AI inspection unmanned aerial vehicle control method and system based on real-time dynamic measurement technology

Abstract

The application provides an AI inspection unmanned aerial vehicle control method and system based on real-time dynamic measurement technology, and relates to the technical field of unmanned aerial vehicle inspection. In the case of stable unmanned aerial vehicle body, the method periodically acquires a multi-dimensional environment state vector in the unmanned aerial vehicle inspection process, and determines an environment stability score accordingly. Based on the environment stability score, it is determined whether to trigger an adjustment mechanism for the first configuration parameter of the camera, and in the case of triggering, based on the preset second configuration parameter, the environment stability score and the target distance of the unmanned aerial vehicle from the shooting target, the target focal length value and the target exposure time of the camera are determined, and finally the camera is controlled to shoot the inspection image. Therefore, by calculating the multi-dimensional environment state vector in real time, the environment stability is dynamically evaluated, and the intelligent parameter adjustment mechanism is triggered based on the environment stability score, the adaptive optimization of focal length and exposure time is realized, and the inspection image quality is improved.

Description

Technical Field

[0001] This application relates to the field of drone inspection technology, and in particular to an AI inspection drone control method and system based on real-time dynamic measurement technology. Background Technology

[0002] As a core tool for modern industrial monitoring and infrastructure maintenance, drone inspection technology plays an irreplaceable role in key areas such as power line inspection, pipeline monitoring, and building inspection. With the continuous improvement of industrial automation, the requirements for inspection accuracy and efficiency are becoming increasingly stringent, making intelligent inspection technology based on industrial monitoring software platforms a crucial support for ensuring safe industrial operation.

[0003] Current drone inspection solutions generally employ a static observation mode with preset parameters. Key parameters such as camera focal length and exposure time are fixed before flight, making dynamic monitoring and control impossible based on actual environmental changes. This fixed parameter configuration leads to a significant deterioration in image quality when lighting conditions change, target distance shifts, or environmental interference intensifies, greatly reducing inspection effectiveness. Furthermore, existing solutions and their industrial monitoring software lack real-time environmental perception capabilities, making it difficult to cope with complex and ever-changing operational scenarios. Summary of the Invention

[0004] This application provides an AI inspection drone control method and system based on real-time dynamic measurement technology, which can improve the quality of inspection images, thereby improving inspection efficiency and reliability.

[0005] The first aspect of this application provides an AI inspection drone control method based on real-time dynamic measurement technology, the method comprising:

[0006] With the drone body stable, a multi-dimensional environmental state vector is periodically acquired during the drone inspection process; the multi-dimensional environmental state vector includes the rate of change of light intensity, the degree of environmental noise interference, and the fluctuation value of the distance between the drone and the target being photographed.

[0007] An environmental stability score is determined based on the rate of change of light intensity, the fluctuation value of the target distance, and the degree of environmental noise interference.

[0008] Based on the environmental stability score, it is determined whether to trigger an adjustment mechanism for the camera's first configuration parameters; wherein the first configuration parameters include focal length and exposure time.

[0009] If it is determined that an adjustment mechanism for the first configuration parameter of the camera is triggered, the target focal length value and target exposure time of the camera are determined based on the preset second configuration parameter, the environmental stability score and the target distance between the UAV and the target being photographed.

[0010] The camera is controlled to perform a shooting action according to the target focal length value and the target exposure time to obtain the inspection image corresponding to the shooting target.

[0011] Optionally, the method further includes:

[0012] The initial attitude angle offset of the UAV is obtained; the initial attitude angle offset is obtained by combining the roll angle, pitch angle and yaw angle of the UAV.

[0013] If the initial attitude angle offset is greater than the offset threshold, the flight strategy of the UAV is adjusted.

[0014] Obtain the attitude angle offset and target distance fluctuation value after adjusting the flight strategy;

[0015] If the attitude angle offset is less than the offset threshold and the variance of the target distance fluctuation is less than the fluctuation threshold, the UAV fuselage is determined to be stable.

[0016] Optionally, an environmental stability score is determined based on the rate of change of light intensity, the target distance fluctuation value, and the environmental noise interference level, including:

[0017] An initial environmental stability score is obtained by weighting the initial weights corresponding to the light intensity change rate, the target distance fluctuation value, and the environmental noise interference degree.

[0018] The initial environmental stability score is corrected based on the stability weighting coefficient to obtain the corrected environmental stability score;

[0019] Based on the environmental stability score, determine whether to trigger an adjustment mechanism for the first configuration parameter of the camera, including:

[0020] If the environmental stability score is less than a steady-state threshold, an adjustment mechanism for the first configuration parameter of the camera is triggered; wherein the steady-state threshold is dynamically determined based on the historical environmental stability score collected by the UAV during historical inspections.

[0021] Optionally, based on preset second configuration parameters, the environmental stability score, and the target distance between the UAV and the target, the target focal length value and target exposure time of the camera are determined, including:

[0022] Based on the target distance between the UAV and the target being photographed, the imaging impact factor is determined;

[0023] Based on the imaging impact factor and the environmental stability score, a first illumination compensation value is determined;

[0024] The target exposure time of the camera is determined based on the preset second configuration parameters and the first illumination compensation value;

[0025] Based on a preset focal length distance mapping table, the target focal length value of the camera is determined; wherein, the focal length distance mapping table is used to characterize the correspondence between the distance of the UAV from the target and the focal length of the camera.

[0026] Optionally, the second configuration parameters include focal length adjustment step size and exposure time range;

[0027] After obtaining the inspection image corresponding to the photographed target, the method further includes:

[0028] Determine the image quality parameters of the inspection image; the image quality parameters include sharpness value, brightness distribution value, and contrast value;

[0029] Based on the sharpness value and the brightness distribution value, determine whether to trigger the optimization mechanism for the second configuration parameter;

[0030] If it is determined that an optimization mechanism for the second configuration parameter is triggered, the image quality score of the inspection image is determined based on the sharpness value, the brightness distribution value, and the contrast value.

[0031] Based on the image quality score, the focal length adjustment step size and the exposure time interval are optimized to obtain the optimized step size and optimized exposure interval.

