CN-122018547-A - Dian wall target guidance method and system combined with visual inspection of multi-rotor unmanned aerial vehicle

Abstract

The invention provides a dian wall target guidance method and system combining visual detection of a multi-rotor unmanned aerial vehicle, which are characterized in that firstly, dian wall target area visual information acquired by a multi-rotor unmanned aerial vehicle visual component is acquired and comprises target appearance characteristics and surrounding environment characteristics, then, scene analysis is carried out on dian wall target area visual information to obtain dian wall target spatial distribution characteristics and environment association characteristics, a plurality of groups of initial guidance action parameters are pre-generated based on the spatial distribution characteristics and the environment association characteristics, flight attitude and track parameters are covered, the initial parameters are dynamically adjusted according to real-time change of visual information, target guidance action parameters adapting to current information are obtained, and finally, guidance instructions are generated according to the target guidance action parameters and transmitted to a flight control component to drive guidance action execution, so that accurate guidance of the multi-rotor unmanned aerial vehicle in dian wall scenes is realized.

Inventors

• LI YIFAN
• YANG XU

Assignees

• 戎御科技(太仓)有限公司

Dates

Publication Date

20260512

Application Date

20260205

Claims (10)

1. 1. A dian wall target guidance method in combination with visual inspection of a multi-rotor unmanned aerial vehicle, the method comprising: the method comprises the steps of collecting dian wall target area visual information obtained by a multi-rotor unmanned aerial vehicle visual component, wherein the dian wall target area visual information comprises appearance characteristics of a dian wall target and environmental characteristics of a dian wall target surrounding environment; Performing scene analysis on the dian wall target area visual information to obtain spatial distribution characteristics and environment association characteristics of dian wall targets, wherein the spatial distribution characteristics reflect the position and morphological relationship of dian wall targets in a three-dimensional space, and the environment association characteristics reflect the influence relationship of surrounding environments on guidance actions; generating a plurality of groups of initial guidance action parameters in advance based on the spatial distribution characteristics and the environment association characteristics, wherein the initial guidance action parameters comprise flight attitude parameters and flight trajectory parameters of the multi-rotor unmanned aerial vehicle; Dynamically adjusting the initial guidance action parameters according to the real-time change of the dian wall target area visual information to obtain target guidance action parameters adapting to the current visual information; And generating dian wall target guidance instructions of the multi-rotor unmanned aerial vehicle according to the target guidance action parameters, and transmitting the dian wall target guidance instructions to a flight control assembly of the multi-rotor unmanned aerial vehicle so as to drive the guidance action to be executed.
2. 2. The method for dian wall target guidance in combination with vision inspection of a multi-rotor unmanned aerial vehicle according to claim 1, wherein the performing scene analysis on the dian wall target area vision information to obtain the spatial distribution feature and the environment association feature of the dian wall target comprises: Separating dian the wall target region and the background region from the dian wall target region visual information, and extracting outline edge features and surface texture features of the dian wall target region; Constructing dian a three-dimensional morphological model of the wall target based on the contour edge characteristics, determining dian a convex area position, a concave area position and a plane area range of the wall target through the three-dimensional morphological model, and integrating the convex area position, the concave area position and the plane area range into a space distribution characteristic of the dian wall target; Extracting environmental illumination change characteristics and airflow disturbance characteristics from the dian wall target area visual information, wherein the environmental illumination change characteristics reflect the influence of different illumination intensities on the visual detection result, and the airflow disturbance characteristics reflect the potential influence of surrounding airflows on the flight attitude of the multi-rotor unmanned aerial vehicle; analyzing the association relation between the environment illumination change characteristics and the spatial distribution characteristics, determining the influence degree of illumination intensity change on the spatial distribution characteristic identification accuracy, and generating illumination association sub-characteristics; Analyzing the association relation between the airflow disturbance characteristics and the spatial distribution characteristics, determining the influence degree of the airflow disturbance intensity on the target track of the multi-rotor unmanned aerial vehicle approaching dian walls, and generating airflow association sub-characteristics; integrating the illumination correlation sub-feature and the airflow correlation sub-feature into an environment correlation feature of dian wall targets, and enabling the environment correlation feature and the space distribution feature to be matched in dimension through dimension calibration operation.
