Method and device for detecting internal defects of external thermal insulation layer of unmanned aerial vehicle-mounted outer wall

Abstract

The invention provides an unmanned aerial vehicle-mounted external wall external thermal insulation layer internal defect detection method and device, and the method comprises the steps: setting the spatial position coordinates and boundary dimension of a to-be-detected external wall, and planning the flight path of an unmanned aerial vehicle through a grid method based on the area shape of the detected external wall; the unmanned aerial vehicle performs point-by-point scanning detection on each grid point in sequence according to the flight path to obtain detection data; and performing data processing on the detection data to obtain defect measured data of the to-be-detected outer wall. A computer is adopted to calculate and plan the path of the unmanned aerial vehicle, and the unmanned aerial vehicle and the outer wall detection device are controlled to complete an outer wall detection task. The device and the method have the advantages of high automation degree, labor saving and the like, can realize quantitative evaluation of internal defects, and are small in measurement error.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent detection, and particularly relates to a method and device for detecting internal defects in the external thermal insulation layer of exterior walls by an unmanned aerial vehicle (UAV). Background Art

[0002] In existing external thermal insulation layers of building exterior walls, there are common problems such as easy detachment of facing tiles and insulation boards, which pose potential safety hazards to people's lives and property. Therefore, there is an urgent need for a device that can automatically detect and identify common faults such as cracks, looseness, and hollowing in the external thermal insulation layer of building exterior walls in advance. Currently, there are three methods for detecting external wall faults: manual hammer knocking method, acoustic vibration acquisition method, and infrared thermal image detection method. The implementation and judgment of these three testing methods mainly rely on manual labor, and they have disadvantages such as large workload and low efficiency. They lack the ability to accurately quantitatively evaluate the size and scale of internal defects, and the measurement error is relatively large.

[0003] Among them, the manual knocking method has a large workload, and the judgment of faults is mainly manual, which can only achieve qualitative judgment. Although the acoustic vibration acquisition method can obtain the quantitative defect sound spectrum, the judgment of defects still requires manual participation, and it cannot quantitatively evaluate the size and scale of defects. The infrared thermal imaging method has the advantages of being portable, easy to operate, and high temperature measurement accuracy, but due to factors such as uneven heating and human experience, the measurement error is relatively large, and it still cannot solve the disadvantage of lacking quantitative judgment. Summary of the Invention

[0004] The technical problem to be solved by the present invention is how to avoid manual detection to improve the detection probability, reduce errors, and improve the detection efficiency. In view of this, the present invention provides a method and device for detecting internal defects in the external thermal insulation layer of exterior walls by an unmanned aerial vehicle (UAV).

[0005] The technical solution adopted by the present invention is a method for detecting internal defects in the external thermal insulation layer of exterior walls by an unmanned aerial vehicle (UAV), including:

[0006] Step S1, setting the spatial position coordinates and external dimensions of the exterior wall to be inspected, and planning the UAV flight path using the grid method based on the area shape of the exterior wall to be detected;

[0007] Step S2, the UAV sequentially scans and detects each grid point according to the flight path to obtain detection data;

[0008] Step S3, performing data processing on the detection data to obtain the measured defect data of the exterior wall to be inspected.

[0009] In one embodiment, the setting of the spatial position coordinates and external dimensions of the exterior wall to be inspected, and planning the UAV flight path using the grid method based on the area shape of the exterior wall to be detected, includes:

[0010] The whole is discretely divided according to the grid size w, and each grid is identified by a unique index identifier i. Among them, for any point on the outer wall to be inspected, its coordinates are divided by the set grid size, and the integer part is taken to obtain the grid index;

[0011] The obstacle is represented by a bounding box. The size of each obstacle is increased by L, and its bounding box is calculated. Wherein, the value of L is the width of the current device. Each obstacle is identified by an obstacle index, and the corresponding obstacle index value is the height h of the bounding box, which is saved to all the outer wall grids intersecting the bounding box.

