CN117169086A - Method, medium and system for detecting construction quality of underground waterproof layer of building - Google Patents

Abstract

The invention provides a construction quality detection method, medium and system for the underground waterproof layer of a building, belonging to the technical field of underground waterproof layer construction. The method includes: acquiring multiple hygrometers in a closed basement environment constructed after the completion of waterproof layer construction. Collect the air humidity data in the basement, record the humidity data of air humidity changing with time, and form an air humidity matrix; obtain the reflective images of the basement walls, floors, and ceilings, and establish a set of reflective images that change with time, which is recorded as a reflective image set; For the reflective image, collect reflective point information and establish reflective point data; use the pre-trained water permeability model to calculate the humidity matrix and reflective matrix at each moment to obtain the amount of water permeability at the corresponding moment; based on the leakage location and water permeability The quantity is used as the inspection and evaluation index for the waterproof layer construction in the basement and is sent to the construction personnel. This method has a wide detection range and will not cause damage to the waterproof layer.

Description

A construction quality inspection method, medium and system for the underground waterproof layer of a building

Technical field

The invention belongs to the technical field of underground waterproof layer construction. Specifically, it relates to a construction quality detection method, medium and system for the underground waterproof layer of a building.

Background technique

Basement waterproofing technology is widely used in construction projects. By applying waterproof coating or setting up a waterproof layer on the outside of the basement, it can effectively prevent the leakage and erosion of underground water to the basement and ensure the functionality of the basement. However, due to the uneven construction quality of the waterproof layer, leakage problems occur from time to time, which not only affects the function of the basement, but may also cause foundation stability problems. Therefore, it is very necessary to detect and evaluate the construction quality of basement waterproofing layer.

Traditional waterproof layer quality inspection mostly relies on manual visual inspection and simple instrument testing. These methods have problems such as large detection blind areas and difficulty in ensuring accuracy. For example, the visual inspection method: workers visually inspect the waterproof layer to look for obvious defects, such as cracks, potholes, damage, etc. This is the most basic quality inspection method. Tapping test method: Use a small hammer to tap the waterproof layer, and use the sound to determine whether there are holes inside the waterproof layer. This method can identify local obvious quality problems. Water penetration test method: Spray a water column on a local area of the waterproof layer and observe whether the water penetrates to determine the waterproofing effect. But the test is more destructive. Current testing method: Use a current tester to apply current to both ends of the waterproof layer to test the resistance value. Abnormal resistance value indicates a damaged point. Water pressure testing method: Establish a water pressure testing system to maintain the waterproof layer under a certain water pressure for a period of time to check whether leakage occurs. This method is more accurate but expensive.

The above methods have problems such as limited detection range or damage to the waterproof layer.

Contents of the invention

In view of this, the present invention provides a construction quality detection method, medium and system for the underground waterproof layer of a building, which can solve the problem of limited detection range or limited detection range of the waterproof layer in the existing technology when detecting the construction quality of the underground waterproof layer in saline-alkali geology. Technical issues causing disruption.

The present invention is implemented as follows:

A first aspect of the present invention provides a method for testing the construction quality of underground waterproofing layers of buildings, which includes the following steps:

S10. In the closed basement environment constructed after the construction of the waterproof layer, obtain the air humidity data in the basement collected by multiple hygrometers, record the humidity data of the air humidity changing with time, and form an air humidity matrix, wherein the multiple hygrometers The hygrometers are evenly distributed in the basement, and the humidity data includes the spatial coordinates of the hygrometer and the humidity value;

S20. Obtain the reflective images of the basement wall, floor and ceiling, and establish a set of reflective images that change over time, which is recorded as a reflective image set;

S30. For the reflective image, collect reflective point information and establish reflective point data. The reflective point data includes the spatial coordinates, area, and reflective intensity of the geometric center of the reflective point; collect all reflective points at each moment in the reflective image. The reflective point data generated by the image generates a reflective matrix and determines the leakage location based on the reflective area;

S40. Use the pre-trained water permeability model to calculate the humidity matrix and reflection matrix at each moment to obtain the water permeability amount at the corresponding moment;

S50. Use the leakage location and water permeability as the basement waterproofing layer construction inspection and evaluation indicators, and send them to the construction personnel.

On the basis of the above technical solution, the construction quality detection method of the underground waterproof layer of a building of the present invention can also be improved as follows:

Wherein, the plurality of hygrometers at least include a plurality of hygrometers arranged close to each wall of the basement.