3. 3. The method for dian wall target guidance in combination with vision inspection of a multi-rotor unmanned aerial vehicle according to claim 2, wherein separating dian wall target area and background area from the dian wall target area vision information, extracting contour edge features and surface texture features of the dian wall target area, comprises: acquiring an image pixel set corresponding to the dian wall target area visual information, wherein the image pixel set comprises color characteristics and brightness characteristics of each pixel; analyzing the color feature difference of each pixel in the image pixel set, and determining the pixel range of which the color feature belongs to dian wall targets and the pixel range of which the color feature belongs to background areas; Analyzing the brightness characteristic difference of each pixel in the image pixel set, and determining the pixel range of which the brightness characteristic belongs to a dian wall target and the pixel range of which the brightness characteristic belongs to a background area; combining the pixel range divided by the color features and the pixel range divided by the brightness features, determining a pixel set which simultaneously meets dian wall target color features and brightness features, and taking the region corresponding to the pixel set as a preliminary dian wall target region; Taking the area except the preliminary dian wall target area in the image pixel set as a preliminary background area; analyzing boundary pixels of the preliminary dian wall target area and the preliminary background area, judging whether color features and brightness features of the boundary pixels are closer to the dian wall target area or the background area, and adjusting the boundary between the dian wall target area and the background area according to a judging result to obtain a final dian wall target area and the background area; extracting a pixel sequence at the boundary of a final dian wall target area, and generating dian contour edge features of the wall target area according to the position change rule of the pixel sequence, wherein the contour edge features comprise the bending trend and the turning position of the boundary; And extracting a final pixel distribution rule in the dian wall target area, analyzing color transition characteristics and brightness change characteristics between adjacent pixels, and generating surface texture characteristics of the dian wall target area according to the color transition characteristics and the brightness change characteristics, wherein the surface texture characteristics comprise the density degree and trend of textures.
4. 4. The method for dian wall target guidance in combination with visual inspection of a multi-rotor unmanned aerial vehicle according to claim 3, wherein the extracting the pixel distribution rule in the final dian wall target area, analyzing the color transition feature and the brightness change feature between adjacent pixels, and generating the surface texture feature of the dian wall target area according to the color transition feature and the brightness change feature comprises: Dividing a final dian wall target area into a plurality of equal-size pixel subareas, wherein each pixel subarea comprises a preset number of pixels; analyzing pixels in each pixel subarea, extracting the color value and the brightness value of each pixel, and establishing a color value set and a brightness value set of each pixel subarea; Calculating the color value difference between adjacent pixels in each pixel sub-area, and determining the color transition characteristics in the pixel sub-area according to the size and the change frequency of the color value difference, wherein the color transition characteristics comprise transition smoothness and transition direction; Calculating the brightness value difference between adjacent pixels in each pixel sub-area, and determining the brightness change characteristic in the pixel sub-area according to the brightness value difference and the change frequency, wherein the brightness change characteristic comprises a change amplitude and a change period; Counting the color transition characteristics of all the pixel subareas, analyzing the similarity of the color transition characteristics among different pixel subareas, classifying the color transition characteristics with the similarity meeting the preset standard into the same category, and determining the main type of the color transition of the surface of the dian wall target area; counting brightness change characteristics of all pixel subareas, analyzing similarity of the brightness change characteristics among different pixel subareas, classifying the brightness change characteristics with similarity meeting a preset standard into the same category, and determining a main type of dian wall target area surface brightness change; Calculating the area proportion of each type in the whole dian wall target area according to the main type of color transition of the surface of the dian wall target area, and taking the area proportion as a color transition distribution parameter; Calculating the area proportion of each type in the whole dian wall target area according to the main type of the change of the brightness of the surface of the dian wall target area, and taking the area proportion as a brightness change distribution parameter; And integrating the color transition characteristics, the brightness change characteristics, the color transition distribution parameters and the brightness change distribution parameters to generate dian surface texture characteristics of the wall target area, wherein the surface texture characteristics can comprehensively reflect the texture form and the distribution rule of the dian wall target surface.
