Unmanned aerial vehicle control system capable of operating offline

Through image recognition technology of binocular cameras and AI analysis modules, the precise positioning of the drone in remote areas is achieved, the problem of poor GPS signal is solved, and the reliability and safety of the drone control system is improved.

CN119024862APending Publication Date: 2024-11-26JIANGSU FEISUDA AVIATION TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411130271.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

In remote areas or mountainous areas, the signal quality of the GPS positioning system is poor, resulting in inaccurate positioning of the drone and the inability to cruise according to the set track.

Method used

The binocular camera and AI analysis comprehensive module are adopted to realize the precise positioning of the drone through image recognition technology, reduce dependence on the GPS positioning system, and switch control methods when the signal quality is poor to ensure the normal operation of the drone.

Benefits of technology

It improves the reliability and effectiveness of the UAV control system, ensuring that the UAV can achieve automatic positioning and safe flight in areas with poor signal quality, and avoid control failure.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119024862A_ABST
    Figure CN119024862A_ABST
Patent Text Reader

Abstract

The invention relates to the field of unmanned aerial vehicles, in particular to an unmanned aerial vehicle control system capable of operating offline. The technical problems are that an unmanned aerial vehicle control system based on a GPS positioning system is adopted to control the course of an unmanned aerial vehicle so as to realize the automatic cruise function of the unmanned aerial vehicle, the signal quality of the GPS positioning system is poor in a remote zone or a mountainous area, so that the positioning of the unmanned aerial vehicle is inaccurate, and the unmanned aerial vehicle is inconvenient to use. Therefore, an unmanned aerial vehicle control system is difficult to accurately position and control the unmanned aerial vehicle, and the unmanned aerial vehicle cannot cruise according to a set track; according to the technical scheme, the unmanned aerial vehicle control system capable of operating offline comprises an AI analysis and integration module and a signal judgment module; according to the invention, by using the binocular camera technology, accurate positioning of the unmanned aerial vehicle can be realized through the landmark information and the recorded map or landmark information, so that sufficient data support is provided for channel deviation correction of the unmanned aerial vehicle, and the reliability and effectiveness of unmanned aerial vehicle control are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicles, and in particular to an unmanned aerial vehicle control system capable of operating offline. Background Art

[0002] The drone control system is an advanced technology, which consists of hardware and software. The hardware part mainly includes sensors, cameras, motors and batteries. Sensors are used to obtain environmental information, cameras are used to take photos and videos, motors are used to drive and control drones, and batteries provide energy. The software part is responsible for processing and analyzing the acquired information and sending instructions to the hardware part. The drone control system is a set of integrated hardware and software systems that can obtain environmental information in real time and control drones through instructions. The effective operation of the system requires accurate data processing and analysis, as well as stable and reliable hardware and software component support. When controlling the flight attitude of the drone, a drone control system based on the GPS positioning system is used to control the heading of the drone to realize the automatic cruising function of the drone. In remote areas or mountainous areas, the signal quality of the GPS positioning system is poor, which will lead to inaccurate positioning of the drone, making it difficult for the drone control system to accurately locate and control the drone, resulting in the drone being unable to cruise according to the set track. Summary of the invention

[0003] In order to overcome the problem of controlling the flight attitude of the UAV, a UAV control system based on the GPS positioning system is used to control the heading of the UAV to realize the automatic cruising function of the UAV. In remote areas or mountainous areas, the signal quality of the GPS positioning system is poor, which will lead to inaccurate positioning of the UAV, making it difficult for the UAV control system to accurately locate and control the UAV, resulting in the UAV being unable to cruise according to the set track.

