Rapid detection method and device for exterior wall facing of building
By analyzing the amplitude fluctuations, frequency, and energy distribution of acoustic signals, a coupled fluctuation degree sequence was constructed, which solved the problem of misidentification of damp exterior wall finishes and achieved high precision and reliability in detecting hollow exterior wall finishes.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-24
- Publication Date
- 2026-03-24
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing acoustic signal detection methods are prone to misidentifying normal exterior walls as hollow areas when the exterior wall finish is damp, resulting in reduced detection accuracy and an inability to effectively distinguish between damp and hollow areas.
By analyzing the amplitude fluctuation, frequency components, and frequency domain energy distribution of acoustic signals, a coupled fluctuation sequence is constructed. Combining Gaussian significance and difference fusion, the hollow characteristic evaluation value is determined, enabling accurate detection of exterior wall finishes.
It improves the sensitivity and accuracy of detecting hollow areas in exterior wall finishes, effectively distinguishing between damp and hollow areas, and enhancing the reliability and precision of the detection.
Smart Images
Figure CN121725820A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of external wall hollowness detection technology, specifically to a rapid detection method and device for building external wall cladding. Background Technology
[0002] With the development of the construction industry, high-rise buildings are ubiquitous. For aesthetic purposes, some buildings are coated with plaster or covered with tiles on their exterior walls. These materials also have advantages such as corrosion resistance, wind resistance, and rainproofing. Therefore, decorative bricks are used on the exterior walls of modern buildings. However, with long-term erosion from wind and rain, these finishing materials may develop serious hollow spots on the exterior wall surface, causing the plaster or tiles to fall off, creating safety hazards. Therefore, timely detection and repair of hollow spots on the exterior wall finish before they fall off can effectively ensure the safety of the exterior wall finish.
[0003] Acoustic signal detection technology is one of the current methods for rapid detection of exterior wall finishes. By analyzing the acoustic signal characteristics of different detection areas of the exterior wall finish, normal exterior wall finishes and hollow exterior wall finishes can be identified. However, existing methods lack analysis of the coupling of hollow signals and only work well for dry exterior wall finishes. If a local area of the building's exterior wall finish is damp, the moisture on the wall surface will interfere with the acoustic signal characteristics of the normal wall surface. For example, the amplitude and energy of the acoustic signal will decrease, forming a "quasi-hollow" feature. This can lead to the misidentification of a damp, normal exterior wall finish as a hollow exterior wall finish when using acoustic signal detection, thus reducing the accuracy of the detection of hollow exterior wall finishes. Summary of the Invention
[0004] To address the aforementioned technical problems, the purpose of this application is to provide a rapid detection method and apparatus for building exterior wall finishes, the specific technical solution of which is as follows: In a first aspect, embodiments of this application provide a rapid detection method for building exterior wall finishes, the method comprising the following steps: Acoustic signals were collected at various sampling points on the building's exterior wall finish. Analyze the amplitude fluctuation of the acoustic signal, as well as the frequency components and the dispersion of energy distribution in the frequency domain, to determine the coupling fluctuation of the acoustic signal at each sampling point; A sliding window is set for each sampling point. Within the sliding window, based on the positional relationship between the sampling points, a coupling fluctuation sequence is constructed, with the coupling fluctuation of the acoustic signal at the center sampling point within the sliding window as the intermediate element. The hollow characteristic evaluation value of each sampling point is determined by the difference in coupling fluctuation of the acoustic signal between each sampling point and the other sampling points within the sliding window, as well as the degree to which each coupling fluctuation sequence conforms to a normal distribution. Hollow spots are detected in building exterior wall finishes based on the numerical value of the hollow spot characteristic evaluation value.
[0005] In one embodiment, determining the coupling fluctuation of the acoustic signal at each sampling point includes: The acoustic signal at each sampling point is divided into segments, and the mean of all amplitudes of the acoustic signal in each segment is calculated and denoted as the first mean. The mean of the first mean of all segments of the acoustic signal at each sampling point is determined and denoted as the second mean. The dispersion of the first mean of all segments of the acoustic signal at each sampling point is determined. Based on the dispersion and the second mean, the amplitude fluctuation of the acoustic signal at each sampling point is determined. The acoustic signal at each sampling point is transformed in the frequency domain, and the number of peaks in its spectrum and the energy distribution in each sub-band are counted. The energy entropy of the energy in the bandwidth of all sub-bands in the spectrum is calculated. The coupling fluctuation is positively correlated with the amplitude fluctuation, the number of peaks, and the energy entropy.