[0012] In one embodiment, the drone sequentially scans and detects each grid point according to the flight path to obtain detection data. At the same time, the distance between the drone and the wall is L; wherein, if the obstacle index s at the operation point i = 0, the device performs data acquisition at this point. If s i = h, the device is turned off to stop data acquisition, and the distance between the drone and the outer wall is h + L.

[0013] In one embodiment, the drone sequentially scans and detects each grid point according to the flight path to obtain detection data, further including:

[0014] The transmission power P of the outer wall detection device L , the safe distance L between the drone and the outer wall, the pre-calibrated transmission power P 0 and the distance R 0 are adjusted according to the following formula:

[0015]

[0016] In one embodiment, the data processing of the detection data to obtain the measured defect data of the outer wall to be inspected includes:

[0017] Continuously sample the received baseband analog signal of each frequency point to obtain a digital baseband signal sequence;

[0018] Perform windowed inverse Fourier transform on the digital baseband signal sequence respectively to obtain the amplitude and phase angle values of the time-domain discrete signal;

[0019] Perform DC offset removal and band-pass filtering operations on the amplitude sequence of the time-domain discrete signal to obtain scan data, and multiple channels of the scan data constitute two-dimensional test data;

[0020] Perform interface background clutter elimination processing on the two-dimensional test data of each survey line;

[0021] Perform gain adjustment on the current two-dimensional test data to make the target features more prominent;

[0022] Input the two-dimensional test data after gain adjustment into a pre-trained deep learning network for fault identification to obtain a target detection area containing void defects;

[0023] Perform image binarization processing and positioning operations on the target detection area to extract the position, height position, and width of the target.

[0024] In one embodiment, the gain adjustment of the current two-dimensional test data to make the target features more prominent includes:

[0025] For the processing area of the two-dimensional test data, take the maximum value of the absolute value of each row and connect them to obtain a corresponding calculation function;

[0026] Find the extreme value envelope of the calculation function;

[0027] Take the reciprocal of the extreme value envelope and then multiply it by a preset coefficient to obtain a gain function;

[0028] Process the two-dimensional test data using the gain function to obtain the data after gain adjustment.

[0029] Another aspect of the present invention also provides a detection device for internal defects of the external wall thermal insulation layer carried by a drone, including:

[0030] A preprocessing unit configured to set the spatial position coordinates and external dimensions of the external wall to be inspected, and plan the flight path of the drone based on the area shape of the external wall to be detected;

[0031] An acquisition unit configured to enable the drone to sequentially scan and detect each grid point according to the flight path to obtain detection data;

[0032] A data processing unit configured to process the detection data to obtain the measured defect data of the external wall to be inspected.

[0033] Another aspect of the present invention also provides an electronic device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the detection method for internal defects of the external wall thermal insulation layer carried by a drone as described in any one of the above.

[0034] Another aspect of the present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the detection method for internal defects of the external wall thermal insulation layer carried by a drone as described in any one of the above.

[0035] Compared with the prior art, the present invention has at least the following advantages:

[0036] The present invention provides a method and device for detecting internal defects in the external thermal insulation layer of an exterior wall carried by an unmanned aerial vehicle (UAV). A computer is used to calculate and plan the UAV path, and the UAV is controlled to automatically complete the exterior wall detection task with an exterior wall detection device. This device and method have the advantages of high automation and labor saving, can achieve quantitative evaluation of internal defects, and have small measurement errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 FIG. is a schematic flow chart of a method for detecting internal defects in the external thermal insulation layer of an exterior wall carried by a UAV according to an embodiment of the present invention;

[0038] Figure 2 FIG. is a schematic structural diagram of the implementation equipment according to an embodiment of the present invention;

[0039] Figure 3 FIG. is a schematic diagram of the exterior wall after gridization and index marking according to an embodiment of the present invention;

[0040] Figure 4 FIG. is a schematic diagram of the exterior wall after gridization and obstacle index marking according to an embodiment of the present invention;

[0041] Figure 5 FIG. is a schematic diagram of the composition of a device for detecting internal defects in the external thermal insulation layer of an exterior wall carried by a UAV according to an embodiment of the present invention;

[0042] Figure 6 FIG. is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] To further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined purpose, the present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments.