[0004] The technical solution of the present invention is: a drone control system that can be operated offline, comprising:

[0005] The data acquisition module is used to collect environmental data around the drone during its flight;

[0006] The data processing module is used to process the data collected by the data acquisition module to improve the quality and recognizability of the data collected by the data acquisition module;

[0007] The AI ​​analysis and synthesis module is used to analyze the data processed by the data processing module, identify the flight status of the drone and the surrounding environment information, and control the flight attitude of the drone based on the identified information;

[0008] A data transmission module, used to realize data transmission between the UAV and the human manual control terminal of the UAV;

[0009] A data storage module is used to store information sent by the manual control terminal and to store trajectory-related data such as map data;

[0010] The signal judgment module is used to judge the stability and reliability of the data transmission module, and send instructions to the AI ​​analysis and synthesis module to switch the drone control mode.

[0011] Preferably, by setting up a signal judgment module, the signal quality of the data transmission module can be effectively judged in real time, and the control mode of the UAV can be intelligently switched according to the judgment result of the signal quality, thereby ensuring that the UAV will not suddenly lose control when encountering a poor communication signal, resulting in UAV control failure, thereby ensuring the safety and normal operation of the UAV and improving the reliability of the UAV control system.

[0012] Preferably, the data acquisition module includes a GPS positioning module, an image acquisition module and an environment detection module. The GPS positioning module is used to collect the geographic location information data of the drone, the image acquisition module is used to collect image data of the surrounding and lower environment of the drone, and the environment detection module is used to collect data information of the surrounding environment of the drone; the data acquisition module also includes a three-axis acceleration sensor, which is used to collect acceleration information of the drone in three-dimensional space.

[0013] Preferably, the image acquisition module is a binocular camera, which is arranged in multiple groups, and one of the multiple groups of binocular cameras is arranged at the bottom of the drone, for collecting image data of the surface below the drone, so as to collect image data of landmarks on the surface as analysis data for offline positioning of the drone; the environment detection module includes multiple sensors such as temperature and humidity sensors, wind speed sensors, light intensity sensors, particle concentration sensors and barometers.

[0014] Preferably, when the data processing module processes the data collected by the data acquisition module to improve the quality and recognizability of the data collected by the data acquisition module, the data collected by the image acquisition module is processed by the following steps:

[0015] S11: performing frame processing on the image data collected by the image acquisition module;

[0016] S12: grayscale processing is performed on the image data after the frame processing to obtain a grayscale image, wherein the principle formula of the image grayscale processing is: Y=0.299×R+0.587×G+0.144×B; wherein Y is the brightness of the grayscale image after conversion, that is, the grayscale value, and R, G and B are the values ​​of the three primary colors of red, green and blue of the color image before conversion respectively;

[0017] S13: Using an image edge detection algorithm based on a gradient algorithm to perform edge detection on the grayscale image, and performing image segmentation based on the detection result.

[0018] Preferably, when the AI ​​analysis and synthesis module analyzes the data processed by the data processing module and identifies the flight status of the drone and the surrounding environment information, the real-time position of the drone is identified through the following steps:

[0019] S21: recognizing the grayscale image after image segmentation through an image recognition algorithm based on a convolutional neural network, recognizing the meaning of each part of the grayscale image, and selecting a special area according to the set recognition content, wherein the special area is an image of the object used as a positioning reference point;

[0020] S22: dividing the image captured by the binocular vision camera into a left-eye view and a right-eye view, and placing the pixel points of all special areas of the left-eye view and the right-eye view into the same rectangular coordinate system; wherein the origin of the rectangular coordinate system corresponds to a pixel point at the same position in the left-eye view and the right-eye view;

[0021] S23: Calculate the disparity of the pixel point at each position in the binocular vision camera according to the coordinates of the pixel points in all face regions in the left eye view and the right eye view; wherein the disparity of the pixel point at each position in the binocular vision camera is calculated according to the following formula:

[0022] d=|x l -x r |,

[0023] Among them, d is the disparity of the pixel point at that position in the binocular vision camera, x l is the horizontal coordinate of the pixel point at this position in the left eye view, x r is the horizontal coordinate of the pixel point at that position in the right eye view;

[0024] S24: Calculate the depth of the pixel at the position in space according to the disparity; wherein the calculation formula is: Among them, f is the focal length of the binocular vision camera, b is the camera baseline distance;

[0025] S25: establishing a three-dimensional coordinate system, and establishing a three-dimensional space model of the space where the drone is located according to the horizontal coordinate, the vertical coordinate and the calculated depth in the space of each pixel point;

[0026] S26: Determine the precise real-time position of the drone based on the location information of the positioning point stored in the data storage module.