5. 5. The method for dian wall target guidance in combination with vision inspection of a multi-rotor unmanned aerial vehicle of claim 1, wherein the pre-generating of the plurality of sets of initial guidance motion parameters based on the spatial distribution feature and the environmental correlation feature comprises: Extracting the positions of the raised areas and the positions of the recessed areas in the spatial distribution characteristics, and determining the range of an obstacle area to be avoided and the range of a safe area capable of flying stably when the multi-rotor unmanned aerial vehicle approaches to a dian wall target; According to the obstacle area range and the safety area range, preliminarily setting a flying height range and a horizontal flying direction range of the multi-rotor unmanned aerial vehicle as basic ranges of initial flying track parameters; Extracting illumination correlation sub-features in the environment correlation features, determining effective detection distances of the multi-rotor unmanned aerial vehicle vision components under different illumination intensities, and adjusting a preliminarily set flight height range and a horizontal flight direction range according to the effective detection distances to obtain flight track sub-parameters adapting to illumination conditions; extracting airflow-related sub-features in the environment-related features, determining attitude stability thresholds of the multi-rotor unmanned aerial vehicle under different airflow disturbance intensities, and setting a rotor rotation speed adjusting range and a fuselage inclination angle range of the multi-rotor unmanned aerial vehicle according to the attitude stability thresholds to serve as a basic range of initial flight attitude parameters; Performing association matching on the flight trajectory sub-parameters adapting to the illumination conditions and the basic range of the initial flight attitude parameters to generate a first group of initial guidance action parameters, wherein the flight trajectory parameters and the flight attitude parameters in the first group of initial guidance action parameters are mutually adapted to correspond to the current illumination and airflow conditions; Adjusting the dividing standard of the obstacle area range and the safety area range, resetting the flying height range and the horizontal flying direction range, and generating a second group of initial guidance action parameters by combining the adjusted illumination-related sub-features and the airflow-related sub-features; And repeatedly adjusting the dividing standard of the obstacle area range and the safety area range and the analysis dimensionalities of the illumination-related sub-feature and the airflow-related sub-feature to generate at least three groups of initial guidance action parameters, wherein each group of initial guidance action parameters corresponds to different environment adaptation scenes and target approaching strategies.
6. 6. The method for achieving dian wall target guidance in combination with visual inspection of a multi-rotor unmanned aerial vehicle according to claim 5, wherein the preliminarily setting a flying height range and a horizontal flying direction range of the multi-rotor unmanned aerial vehicle according to the obstacle area range and the safety area range as basic ranges of initial flying track parameters comprises: Mapping the obstacle region range into a flight coordinate system of the multi-rotor unmanned aerial vehicle, and determining a three-dimensional coordinate range of the obstacle region in the flight coordinate system, wherein the flight coordinate system takes the current position of the multi-rotor unmanned aerial vehicle as an origin, the vertical direction is a Z axis, and the horizontal direction is an X axis and a Y axis; Mapping the safety area range into a flight coordinate system of the multi-rotor unmanned aerial vehicle, and determining a three-dimensional coordinate range of the safety area in the flight coordinate system; Analyzing a Z-axis coordinate range of an obstacle region in a flight coordinate system, and determining a flight height section to be avoided of the multi-rotor unmanned aerial vehicle, wherein the flight height section to be avoided is a section corresponding to the Z-axis coordinate range of the obstacle region; Analyzing a Z-axis coordinate range of a safety area in a flight coordinate system, determining a flyable height interval of the multi-rotor unmanned aerial vehicle, and taking the flyable height interval as a basis of a preliminarily set flight height range; Analyzing the X-axis and Y-axis coordinate positions of the obstacle region in a flight coordinate system, calculating horizontal direction angles from the current position of the multi-rotor unmanned aerial vehicle to each point of the obstacle region, and determining a horizontal flight direction interval to be avoided by the multi-rotor unmanned aerial vehicle according to the distribution range of the horizontal direction angles, wherein the horizontal flight direction interval to be avoided is an interval corresponding to the distribution range of the horizontal direction angles; Analyzing the X-axis and Y-axis coordinate positions of the safety area in a flight coordinate system, calculating the horizontal direction angles from the current position of the multi-rotor unmanned aerial vehicle to each point of the safety area, determining the horizontal direction interval which can be flown by the multi-rotor unmanned aerial vehicle according to the distribution range of the horizontal direction angles, and taking the horizontal direction interval which can be flown as the basis of the preliminarily set horizontal flight direction range; Combining the maximum flying height limit and the minimum flying height limit of the multi-rotor unmanned aerial vehicle, adjusting the basis of the preliminarily set flying height range, comparing the basis of the preliminarily set flying height range with the maximum flying height limit and the minimum flying height limit in the adjusting process, reserving a height section within the limiting range, and removing the height section exceeding any limiting range; Combining the maximum horizontal flight direction adjustment angle limit of the multi-rotor unmanned aerial vehicle, adjusting the basis of the primarily set horizontal flight direction range, comparing the primarily set horizontal flight direction range basis with the maximum horizontal flight direction adjustment angle limit in the adjustment process, reserving a direction interval within the limit range, and removing the direction interval exceeding the limit range; And integrating the adjusted flying height range and the adjusted horizontal flying direction range to serve as a basic range of initial flying track parameters, wherein the basic range of the initial flying track parameters can enable the multi-rotor unmanned aerial vehicle to approach dian wall targets through a safety area while avoiding an obstacle area.