[0044] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. It should also be understood that terms (such as those defined in a common dictionary) should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

[0045] It should be noted that, without conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The present application will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0046] An embodiment of the present invention provides a method for detecting internal defects in the external thermal insulation layer of an exterior wall carried by a UAV, as Figure 1 shown, including:

[0047] Step S1, set the spatial position coordinates and external dimensions of the exterior wall to be inspected, and plan the UAV flight path using the grid method based on the area shape of the exterior wall to be detected;

[0048] Step S2, the UAV sequentially scans and detects each grid point according to the flight path to obtain detection data;

[0049] Step S3, perform data processing on the detection data to obtain the actual measured data of the defects of the exterior wall to be inspected.

[0050] In this embodiment, the setting of the spatial position coordinates and external dimensions of the exterior wall to be inspected, and planning the UAV flight path using the grid method based on the area shape of the exterior wall to be detected includes:

[0051] Discretely divide the whole according to the grid size w, and each grid is identified by a unique index identifier i. Among them, for any point on the exterior wall to be inspected, divide its coordinates by the set grid size and take the integer part to obtain the grid index;

[0052] Represent the obstacles using bounding boxes, increase the size of each obstacle by L, and calculate its bounding box, where L takes the value of the width of the current device. Each obstacle is identified by an obstacle index, and the corresponding obstacle index value is set to the height h of the bounding box and saved to all the exterior wall grids intersecting with the bounding box.

[0053] In this embodiment, the UAV sequentially scans and detects each grid point according to the flight path to obtain detection data, and at the same time, the distance between the UAV and the wall is L; among them, if the obstacle index s i = 0, the device performs data acquisition at this point. If s i = h, turn off the device to stop data acquisition, and the distance between the UAV and the exterior wall is h + L.

[0054] In this embodiment, the UAV sequentially scans and detects each grid point according to the flight path to obtain detection data, and further includes:

[0055] The transmission power P L of the exterior wall detection device, the safe distance L between the UAV and the exterior wall, the pre-calibrated transmission power P 0 and the distance R 0 are adjusted according to the following formula:

[0056]

[0057] In this embodiment, the performing data processing on the detection data to obtain the actual measured data of the defects of the exterior wall to be inspected includes:

[0058] Continuously sample the received baseband analog signal for each frequency point to obtain a digital baseband signal sequence;

[0059] Perform windowed inverse Fourier transform on the digital baseband signal sequence respectively to obtain the amplitude and phase angle values of the time-domain discrete signal;

[0060] Perform DC offset removal and band-pass filtering operations on the amplitude sequence of the time-domain discrete signal to obtain scan data, and multiple channels of the scan data constitute two-dimensional test data;

[0061] Perform interface background clutter elimination processing on the two-dimensional test data of each survey line;

[0062] Adjust the gain of the current two-dimensional test data to make the target features more prominent;

[0063] Input the gain-adjusted two-dimensional test data into a pre-trained deep learning network for fault identification to obtain a target detection area containing void defects;

[0064] Perform image binarization processing and positioning operations on the target detection area to extract the position, height position, and width of the target.

[0065] In this embodiment, the adjusting the gain of the current two-dimensional test data to make the target features more prominent includes:

[0066] For the processing area of the two-dimensional test data, take the maximum value of the absolute value of each row and connect them to obtain a corresponding calculation function;

[0067] Find the extreme value envelope of the calculation function;

[0068] Take the reciprocal of the extreme value envelope and multiply it by a preset coefficient to obtain a gain function;

[0069] Process the two-dimensional test data with the gain function to obtain the gain-adjusted data.