[0027] Preferably, by setting up a binocular camera and image recognition of the binocular camera, the relative position between the UAV and the feature object can be effectively and accurately identified through the images collected by the binocular camera, thereby reducing the UAV's dependence on the GPS positioning system when cruising. Therefore, when the UAV is working in remote areas or other areas with poor signal quality, the UAV can achieve automatic positioning through the entered map data and other geographic data, thereby ensuring the normal operation of the UAV and improving the reliability of the control system.

[0028] Preferably, the AI ​​analysis and synthesis module includes the following control modes when controlling the flight attitude of the drone according to the identified information:

[0029] A11: The drone operator controls the flight trajectory of the drone by manual control, so that the drone flies according to the operator's control instructions;

[0030] A12: The drone operator inputs the destination information into the manual control terminal, which automatically generates the drone inspection path through the built-in algorithm and controls the drone's flight attitude in real time according to the drone's GPS positioning to complete the drone's automatic cruise.

[0031] A13: When the drone loses signal connection with the manual control terminal or the signal connection strength is weak, the AI ​​analysis and synthesis module controls the drone to fly in real time and correct deviation according to the path information stored in the data storage module;

[0032] A14: When the drone loses signal connection with the manual control terminal or the signal connection strength is weak, the AI ​​analysis and synthesis module controls the drone to automatically return based on the drone's historical flight trajectory recorded in the data storage module.

[0033] As a preference, the AI ​​analysis and synthesis module also includes the following functions:

[0034] A21: Calculate the acceleration of the drone in the forward, vertical and radial directions based on the data collected by the three-axis acceleration sensor, and correct the flight attitude of the drone based on the calculation results;

[0035] A22: When the UAV is flying according to the command of the manual control terminal, the image data of the ground surface below the UAV is intermittently recognized, and areas or objects with obvious features in the ground surface image are selected and marked as landmarks on the ground surface;

[0036] A23: The maximum flight speed of the drone can be intelligently limited based on data from multiple sensors such as temperature and humidity sensors, wind speed sensors, light intensity sensors, particle concentration sensors, and barometers.

[0037] Preferably, the signal determination module comprises the following steps when determining the stability and reliability of the data transmission module:

[0038] S31: The signal determination module intermittently sends verification information with time information to the manual control terminal through the data transmission module;

[0039] S32: At the same time, the manual control terminal intermittently sends verification information with time information to the signal judgment module through the data transmission module;

[0040] S33: The manual control terminal and the signal judgment module simultaneously verify the received verification information, verify the deviation between the received information and the existing time, and simultaneously verify the integrity of the received verification information;

[0041] S34: Perform a weighted average based on the two verification results obtained in step S33 to judge the signal quality, and divide the signal quality into four levels: good signal quality, poor signal quality, bad signal quality and signal interruption according to the set rules.

[0042] Preferably, when the signal judgment module sends instructions to the AI ​​analysis and synthesis module to switch the drone control mode, the following methods are included:

[0043] A31: Read the real-time signal quality level. When the signal quality is at a good level, the control mode is A11 or A12.

[0044] A32: Read the real-time signal quality level. When the control signal is in poor quality, the control mode is A11 or A12, and the information is sent to the AI ​​analysis and synthesis module, which is used to locate and correct the drone in real time.

[0045] A33: Read the real-time signal quality level. When the signal quality is at a poor level, send an alert to the manual control terminal and control the drone to continue flying or return according to the rotation of the drone operator.

[0046] A34: Read the real-time signal quality level. When the signal quality is at the signal interruption level, the control mode is A13 or A14.