Further, the steps of obtaining reflective images of basement walls, floors and ceilings and establishing a collection of reflective images changing over time specifically include:

Set up multiple imaging devices in the basement to regularly collect images of walls, floors and ceilings;

Preprocess the collected images, including denoising, enhancement, and correction;

Convert the preprocessed image to HSV space, and extract the reflective area based on the brightness threshold;

Perform morphological analysis on the extracted reflective areas, remove noise areas, and obtain effective reflective areas;

Calculate the geometric characteristics of each reflective area;

Retain appropriate reflective areas and obtain an effective set of reflective points in the image;

Merge the sets of reflective points in images at different times to establish a set of reflective points that change with time.

Further, the device for collecting reflective images includes a camera and a light source. When the camera takes pictures of the walls or the ground in the basement, the shooting angle is 45° with the wall or the ground; when the camera takes pictures of the walls or the ground of the basement, the The angle between the line connecting the light source and the shooting point and the shooting angle of the camera is greater than 30°.

Further, the steps of collecting reflective point information and establishing reflective point data specifically include:

Calculate the area, intensity, boundary contour and geometric center coordinates of each reflective point based on the reflective image;

Construct a feature matrix of reflective points at each moment, including area, intensity, contour, and coordinate information;

Merge the feature matrices at different times to obtain the global reflective feature matrix;

Match the reflective points on the global reflective feature matrix;

Use contour and coordinate information to match and track the movement of reflective points;

Determine the leakage area through the trace of reflective dots.

Among them, the steps of establishing and training the water permeable model specifically include:

Step 1. Establish a training data set, including the humidity matrix, reflective matrix obtained during the inspection of multiple historical underground waterproof layer construction projects, and the water permeability obtained by collecting water that penetrated into the basement;

Step 2. Use convolutional neural network to establish a prototype of the water permeability model;

Step 3. Use the training data set to train the prototype of the water permeability model to obtain the water permeability model; the input of the training is the historical humidity matrix and the reflective matrix, and the output of the training is the water permeability corresponding to the historical humidity matrix and the reflective matrix.

Among them, the step of using a pre-trained water permeability model to calculate the humidity matrix and reflection matrix at each moment to obtain the water permeability amount at the corresponding moment specifically includes:

Normalize the input data;

Input the humidity matrix and reflectance matrix into the water permeability model;

The model outputs the predicted water penetration at each moment.

Among them, the leakage location and water permeability are used as the detection and evaluation indicators of the waterproof layer construction in the basement, and are sent to the construction personnel in a graphical display.

A second aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores program instructions, and when the program instructions are run, they are used to perform the above-mentioned construction of an underground waterproof layer of a building. Quality testing methods.

A third aspect of the present invention provides a construction quality inspection system for underground waterproofing layers of buildings, which includes the above-mentioned computer-readable storage medium.

Compared with the existing technology, the beneficial effects of the construction quality detection method, medium and system of the underground waterproof layer of a building provided by the present invention are: the present invention constructs a three-dimensional humidity field by setting up a distributed hygrometer array in the basement, which can effectively Monitor the humidity distribution inside the basement and locate problem areas. At the same time, reflective information is obtained through image processing, and a water permeability evaluation index combined with a physical model of leakage is established to accurately detect leaks in the waterproof layer. The organic fusion of the two types of information forms a waterproof layer construction quality inspection method with complete information, controllable process, and intuitive results; this method has a wide detection range and will not cause damage to the waterproof layer. Specifically, the technical effects of the present invention include the following:

1. Wide detection coverage to avoid missed detections

This invention uses a distributed humidity sensor array and multi-angle camera equipment to monitor the entire basement space in an all-round way, covering every corner and avoiding the missed detection problem of traditional manual detection. Any small-scale leakage can be recorded in detail, fundamentally eliminating blind spots in missed detection.

2. The detection accuracy is high and quantitative analysis can be carried out

The invention is based on an image processing algorithm, which can accurately extract the characteristics of the reflective area and calculate the amount of leakage. Compared with simple qualitative visual detection, the present invention realizes quantitative monitoring and analysis of the leakage process, greatly improving detection accuracy.