[0032] Optionally, the inspection image includes multiple images; based on the sharpness value and the brightness distribution value, determining whether to trigger the optimization mechanism for the second configuration parameter includes:

[0033] Determine the volatility of the sharpness values ​​of multiple images, and if the volatility is greater than a change threshold, determine whether to trigger an optimization mechanism for the focal length adjustment step size.

[0034] Images whose brightness distribution values ​​are not concentrated within a preset brightness range are identified as poor-quality images;

[0035] If the proportion of poor-quality images to the total number of images is greater than a certain threshold, an optimization mechanism for the exposure time interval is triggered.

[0036] Optionally, based on the image quality score, the focal length adjustment step size is optimized to obtain an optimized step size, including:

[0037] The average quality score is obtained by determining the average quality score of the multiple images.

[0038] Based on the target distance between the UAV and the photographed target during the acquisition of the inspection images, the imaging influence factor is determined;

[0039] The optimized step size is determined based on the initial step size, the difference between the average quality score and the standard quality score, and the imaging influence factor.

[0040] Optionally, based on the image quality score, the exposure time interval is optimized to obtain an optimized exposure interval, including:

[0041] Based on the difference between the average quality score and the standard quality score, a second illumination compensation value is determined using a preset illumination compensation model.

[0042] The target exposure time is corrected based on the second illumination compensation value to obtain the corrected target exposure time;

[0043] The optimized exposure range is determined based on the corrected target exposure time and the optimized step size.

[0044] Optionally, after optimizing the focal length adjustment step size and the exposure time interval based on the image quality score to obtain an optimized step size and an optimized exposure interval, the method further includes:

[0045] Construct a database linking the shooting environment and camera configuration. Any historical data in the database includes the light intensity when the inspection image was captured, the target distance between the UAV and the target being photographed, the optimization step size, and the optimization exposure range.

[0046] The method further includes:

[0047] Obtain environmental parameters of the current shooting environment, including the current light intensity and the current target distance;

[0048] Determine one or more historical related data that match the environmental parameters from the related database;

[0049] Based on the optimization step size and the optimization exposure interval of one or more of the historical associated data, the target focal length adjustment step size and the target exposure time interval are determined.

[0050] Based on the target focal length adjustment step size and the target exposure time interval, the preset second configuration parameters are obtained.

[0051] Based on the same inventive concept, a second aspect of the present application provides an AI inspection drone control system based on real-time dynamic measurement technology, the system comprising:

[0052] A light sensor is used to collect the rate of change in light intensity during drone inspections.

[0053] An acoustic sensor is used to collect the environmental noise interference level during the UAV inspection process;

[0054] A distance measuring device is used to collect the target distance fluctuation value between the UAV and the target being photographed;

[0055] A camera is used to capture inspection images corresponding to the target being photographed;

[0056] A flight controller for executing the AI ​​inspection drone control method based on real-time dynamic measurement technology as proposed in the first aspect of this application.

[0057] Based on the same inventive concept, a third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When executed by the processor, the program implements the AI ​​inspection drone control method based on real-time dynamic measurement technology proposed in the first aspect of this application.

[0058] Compared with the prior art, this application has the following advantages:

[0059] This application provides an AI inspection drone control method based on real-time dynamic measurement technology. Under stable drone conditions, a multi-dimensional environmental state vector is periodically acquired during the drone's inspection process. This vector includes the rate of change of light intensity, environmental noise interference, and the fluctuation value of the distance between the drone and the target being photographed. An environmental stability score is determined based on these parameters. The environmental stability score is then used to determine whether to trigger an adjustment mechanism for the camera's first configuration parameters, including focal length and exposure time. If the adjustment mechanism is triggered, the target focal length and exposure time of the camera are determined based on preset second configuration parameters, the environmental stability score, and the distance between the drone and the target. The camera is then controlled to perform a shooting action according to the target focal length and exposure time to obtain the inspection image corresponding to the target. Therefore, by calculating the multi-dimensional environmental state vector in real time, the environmental stability is dynamically evaluated, and an intelligent parameter adjustment mechanism is triggered based on the environmental stability score to achieve adaptive optimization of focal length and exposure time, thereby improving the quality of inspection images. It can be applied to scenarios that require high-precision image acquisition, such as power inspection, pipeline inspection, and building monitoring, thereby improving inspection efficiency and reliability. Attached Figure Description

[0060] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0061] Figure 1 This is a flowchart of an AI inspection drone control method based on real-time dynamic measurement technology in one embodiment of this application;

[0062] Figure 2 This is a flowchart of camera configuration parameter optimization based on image quality in one embodiment of this application;

[0063] Figure 3 This is a flowchart of setting the second configuration parameter in one embodiment of this application;

[0064] Figure 4 This is a schematic diagram of the structure of an AI inspection drone control method based on real-time dynamic measurement technology in one embodiment of this application;

[0065] Figure 5 This is a schematic diagram of the structure of an electronic device according to one embodiment of this application. Detailed Implementation

[0066] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0067] The fundamental challenge of drone inspection lies in the dynamic changes in environmental conditions and target characteristics. During inspection, environmental factors such as light intensity, weather conditions, and target distance are constantly changing, while different inspection targets possess unique material, shape, and reflective properties. This dual dynamism makes fixed observation parameters unsuitable for all scenarios. The dynamic changes in environment and target further complicate the optimization of observation parameters. Traditional parameter adjustment methods cannot quickly determine the optimal combination of multiple parameters such as camera focal length and exposure time under limited computational resources and time constraints. The lag and inaccuracy of observation parameter configuration directly affect data acquisition quality, rendering subsequent defect identification and condition assessment unreliable.

[0068] Therefore, how to construct an intelligent monitoring and control mechanism that can perceive changes in environmental conditions and target characteristics in real time and dynamically optimize the configuration of observation parameters accordingly, and integrate it into an advanced industrial monitoring software platform to achieve adaptive assurance of data acquisition quality during the inspection process, has become a key issue in improving the performance of UAV inspection systems.