7. 7. The method for achieving dian wall target guidance in combination with vision inspection of multi-rotor unmanned aerial vehicle according to claim 1, wherein dynamically adjusting the initial guidance action parameter according to the real-time change of the dian wall target area vision information to obtain the target guidance action parameter adapted to the current vision information comprises: Acquiring the updated dian wall target area visual information of the multi-rotor unmanned aerial vehicle visual component in real time, comparing the updated dian wall target area visual information with dian wall target area visual information at the previous moment, and extracting visual information change characteristics; The visual information change features comprise dian offset features of the wall target contour edge, definition change features of surface textures, fluctuation features of ambient illumination intensity and image blurring features generated by air flow disturbance; judging the position change trend of the dian wall target relative to the multi-rotor unmanned aerial vehicle according to dian wall target outline edge offset characteristics in the visual information change characteristics, and determining whether the horizontal flight direction parameters in the initial guidance action parameters need to be adjusted or not; If the horizontal flight direction parameters are judged to be adjusted, calculating the adjustment angle of the horizontal flight direction based on the offset direction and the offset amplitude of the profile edge offset characteristic, and correcting the horizontal flight direction parameters in the initial guidance action parameters; Judging the distance change trend of the multi-rotor unmanned aerial vehicle and dian wall targets according to the surface texture definition change characteristics in the visual information change characteristics, and determining whether the flight height parameters and the flight speed parameters in the initial guidance action parameters need to be adjusted or not; If the flying height parameter and the flying speed parameter are judged to be required to be adjusted, calculating the adjusting amplitude of the flying height and the adjusting speed of the flying speed based on the change degree of the surface texture definition, and correcting the flying height parameter and the flying speed parameter in the initial guidance action parameters; according to the environmental illumination intensity fluctuation characteristics in the visual information change characteristics and the illumination correlation sub-characteristics in the environment correlation characteristics, judging the influence change of illumination conditions on visual detection precision, and determining whether the flight attitude stability parameters in the initial guidance action parameters need to be adjusted; if the flight attitude stability parameters are judged to be required to be adjusted, calculating the adjustment amplitude of the rotating speed of the rotor wing based on the frequency and the amplitude of the illumination intensity fluctuation, and correcting the flight attitude stability parameters in the initial guidance action parameters; According to the image fuzzy characteristic generated by the air flow disturbance in the visual information change characteristic, and combining the air flow correlation sub-characteristic in the environment correlation characteristic, judging the change trend of the air flow disturbance intensity, and determining whether the body inclination angle parameter in the initial guidance action parameter needs to be adjusted; If the machine body inclination angle parameter is judged to be required to be adjusted, calculating an adjustment range of the machine body inclination angle based on the blurring degree and the duration time of the image blurring characteristic, and correcting the machine body inclination angle parameter in the initial guidance action parameter; And integrating the corrected flight trajectory parameters with the flight attitude parameters to obtain target guidance action parameters adapting to the current visual information, wherein the target guidance action parameters can enable the multi-rotor unmanned aerial vehicle to stably approach dian wall targets in a scene corresponding to the current visual information.
8. 8. The method for dian wall target guidance in combination with a visual inspection multi-rotor unmanned aerial vehicle according to claim 7, wherein the determining dian wall target position change trend relative to the multi-rotor unmanned aerial vehicle according to dian wall target contour edge offset feature in the visual information change feature, determining whether to adjust a horizontal flight direction parameter in initial guidance motion parameters comprises: extracting dian wall target contour edge offset features in the visual information change features, wherein the contour edge offset features comprise X-axis offset and Y-axis offset of the contour edge in an image coordinate system; Establishing a mapping relation between an image coordinate system and a multi-rotor unmanned aerial vehicle flight coordinate system, and converting X-axis offset and Y-axis offset of a contour edge in the image coordinate system into X-axis offset and Y-axis offset in the flight coordinate system; analyzing the change rule of the X-axis offset in the converted flight coordinate system, and judging dian a position change trend of the wall target relative to the multi-rotor unmanned aerial vehicle in the X-axis direction of the flight coordinate system, wherein if the X-axis offset is continuously increased, the wall target is indicated dian to be far away from the preset approaching direction of the multi-rotor unmanned aerial vehicle in the X-axis direction; analyzing a Y-axis offset change rule in the converted flight coordinate system, and judging dian a position change trend of the wall target relative to the