[0070] Next, the method provided in this embodiment will be described in detail with reference to the accompanying drawings.

[0071] (1) Overall device composition

[0072] The device designed by the present invention consists of a drone platform, an exterior wall detection device, and a display control terminal.

[0073] Reference Figure 2 , the drone platform can select a commercially mature drone system, which mainly includes modules such as a drone and a high-precision positioning RTK. The drone realizes high-precision positioning in the RTK mode. During operation, an RTK mobile terminal is carried on the drone platform, and an RTK fixed terminal is set on the ground to achieve high-precision positioning of the drone platform. The exterior wall detection device is carried on the drone platform.

[0074] The display control terminal can be implemented on a portable computer or an embedded device. The display control terminal is connected via a wireless network. The display control terminal implements display control and data processing. The display control mainly controls the drone platform and the external wall detection device. The data processing operations perform data processing operations such as DC bias removal, background clutter elimination, noise suppression, gain compensation, and automatic target recognition based on a deep learning network on the collected data.

[0075] The external wall detection device consists of a transmitting and receiving antenna, a transceiver module, and a sampling and preprocessing module.

[0076] The transmitting and receiving antenna is used for the transmission and reception of ultra-wideband signals. The transmitting channel in the transceiver module includes a power amplifier, upconverter, attenuator, and filter, etc. Under the action of the timing signal and control command generated by the sampling and preprocessing module, the transmitting channel generates a stepped-frequency continuous wave signal that meets the transmit power requirements. The stepped-frequency range can reach 0.4 GHz to 12 GHz. The operating frequency range, pulse width, operating cycle duration, transmit frequency point power, and frequency step increment of the signal transmitted by the transmitting channel can all be programmed and controlled online in real time. The receiving channel in the transceiver module includes a low-noise amplifier, downconverter, attenuator, and channel filter, etc. The receiving channel amplifies, filters, and downconverts the 0.8 GHz to 12 GHz radio frequency echo signal received from the antenna and outputs an intermediate-frequency analog signal to the digital sampling of the sampling and preprocessing module. Parameters such as the intermediate-frequency bandwidth and channel gain of the receiving channel can also be programmed and controlled online in real time. The sampling and preprocessing module is mainly composed of an ADC (analog-to-digital conversion) and an FPGA (field programmable gate array) to realize digital acquisition, downconversion, filtering, and decimation of the input intermediate-frequency analog signal, and then output the baseband signal to the display control terminal.

[0077] (2) Working process

[0078] Defect detection of the external wall insulation layer of a specified area is achieved through the following process

[0079] Step 1: Set the spatial position coordinates and external dimensions of the external wall to be inspected on the display control terminal. According to the area shape of the external wall to be detected, the grid method is used to plan the flight path of the drone. At the same time, set the working parameters such as the transmit power, sweep frequency range, frequency step increment, transmit sub-pulse width, cycle duration, and accumulation cycle number of the external wall detection device.

[0080] Reference Figures 3 to 4 , the grid method route planning algorithm process is as follows:

[0081] 1) Discretely divide the whole according to the grid size w, and each grid is identified by a unique index identifier i. For any point on the exterior wall to be inspected, divide its coordinates by the set grid size and take the integer part to obtain the grid index.

[0082] 2) Represent obstacles such as windows and balconies on it using bounding boxes. Increase the size of each obstacle by L and calculate its bounding box. L generally takes the value of the width of the device. Each obstacle is identified by an obstacle index. Take the corresponding obstacle index value as the height h of the bounding box and save it to all the exterior wall grids that intersect with the bounding box.