[0006] In one embodiment, the ratio of the number of spikes at each sampling point to the maximum value among all the number of spikes at all sampling points is calculated, and multiplied by the amplitude fluctuation and the energy entropy to obtain the coupling fluctuation of the acoustic signal at each sampling point.
[0007] In one embodiment, the amplitude fluctuation of the acoustic signal at each sampling point is the ratio of the dispersion to the second mean.
[0008] In one embodiment, determining the respective coupled volatility sequences includes: Within the sliding window of any sampling point, the coupling volatility corresponding to the horizontal sampling point, vertical sampling point, 45° oblique sampling point, and 135° oblique sampling point of any sampling point are sequentially arranged in order of position to form each coupling volatility sequence, wherein each coupling volatility sequence includes the coupling volatility corresponding to any sampling point.
[0009] In one embodiment, determining the hollow feature evaluation value of each sampling point includes: The Gaussian significance of the acoustic signal at each sampling point is calculated by determining the degree to which the coupling fluctuation sequence corresponding to the acoustic signal at each sampling point conforms to a normal distribution; the differences between each sampling point and all other sampling points within its sliding window are then fused to obtain a first fusion value. The hollow drum feature evaluation value is determined by combining the first fusion value, the Gaussian saliency, and the coupling fluctuation of the acoustic signal at each sampling point.
[0010] In one embodiment, the hollow drum feature evaluation value is the product of the first fusion value, the Gaussian saliency, and the coupling fluctuation of the acoustic signal at each sampling point.
[0011] In one embodiment, determining the Gaussian saliency includes: For each sampling point, calculate the first-order difference sequence of each coupled fluctuation sequence. Take the negative of the second half of the elements in the first-order difference sequence and denote the entire first-order difference sequence as a Gaussian sequence. Calculate the ratio of the total number of positive elements in the Gaussian sequence of all coupled fluctuation sequences of each sampling point to the total number of elements in all Gaussian sequences, and use this ratio as the Gaussian saliency of the acoustic signal at each sampling point.
[0012] In one embodiment, the hollow detection of the building exterior wall cladding includes: Threshold segmentation is performed on the hollow feature evaluation values of all sampling points of the building exterior wall cladding. Sampling points with hollow feature evaluation values greater than or equal to the segmentation threshold are marked as 1, and sampling points with hollow feature evaluation values less than the segmentation threshold are marked as 0. Connected component extraction is performed on the marked values of all sampling points of the building exterior wall cladding to obtain each hollow area of the building exterior wall cladding.
[0013] Secondly, embodiments of this application also provide a rapid detection device for building exterior wall finishes, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of any of the methods described above.
[0014] This application has at least the following beneficial effects: This application analyzes the amplitude fluctuation, frequency components, and frequency domain energy distribution of acoustic signals to determine the coupling fluctuation of acoustic signals at each sampling point. This reflects the stability of the acoustic signals at each sampling point under coupling effects, helping to distinguish the differences in acoustic signal coupling effects between hollow areas and damp areas of building exterior wall finishes, and improving the sensitivity of hollow area detection. Furthermore, by constructing a coupling fluctuation sequence, the hollow characteristic evaluation value of each sampling point is determined, reflecting the probability that each sampling point belongs to a hollow area. This solves the interference caused by the "quasi-hollow" characteristics of acoustic signals in damp areas of exterior wall finishes on hollow detection, improving the accuracy and reliability of hollow detection of building exterior wall finishes, and providing accurate feedback for building safety monitoring. Attached Figure Description
[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0016] Figure 1 A flowchart illustrating the steps of a rapid detection method for building exterior wall finishes provided in one embodiment of this application; Figure 2A flowchart for detecting hollow areas in building exterior wall finishes. Detailed Implementation
[0017] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a rapid detection method and apparatus for building exterior wall cladding proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.