3. Fast detection speed and real-time feedback

The image processing and model prediction process in the present invention can be quickly completed in the computer, and the results are fed back in real time, and the entire detection cycle is short. However, traditional manual detection takes too long and cannot achieve real-time performance. Rapid detection can provide support for construction personnel’s subsequent processing decisions.

4. Accurate leak location and direct repair guidance

The invention can not only quantitatively detect leakage, but also accurately locate the leakage location by constructing a three-dimensional humidity field, allowing maintenance personnel to directly deal with the problem area without expanding the maintenance scope, thus saving maintenance costs.

5. The process is controllable and can be continuously optimized

This method digitizes the entire detection process, making each step clear and controllable. The detection effect can also be improved by continuously optimizing algorithms and models, but the effect of traditional methods is difficult to continue to improve.

Description of drawings

In order to explain the technical solutions of the embodiments of the present invention more clearly, the drawings needed to be used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. , for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without exerting creative labor.

Figure 1 is a flow chart of a construction quality inspection method for the underground waterproof layer of a building provided by the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention.

As shown in Figure 1, it is a flow chart of a construction quality inspection method for the underground waterproof layer of a building provided by the first aspect of the present invention. This method includes the following steps:

S10. In the closed basement environment constructed after the construction of the waterproof layer, obtain the air humidity data in the basement collected by multiple hygrometers, record the humidity data of the air humidity changing with time, and form an air humidity matrix, in which multiple hygrometers Evenly distributed in the basement, the humidity data includes the spatial coordinates of the hygrometer and the humidity value;

S20. Obtain the reflective images of the basement wall, floor and ceiling, and establish a set of reflective images that change over time, which is recorded as a reflective image set;

S30. For the reflective image, collect reflective point information and establish reflective point data. The reflective point data includes the spatial coordinates, area, and reflective intensity of the geometric center of the reflective point; generate reflective point data from all reflective images at each moment in the reflective image. Generate a reflective matrix and determine the leakage location based on the reflective area;

S40. Use the pre-trained water permeability model to calculate the humidity matrix and reflection matrix at each moment to obtain the water permeability amount at the corresponding moment;

S50. Use the leakage location and water permeability as the basement waterproofing layer construction inspection and evaluation indicators, and send them to the construction personnel.

In the above technical solution, the plurality of hygrometers at least include a plurality of hygrometers arranged close to each wall of the basement.

Specifically, the implementation of step S10 is described as follows:

After the construction of the waterproof layer is completed, n hygrometers are evenly arranged in the basement to obtain air humidity data in the basement. Assume that the number of hygrometers is 1, 2,...,n, then the spatial coordinates of the i-th hygrometer are (x _i , y _i , z _i ). The data of each hygrometer is collected at time t ₁ , t ₂ ,..., t _m , where t _j represents the jth time point. The humidity value measured by the i-th hygrometer at time t _j is h _ij . Then an n×m humidity matrix H can be established:

Among them, the element h _ij in the i-th row and j-th column represents the humidity value of the i-th hygrometer at time t _j . In this way, the humidity matrix H contains the humidity values measured by all hygrometers at various time points.

In order to evaluate the construction quality of the waterproof layer, it is necessary to analyze the change pattern of humidity data. The average humidity at each moment can be calculated

Through the average humidity change curve, the humidity change trend in the basement can be judged. If the average humidity continues to rise, it means that the basement is gradually becoming humid and there may be a problem with the waterproofing layer.

In addition, the data change curve of each hygrometer can also be calculated to analyze whether the hygrometers at different locations have significantly different change patterns. Define the humidity change curve of the i-th hygrometer as h _i :

h _i =[h _i1 , h _i2 ,…, h _im ];

If the data fluctuations of some hygrometers are significantly greater than that of other hygrometers, that location may be a leak point.

Through the above analysis of the humidity matrix H, the construction quality of the waterproof layer in the basement can be preliminarily judged. If the average humidity is stable and the changing trends of each hygrometer reading are basically the same, it means that the waterproof layer is of good quality and can effectively block groundwater; if the average humidity continues to rise and individual hygrometer readings fluctuate violently, it means that there is a problem with the waterproof layer and further detection and location is needed. Leakage area.