[0069] Therefore, this application proposes an AI inspection drone control method based on real-time dynamic measurement technology. Please refer to [link / reference] for details. Figure 1 , Figure 1 This is a flowchart illustrating an AI inspection drone control method based on real-time dynamic measurement technology, as proposed in an embodiment of this application. Figure 1 As shown, the method includes the following steps:

[0070] S101: Under stable conditions, periodically acquire multi-dimensional environmental state vectors during the drone inspection process; the multi-dimensional environmental state vectors include the rate of change of light intensity, environmental noise interference, and the fluctuation value of the distance between the drone and the target being photographed.

[0071] In this embodiment, during the UAV inspection process, it is first necessary to maintain fuselage stability to ensure flight safety, and then execute the planned inspection tasks. Flotation stability refers to the UAV being in a hovering state.

[0072] When the drone's fuselage is stable, a multi-dimensional environmental state vector is further acquired during the inspection process. This multi-dimensional environmental state vector reflects the stability of the current flight environment. Specifically, the multi-dimensional environmental state vector includes: the rate of change of light intensity, the degree of environmental noise interference, and the fluctuation value of the distance between the drone and the target being photographed.

[0073] The rate of change of light intensity refers to the magnitude of change in ambient light intensity per unit time. For example, ambient light intensity data can be collected by a light sensor mounted on a drone at a preset sampling frequency. For instance, if the light intensity changes from 500 lux to 800 lux in 30 seconds, the rate of change of light intensity can be calculated as (800-500) / 30 = 10 lux / s. However, the rate of change of light intensity may change abruptly, for example, when suddenly obscured by clouds.

[0074] Furthermore, the rate of change of light intensity reflects the dynamic fluctuations of lighting conditions, that is, the stability of lighting conditions. For example, assuming the light sensor samples 10 times per second and continuously collects data for 60 seconds during a flight mission, 600 light intensity data points are obtained. These 600 light intensity data points are then analyzed using a time series analysis algorithm to understand the trend of the rate of change of light intensity. For instance, setting the sliding window size to 6 seconds and sliding it 10 times over the 600 light intensity data points, calculating the rate of change of light intensity using 6 data points within the sliding window each time, yields 10 rates of change of light intensity. Then, window evaluation can be performed: for example, if the rate of change of light intensity exceeds 2.0 lux / s in 8 of the windows, it indicates that the current lighting conditions are unstable.

[0075] Environmental noise interference level refers to the degree of interference that environmental noise causes to the drone's sensors and communication links, used to quantify the severity of the environment. For example, environmental noise values ​​can be collected using acoustic sensors, and then a pre-defined noise impact model can be used to calculate the interference level caused by the environmental noise value to the drone. This noise impact model can be trained using a deep learning network, based on sample data consisting of numerous sample environmental noise values ​​and sample interference levels. For example, if the environmental noise value is 45 dB, the environmental noise interference level calculated using the noise impact model is 0.6 (range 0 to 1). A higher environmental noise interference level indicates a more severe environment. For example, the environmental noise interference level is higher in strong winds or when flying in areas with electromagnetic interference.

[0076] The target distance fluctuation value between the drone and the target refers to the short-term fluctuation range of the distance between the drone and the target, reflecting the stability of the environment. For example, target distance information can be obtained through a laser distance measurement device, set to measure 5 times per second, with a measurement range of 0.5 meters to 50 meters and an accuracy of 0.01 meters. Assuming that the target distance data fluctuates from 20.5 meters to 22.3 meters within 10 seconds during a certain flight, the calculated target distance fluctuation value is: 22.3 - 20.5 = 1.8 meters.

[0077] Furthermore, the standard deviation algorithm was used to analyze the fluctuations in the 50 collected target distance data points. A standard deviation greater than 0.5 meters was considered a significant distance fluctuation. For example, target distance fluctuations are typically larger during windy weather.

[0078] S102: Determine the environmental stability score based on the rate of change of light intensity, the fluctuation value of target distance, and the degree of environmental noise interference.

[0079] In this embodiment, the environmental stability score is used to quantify the stability of the current flight environment. Specifically, the process mainly includes:

[0080] S102-1: The initial environmental stability score is obtained by weighting the initial weights corresponding to the rate of change of light intensity, the fluctuation value of target distance, and the degree of environmental noise interference.

[0081] First, based on the current inspection scenario, initial weights are determined for the rate of change in light intensity, the fluctuation value of target distance, and the degree of environmental noise interference, resulting in an initial weight set. The initial weight set corresponds to the inspection scenario; different inspection scenarios correspond to different initial weight sets.

[0082] For example, in the scenario of high-voltage power line inspection, due to the environmental characteristics of the power lines swinging, the target distance generally fluctuates greatly. Therefore, in this scenario, the initial weight corresponding to the target distance fluctuation value can be set relatively large, for example, the initial weight corresponding to the target distance fluctuation value is 50%, the initial weight corresponding to the light intensity change rate is 30%, and the initial weight corresponding to the environmental noise interference degree is 20%.

[0083] In the scenario of inspecting photovoltaic panel arrays, due to the environmental characteristics of clouds / panel reflection, the reflection is strong and the light intensity changes rapidly. Therefore, in this scenario, the initial weight corresponding to the rate of change of light intensity can be set relatively large. For example, the initial weight corresponding to the rate of change of light intensity is 60%, the initial weight corresponding to the target distance fluctuation value is 15%, and the initial weight corresponding to the environmental noise interference degree is 25%.

[0084] Then, the light intensity change rate, target distance fluctuation value, and environmental noise interference degree are normalized respectively (for example, normalized to between 0 and 1; in the above example, the environmental noise interference degree is already between 0 and 1, so no further normalization is needed). The three are transformed to the same dimension, and the light intensity change rate, target distance fluctuation value, and environmental noise interference degree are weighted according to the initial weight to obtain the initial environmental stability score.

[0085] Specifically, the weighted formula is as follows:

[0086]

[0087] in, This represents the initial environmental stability score. This represents the normalized rate of change of light intensity. This represents the normalized target distance fluctuation value. This represents the normalized environmental noise interference level. This represents the initial weight corresponding to the rate of change of light intensity. This represents the initial weight corresponding to the target distance fluctuation value. This represents the initial weight corresponding to the environmental noise interference level.