[0083] The obstacle index identifier takes values according to the following relationship:

[0084] s i = 0, corresponding grid has no obstacle

[0085] s i = h, corresponding grid has an obstacle, and the height of the obstacle is h

[0086] The specific implementation method is as follows:

[0087] If the length and width dimensions of the exterior wall to be inspected are a and b respectively, and the grid size is w, then the number of grids obtained after discretization is M*N, where M = ceil(a / w), N = ceil(b / w), and ceil represents the operation of taking the integer part downward. Number the spatial grid from top to bottom and from left to right, then the value range of the obtained number i is 1 to ceil(a / w)*ceil(b / w). It is stipulated that any grid f i belongs to the corresponding coordinates (x i , y i ), and the coordinates and serial numbers of the grid take values according to the following relationship:

[0088] x i = w*(ceil((i - 1) / N)+0.5), y i = w*(mod((i - 1) / N)+0.5)

[0089] Increase the size of the obstacle by L and calculate its bounding box. L takes the value of the width of the device. Each obstacle is identified by an obstacle index. Take the corresponding obstacle index value as the height h of the bounding box and save it to all the exterior wall grids that intersect with the bounding box. If the grid has no obstacle, the index identifier s corresponding to the grid i = 0.

[0090] Step 2: The drone sequentially scans and detects each point of the grids numbered from 1 to ceil(a / w)*ceil(b / w) according to the planned route, and at the same time, the drone keeps a distance L from the wall. If the obstacle index s at the working point i= 0, the device normally performs data acquisition at this point. If s i = h, the device is turned off to stop data acquisition, and the distance between the drone and the outer wall is h + L. Repeat the second step continuously until the outer wall detection task ends.

[0091] In actual work, the transmission power P of the outer wall detection device L , the safe distance L between the drone and the outer wall, the pre-calibrated transmission power P 0 and the distance R 0 , are adjusted according to the following formula:

[0092]

[0093] Step 3: After the drone completes all detection operations on the outer wall to be inspected. The display terminal control processes the collected data according to the following process to obtain the measured defect data of the outer wall to be inspected.

[0094] The specific data processing methods and processes are as follows:

[0095] 1) Use the sampling and preprocessing module to continuously sample the received baseband analog signal at each frequency point to obtain a digital baseband signal sequence S 1 , S 2 ,…, S N .

[0096] 2) Perform windowed inverse Fourier transform on the sequences S 1 , S 2 ,…, S N respectively to obtain the amplitudes A 1 , A 2 ,…, A N and phase angles φ 1 , φ 2 ,…, φ N values of the time-domain discrete signals;

[0097] 3) Perform DC offset removal and band-pass filtering operations on the amplitude sequences A 1 , A 2 ,…, A N at each operation point.

[0098] DC offset removal:

[0099]

[0100] where A(n) is the original one-dimensional scan data before processing. A'(n) is the processed one-dimensional scan data. N is the number of data samples.

[0101] Band-pass filtering operation:

[0102] B(f) = A(f)H(f),

[0103] Among them, H(f) is the frequency characteristic function of the band-pass filter. B(f) is the one-dimensional scanned data in the frequency domain after filtering, and A(f) is the one-dimensional scanned data in the frequency domain before filtering.

[0104] 4) Perform interface background clutter elimination processing on the two-dimensional test data of each survey line.

[0105] Subtract the same background value from each one-dimensional scanned data to eliminate background clutter

[0106]

[0107] Among them, r j '(n) represents the one-dimensional scanned data of the jth channel. n is the sample point of the data. B j (n) represents the one-dimensional scanned data of each channel before filtering; M represents the total number of channels in the two-dimensional test data.

[0108] 5) Perform gain adjustment on the two-dimensional test data after background clutter elimination to make the target features more prominent.

[0109] The gain adjustment method is as follows: for the processing area of the two-dimensional image, take the maximum value r of the absolute value of each row max '(n), connect them to obtain the function f(n), find the extreme envelope C(n) of the function f(n), then take the reciprocal of C(n), and then multiply by a certain coefficient to obtain the gain function W(n). Use this gain function to process the two-dimensional data to obtain the data after gain adjustment.