[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.
[0019] The following description, in conjunction with the accompanying drawings, details the specific scheme of a rapid detection method and device for building exterior wall finishes provided in this application.
[0020] Please see Figure 1 The diagram illustrates a flowchart of a rapid detection method for building exterior wall finishes according to an embodiment of this application. The method includes the following steps: S1 collects acoustic signals at various sampling points on the building's exterior wall finish and performs preprocessing.
[0021] This embodiment employs an onboard acoustic signal collection device on a negative pressure wall-climbing robot to automatically collect acoustic signals generated by the friction between the ball bearings and the building wall. The acoustic signal collection device mainly consists of an array of sound sensors. The robot's travel path is pre-defined with several 5cm × 5cm grids based on the size and shape of the building's exterior wall. Each time, the robot travels along the center line of the grid in an S-shaped path, collecting acoustic signals from all grids and recording them as the acoustic signal at each sampling point. Each grid is marked as a sampling point. The acoustic signal sampling frequency is set to 20kHz, and the robot's travel speed is controlled to ensure that the acoustic signal collection time within each grid is 1 second. The grid size, acoustic signal sampling frequency, and the acoustic signal collection time for each grid can be set by the implementer according to the actual situation.
[0022] It should be noted that in this embodiment, the wall-climbing robot moves vertically along the vertical center line of the grid, and when it needs to turn at the edge of the wall, it makes an S-shaped turn along the horizontal center line of the grid. Implementers can set the data collection path of the wall-climbing robot according to actual conditions, for example, moving horizontally along the horizontal center line of the grid.
[0023] To avoid excessive noise in the acoustic signal of the building's exterior wall finish due to ambient noise, this embodiment uses a wavelet threshold denoising algorithm to denoise the acoustic signal collected from each grid, obtaining an acoustic signal with as little noise as possible. Wavelet threshold denoising is a well-known existing technology, and implementers can choose other feasible signal denoising algorithms. This embodiment does not impose any restrictions on this.
[0024] S2, analyze the amplitude fluctuation of the sound signal, as well as the frequency components of the sound signal and the dispersion of its energy distribution in the frequency domain, to determine the coupling fluctuation of the sound signal at each sampling point.
[0025] Currently, when using acoustic signals to detect hollow areas in building exterior wall finishes, the main approach is to analyze simple differences in acoustic signal amplitude and energy at each detection point. For example, in hollow areas of building exterior wall finishes, the presence of air (with greater damping) absorbs some of the sound wave signal, resulting in lower amplitude and faster energy attenuation of the collected sound signal. Conversely, the acoustic signals from normal exterior wall finishes show the opposite behavior, and statistical analysis can quickly identify hollow acoustic signals. However, this detection method lacks in-depth feature extraction, such as differences in acoustic signal coupling characteristics. Consequently, current methods for detecting hollow building exterior wall finishes are primarily effective for dry finishes, but less effective for areas with dampness, potentially misidentifying damp finishes as hollow ones. This is because damp areas contain a large amount of moisture, which increases the damping of the exterior wall finish and the absorption rate of the sound signal, causing the amplitude and energy of the acoustic signals collected from damp finishes to exhibit "hollow-like" characteristics.
[0026] Although the acoustic signal of damp exterior wall finishes exhibits a "quasi-hollow" characteristic, the internal structure of a normal damp exterior wall finish remains dense, while a hollow exterior wall finish lacks rigid support. This leads to varying degrees of coupling when the ball bearings of the wall-climbing robot travel to the hollow area. In damp, supported exterior wall finishes, due to their dense internal structure, the ball bearings maintain stable contact with the wall, resulting in relatively stable acoustic signal coupling. The amplitude and energy of the acoustic signal decay gradually, with a concentrated dominant frequency and a single peak. However, in hollow areas, the lack of dense, rigid base support causes significant wall vibration when the wall-climbing robot travels to these areas. The coupling between the ball bearings and the wall in these areas can cause the ball bearings to bounce, resulting in poor coupling stability and dispersed amplitude, peaks, and energy distribution of the collected acoustic signal. Therefore, this embodiment, by analyzing the coupling stability of the acoustic signal, helps improve the accuracy of hollow detection on damp exterior wall finishes.