[0088] Assuming the normalized rate of change of illumination intensity, target distance fluctuation, and environmental noise interference are 0.8, 0.42, and 0.6 respectively, and the corresponding initial weights are 0.5, 0.3, and 0.2 respectively, then the weighted initial environmental stability score is 0.8*0.5 + 0.42*0.3 + 0.6*0.2 = 0.646.

[0089] S102-2: The initial environmental stability score is corrected based on the stability weight coefficient to obtain the corrected environmental stability score.

[0090] In this embodiment, due to the inherent accuracy errors of the light sensor, laser distance measurement device, and acoustic sensor, the collected light intensity change rate, target distance fluctuation value, and environmental noise interference may not be accurate enough. Therefore, after obtaining the initial environmental stability score, this embodiment further corrects the initial environmental stability score according to the stability weight coefficient to obtain a more accurate environmental stability score, thereby more realistically reflecting the current environmental stability.

[0091] For example, if the stability weighting coefficient is set to 0.8, the final corrected environmental stability score will be... for: =0.646*0.8=0.5168.

[0092] This embodiment uses multi-dimensional data on illumination, distance, and noise to comprehensively determine the environmental stability score, which has a strong anti-interference capability.

[0093] S103: Based on the environmental stability score, determine whether to trigger the adjustment mechanism for the camera's first configuration parameters; wherein the first configuration parameters include focal length and exposure time.

[0094] In this embodiment, a threshold judgment is made on the environmental stability score to determine whether to trigger the adjustment mechanism for the first configuration parameters of the camera (i.e., focal length and exposure time).

[0095] Specifically, the process mainly includes:

[0096] If the environmental stability score is less than the steady-state threshold, the adjustment mechanism for the first configuration parameter of the camera is determined to be triggered; if the environmental stability score is greater than or equal to the steady-state threshold, the adjustment mechanism for the first configuration parameter of the camera is determined not to be triggered.

[0097] The steady-state threshold is dynamically determined based on the historical environmental stability scores collected during the drone's historical inspections. For example, the average of the historical environmental stability scores collected multiple times recently (e.g., within a week) along the same inspection route can be used to determine whether to adjust the camera's focal length and exposure time under the current environment.

[0098] The adjustment mechanism is used to adjust the current camera's focal length and exposure time.

[0099] For example, suppose the steady-state threshold set based on multiple historical environmental stability scores is 0.8. If the current environmental stability score is 0.5168, then the current environmental conditions are considered to have changed significantly, triggering an adjustment mechanism for the camera's focal length and exposure time.

[0100] Simultaneously, the multidimensional environmental state vector, environmental stability score, and adjustment judgment results collected this time are uploaded to the cloud analysis platform to form a complete logical chain from data collection to adjustment judgment. This is used for dynamic updates of steady-state thresholds during the next inspection process and to ensure that the basis for subsequent task adjustments is accurate.

[0101] S104: If the adjustment mechanism for the first configuration parameter of the camera is triggered, the target focal length value and target exposure time of the camera are determined based on the preset second configuration parameter, the environmental stability score and the target distance between the UAV and the target to be photographed.

[0102] In this embodiment, the second configuration parameters include the camera's current focal length adjustment step size and exposure time range, which are used to assist in determining and adjusting the camera's target focal length value and target exposure time.

[0103] Specifically, the calculation process for the target focal length and target exposure time is as follows:

[0104] S104-1: Determine the imaging impact factor based on the target distance between the UAV and the target being photographed.

[0105] The imaging influence factor reflects the degree of influence of target distance on image quality, ranging from [0,1]. A larger imaging influence factor indicates a higher degree of influence of target distance on image quality; a smaller imaging influence factor indicates a lower degree of influence of target distance on image quality.

[0106] In practice, the formula for calculating the imaging impact factor is as follows:

[0107]

[0108] in, Indicates the imaging impact factor. Indicates the current focal length. Indicates the distance to the target. This indicates the radius of the permissible circle of confusion, which is typically twice the size of a camera sensor pixel.

[0109] For example, suppose the current focal length The distance is 35mm, the target distance. The allowable radius of the dispersion circle is 10m. If the value is 7.52 μm, then the imaging influence factor can be calculated. The value is 16.3. Then, the imaging impact factor... Normalization is performed to bring the factor to a value between 0 and 1, thus obtaining the imaging impact factor. It is 0.8.

[0110] S104-2: Determine the first illumination compensation value based on the imaging impact factor and environmental stability score.

[0111] Specifically, the calculation method for the first illumination compensation value is as follows:

[0112]

[0113] in, This represents the first illumination compensation value. This represents the environmental stability score. Indicates the imaging impact factor. This represents the base compensation coefficient, A, which is determined based on the camera sensor calibration (e.g., under standard illumination of 1000 lux). =2).

[0114] For example, suppose the environmental stability score The imaging impact factor is 0.7. The basic compensation coefficient is 0.8. If the value is 2, then the first illumination compensation value is... for: =0.7×0.8×2.0=1.12 lux, which means that 1.12 lux of brightness compensation is needed.

[0115] S104-3: Determine the target exposure time of the camera based on the preset second configuration parameters and the first illumination compensation value.

[0116] Specifically, the target exposure time is calculated as follows:

[0117]

[0118] in, Indicates the target exposure time. Indicates the initial exposure time. This represents the first illumination compensation value. This indicates the reference illumination.

[0119] It should be noted that the above initial exposure time This is a preset value used to provide a starting point for exposure adjustments, ensuring that the initial shooting brightness is appropriate. The initial exposure time is within the preset exposure time range, for example, if the preset exposure time range is [1 / 260s, 1 / 140s], the initial exposure time... It is 1 / 200s.

[0120] When there are significant changes in lighting conditions and low environmental stability, the actual exposure time needs to be adjusted to ensure the quality of image acquisition.

[0121] For example, assuming the exposure time range is [1 / 1000s, 1 / 30s], the initial exposure time... The first illumination compensation value is 1 / 500 of a second. 25 lux, reference illumination If the value is 100 lux, then the calculated target exposure time is 1 / 500 × (1 + 25 / 100) = 1 / 400 second.