[0110] 6) Input the two-dimensional test data after gain adjustment into a pre-trained deep learning network for fault identification to obtain the target detection area containing cavity defects.

[0111] 7) For the target detection area after the above identification, perform image binarization processing and positioning operations to extract the position (X c , Y c ), height position H, width W and other parameters of the target.

[0112] The calculation formulas for each parameter are as follows

[0113] Center position:

[0114]

[0115] Height:

[0116] H = y max -y min ;

[0117] Width:

[0118] W = xmax -x min

[0119] where x j , y j are the pixel coordinates of the target area.

[0120] Compared with the prior art, this embodiment has at least the following effects:

[0121] 1) In the present invention, through the exterior wall detection device carried on the UAV platform, a computer is used to calculate and plan the UAV path, and the UAV is controlled to automatically complete the exterior wall detection task with the exterior wall detection device;

[0122] 2) Specifically, this device and method have the characteristics of high automation and labor saving, can realize the quantitative evaluation of internal defects, and have the advantages of small measurement error;

[0123] 3) The embodiment of the present invention can be used to solve the problems of the traditional exterior wall defect detection method, which mainly relies on manual qualitative judgment for defect judgment, has a large task volume, cannot quantitatively evaluate defects, and has low measurement accuracy.

[0124] The second embodiment of the present invention corresponds to the first embodiment. As Figure 5 shown, this embodiment introduces a detection device for internal defects of the exterior wall thermal insulation layer carried by a UAV, which includes the following components:

[0125] A preprocessing unit configured to set the spatial position coordinates and external dimensions of the exterior wall to be inspected, and plan the UAV flight path using the grid method based on the area and shape of the exterior wall to be detected;

[0126] An acquisition unit configured to enable the UAV to sequentially scan and detect each grid point according to the flight path to obtain detection data;

[0127] A data processing unit configured to process the detection data to obtain the measured defect data of the exterior wall to be inspected.

[0128] The third embodiment of the present invention, an electronic device, as Figure 6 shown, can be understood as an entity device, including a processor and a memory storing processor-executable instructions. When the instructions are executed by the processor, the following operations are performed:

[0129] Step S1, set the spatial position coordinates and external dimensions of the exterior wall to be inspected, and plan the UAV flight path using the grid method based on the area and shape of the exterior wall to be detected;

[0130] Step S2, enable the UAV to sequentially scan and detect each grid point according to the flight path to obtain detection data;

[0131] Step S3: Process the detection data to obtain the measured defect data of the exterior wall to be inspected.

[0132] In the fourth embodiment of the present invention, the process of the method for detecting internal defects in the exterior wall thermal insulation layer carried by the unmanned aerial vehicle in this embodiment is the same as that in the first, second, or third embodiment. The difference is that in engineering implementation, this embodiment can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases, the former is a better implementation method. Based on such an understanding, the method of the present invention can be embodied in the form of a computer software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions for causing a device to execute the method described in the embodiments of the present invention.

[0133] Through the description of the specific implementation manners, it should be possible to have a more in-depth and specific understanding of the technical means and effects adopted by the present invention to achieve the predetermined purpose. However, the attached drawings are only for reference and illustration, and are not used to limit the present invention.

Claims

1. A method for detecting internal defects of an external wall external insulation layer by an unmanned aerial vehicle, characterized in that: include: Step S1, setting the spatial position coordinates and external dimensions of the external wall to be inspected, and planning the flight path of the drone using a grid method based on the area and shape of the external wall to be inspected; Step S2, the UAV scans and detects each grid point point by point according to the flight path to obtain detection data; Step S3, processing the detection data to obtain actual defect measurement data of the exterior wall to be inspected.