[0047] Preferably, when the UAV is flying according to the instruction of the manual control terminal, the AI ​​analysis and synthesis module intermittently recognizes the image data of the surface below the UAV through the following steps, and selects areas or objects with obvious features in the surface image to mark as landmarks on the surface:

[0048] S41: according to the set time interval, start the binocular camera at the bottom of the drone, collect the surface image below the drone, and record the current GPS coordinates;

[0049] S42: Analyze and identify the image, and randomly select objects or areas with identifiable features as landmarks;

[0050] S43: Calculate the location of the landmark through steps S21 to S26, record and store it, and complete the marking of the landmark.

[0051] Beneficial effects of the present invention:

[0052] 1. Compared with the prior art, when controlling the flight attitude of a drone, a drone control system based on a GPS positioning system is used to control the heading of the drone to realize the automatic cruise function of the drone. In remote areas or mountainous areas, the signal quality of the GPS positioning system is poor, which will lead to inaccurate positioning of the drone, making it difficult for the drone control system to accurately locate and control the drone, resulting in the drone being unable to cruise according to the set track; the drone control system uses binocular camera technology to achieve accurate positioning of the drone through landmark information and recorded maps or landmark information, thereby providing sufficient data support for the drone's course correction and improving the reliability and effectiveness of drone control;

[0053] 2. By setting up binocular cameras and image recognition of binocular cameras, the relative position between the UAV and the characteristic object can be accurately identified through the images collected by the binocular cameras, reducing the dependence of the UAV on the GPS positioning system when cruising. When the UAV is working in remote areas or other areas with poor signal quality, the UAV can realize automatic positioning through the input map data and other geographic data, ensuring the normal operation of the UAV and improving the reliability of the control system;

[0054] 3. By setting up the signal judgment module, the signal quality of the data transmission module can be effectively judged in real time, and the control mode of the drone can be intelligently switched according to the judgment result of the signal quality, so as to ensure that the drone will not suddenly lose control when encountering a poor communication signal, resulting in drone control failure, thereby ensuring the safety and normal operation of the drone and improving the reliability of the drone control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 Shown is a schematic diagram of the structure of the offline-operable drone control system of the present invention;

[0056] Figure 2 The following is a schematic diagram of the working process of the signal judgment module in the UAV control system capable of offline operation of the present invention;

[0057] Figure 3 Shown is a partial workflow diagram of the AI ​​analysis and synthesis module in the offline-operable drone control system of the present invention. DETAILED DESCRIPTION

[0058] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0059] See also Figure 1 The present invention provides an embodiment: a drone control system capable of operating offline, comprising:

[0060] The data acquisition module is used to collect environmental data around the drone during its flight;

[0061] The data processing module is used to process the data collected by the data acquisition module to improve the quality and recognizability of the data collected by the data acquisition module;

[0062] The AI ​​analysis and synthesis module is used to analyze the data processed by the data processing module, identify the flight status of the drone and the surrounding environment information, and control the flight attitude of the drone based on the identified information;

[0063] A data transmission module, used to realize data transmission between the UAV and the human manual control terminal of the UAV;

[0064] A data storage module is used to store information sent by the manual control terminal and to store trajectory-related data such as map data;

[0065] The signal judgment module is used to judge the stability and reliability of the data transmission module, and send instructions to the AI ​​analysis and synthesis module to switch the drone control mode.

[0066] Preferably, by setting up a signal judgment module, the signal quality of the data transmission module can be effectively judged in real time, and the control mode of the UAV can be intelligently switched according to the judgment result of the signal quality, thereby ensuring that the UAV will not suddenly lose control when encountering a poor communication signal, resulting in UAV control failure, thereby ensuring the safety and normal operation of the UAV and improving the reliability of the UAV control system.

[0067] Preferably, the data acquisition module includes a GPS positioning module, an image acquisition module and an environment detection module. The GPS positioning module is used to collect the geographic location information data of the drone, the image acquisition module is used to collect image data of the surrounding and lower environment of the drone, and the environment detection module is used to collect data information of the surrounding environment of the drone; the data acquisition module also includes a three-axis acceleration sensor, which is used to collect acceleration information of the drone in three-dimensional space.