[0122] S104-4: Determine the target focal length value of the camera based on a preset focal length distance mapping table; wherein, the focal length distance mapping table is used to characterize the correspondence between the distance between the UAV and the target and the focal length of the camera.

[0123] The focal length distance mapping table is derived from optical analysis, image resolution requirements, and relevant historical data, through expert experience analysis. Different inspection scenarios correspond to different focal length distance mapping tables, and even within the same inspection scenario, the appropriate focal length varies depending on the distance between the drone and the target.

[0124] For example, in a high-voltage line inspection scenario, when the distance between the drone and the target is 10 meters, the appropriate target focal length is 30 millimeters.

[0125] Finally, the target focal length was set to 30 mm, the target exposure time was set to 1 / 480 second, and the parameters were sent to the camera control module in real time to ensure that the image quality was adapted to the current flight environment.

[0126] S105: Controls the camera to perform shooting actions according to the target focal length value and target exposure time, and obtains the inspection image corresponding to the shooting target.

[0127] In this embodiment, after determining the target focal length and target exposure time, the camera's focal length can be adjusted from the current 27 mm to 30 mm and the exposure time can be adjusted from 1 / 500 second to 1 / 480 second based on the focal length adjustment step size in the second configuration parameters.

[0128] For example, during focus adjustment, the lens's response speed is first acquired, for example, 2 mm / s. Next, the required movement distance is calculated: if the current focus is 27 mm and the target focus is 30 mm, then a 3 mm adjustment is needed, taking 3 ÷ 2 = 1.5 seconds. Then, a pulse-width modulation signal is sent via the servo motor driver to control the lens to move to 30 mm, and encoder feedback is monitored in real time to ensure the error is less than 0.05 mm.

[0129] Finally, based on the adjusted target focal length and exposure time, a shooting action is performed to acquire inspection images corresponding to the target. For example, shooting at a rate of 30 frames per second yields multiple consecutive inspection images.

[0130] This embodiment dynamically evaluates environmental stability by calculating a multi-dimensional environmental state vector in real time, and triggers an intelligent parameter adjustment mechanism based on the environmental stability score to achieve adaptive optimization of focal length and exposure time, thereby improving the quality of inspection images. It is applicable to scenarios that require high-precision image acquisition, such as power line inspection, pipeline inspection, and building monitoring, thereby improving inspection efficiency and reliability.

[0131] Optionally, before acquiring the inspection images, the above method further includes:

[0132] Obtain the initial attitude angle offset of the UAV; the initial attitude angle offset is obtained by combining the roll angle, pitch angle and yaw angle of the UAV; if the initial attitude angle offset is greater than the offset threshold, adjust the flight strategy of the UAV; obtain the adjusted attitude angle offset and the adjustment target distance fluctuation value after adjusting the flight strategy; if the adjusted attitude angle offset is less than the offset threshold and the variance of the adjustment target distance fluctuation value is less than the fluctuation threshold, determine the stability of the UAV fuselage.

[0133] In this embodiment, the initial attitude angle offset is used to quantify the degree of fuselage instability. Specifically, the initial attitude angle offset is calculated as follows:

[0134]

[0135] in, This represents the initial attitude angle offset. Indicates the roll angle. Indicates pitch angle, The yaw angle is represented by the unit for all three values, which are degrees.

[0136] A larger initial attitude angle deviation indicates greater instability of the drone, which is detrimental to photography. Therefore, to ensure image quality, when the initial attitude angle deviation exceeds a threshold, the drone's flight strategy is adjusted to maintain stability. This can be achieved through PID control to adjust motor speeds to compensate for the attitude angle (e.g., increasing the power of the left motor to correct right roll) and reducing speed to enhance stability.

[0137] After adjustments, the attitude angle offset and target distance fluctuation are checked again, and the aircraft stability is assessed based on both. Ultimately, if the attitude angle offset is less than the offset threshold and the variance of the target distance fluctuation is less than the fluctuation threshold, the UAV is considered stable and can continue its inspection mission. If either condition is not met, the flight strategy is adjusted, and the aircraft stability is reassessed until the aircraft is stable.

[0138] The offset threshold and fluctuation threshold can be set based on historical data corresponding to different inspection scenarios.

[0139] This embodiment collects the roll, pitch, and yaw angles of the UAV in real time to construct an initial attitude angle offset, determining whether to trigger adjustments to the flight strategy to adapt to disturbances such as wind and load changes. Simultaneously, a dual-condition judgment mechanism based on attitude offset and distance fluctuation is used to measure aircraft stability. Only when aircraft stability is ensured is the inspection and photography task performed to guarantee image clarity.

[0140] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating camera configuration parameter optimization based on image quality in one embodiment of this application. Figure 2 As shown, after obtaining the inspection image corresponding to the captured target in step S105, the method further includes:

[0141] S201: Determine the image quality parameters of the inspection image; the image quality parameters include sharpness value, brightness distribution value and contrast value.

[0142] In this embodiment, multiple inspection images are captured for the same target. For example, if the shooting frequency is 30 frames per second and image quality analysis is performed every 2 seconds, then 60 inspection images will be acquired.

[0143] Specifically, image quality is reflected through image quality parameters. These parameters mainly include three quality indicators: sharpness value, brightness distribution value, and contrast value.

[0144] The sharpness value reflects the clarity of image edges and the richness of detail, and is typically quantized using gradient operations. Specifically, the sharpness value of each inspected image can be calculated using the Laplacian transform method. A higher sharpness value indicates more pronounced edge abrupt changes in the inspected image, and thus a clearer image.

[0145] Brightness distribution values ​​describe the overall brightness and darkness distribution characteristics of an image and can be statistically analyzed using histograms. Contrast values ​​measure the degree of difference between bright and dark areas of an image and can be calculated using methods such as Michelson contrast ratio and standard deviation contrast ratio.

[0146] S202: Based on the sharpness value and brightness distribution value, determine whether to trigger the optimization mechanism for the second configuration parameter; the second configuration parameter includes the focal length adjustment step size and the exposure time range.

[0147] In this embodiment, the optimization mechanism refers to optimizing the camera's focal length adjustment step size and exposure time range under the current environmental conditions.