[0027] Specifically, for the acoustic signal at each sampling point, noise cannot be completely removed. High-frequency transient noise in the environment may cause a momentary surge in amplitude, resulting in some spurious fluctuations. The amplitude then quickly returns to the normal range. If all amplitude points in the acoustic signal are directly analyzed for amplitude fluctuations, it may be impossible to distinguish between spurious and genuine fluctuations. Therefore, this embodiment divides the acoustic signal at each sampling point into segments. Specifically, a sliding window with a length of 10ms and a sliding step of 5ms is set. Each sliding window corresponds to one segment of the acoustic signal. The mean of all amplitudes in each sliding window is obtained to smooth out spurious amplitude fluctuations, thus constructing an amplitude sequence for the acoustic signal at each sampling point. ,in, This represents the amplitude sequence of the acoustic signal at the i-th sampling point. Indicates the number of sliding windows. The first mean value is the average of all acoustic signal amplitudes within the m-th sliding window of the acoustic signal at the i-th sampling point. The length and step size of the sliding window can be set by the implementer according to the actual situation; this embodiment does not impose any restrictions on this.
[0028] Furthermore, the mean of the amplitude sequence of the acoustic signal at each sampling point is calculated and denoted as the second mean. The dispersion of the amplitude sequence of the acoustic signal at each sampling point is also calculated. Based on the dispersion and the second mean, the amplitude fluctuation of the acoustic signal at each sampling point is determined. The dispersion can be calculated using methods such as standard deviation or coefficient of variation; this embodiment does not impose any limitations on this.
[0029] In this embodiment, the aforementioned dispersion can specifically be the standard deviation. Specifically, the standard deviation of the amplitude sequence can be calculated first, and then the ratio of the standard deviation to the second mean can be used as the amplitude fluctuation of the sound signal at each sampling point to represent the degree of fluctuation of the sound signal amplitude. The sound signal in the hollow area has a sawtooth waveform due to the bouncing vibration of the ball bearings, and the amplitude fluctuation is relatively large.
[0030] Furthermore, the acoustic signal collected at each sampling point is used as input, and a Fast Fourier Transform (FFT) is used to output the corresponding spectrum. Then, the AMPD (Automatic multiscale-based peak detection) algorithm is used to obtain the peak points of the spectrum. To filter out prominent peaks, the amplitude of all peak points is used as input to the Otsu's inter-class variance algorithm to obtain a threshold. Peak points with amplitudes greater than the threshold are selected, which represent the number of peaks in the spectrum corresponding to that acoustic signal. Similarly, the number of peaks for each sampling point is calculated. A higher number of peaks indicates a "split" of the main frequency, resulting in energy distribution across multiple frequency bands, higher energy dispersion, and a larger energy entropy value. Specifically, in this embodiment, the sub-band bandwidth is set to 100Hz to divide the effective frequency. The effective frequency range in this embodiment is [value missing]. The energy within each sub-band bandwidth of the spectrum is obtained, and then the energy entropy of the energy within all sub-band bandwidths of the spectrum is calculated. The Fast Fourier Transform, the AMPD peak detection algorithm, the calculation of energy within the sub-band bandwidth of the spectrum, and the entropy calculation are all existing well-known techniques, and the specific process will not be described in detail.
[0031] Based on the above analysis, this embodiment calculates the coupling fluctuation of the acoustic signal at each sampling point to characterize the stability of the acoustic signal under coupling effects. The specific expression is as follows: In the formula, This represents the coupling fluctuation of the acoustic signal at the i-th sampling point. This represents the amplitude fluctuation of the acoustic signal at the i-th sampling point. This represents the number of spikes in the acoustic signal at the i-th sampling point. This represents the maximum number of spikes in the acoustic signal across all sampling points. This represents the energy entropy of the acoustic signal at the i-th sampling point. The greater the amplitude fluctuation, the greater the number of peaks, and the greater the energy entropy of the acoustic signal, the greater the coupling fluctuation. The larger the value, the more likely the sampling point of the acoustic signal belongs to a hollow region. It should be noted that in the calculation of coupling fluctuation, when... When, the coupling fluctuation of the acoustic signal at the i-th sampling point is calculated as follows: .