[0068] Preferably, the image acquisition module is a binocular camera, which is arranged in multiple groups, and one of the multiple groups of binocular cameras is arranged at the bottom of the drone, for collecting image data of the surface below the drone, so as to collect image data of landmarks on the surface as analysis data for offline positioning of the drone; the environment detection module includes multiple sensors such as temperature and humidity sensors, wind speed sensors, light intensity sensors, particle concentration sensors and barometers.

[0069] Preferably, when the data processing module processes the data collected by the data acquisition module to improve the quality and recognizability of the data collected by the data acquisition module, the data collected by the image acquisition module is processed by the following steps:

[0070] S11: performing frame processing on the image data collected by the image acquisition module;

[0071] S12: grayscale processing is performed on the image data after the frame processing to obtain a grayscale image, wherein the principle formula of the image grayscale processing is: Y=0.299×R+0.587×G+0.144×B; wherein Y is the brightness of the grayscale image after conversion, that is, the grayscale value, and R, G and B are the values ​​of the three primary colors of red, green and blue of the color image before conversion respectively;

[0072] S13: Using an image edge detection algorithm based on a gradient algorithm to perform edge detection on the grayscale image, and performing image segmentation based on the detection result.

[0073] Preferably, when the AI ​​analysis and synthesis module analyzes the data processed by the data processing module and identifies the flight status of the drone and the surrounding environment information, the real-time position of the drone is identified through the following steps:

[0074] S21: recognizing the grayscale image after image segmentation through an image recognition algorithm based on a convolutional neural network, recognizing the meaning of each part of the grayscale image, and selecting a special area according to the set recognition content, wherein the special area is an image of the object used as a positioning reference point;

[0075] S22: dividing the image captured by the binocular vision camera into a left-eye view and a right-eye view, and placing the pixel points of all special areas of the left-eye view and the right-eye view into the same rectangular coordinate system; wherein the origin of the rectangular coordinate system corresponds to a pixel point at the same position in the left-eye view and the right-eye view;

[0076] S23: Calculate the disparity of the pixel point at each position in the binocular vision camera according to the coordinates of the pixel points in all face regions in the left eye view and the right eye view; wherein the disparity of the pixel point at each position in the binocular vision camera is calculated according to the following formula:

[0077] d=|x l -x r |,

[0078] Among them, d is the disparity of the pixel point at that position in the binocular vision camera, x l is the horizontal coordinate of the pixel point at this position in the left eye view, x r is the horizontal coordinate of the pixel point at that position in the right eye view;

[0079] S24: Calculate the depth of the pixel at the position in space according to the disparity; wherein the calculation formula is: Among them, f is the focal length of the binocular vision camera, b is the camera baseline distance;

[0080] S25: establishing a three-dimensional coordinate system, and establishing a three-dimensional space model of the space where the drone is located according to the horizontal coordinate, the vertical coordinate and the calculated depth in the space of each pixel point;

[0081] S26: Determine the precise real-time position of the drone based on the location information of the positioning point stored in the data storage module.

[0082] Preferably, by setting up a binocular camera and image recognition of the binocular camera, the relative position between the UAV and the feature object can be effectively and accurately identified through the images collected by the binocular camera, thereby reducing the UAV's dependence on the GPS positioning system when cruising. Therefore, when the UAV is working in remote areas or other areas with poor signal quality, the UAV can achieve automatic positioning through the entered map data and other geographic data, thereby ensuring the normal operation of the UAV and improving the reliability of the control system.

[0083] Preferably, the AI ​​analysis and synthesis module includes the following control modes when controlling the flight attitude of the drone according to the identified information:

[0084] A11: The drone operator controls the flight trajectory of the drone by manual control, so that the drone flies according to the operator's control instructions;

[0085] A12: The drone operator inputs the destination information into the manual control terminal, which automatically generates the drone inspection path through the built-in algorithm and controls the drone's flight attitude in real time according to the drone's GPS positioning to complete the drone's automatic cruise.