[0148] Specifically, the triggering process of the optimization mechanism for the second configuration parameter mainly includes:

[0149] S202-1: Determine the volatility of sharpness values ​​for multiple images, and if the volatility exceeds a change threshold, determine the trigger for an optimization mechanism targeting the focal length adjustment step size.

[0150] The higher the volatility of the sharpness value, the worse the focus stability. In this case, the step size can be reduced to avoid over-adjustment. If the volatility is extremely low, the step size can be appropriately increased to speed up focusing. Therefore, the volatility of the sharpness value can be used to determine whether the optimization mechanism for adjusting the focus size is triggered. Specifically, this determination process mainly includes:

[0151] First, based on the sharpness values ​​of N consecutive inspection images, calculate the standard deviation and mean. Then, divide the standard deviation by the mean to obtain the volatility. For example, if the sharpness value sequence of 5 consecutive images is [50, 52, 48, 55, 45], then its mean is 50, the standard deviation is 3.5, and the final volatility is 7%.

[0152] If the volatility is greater than the change threshold, it indicates that the focal length is unstable (e.g., lens shake caused by wind), thus triggering an optimization mechanism for the focal length adjustment step size. If the volatility is less than or equal to the change threshold, there is no need to optimize the current focal length adjustment step size.

[0153] S202-2: Images whose brightness distribution values ​​are not concentrated within the preset brightness range are identified as poor quality images;

[0154] S202-3: If the proportion of poor-quality images in the total number of images is greater than the proportion threshold, determine to trigger the optimization mechanism for the exposure time interval.

[0155] The concentration of brightness distribution values ​​reflects whether the exposure time is reasonable. If the exposure time is too long, the brightness distribution of the inspected image will be biased towards high grayscale; if the exposure time is too short, the brightness distribution will be biased towards low grayscale. Therefore, the brightness distribution values ​​can be used to determine whether to trigger an optimization mechanism for the exposure time range. Specifically, this determination process mainly includes:

[0156] First, based on the requirements of the current inspection scenario, determine the required preset brightness range, such as [100, 150] grayscale values, to avoid overexposure / underexposure. Then, determine whether the brightness distribution value of each inspection image is concentrated within the preset brightness range. If so, the inspection image is considered a qualified image; otherwise, it is considered a poor-quality image.

[0157] Then, the percentage of all poor-quality images out of the total number of images is calculated. If the percentage is greater than a threshold, it indicates that the current exposure strategy is ineffective, and an optimization mechanism for the exposure time interval is triggered. If the percentage is less than or equal to the threshold, no optimization of the exposure time interval is required.

[0158] S203: If it is determined that the optimization mechanism for the second configuration parameter is triggered, the image quality score of the inspection image is determined based on the sharpness value, brightness distribution value and contrast value.

[0159] In this embodiment, for each inspection image, the image quality score of the inspection image can be obtained by weighting its corresponding sharpness value, brightness distribution value and contrast value, and the image quality score is converted into a percentage system (the lowest is 0 points and the highest is 100 points).

[0160] S204: Based on the image quality score, optimize the focus adjustment step size and exposure time range to obtain the optimized step size and optimized exposure range.

[0161] The optimization process for the focal length adjustment step size mainly includes:

[0162] The average quality score is obtained by determining the average quality score of multiple images; the imaging influence factor is determined based on the target distance between the UAV and the target being photographed during the inspection image acquisition; and the optimized step size is determined based on the initial step size, the difference between the average quality score and the standard quality score, and the imaging influence factor.

[0163] In this embodiment, the average quality score of multiple inspection images is first calculated. For example, if the image quality score sequence of five consecutive inspection images is [78, 75, 80, 80, 82], then its average quality score is 79. Next, the difference between the average quality score and the standard quality score is calculated. For example, if the standard quality score is 82, then the difference is 3 points. The standard quality score refers to the minimum quality score that an inspection image acquired in the current inspection scenario must meet. It is easy to understand that, given a fixed standard quality score, the larger the difference, the worse the quality of the currently acquired inspection image.

[0164] Then, based on the target distance between the UAV and the target being photographed, the imaging influence factor is determined. The imaging influence factor reflects the degree of influence of the target distance on the image quality, ranging from [0,1]. The larger the imaging influence factor, the greater the influence of the target distance on the image quality; the smaller the imaging influence factor, the lower the influence of the target distance on the image quality. For example, when the target distance between the UAV and the target being photographed is 10 meters, the imaging influence factor is determined to be 0.8.

[0165] Finally, the optimization step size is determined using the following formula:

[0166]

[0167] in, Indicates the optimization step size. This represents the initial step size, i.e., the focus adjustment step size during image capture. This represents the sensitivity coefficient, with values ​​ranging from [0,1] (e.g., 0.5), used to control the adjustment intensity. This represents the difference between the average mass fraction and the standard mass fraction. Indicates standard mass fraction. This indicates the imaging impact factor.

[0168] Therefore, if If the value is negative, meaning the average quality score is lower than the standard quality score, then the focal length adjustment step size is increased to accelerate the correction; Imaging Influence Factor When the distance between the drone and the target is small (i.e., when the distance between the drone and the target is far), the focal length adjustment step size is reduced to avoid over-adjustment.

[0169] For example, suppose the initial step size value is... The sensitivity coefficient is 0.1 mm. The value is 0.5, representing the difference between the average mass fraction and the standard mass fraction. The standard mass fraction is -5.4. The imaging impact factor is 90. If the value is 0.61, then the calculated optimal step size is... It is 0.059mm.

[0170] The optimization process for the exposure time interval mainly includes:

[0171] Based on the difference between the average quality score and the standard quality score, a second illumination compensation value is determined using a preset illumination compensation model; the target exposure time is corrected based on the second illumination compensation value to obtain the corrected target exposure time; and the optimized exposure range is determined based on the corrected target exposure time and the optimization step size.

[0172] In this embodiment, the preset illumination compensation model can be a linear model, used to convert the difference in image quality into a second illumination compensation value. For example:

[0173]

[0174] in, This represents the second illumination compensation value. This indicates the scaling factor (e.g., -0.05, with a negative sign indicating reverse compensation). This represents the difference between the average mass score and the standard mass score.