multi-rotor unmanned aerial vehicle in the Y-axis direction of the flight coordinate system, wherein if the Y-axis offset is continuously increased, the wall target is indicated dian to be far away from the preset approaching direction of the multi-rotor unmanned aerial vehicle in the Y-axis direction; If dian the wall target is far away from the preset approaching direction of the multi-rotor unmanned aerial vehicle in the X-axis direction or the Y-axis direction, and the change rate of the offset exceeds a preset trend threshold, determining that the horizontal flight direction parameter in the initial guidance action parameters needs to be adjusted so as to correct the flight direction of the multi-rotor unmanned aerial vehicle, so that the multi-rotor unmanned aerial vehicle can continue to fly along the direction approaching to the dian wall target; If dian wall targets are close to the preset approaching direction of the multi-rotor unmanned aerial vehicle in the X-axis direction and the Y-axis direction, or the change rate of the offset does not exceed a preset trend threshold, determining that the horizontal flight direction parameter in the initial guidance action parameters does not need to be adjusted, maintaining the current horizontal flight direction parameter, enabling the multi-rotor unmanned aerial vehicle to approach the dian wall target, and recording the current state of the parameter when maintaining the current parameter.
9. 9. The dian wall target guidance method for a multi-rotor unmanned aerial vehicle in combination with vision inspection as claimed in claim 1, wherein the generating dian wall target guidance command for a multi-rotor unmanned aerial vehicle based on the target guidance motion parameters, transmitting the dian wall target guidance command to a flight control assembly of the multi-rotor unmanned aerial vehicle to drive the guidance motion to perform comprises: Extracting flight track parameters in the target guidance action parameters, wherein the flight track parameters comprise flight height parameters, horizontal flight direction parameters and flight speed parameters; converting the flying height parameter into a lifting force adjusting instruction of a rotor wing of the multi-rotor unmanned aerial vehicle, wherein the lifting force adjusting instruction can control the lifting force output by the rotor wing to maintain or adjust the flying height; Converting the horizontal flight direction parameters into direction control instructions of the multi-rotor unmanned aerial vehicle, wherein the direction control instructions can control the rotation speed difference of different rotors so as to adjust the flight direction; Converting the flying speed parameter into a power adjusting instruction of the multi-rotor unmanned aerial vehicle, wherein the power adjusting instruction can control the rotating speed level of the whole rotor to adjust the flying speed; Extracting flight attitude parameters in the target guidance action parameters, wherein the flight attitude parameters comprise rotor rotation speed adjusting range parameters, fuselage inclination angle parameters and attitude stability threshold parameters; converting the rotor rotation speed adjusting range parameter into a rotation speed limiting instruction of the multi-rotor unmanned aerial vehicle, wherein the rotation speed limiting instruction can limit the maximum rotation speed and the minimum rotation speed of each rotor so as to enable the gesture to be stable; Converting the fuselage inclination angle parameters into attitude adjustment instructions of the multi-rotor unmanned aerial vehicle, wherein the attitude adjustment instructions can control balance components at all parts of the fuselage to maintain or adjust the fuselage inclination angle; converting the gesture stability threshold parameter into a gesture monitoring instruction of the multi-rotor unmanned aerial vehicle, wherein the gesture monitoring instruction can set an allowable range of gesture deviation to trigger subsequent gesture fine adjustment; Integrating the lift force adjusting instruction, the direction control instruction, the power adjusting instruction, the rotating speed limiting instruction, the gesture adjusting instruction and the gesture monitoring instruction to generate dian wall target guidance instructions of the multi-rotor unmanned aerial vehicle, wherein the execution time sequence of each sub-instruction in the dian wall target guidance instructions is matched with the flight phase of the multi-rotor unmanned aerial vehicle, the execution sequence of each sub-instruction is determined through time sequence arrangement operation during integration, and the time sequence arrangement operation refers to the task requirements of the multi-rotor unmanned aerial vehicle in different flight phases and prioritizes the sub-instructions related to the core tasks of the current phase; And establishing a communication link between the dian wall target guidance instruction and the multi-rotor unmanned aerial vehicle flight control assembly, and transmitting the dian wall target guidance instruction to the flight control assembly through the communication link so that after the flight control assembly receives the dian wall target guidance instruction, analyzing the content and the execution time sequence of each sub instruction, and driving a corresponding flight execution component to execute actions according to instruction requirements to realize dian wall target guidance of the multi-rotor unmanned aerial vehicle.
10. 10. A dian wall target guidance system for a combined vision inspection multi-rotor unmanned aerial vehicle, the dian wall target guidance system for a combined vision inspection multi-rotor unmanned aerial vehicle comprising a processor and a memory, the memory being coupled to the processor, the memory being configured to store programs, instructions or code, the processor being configured to execute the programs, instructions or code in the memory to implement the dian wall target guidance method for a combined vision inspection multi-rotor unmanned aerial vehicle as claimed in any one of claims 1 to 9.