In addition, the humidity matrix H also contains the spatial coordinate information of the hygrometer. In order to locate the leakage area more accurately, you can use the coordinate information of the hygrometer and use the spatial interpolation method to interpolate a three-dimensional humidity distribution field H (x, y, z) inside the basement:

H(x,y,z)＝f(h ₁₁ ,h ₁₂ ,…,h _nm ;x ₁ ,y ₁ ,z ₁ ;x ₂ ,y ₂ ,z ₂ ;…;x _n ,y _n ,z _n );

Among them, F represents the spatial interpolation function. According to the coordinates and humidity value of the sampling point, the humidity field is interpolated in the entire three-dimensional space. Commonly used spatial interpolation methods include inverse distance weighted interpolation, Kriging interpolation, radial basis function interpolation, etc.

For example, using a simple inverse distance weighted interpolation method, the humidity field can be expressed as:

Where p usually takes 2, |·| represents the Euclidean distance. In this way, the three-dimensional humidity distribution inside the basement can be obtained.

According to the three-dimensional humidity field H (x, y, z), the area with greater humidity can be intuitively determined to locate possible leakage points. Humidity distribution maps of slices at different heights can also be drawn to analyze the spatial distribution of humidity.

The above method makes use of the hygrometer distribution coordinate information contained in the humidity matrix. By establishing a three-dimensional humidity field, it can effectively analyze the humidity distribution inside the basement, detect areas with poor quality waterproofing layers, and provide information for subsequent accurate leak point location. in accordance with.

Further, in the above technical solution, the steps of obtaining reflective images of the basement walls, floors and ceilings and establishing a collection of reflective images changing over time include:

Set up multiple imaging devices in the basement to regularly collect images of walls, floors and ceilings;

Preprocess the collected images, including denoising, enhancement, and correction;

Convert the preprocessed image to HSV space, and extract the reflective area based on the brightness threshold;

Perform morphological analysis on the extracted reflective areas, remove noise areas, and obtain effective reflective areas;

Calculate the geometric characteristics of each reflective area;

Retain appropriate reflective areas and obtain an effective set of reflective points in the image;

Merge the sets of reflective points in images at different times to establish a set of reflective points that change with time.

Specifically, the implementation of step S20 is described as follows:

Set up multiple image acquisition devices (such as cameras) in the basement to obtain images of the basement walls, floors and ceilings. Assuming that the total number of images collected is n and taken at times t ₁ , t ₂ ,..., t _M , an image sequence can be established:

I＝{I ₁ ,I ₂ ,…,I _N };

Among them, the image I _k represents the time The k-th image taken, m _k ∈[1,M].

Preprocess each image I _k :

Next, detect and extract reflective areas from the preprocessed image:

1) Convert to HSV space

2) Extract reflective areas with higher brightness according to the threshold value

R_k＝{(x,y)|I^″{HSV}_k(x,y,3)>T_″{thresh}};

Where T _thresh represents the brightness threshold.

3) Perform morphological operations to extract connected areas and obtain the reflective area R′ _k .

4) Calculate the geometric characteristics of each reflective area, including area _Si , perimeter _Li , circularity, etc.

5) Retain reflective areas with larger area and higher circularity to generate a set of reflective areas Where r _i represents the i-th retained reflective area, and N _k is the number of reflective areas in the k-th image.

6) Calculate the geometric center coordinates (x _i , y _i ) of each reflective area as the reflective point.

Through the above image processing, the set of all reflective points in each image I _k can be obtained:

Finally, merge the reflective point sets in all images I _k :

P＝{P ₁ ,P ₂ ,…,P _N };

The above method uses image processing technologies such as image preprocessing, reflective area extraction, and geometric feature calculation to obtain the reflective point set of reflective images that change over time, providing data support for subsequent reflective information analysis.

Further, in the above technical solution, the device for collecting reflective images includes a camera and a light source. When the camera shoots the wall or floor of the basement, the shooting angle is 45° with the wall or ground; when the camera shoots the wall or floor of the basement, the light source The angle between the line connecting the shooting point and the camera shooting angle is greater than 30°.

Specifically, the arrangement of cameras and light sources is described as follows:

1.Camera arrangement

- Arrange multiple cameras in the basement to achieve a full range of shooting effects. Camera placement should cover walls, floors, and ceilings.

-The distance between the camera and the subject should be moderate, not only to ensure image clarity, but also to maximize the viewing range. Generally, a distance of 3-5 meters is more suitable.

- Try to keep the camera at a 45° angle to the wall to reduce the impact of reflection on the image.