[0175] For example, suppose =-5.4 (Inspection image quality is low) =-0.05, then the second illumination compensation value is obtained through the illumination compensation model. A value of +0.27 indicates that a 27% increase in illumination compensation is needed to improve image quality to meet the requirements of standard image quality.

[0176] Then, based on the second illumination compensation value, the target exposure time used during image acquisition is corrected to obtain the corrected target exposure time. The correction process can refer to the calculation method of the target exposure time described above, and will not be repeated here.

[0177] Assuming the corrected target exposure time is 1 / 157 second and the optimization step size is 0.15, the final optimized exposure range is adjusted as follows: .

[0178] This embodiment calculates an image quality score based on sharpness, brightness distribution, and contrast values, and determines whether configuration parameters need optimization based on these indicators. If optimization is required, the focus adjustment step size and exposure time range are optimized based on the difference between the actual image quality and the required standard image quality, ultimately yielding an optimized step size and optimized exposure range. Subsequently, based on the optimized step size and optimized exposure range, inspection image acquisition can be performed again to ensure that the image quality consistently approaches the standard value.

[0179] Please refer to Figure 3 , Figure 3This is a flowchart illustrating the setting of a second configuration parameter in one embodiment of this application. For example... Figure 3 As shown, after optimizing the focus adjustment step size and exposure time range based on the image quality score in step S204 to obtain the optimized step size and optimized exposure range, the method further includes:

[0180] S301: Build a database linking the shooting environment and camera configuration. Any historical data in the database includes the light intensity when the inspection image was captured, the target distance between the drone and the target, the optimization step size, and the optimization exposure range.

[0181] In this embodiment, for inspection images that fail to meet quality standards, the shooting environment at the time of shooting—light intensity, the distance between the drone and the target—and the camera configuration parameters obtained based on image quality analysis—optimization step size and optimization exposure range are linked and stored to form a historical correlation data. Then, based on the historical correlation data under different shooting environment conditions, a correlation database is formed, and the data in the correlation database is stored in a cloud-based log database to form a parameter optimization traceability chain to support subsequent environmental adaptability analysis.

[0182] S302: Obtain environmental parameters of the current shooting environment, including the current light intensity and the current target distance.

[0183] S303: Determine one or more historical related data that match the environmental parameters from the related database.

[0184] In this embodiment, during the subsequent actual inspection and shooting process, the current light intensity and current target distance can be matched with the historical related data in the associated database to find one or more historical related data that are similar to the current shooting environment.

[0185] Specifically, Euclidean distance can be used to assess the similarity between current parameters and historical data:

[0186]

[0187] in, Indicates similarity. Indicates the current light intensity. The light intensity represents the historical correlation data. This is a reference value for illumination (e.g., 1000 lux). Indicates the current target distance. Indicates the target distance of historical related data. This is a distance reference value (e.g., 10m).

[0188] Finally, select the top k historical related data with the highest similarity (e.g., k=3).

[0189] S304: Determine the target focal length adjustment step size and target exposure time range based on the optimization step size and optimization exposure range of one or more historical related data.

[0190] In this embodiment, after obtaining one or more historical correlation data that match the current environmental parameters, the target focal length adjustment step and target exposure time range can be calculated by weighted average method based on their respective optimization step size and optimization exposure range.

[0191] For example, suppose the current light intensity is 1000 lux, the current target distance is 10 meters, and two similar historical data entries are matched from the correlation database:

[0192]

[0193] Therefore, the target focal length adjustment step is calculated to be (0.07 + 0.09) / 2 = 0.08 mm, and the target exposure time interval is: .

[0194] S305: Based on the target focal length adjustment step size and the target exposure time interval, the preset second configuration parameters are obtained.

[0195] In this embodiment, during the subsequent actual inspection and shooting process, the appropriate target focal length adjustment step and target exposure time range can be determined based on historical data under historical environments similar to the current shooting environment. Then, the above steps S101-S105 are run to capture inspection images that meet the standard quality requirements, thereby accelerating the inspection efficiency.

[0196] This embodiment constructs a relational database by associating and storing historical shooting environments with the corresponding optimized configuration parameters of the camera. In subsequent inspection applications, historical data from historical environments similar to the current shooting environment can be used to predict the target focal length adjustment step and target exposure time range suitable for the current shooting environment. This provides an effective basic parameter for image capture, improves the efficiency of real-time camera adjustments, and enhances the quality of inspection images.

[0197] Secondly, based on the same inventive concept, and referring to... Figure 4 This application provides an AI inspection drone control system based on real-time dynamic measurement technology. The drone control system 400 includes:

[0198] A light sensor is used to collect the rate of change in light intensity during drone inspections.

[0199] Acoustic sensors are used to collect environmental noise interference levels during drone inspections.

[0200] Distance measuring device is used to collect the fluctuation value of the target distance between the drone and the target being photographed;

[0201] Camera, used to capture inspection images corresponding to the target;

[0202] A flight controller for executing the AI ​​inspection drone control method based on real-time dynamic measurement technology as proposed in the first aspect of this application.

[0203] Thirdly, based on the same inventive concept, and referring to... Figure 5 This application provides an electronic device 500, including a processor 501 and a memory 502; the memory 502 stores machine-executable instructions that can be executed by the processor 501, and the processor 501 is used to execute the machine-executable instructions to implement the AI ​​inspection drone control method based on real-time dynamic measurement technology proposed in the first aspect of this application.

[0204] It should be noted that the specific implementation of the electronic device 500 in this application embodiment refers to the specific implementation of the AI ​​inspection drone control method based on real-time dynamic measurement technology proposed in the first aspect of the above-mentioned application embodiment, and will not be repeated here.