[0086] A13: When the drone loses signal connection with the manual control terminal or the signal connection strength is weak, the AI ​​analysis and synthesis module controls the drone to fly in real time and correct deviation according to the path information stored in the data storage module;

[0087] A14: When the drone loses signal connection with the manual control terminal or the signal connection strength is weak, the AI ​​analysis and synthesis module controls the drone to automatically return based on the drone's historical flight trajectory recorded in the data storage module.

[0088] As a preference, the AI ​​analysis and synthesis module also includes the following functions:

[0089] A21: Calculate the acceleration of the drone in the forward, vertical and radial directions based on the data collected by the three-axis acceleration sensor, and correct the flight attitude of the drone based on the calculation results;

[0090] A22: When the UAV is flying according to the command of the manual control terminal, the image data of the ground surface below the UAV is intermittently recognized, and areas or objects with obvious features in the ground surface image are selected and marked as landmarks on the ground surface;

[0091] A23: The maximum flight speed of the drone can be intelligently limited based on data from multiple sensors such as temperature and humidity sensors, wind speed sensors, light intensity sensors, particle concentration sensors, and barometers.

[0092] See also Figure 2 In this embodiment, the signal determination module includes the following steps when determining the stability and reliability of the data transmission module:

[0093] S31: The signal determination module intermittently sends verification information with time information to the manual control terminal through the data transmission module;

[0094] S32: At the same time, the manual control terminal intermittently sends verification information with time information to the signal judgment module through the data transmission module;

[0095] S33: The manual control terminal and the signal judgment module simultaneously verify the received verification information, verify the deviation between the received information and the existing time, and simultaneously verify the integrity of the received verification information;

[0096] S34: Perform a weighted average based on the two verification results obtained in step S33 to judge the signal quality, and divide the signal quality into four levels: good signal quality, poor signal quality, bad signal quality and signal interruption according to the set rules.

[0097] Preferably, when the signal judgment module sends instructions to the AI ​​analysis and synthesis module to switch the drone control mode, the following methods are included:

[0098] A31: Read the real-time signal quality level. When the signal quality is at a good level, the control mode is A11 or A12.

[0099] A32: Read the real-time signal quality level. When the control signal is in poor quality, the control mode is A11 or A12, and the information is sent to the AI ​​analysis and synthesis module, which is used to locate and correct the drone in real time.

[0100] A33: Read the real-time signal quality level. When the signal quality is at a poor level, send an alert to the manual control terminal and control the drone to continue flying or return according to the rotation of the drone operator.

[0101] A34: Read the real-time signal quality level. When the signal quality is at the signal interruption level, the control mode is A13 or A14.

[0102] See also Figure 3In this embodiment, when the UAV flies according to the instructions of the manual control terminal, the AI ​​analysis and synthesis module intermittently recognizes the image data of the ground surface below the UAV through the following steps, and selects areas or objects with obvious features in the ground surface image to mark as landmarks on the ground surface:

[0103] S41: according to the set time interval, start the binocular camera at the bottom of the drone, collect the surface image below the drone, and record the current GPS coordinates;

[0104] S42: Analyze and identify the image, and randomly select objects or areas with identifiable features as landmarks;

[0105] S43: Calculate the location of the landmark through steps S21 to S26, record and store it, and complete the marking of the landmark.

[0106] Through the above steps, compared with the prior art, when controlling the flight attitude of the UAV, a UAV control system based on the GPS positioning system is used to control the heading of the UAV to realize the automatic cruising function of the UAV. In remote areas or mountainous areas, the signal quality of the GPS positioning system is poor, which will lead to inaccurate positioning of the UAV, making it difficult for the UAV control system to accurately locate and control the UAV, resulting in the UAV being unable to cruise according to the set track; the UAV control system uses binocular camera technology to achieve accurate positioning of the UAV through landmark information and entered maps or landmark information, thereby providing sufficient data support for the UAV's course correction and improving the reliability and effectiveness of UAV control.