2. Arrangement of lighting sources

-Use LED light source as the main lighting source, with a color temperature of 5000-6000K. The LED light source beam is more concentrated and easy to control.

-Light sources should be arranged evenly to avoid overly bright or dark areas. Multi-point lighting can be used.

- The angle between the line connecting the light source and the shooting point and the shooting angle of the camera should not be too small, usually above 30°, otherwise strong reflections will easily appear in the image.

-Additional lateral light sources are arranged specifically to form reflections. These light sources are at an angle to the main lighting.

-It is best to use harder light for all light sources to avoid forming a large area of reflection.

3.Others

-Use moisture-proof cameras and light sources to adapt to the hot and humid environment in the basement.

-Perform regular maintenance and calibration of equipment to ensure image quality.

-In addition to visible light photography, you can also try to use thermal imaging technology to obtain thermal information about water leakage.

Through reasonable equipment layout, clear images containing obvious reflections can be obtained, providing a basis for subsequent reflection analysis. At the same time, you also need to pay attention to the control of the light source and environment to ensure image quality.

Further, in the above technical solution, the steps of collecting reflective point information and establishing reflective point data specifically include:

Calculate the area, intensity, boundary contour and geometric center coordinates of each reflective point based on the reflective image;

Construct a feature matrix of reflective points at each moment, including area, intensity, contour, and coordinate information;

Merge the feature matrices at different times to obtain the global reflective feature matrix;

Match the reflective points on the global reflective feature matrix;

Use contour and coordinate information to match and track the movement of reflective points;

Determine the leakage area through the trace of reflective dots.

In step S20, the set of reflective points P={P ₁ , P ₂ ,..., P _N } extracted from the image at each time has been obtained. For each reflective point p _i =(x _i ,y _i )∈P _k , the following characteristics can be calculated:

1) Reflective area: A _i

Extract the number of pixels in the reflective area r _i to represent the reflective area:

A _i = N _pix (r _i );

2) Reflective intensity I _i :

Calculate the average brightness of the reflective area r _i , indicating the intensity of reflection:

where I(p) represents the brightness value of pixel p.

3) Boundary contour: B _i

Extract the boundary contour of the reflective area r _i to represent the shape of the reflective area.

4) Geometric center (x _c ,y _c ):

Calculate the geometric center coordinates of the reflective area:

The above characteristics comprehensively reflect the size, intensity and shape information of the reflective points. For time t _k , a reflective feature matrix F _k can be constructed:

The first to third columns respectively store the area, intensity and boundary contour of each reflective point, and the fourth and fifth columns store the coordinates of the geometric center.

Then, combine the reflective feature matrices at each time t _k to construct a global reflective feature matrix:

F＝[F ₁ ,F ₂ ,…,F _N ];

Through the global reflective feature matrix F, the detailed information of reflective points at different times is recorded. Reflective point matching can be performed based on F, and changes in reflective points between different moments can be tracked to locate the leakage area.

Specifically, the following methods can be used to match reflective points:

1) Use boundary contour B _i for matching. Calculate the similarity between the contours of reflective areas at different times, and the ones with high similarity correspond to matching reflective points.

2) Perform nearest neighbor matching based on geometric center coordinates. Calculate the Euclidean distance between the reflective geometric centers at different times, and the nearest corresponding matching reflective point.

3) Combine area and intensity changes for matching. Reflective points with consistent changing trends in area and intensity are matched accordingly.

Through the matching results, the movement trajectory of each reflective point can be tracked, its change pattern can be judged, and the leakage location can be found.

Among them, in the above technical solution, the steps of establishing and training the water permeable model specifically include:

Step 1. Establish a training data set, including the humidity matrix, reflective matrix obtained during the inspection of multiple historical underground waterproof layer construction projects, and the water permeability obtained by collecting water that penetrated into the basement;

Step 2. Use convolutional neural network to establish a prototype of the water permeability model;

Step 3. Use the training data set to train the prototype of the water permeability model to obtain the water permeability model; the input of the training is the historical humidity matrix and the reflective matrix, and the output of the training is the water permeability corresponding to the historical humidity matrix and the reflective matrix.

Specifically, the steps for establishing and training the permeable model are described as follows:

Step 1. Create a training data set

Collect inspection data from multiple historical underground waterproof layer construction projects, including:

-Humidity matrix HL: Set up a hygrometer in the basement to record the humidity values HL _it at different locations and at different times.