[0205] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0206] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0207] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0208] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0209] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0210] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0211] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0212] The above provides a detailed description of the AI ​​inspection drone control method and system based on real-time dynamic measurement technology provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A control method for an AI-powered inspection drone based on real-time dynamic measurement technology, characterized in that, The method includes: With the drone body stable, a multi-dimensional environmental state vector is periodically acquired during the drone inspection process; the multi-dimensional environmental state vector includes the rate of change of light intensity, the degree of environmental noise interference, and the fluctuation value of the distance between the drone and the target being photographed. An environmental stability score is determined based on the rate of change of light intensity, the fluctuation value of the target distance, and the degree of environmental noise interference. Based on the environmental stability score, it is determined whether to trigger an adjustment mechanism for the camera's first configuration parameters; wherein the first configuration parameters include focal length and exposure time. If it is determined that an adjustment mechanism for the first configuration parameter of the camera is triggered, the target focal length value and target exposure time of the camera are determined based on the preset second configuration parameter, the environmental stability score and the target distance between the UAV and the target being photographed. The camera is controlled to perform a shooting action according to the target focal length value and the target exposure time to obtain the inspection image corresponding to the shooting target; Specifically, based on preset second configuration parameters, the environmental stability score, and the target distance between the drone and the target, the target focal length value and target exposure time of the camera are determined, including: Based on the target distance between the UAV and the target being photographed, the imaging impact factor is determined; Based on the imaging impact factor and the environmental stability score, a first illumination compensation value is determined; The target exposure time of the camera is determined based on the preset second configuration parameters and the first illumination compensation value; Based on a preset focal length distance mapping table, the target focal length value of the camera is determined; wherein, the focal length distance mapping table is used to characterize the correspondence between the distance of the UAV from the target and the focal length of the camera.

2. The method according to claim 1, characterized in that, The method further includes: The initial attitude angle offset of the UAV is obtained; the initial attitude angle offset is obtained by combining the roll angle, pitch angle and yaw angle of the UAV. If the initial attitude angle offset is greater than the offset threshold, the flight strategy of the UAV is adjusted. Obtain the attitude angle offset and target distance fluctuation value after adjusting the flight strategy; If the attitude angle offset is less than the offset threshold and the variance of the target distance fluctuation is less than the fluctuation threshold, the UAV fuselage is determined to be stable.

3. The method according to claim 1, characterized in that, Based on the rate of change of light intensity, the fluctuation value of the target distance, and the degree of environmental noise interference, an environmental stability score is determined, including: An initial environmental stability score is obtained by weighting the initial weights corresponding to the light intensity change rate, the target distance fluctuation value, and the environmental noise interference degree. The initial environmental stability score is corrected based on the stability weighting coefficient to obtain the corrected environmental stability score; Based on the environmental stability score, determine whether to trigger an adjustment mechanism for the first configuration parameter of the camera, including: If the environmental stability score is less than a steady-state threshold, an adjustment mechanism for the first configuration parameter of the camera is triggered; wherein the steady-state threshold is dynamically determined based on the historical environmental stability score collected by the UAV during historical inspections.

4. The method according to any one of claims 1-3, characterized in that, The second configuration parameters include the focal length adjustment step size and the exposure time range; After obtaining the inspection image corresponding to the photographed target, the method further includes: Determine the image quality parameters of the inspection image; the image quality parameters include sharpness value, brightness distribution value, and contrast value; Based on the sharpness value and the brightness distribution value, determine whether to trigger the optimization mechanism for the second configuration parameter; If it is determined that an optimization mechanism for the second configuration parameter is triggered, the image quality score of the inspection image is determined based on the sharpness value, the brightness distribution value, and the contrast value. Based on the image quality score, the focal length adjustment step size and the exposure time interval are optimized to obtain the optimized step size and optimized exposure interval.

5. The method according to claim 4, characterized in that, The inspection images include multiple images; Based on the sharpness value and the brightness distribution value, determine whether to trigger the optimization mechanism for the second configuration parameter, including: Determine the volatility of the sharpness values ​​of multiple images, and if the volatility is greater than a change threshold, determine whether to trigger an optimization mechanism for the focal length adjustment step size. Images whose brightness distribution values ​​are not concentrated within a preset brightness range are identified as poor-quality images; If the proportion of poor-quality images to the total number of images is greater than a certain threshold, an optimization mechanism for the exposure time interval is triggered.

6. The method according to claim 4, characterized in that, Based on the image quality score, the focal length adjustment step size is optimized to obtain an optimized step size, including: The average quality score is obtained by determining the average quality score of the multiple images. Based on the target distance between the UAV and the photographed target during the acquisition of the inspection images, the imaging influence factor is determined; The optimized step size is determined based on the initial step size, the difference between the average quality score and the standard quality score, and the imaging influence factor.

7. The method according to claim 6, characterized in that, Based on the image quality score, the exposure time interval is optimized to obtain an optimized exposure interval, including: Based on the difference between the average quality score and the standard quality score, a second illumination compensation value is determined using a preset illumination compensation model. The target exposure time is corrected based on the second illumination compensation value to obtain the corrected target exposure time; The optimized exposure range is determined based on the corrected target exposure time and the optimized step size.

8. The method according to claim 4, characterized in that, After optimizing the focal length adjustment step size and the exposure time interval based on the image quality score to obtain the optimized step size and optimized exposure interval, the method further includes: Construct a database linking the shooting environment and camera configuration. Any historical data in the database includes the light intensity when the inspection image was captured, the target distance between the UAV and the target being photographed, the optimization step size, and the optimization exposure range. The method further includes: Obtain environmental parameters of the current shooting environment, including the current light intensity and the current target distance; Determine one or more historical related data that match the environmental parameters from the related database; Based on the optimization step size and the optimization exposure interval of one or more of the historical associated data, the target focal length adjustment step size and the target exposure time interval are determined. Based on the target focal length adjustment step size and the target exposure time interval, the preset second configuration parameters are obtained.

9. A control system for an AI-powered inspection drone based on real-time dynamic measurement technology, characterized in that, The system includes: A light sensor is used to collect the rate of change in light intensity during drone inspections. An acoustic sensor is used to collect the environmental noise interference level during the UAV inspection process; A distance measuring device is used to collect the target distance fluctuation value between the UAV and the target being photographed; A camera is used to capture inspection images corresponding to the target being photographed; A flight controller for executing the AI ​​inspection drone control method based on real-time dynamic measurement technology as described in any one of claims 1 to 8.

Images