[0032] S3. Set a sliding window for each sampling point. Within the sliding window, construct a series of coupling fluctuations with the coupling fluctuation of the acoustic signal at the center sampling point in the sliding window as the intermediate element, based on the positional relationship between the sampling points. Determine the hollow characteristic evaluation value of each sampling point by the difference in coupling fluctuation of the acoustic signal between each sampling point and the other sampling points in the sliding window, as well as the degree to which each coupling fluctuation sequence conforms to a normal distribution.
[0033] In actual building exterior wall cladding applications, there may be uneven or thin walls, which may cause unstable coupling to occur even on normal walls. This can be easily confused with the unstable coupling signal of hollow exterior wall cladding. Therefore, using only coupling fluctuation to detect hollow exterior wall cladding may not be accurate enough.
[0034] Even if unstable coupling occurs on a normal wall surface, the resulting coupling fluctuation is relatively large. However, due to the uniformity inside the wall, the difference in coupling fluctuation between adjacent sampling points is small. In contrast, the hollow areas of building exterior wall finishes are irregular. Usually, the coupling instability is relatively large at the center of the hollow area because there is a lack of rigid support around it. The instability gradually decreases towards the surrounding area, which conforms to the characteristics of a two-dimensional Gaussian distribution.
[0035] Specifically, due to the rectangular shape of the wall surface, a coupling fluctuation matrix is constructed based on the sampling point locations of all sampling points on the building's exterior wall surface. Since some walls may have locations that climbing robots cannot detect, such as windows, the corresponding positions in the coupling fluctuation matrix are filled with 0 values. In this embodiment, each sampling point is used as the center to set... A sliding window with a step size of 1 is used. For missing elements in the sliding window at sampling points at the boundary positions, a value of 0 is used to fill the missing elements, ensuring that the elements in the sliding window at each sampling point are equal. The size of the sliding window can be set by the implementer according to the actual situation; this embodiment does not impose any restrictions on it.
[0036] For the sliding window of the i-th sampling point, the coupling fluctuations of the acoustic signals of the i-th sampling point and all sampling points in the horizontal direction are arranged into a coupling fluctuation sequence according to the positional order between the sampling points. That is, the coupling fluctuation of the acoustic signal of the i-th sampling point is located in the middle of the coupling fluctuation sequence. Similarly, the coupling fluctuations of the acoustic signals of the i-th sampling point and all sampling points in the vertical direction are arranged into a coupling fluctuation sequence according to the positional order between the sampling points. The coupling fluctuations of the acoustic signals of the i-th sampling point and all sampling points in the 45° diagonal direction are arranged into a coupling fluctuation sequence according to the positional order between the sampling points. The coupling fluctuations of the acoustic signals of the i-th sampling point and all sampling points in the 135° diagonal direction are arranged into a coupling fluctuation sequence according to the positional order between the sampling points. Thus, each sampling point can obtain 4 coupling fluctuation sequences. In this embodiment, each coupling fluctuation sequence contains 5 elements, with the middle element being the coupling fluctuation of the acoustic signal of the i-th sampling point.
[0037] Furthermore, each coupled volatility sequence at each sampling point is used as input. The first-order difference sequence of each coupled volatility sequence is calculated. The negative values of the latter half of the elements in the first-order difference sequence are taken, and the entire first-order difference sequence after taking the negative values is denoted as a Gaussian sequence. The ratio of the total number of positive elements in the Gaussian sequence of all coupled volatility sequences at each sampling point to the total number of elements in the Gaussian sequence of all coupled volatility sequences at each sampling point is calculated as the Gaussian significance of the acoustic signal at each sampling point, reflecting the degree to which the coupled volatility of the acoustic signal at the sampling points within the sliding window conforms to a Gaussian distribution. Specifically, the coupled volatility sequence constructed through the center of the sliding window cannot contain the coupled volatility values of all sampling points within the sliding window. Since the unincluded coupled volatility values are only a minority and are discretely distributed around the outermost ring of the sliding window, the distribution characteristics within the inner ring of the sliding window are unaffected. Therefore, this embodiment calculates the distribution of a portion of the coupled volatility within the sliding window to evaluate whether the overall sliding window conforms to a Gaussian-like distribution.