[0107] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge scope of those skilled in the art without departing from the spirit of the present invention.

Claims

1. A drone control system that can run offline; Features: Included are: The data acquisition module is used to collect environmental data around the drone during its flight; The data processing module is used to process the data collected by the data acquisition module to improve the quality and recognizability of the data collected by the data acquisition module; The AI ​​analysis and synthesis module is used to analyze the data processed by the data processing module, identify the flight status of the drone and the surrounding environment information, and control the flight attitude of the drone based on the identified information; A data transmission module, used to realize data transmission between the UAV and the human manual control terminal of the UAV; A data storage module is used to store information sent by the manual control terminal and to store trajectory-related data such as map data; The signal judgment module is used to judge the stability and reliability of the data transmission module, and send instructions to the AI ​​analysis and synthesis module to switch the drone control mode.

2. The drone control system capable of operating offline according to claim 1, Features: The data acquisition module includes a GPS positioning module, an image acquisition module and an environment detection module. The GPS positioning module is used to collect the geographic location information data of the drone, the image acquisition module is used to collect image data of the surrounding and lower environment of the drone, and the environment detection module is used to collect data information of the surrounding environment of the drone; the data acquisition module also includes a three-axis acceleration sensor, which is used to collect acceleration information of the drone in three-dimensional space.

3. The drone control system capable of operating offline according to claim 2, Features: The image acquisition module is a binocular camera, which is arranged in multiple groups, and one of the multiple groups of binocular cameras is arranged at the bottom of the drone, and is used to collect image data of the ground surface below the drone, so as to collect image data of landmarks on the ground surface as analysis data for offline positioning of the drone; The environment detection module includes multiple sensors such as temperature and humidity sensors, wind speed sensors, light intensity sensors, particle concentration sensors and barometers.

4. The drone control system capable of operating offline according to claim 3, Features: When the data processing module processes the data collected by the data acquisition module to improve the quality and recognizability of the data collected by the data acquisition module, the data collected by the image acquisition module is processed through the following steps: S11: performing frame processing on the image data collected by the image acquisition module; S12: grayscale processing is performed on the image data after the frame processing to obtain a grayscale image, wherein the principle formula of the image grayscale processing is: Y=0.299×R+0.587×G+0.144×B; wherein Y is the brightness of the grayscale image after conversion, that is, the grayscale value, and R, G and B are the values ​​of the three primary colors of red, green and blue of the color image before conversion respectively; S13: Using an image edge detection algorithm based on a gradient algorithm to perform edge detection on the grayscale image, and performing image segmentation based on the detection result.

5. The drone control system capable of operating offline according to claim 4, Features: When the AI ​​analysis and synthesis module analyzes the data processed by the data processing module and identifies the flight status of the drone and the surrounding environment information, it identifies the real-time location of the drone through the following steps: S21: recognizing the grayscale image after image segmentation through an image recognition algorithm based on a convolutional neural network, recognizing the meaning of each part of the grayscale image, and selecting a special area according to the set recognition content, wherein the special area is an image of the object used as a positioning reference point; S22: dividing the image captured by the binocular vision camera into a left-eye view and a right-eye view, and placing the pixel points of all special areas of the left-eye view and the right-eye view into the same rectangular coordinate system; wherein the origin of the rectangular coordinate system corresponds to a pixel point at the same position in the left-eye view and the right-eye view; S23: Calculate the disparity of the pixel point at each position in the binocular vision camera according to the coordinates of the pixel points in all face regions in the left eye view and the right eye view; wherein the disparity of the pixel point at each position in the binocular vision camera is calculated according to the following formula: d=|x l -x r |, Among them, d is the disparity of the pixel point at that position in the binocular vision camera, x l is the horizontal coordinate of the pixel point at this position in the left eye view, x r is the horizontal coordinate of the pixel point at that position in the right eye view; S24: Calculate the depth of the pixel point at the position in space according to the disparity; wherein the calculation formula is: Among them, f is the focal length of the binocular vision camera, b is the camera baseline distance; S25: establishing a three-dimensional coordinate system, and establishing a three-dimensional space model of the space where the drone is located according to the horizontal coordinate, the vertical coordinate and the calculated depth in the space of each pixel point; S26: Determine the precise real-time position of the drone based on the location information of the positioning point stored in the data storage module.