2. The method for detecting internal defects of an external wall external insulation layer carried by an unmanned aerial vehicle according to claim 1 is characterized in that: The method of setting the spatial position coordinates and the external dimensions of the external wall to be inspected and planning the flight path of the drone using a grid method based on the area and shape of the external wall to be inspected includes: The entire grid is divided into discrete parts according to the grid size w, and each grid is identified by a unique index identifier i. For any point on the exterior wall to be inspected, its coordinates are divided by the set grid size, and the integer part is taken to obtain the grid index; The obstacles are represented by bounding boxes, the size of each obstacle is increased by L, and its bounding box is calculated, where L is the width of the current device, each obstacle is identified by an obstacle index, and the corresponding obstacle index is taken as the height h of the bounding box, which is saved in all external wall grids that intersect with the bounding box.

3. The method for detecting internal defects of an external wall external insulation layer carried by an unmanned aerial vehicle according to claim 2 is characterized in that: The UAV scans and detects each grid point one by one according to the flight path to obtain detection data. At the same time, the distance between the UAV and the wall is L. If the obstacle index s i = 0, the device collects data at this point. If s i =h, turn off the device and stop data collection, and the distance between the drone and the outer wall is h+L.

4. The method for detecting internal defects of an external wall external insulation layer carried by an unmanned aerial vehicle according to claim 3 is characterized in that: The UAV scans and detects each grid point point by point in sequence according to the flight path to obtain detection data, and further includes: External wall detection equipment transmission power P L , the safe distance L between the drone and the outer wall, the pre-calibrated transmission power P0 and distance R0, and adjust according to the following formula:

5. The method for detecting internal defects of an external wall external insulation layer carried by an unmanned aerial vehicle according to claim 4 is characterized in that: The processing of the detection data to obtain actual defect measurement data of the exterior wall to be inspected includes: Continuously sample the received baseband analog signal at each frequency point to obtain a digital baseband signal sequence; Performing windowed inverse Fourier transform on the digital baseband signal sequence to obtain the amplitude and phase angle values ​​of the time domain discrete signal; Performing DC offset removal and bandpass filtering operations on the amplitude sequence of the time-domain discrete signal to obtain scanning data, wherein multiple channels of the scanning data constitute two-dimensional test data; Complete interface background clutter elimination processing for the two-dimensional test data of each measuring line; Perform gain adjustment on the current two-dimensional test data to make the target features more prominent; The gain-adjusted two-dimensional test data is input into the pre-trained deep learning network for fault identification, and the target detection area containing the void defect is obtained; The target detection area is subjected to image binarization processing and positioning operation to extract the position, height position and width of the target.

6. The method for detecting internal defects of an external wall external insulation layer carried by an unmanned aerial vehicle according to claim 5 is characterized in that: The gain adjustment of the current two-dimensional test data to make the target feature more prominent includes: For the processing area of ​​the two-dimensional test data, take the maximum absolute value of each row and connect them to obtain the corresponding calculation function; Calculating the extreme value envelope of the calculation function; Taking the inverse of the extreme value envelope and multiplying it by a preset coefficient to obtain a gain function; The two-dimensional test data is processed using the gain function to obtain gain-adjusted data.

7. An unmanned aerial vehicle-mounted device for detecting internal defects of an external wall thermal insulation layer, characterized in that: include: The preprocessing unit is configured to set the spatial position coordinates and the external dimensions of the external wall to be inspected, and plan the flight path of the UAV by using a grid method based on the area and shape of the external wall to be inspected; The acquisition unit is configured as a drone that scans and detects each grid point point by point according to the flight path to obtain detection data; The data processing unit is configured to process the detection data to obtain actual defect measurement data of the exterior wall to be inspected.

8. An electronic device, characterized in that: The electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the method for detecting internal defects of the external insulation layer of an external wall carried by an unmanned aerial vehicle as claimed in any one of claims 1 to 7 are implemented.

9. A computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method for detecting internal defects of an external insulation layer of an external wall carried by an unmanned aerial vehicle as claimed in any one of claims 1 to 7.

Images