[0038] Based on the above analysis, a hollow feature evaluation value is calculated at each sampling point to characterize the probability that the sampling point belongs to a hollow area. Specifically: The difference in coupling fluctuation of the acoustic signal at each sampling point and all other sampling points within its sliding window is fused to obtain the first fused value; The hollow drum feature evaluation value is determined by combining the first fusion value, the Gaussian saliency, and the coupling fluctuation of the acoustic signal at each sampling point.
[0039] It should be noted that difference indicates the degree of difference between two variables, and can be calculated using methods such as the absolute value of the difference, the square of the difference, or the ratio. Fusion indicates combining multiple variables, and can be calculated using methods such as multiplication, addition, a combination of addition and multiplication, or taking the mean.
[0040] In this embodiment, the specific expression for the hollow feature evaluation value of each sampling point is as follows: In the formula, This represents the hollow feature evaluation value of the i-th sampling point. This represents the coupling fluctuation of the acoustic signal at the i-th sampling point. This represents the mean of the absolute values of the differences between the coupling fluctuation of the acoustic signal at the i-th sampling point and the coupling fluctuation of the acoustic signals at all other sampling points within its sliding window. This indicates taking the absolute value. This represents the Gaussian significance of the acoustic signal at the i-th sampling point. The higher the coupling volatility of the acoustic signal at the i-th sampling point, the greater the Gaussian significance, and the more the coupling volatility sequence conforms to the two-dimensional Gaussian distribution with the characteristic of rising and then falling. The larger the value, the more likely the sampling point is to be a hollow area. Among them, [the value is missing here]. This is recorded as the first fusion value.
[0041] S4, based on the numerical value of the hollow feature evaluation value, performs hollow detection on the building exterior wall cladding.
[0042] The hollow feature evaluation values of all sampling points of the building exterior wall cladding are used as input. A threshold segmentation algorithm is used to output a segmentation threshold. If a hollow feature evaluation value is greater than or equal to the segmentation threshold, the corresponding sampling point is determined to belong to the hollow building exterior wall cladding area and is marked as 1. Sampling points with hollow feature evaluation values less than the segmentation threshold are determined to belong to the normal building exterior wall cladding area and are marked as 0. A hollow label matrix is constructed. In this embodiment, the threshold segmentation algorithm adopts the maximum inter-class variance algorithm. Implementers can use other existing feasible threshold segmentation algorithms, and this embodiment does not impose any restrictions on this.
[0043] Connected component extraction is performed on the hollow marking matrix to obtain each connected component with a marking value of 1. These connected components are then used as hollow areas in the building exterior wall cladding, thus achieving hollow detection of the building exterior wall cladding. Connected component extraction is a well-known existing technology, and this embodiment will not elaborate on it in detail. The flowchart for detecting hollow areas in building exterior wall cladding is as follows. Figure 2 As shown.
[0044] Based on the same inventive concept as the above method, this application embodiment also provides a rapid detection device for building exterior wall cladding, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described rapid detection methods for building exterior wall cladding.
[0045] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, specific embodiments of this specification have been described above. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.
[0046] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
[0047] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this application should be included within the protection scope of this application.
Claims
1. A rapid detection method for building exterior wall finishes, characterized in that, The method includes the following steps: Acoustic signals were collected at various sampling points on the building's exterior wall finish. Analyze the amplitude fluctuation of the acoustic signal, as well as the frequency components and the dispersion of energy distribution in the frequency domain, to determine the coupling fluctuation of the acoustic signal at each sampling point; A sliding window is set for each sampling point. Within the sliding window, based on the positional relationship between the sampling points, a coupling fluctuation sequence is constructed, with the coupling fluctuation of the acoustic signal at the center sampling point within the sliding window as the intermediate element. The hollow characteristic evaluation value of each sampling point is determined by the difference in coupling fluctuation of the acoustic signal between each sampling point and the other sampling points within the sliding window, as well as the degree to which each coupling fluctuation sequence conforms to a normal distribution. Hollow spots are detected in building exterior wall finishes based on the numerical value of the hollow spot characteristic evaluation value.