6. The drone control system capable of operating offline according to claim 5, Features: The AI ​​analysis and synthesis module controls the flight attitude of the drone based on the identified information, including the following control modes: A11: The drone operator controls the flight trajectory of the drone by manual control, so that the drone flies according to the operator's control instructions; A12: The drone operator inputs the destination information into the manual control terminal, which automatically generates the drone inspection path through the built-in algorithm and controls the drone's flight attitude in real time according to the drone's GPS positioning to complete the drone's automatic cruise. A13: When the drone loses signal connection with the manual control terminal or the signal connection strength is weak, the AI ​​analysis and synthesis module controls the drone to fly in real time and correct deviation according to the path information stored in the data storage module; A14: When the drone loses signal connection with the manual control terminal or the signal connection strength is weak, the AI ​​analysis and synthesis module controls the drone to automatically return based on the drone's historical flight trajectory recorded in the data storage module.

7. The drone control system capable of operating offline according to claim 6, Features: The AI ​​analysis and synthesis module also includes the following functions: A21: Calculate the acceleration of the drone in the forward, vertical and radial directions based on the data collected by the three-axis acceleration sensor, and correct the flight attitude of the drone based on the calculation results; A22: When the UAV is flying according to the command of the manual control terminal, the image data of the ground surface below the UAV is intermittently recognized, and areas or objects with obvious features in the ground surface image are selected and marked as landmarks on the ground surface; A23: The maximum flight speed of the drone can be intelligently limited based on data from multiple sensors such as temperature and humidity sensors, wind speed sensors, light intensity sensors, particle concentration sensors, and barometers.

8. The drone control system capable of operating offline according to claim 7, Features: The signal judgment module includes the following steps when judging the stability and reliability of the data transmission module: S31: The signal determination module intermittently sends verification information with time information to the manual control terminal through the data transmission module; S32: At the same time, the manual control terminal intermittently sends verification information with time information to the signal judgment module through the data transmission module; S33: The manual control terminal and the signal judgment module simultaneously verify the received verification information, verify the deviation between the received information and the existing time, and simultaneously verify the integrity of the received verification information; S34: Perform a weighted average based on the two verification results obtained in step S33 to judge the signal quality, and divide the signal quality into four levels: good signal quality, poor signal quality, bad signal quality and signal interruption according to the set rules.

9. The drone control system capable of operating offline according to claim 8, Features: The signal judgment module sends instructions to the AI ​​analysis and synthesis module to switch the drone control mode, including the following methods: A31: Read the real-time signal quality level. When the signal quality is at a good level, the control mode is A11 or A12. A32: Read the real-time signal quality level. When the control signal is in poor quality, the control mode is A11 or A12, and the information is sent to the AI ​​analysis and synthesis module, which is used to locate and correct the drone in real time. A33: Read the real-time signal quality level. When the signal quality is at a poor level, send an alert to the manual control terminal and control the drone to continue flying or return according to the rotation of the drone operator. A34: Read the real-time signal quality level. When the signal quality is at the signal interruption level, the control mode is A13 or A14.

10. The offline operable drone control system according to claim 9, Features: When the UAV is flying according to the instructions of the manual control terminal, the AI ​​analysis and synthesis module intermittently recognizes the image data of the ground surface below the UAV through the following steps, and selects areas or objects with obvious features in the ground surface image to mark as landmarks on the ground surface: S41: according to the set time interval, start the binocular camera at the bottom of the drone, collect the surface image below the drone, and record the current GPS coordinates; S42: Analyze and identify the image, and randomly select objects or areas with identifiable features as landmarks; S43: Calculate the location of the landmark through steps S21 to S26, record and store it, and complete the marking of the landmark.