instruction.

Images