2. The rapid detection method for building exterior wall finishes as described in claim 1, characterized in that, Determining the coupling fluctuation of the acoustic signal at each sampling point includes: The acoustic signal at each sampling point is divided into segments, and the mean of all amplitudes of the acoustic signal in each segment is calculated and denoted as the first mean. The mean of the first mean of all segments of the acoustic signal at each sampling point is determined and denoted as the second mean. The dispersion of the first mean of all segments of the acoustic signal at each sampling point is determined. Based on the dispersion and the second mean, the amplitude fluctuation of the acoustic signal at each sampling point is determined. The acoustic signal at each sampling point is transformed in the frequency domain, and the number of peaks in its spectrum and the energy distribution in each sub-band are counted. The energy entropy of the energy in the bandwidth of all sub-bands in the spectrum is calculated. The coupling fluctuation is positively correlated with the amplitude fluctuation, the number of peaks, and the energy entropy.
3. The rapid detection method for building exterior wall finishes as described in claim 2, characterized in that, Calculate the ratio of the number of spikes at each sampling point to the maximum value among all sampling points, and multiply it by the amplitude fluctuation and the energy entropy to obtain the coupling fluctuation of the acoustic signal at each sampling point.
4. The rapid detection method for building exterior wall finishes as described in claim 2, characterized in that, The amplitude fluctuation of the acoustic signal at each sampling point is the ratio of the dispersion to the second mean.
5. The rapid detection method for building exterior wall finishes as described in claim 1, characterized in that, The determination of each coupled volatility sequence includes: Within the sliding window of any sampling point, the coupling volatility corresponding to the horizontal sampling point, vertical sampling point, 45° oblique sampling point, and 135° oblique sampling point of any sampling point are sequentially arranged in order of position to form each coupling volatility sequence, wherein each coupling volatility sequence includes the coupling volatility corresponding to any sampling point.
6. The rapid detection method for building exterior wall finishes as described in claim 1, characterized in that, The determination of the hollow feature evaluation value of each sampling point includes: The Gaussian significance of the acoustic signal at each sampling point is calculated by determining the degree to which the coupling fluctuation sequence corresponding to the acoustic signal at each sampling point conforms to a normal distribution; the differences between each sampling point and all other sampling points within its sliding window are then fused to obtain a first fusion value. The hollow drum feature evaluation value is determined by combining the first fusion value, the Gaussian saliency, and the coupling fluctuation of the acoustic signal at each sampling point.
7. The rapid detection method for building exterior wall finishes as described in claim 6, characterized in that, The hollow drum feature evaluation value is the product of the first fusion value, the Gaussian saliency, and the coupling fluctuation of the acoustic signal at each sampling point.
8. A rapid detection method for building exterior wall finishes as described in claim 6, characterized in that, The determination of the Gaussian significance includes: For each sampling point, calculate the first-order difference sequence of each coupled fluctuation sequence. Take the negative of the second half of the elements in the first-order difference sequence and denote the entire first-order difference sequence as a Gaussian sequence. Calculate the ratio of the total number of positive elements in the Gaussian sequence of all coupled fluctuation sequences of each sampling point to the total number of elements in all Gaussian sequences, and use this ratio as the Gaussian saliency of the acoustic signal at each sampling point.
9. A rapid detection method for building exterior wall finishes as described in claim 1, characterized in that, The process of detecting hollow areas in the exterior wall finishes of buildings includes: Threshold segmentation is performed on the hollow feature evaluation values of all sampling points of the building exterior wall cladding. Sampling points with hollow feature evaluation values greater than or equal to the segmentation threshold are marked as 1, and sampling points with hollow feature evaluation values less than the segmentation threshold are marked as 0. Connected component extraction is performed on the marked values of all sampling points of the building exterior wall cladding to obtain each hollow area of the building exterior wall cladding.
10. A rapid detection device for building exterior wall finishes, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-9.