Description

Dian wall target guidance method and system combined with visual inspection of multi-rotor unmanned aerial vehicle Technical Field The invention relates to the technical field of computer vision, in particular to a dian wall target guidance method and system for a multi-rotor unmanned aerial vehicle with combination of vision detection. Background In the technical field of the existing multi-rotor unmanned aerial vehicle target guidance, a plurality of limitations exist in a guidance method aiming at a target close to a wall body (dian walls). Conventional guidance methods mostly rely on single sensor data, for example, only relying on distance sensors to obtain distance information between the drone and the target for guidance. However, in dian wall scenarios, distance information alone is far from sufficient to achieve accurate guidance. The environment of dian wall target area is complicated changeable, and the factors such as barrier distribution, illumination condition change and the material and the shape of wall all can all produce apparent influence to unmanned aerial vehicle's guidance. The existing guidance method does not fully consider the appearance characteristics of dian wall targets and the environmental characteristics of the surrounding environment, cannot accurately analyze the position and morphological relationship of dian wall targets in a three-dimensional space, and cannot effectively evaluate the influence of the surrounding environment on the guidance action. In the guiding process, the unmanned aerial vehicle is difficult to dynamically adjust the guiding action according to the scene which changes in real time, the guiding deviation is easy to occur, the aim of dian walls cannot be accurately reached, and the accuracy and the reliability of guiding are reduced. Disclosure of Invention In view of the above-mentioned problems, in combination with the first aspect of the present invention, an embodiment of the present invention provides a dian wall target guidance method in combination with a visual inspection of a multi-rotor unmanned aerial vehicle, the method including: the method comprises the steps of collecting dian wall target area visual information obtained by a multi-rotor unmanned aerial vehicle visual component, wherein the dian wall target area visual information comprises appearance characteristics of a dian wall target and environmental characteristics of a dian wall target surrounding environment; Performing scene analysis on the dian wall target area visual information to obtain spatial distribution characteristics and environment association characteristics of dian wall targets, wherein the spatial distribution characteristics reflect the position and morphological relationship of dian wall targets in a three-dimensional space, and the environment association characteristics reflect the influence relationship of surrounding environments on guidance actions; generating a plurality of groups of initial guidance action parameters in advance based on the spatial distribution characteristics and the environment association characteristics, wherein the initial guidance action parameters comprise flight attitude parameters and flight trajectory parameters of the multi-rotor unmanned aerial vehicle; Dynamically adjusting the initial guidance action parameters according to the real-time change of the dian wall target area visual information to obtain target guidance action parameters adapting to the current visual information; And generating dian wall target guidance instructions of the multi-rotor unmanned aerial vehicle according to the target guidance action parameters, and transmitting the dian wall target guidance instructions to a flight control assembly of the multi-rotor unmanned aerial vehicle so as to drive the guidance action to be executed. In still another aspect, an embodiment of the present invention further provides a dian wall target guidance system that incorporates a visual inspection multi-rotor unmanned aerial vehicle, including a processor, a machine-readable storage medium, where the machine-readable storage medium is connected to the processor, and the machine-readable storage medium is used to store a program, an instruction, or a code, and the processor is used to execute the program, the instruction, or the code in the machine-readable storage medium, so as to implement the method described above. Based on the above aspects, the embodiment of the invention acquires the dian wall target area visual information acquired by the multi-rotor unmanned aerial vehicle visual component, covers the appearance characteristics of the dian wall target and the environmental characteristics of the surrounding environment, carries out scene analysis on the dian wall target area visual information to obtain the spatial distribution characteristics and the environmental correlation characteristics of the dian wall target, can deeply understand the position and the