-Reflective matrix FL: Use image processing methods to extract the characteristics of the reflective area at each moment, such as area, intensity, etc.

-Water permeability LL: Collect the leaked water in the basement and measure the water permeability LL _it at different times.

Match the above three parts of data into groups to construct a training set:

D _t = (HL ₁ , FL ₁ , LL ₁ ), (HL ₂ , FL ₂ , LL ₂ ),…, (HL _N , FL _N , LL _N ); D _t needs to be normalized;

In the formula, D _t represents the training data set at time t; the detection data of each historical underground waterproof layer construction project contains multiple D _t , and the time interval for obtaining D _t is 15 minutes; t represents the collection time;

Step 2. Establish a prototype of the permeable model

Construct a water penetration detection network containing convolutional layers and fully connected layers:

-Input layer: receiving humidity matrix HL and reflective matrix FL

-Convolutional layer: extract spatial features

-Pooling layer: dimensionality reduction

-Fully connected layer: merge features and map to water permeability

-Output layer: predict water permeability

Step 3. Model training

Use the normalized historical data set D _t for model training, and optimize the parameters to minimize the loss function:

After training, the water permeability detection model is obtained, and new data can be input to predict the water permeability.

During the training process, parameters adjustment and over-fitting prevention are also required to improve the model's generalization ability. Finally, a water permeability detection model is obtained that maps humidity and reflective information to water permeability.

Among them, in the above technical solution, the pre-trained water permeability model is used to calculate the humidity matrix and reflection matrix at each moment to obtain the water permeability amount at the corresponding moment, which specifically includes:

Normalize the input data;

Input the humidity matrix and reflectance matrix into the water permeability model;

The model outputs the predicted water penetration at each moment.

Specifically, the specific implementation of step S40 is as follows:

Based on the trained water permeable model, the following processing can be performed:

1. Data preprocessing

-Normalize the humidity matrix:

Among them, μ _H and σ _H are the mean and standard deviation of the humidity of the training set respectively.

-Normalize the reflective matrix:

-Splicing humidity matrix and reflection matrix as model input:

2. Prediction of water permeability

Feed the preprocessed input X _k into the trained water permeable model:

The model will output the water permeability prediction at time k

3.Post-processing

-Denormalize the predicted water permeability:

- Smooth filtering of time series predictions.

After the above steps, the water permeability l _k at each moment can be predicted through the water permeability model, reflecting the leakage situation and used as a detection and evaluation index.

Among them, in the above technical solution, the leakage location and water permeability are used as the basement waterproof layer construction detection and evaluation indicators, and the method is sent to the construction personnel for graphic display.

Specifically, the implementation of step S50 is:

In step S40, we obtain the water permeability {l ₁ , l ₂ ,..., l _M } predicted by the model at each moment. Based on the water permeability time series, the following evaluation can be made:

1. Average water permeability

Calculate the average water permeability of the time series:

The average water penetration reflects the overall leakage situation.

2. Water permeability increase

Calculate the water permeability increment at adjacent moments:

Δl _k =l _k -l _k-1 ,k=2,3,…,M;

Monitor the growth trend of water permeability.

3. Degree of volatility

Calculate the standard deviation of water permeability:

The standard deviation can evaluate the degree of fluctuation of water permeability.

4. Extreme value detection

Detect the peak value of water permeability in the timing sequence:

l _max =max{l ₁ ,l ₂ ,…,l _M };

Maximum values may correspond to leakage bursts.

5. Trend analysis

Time series analysis methods, such as the ARIMA model, are used to analyze the changing trend of the water permeability series.

The average water permeability, water permeability increment, fluctuation degree, and extreme value data obtained above are displayed graphically.

A second aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores program instructions, and when the program instructions are run, they are used to perform the above-mentioned construction of an underground waterproof layer of a building. Quality testing methods.

A third aspect of the present invention provides a construction quality inspection system for underground waterproofing layers of buildings, which includes the above-mentioned computer-readable storage medium.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims (10)

1. A method for testing the construction quality of underground waterproofing layers of buildings, which is characterized by including the following steps: S10. In the closed basement environment constructed after the construction of the waterproof layer, obtain the air humidity data in the basement collected by multiple hygrometers, record the humidity data of the air humidity changing with time, and form an air humidity matrix, wherein the multiple hygrometers The hygrometers are evenly distributed in the basement, and the humidity data includes the spatial coordinates of the hygrometer and the humidity value; S20. Obtain the reflective images of the basement wall, floor and ceiling, and establish a set of reflective images that change over time, which is recorded as a reflective image set; S30. For the reflective image, collect reflective point information and establish reflective point data. The reflective point data includes the spatial coordinates, area, and reflective intensity of the geometric center of the reflective point; collect all reflective points at each moment in the reflective image. The reflective point data generated by the image generates a reflective matrix and determines the leakage location based on the reflective area; S40. Use the pre-trained water permeability model to calculate the humidity matrix and reflection matrix at each moment to obtain the water permeability amount at the corresponding moment; S50. Use the leakage location and water permeability as the basement waterproofing layer construction inspection and evaluation indicators, and send them to the construction personnel. 2. A method for testing the construction quality of underground waterproof layers of buildings according to claim 1, characterized in that the plurality of hygrometers at least include a plurality of hygrometers arranged close to each wall of the basement. 3. A method for detecting the construction quality of underground waterproofing layers of buildings according to claim 2, characterized in that the steps of obtaining reflective images of basement walls, floors and ceilings and establishing a collection of reflective images changing over time are specifically include: Set up multiple imaging devices in the basement to regularly collect images of walls, floors and ceilings; Preprocess the collected images, including denoising, enhancement, and correction; Convert the preprocessed image to HSV space, and extract the reflective area based on the brightness threshold; Perform morphological analysis on the extracted reflective areas, remove noise areas, and obtain effective reflective areas; Calculate the geometric characteristics of each reflective area; Retain appropriate reflective areas and obtain an effective set of reflective points in the image; Merge the sets of reflective points in images at different times to establish a set of reflective points that change with time. 4. A method for testing the construction quality of underground waterproof layers of buildings according to claim 3, characterized in that the device for collecting reflective images includes a camera and a light source. When the camera takes pictures of the walls or ground of the basement, the shooting angle is It is at an included angle of 45° with the wall or ground; when the camera shoots the wall or floor of the basement, the angle between the line connecting the light source and the shooting point and the shooting angle of the camera is greater than 30°. 5. A method for testing the construction quality of underground waterproofing layers of buildings according to claim 4, characterized in that the steps of collecting reflective point information and establishing reflective point data specifically include: Calculate the area, intensity, boundary contour and geometric center coordinates of each reflective point based on the reflective image; Construct a feature matrix of reflective points at each moment, including area, intensity, contour, and coordinate information; Merge the feature matrices at different times to obtain the global reflective feature matrix; Match the reflective points on the global reflective feature matrix; Use contour and coordinate information to match and track the movement of reflective points; Determine the leakage area through the trace of reflective dots. 6. A method for testing the construction quality of underground waterproof layers of buildings according to claim 1, characterized in that the steps of establishing and training the water permeable model specifically include: Step 1. Establish a training data set, including the humidity matrix, reflective matrix obtained during the inspection of multiple historical underground waterproof layer construction projects, and the water permeability obtained by collecting water that penetrated into the basement; Step 2. Use convolutional neural network to establish a prototype of the water permeability model; Step 3. Use the training data set to train the prototype of the water permeability model to obtain the water permeability model; the input of the training is the historical humidity matrix and the reflective matrix, and the output of the training is the water permeability corresponding to the historical humidity matrix and the reflective matrix. 7. A method for detecting the construction quality of underground waterproof layers of buildings according to claim 1, characterized in that the humidity matrix and reflection matrix at each moment are calculated using a pre-trained water permeability model to obtain the corresponding The steps to measure the water permeability at all times include: Normalize the input data; Input the humidity matrix and reflectance matrix into the water permeability model; The model outputs the predicted water penetration at each moment. 8. A method for testing the construction quality of underground waterproof layers of buildings according to claim 1, characterized in that the leakage position and water permeability are used as the waterproof layer construction detection and evaluation indicators of the basement and are sent to the construction personnel. For graphic display. 9. A computer-readable storage medium, characterized in that program instructions are stored in the computer-readable storage medium, and when the program instructions are run, they are used to execute the method described in any one of claims 1-8. Construction quality inspection method of underground waterproof layer of building. 10. A construction quality inspection system for underground waterproofing layers of buildings, characterized by comprising the computer-readable storage medium according to claim 9.