CN118641646B - Brick quality detection method and device for building construction - Google Patents

Abstract

The present invention discloses a brick quality detection method and device for construction, and relates to the field of computer vision technology. The brick quality detection method for construction includes: obtaining first ultrasonic images corresponding to each first preset detection point of multiple brick samples; using the first ultrasonic images corresponding to each first preset detection point of the brick sample, determining the local interference degree of each first preset detection point of the brick sample; using the local interference degree of each first preset detection point of the brick sample, determining the overall interference degree of the brick sample; constructing second ultrasonic images corresponding to each second preset detection point of a virtual sample corresponding to the brick sample according to the overall interference degree of each brick sample; training a target model through the first ultrasonic images corresponding to each first preset detection point of the brick sample and the second ultrasonic images corresponding to each first preset detection point of the virtual sample to obtain a brick detection model.

Description

Brick quality detection method and device for building construction

Technical Field

The present invention relates to the technical field of computer vision, and in particular to a brick quality detection method and device for building construction.

Background Art

In the field of construction technology, the quality of bricks is of vital importance. With the increasing requirements for building quality and safety, the application of non-destructive testing technology in brick quality testing is becoming increasingly popular. This method can not only effectively detect internal defects of bricks, but also avoid damage to the brick structure.

In the existing method, ultrasonic equipment is usually used to collect the ultrasonic waveform characteristics of bricks, and then the ultrasonic waveform characteristics are input into a neural network to detect the quality of the bricks.

However, due to the influence of environmental noise during construction, the existing neural network does not fully consider the impact of environmental noise during construction, and the training sample set constructed by it does not eliminate the interference of environmental noise, resulting in low accuracy of brick quality detection.

Summary of the invention

The embodiment of the present invention provides a brick quality detection method for building construction, which can improve the accuracy of brick quality detection.

In one aspect of an embodiment of the present invention, a brick quality detection method for building construction is provided, comprising:

Acquire a first ultrasonic image corresponding to each first preset detection point of a plurality of brick samples;

Determine the local interference degree of each first preset detection point of the brick sample by using the first ultrasonic image corresponding to each first preset detection point of the brick sample;

Determine the overall interference degree of the brick sample by using the local interference degree of each first preset detection point of the brick sample;

According to the overall interference degree of each brick sample, construct a second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the brick sample, and the second preset detection point matches the first preset detection point;

The target model is trained through the first ultrasonic images corresponding to the first preset detection points of the brick sample and the second ultrasonic images corresponding to the first preset detection points of the virtual sample to obtain a brick detection model, so that the brick detection model can be used to detect the quality of bricks in construction.

In one aspect of an embodiment of the present application, a brick quality inspection device for construction is provided, the device comprising: a memory and a program or instruction stored in the memory and executable on a processor, wherein when the program or instruction is executed by the processor, a brick quality inspection method for construction provided in any aspect of the above-mentioned embodiment of the present application is implemented.

In the brick quality detection method for construction provided by the embodiment of the present application, the first ultrasonic image corresponding to each first preset detection point of the brick sample is first used to determine the local interference degree of each first preset detection point of the brick sample. Then, the local interference degree of each first preset detection point of the brick sample is used to determine the overall interference degree of the brick sample. Then, according to the overall interference degree of the brick sample, the second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the brick sample is constructed. Thus, the target model is trained using the first ultrasonic image corresponding to each first preset detection point of the brick sample and the second ultrasonic image corresponding to each first preset detection point of the virtual sample to obtain a brick detection model. In this way, the embodiment of the present application fully considers the influence of the environmental noise of the construction on the ultrasonic data of the brick, and constructs the corresponding virtual sample according to the overall interference degree of the brick sample, so that the ultrasonic image of the virtual sample excludes the influence of the environmental noise as much as possible while meeting the characteristics of the ultrasonic data of the brick, and realizes the expansion of the training sample set of the neural network, which can improve the detection effect of the brick detection model, thereby improving the accuracy of brick quality detection.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

FIG1 is a schematic flow chart of a brick quality detection method for building construction provided by an embodiment of the present invention;

FIG2 is a schematic diagram of a first preset detection point of a brick sample provided by an embodiment of the present invention;

FIG3 is a schematic diagram of the structure of an ultrasonic device provided by an embodiment of the present invention;

FIG4 is a schematic structural diagram of a brick quality inspection device for construction provided by an embodiment of the present invention;

In the figure, 301-brick to be detected, 302-starting probe, 303-terminal probe, 304-first probe connection corresponding to the terminal probe, 305-second probe connection corresponding to the starting probe, 306-controller button of ultrasonic device, 307-controller knob of ultrasonic device, 308-display screen of ultrasonic device.

DETAILED DESCRIPTION

In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a brick quality detection method and device for construction proposed by the present invention, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

Unless defined otherwise, 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 invention belongs.

It should be noted that the acquisition, storage, use, and processing of data in the technical solution of the present invention are in compliance with the relevant provisions of national laws and regulations.

It should be noted that in the embodiments of the present invention, certain software, components, models and other existing solutions in the industry may be mentioned, which should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present invention, but it does not mean that the applicant has or will necessarily use the solution.

In the existing methods, ultrasonic equipment is usually used to collect the ultrasonic waveform features of bricks, and then the ultrasonic waveform features are input into the neural network to detect the quality of bricks. However, due to the influence of environmental noise during construction, the existing neural network does not fully consider the influence of environmental noise during construction, and the training sample set constructed by it does not eliminate the interference of environmental noise, resulting in low accuracy of brick quality detection.

The object of the present invention is to provide a brick quality detection method and device for construction. In the brick quality detection method for construction provided by the embodiment of the present invention, the local interference degree of each first preset detection point of the brick sample is first determined by using the first ultrasonic image corresponding to each first preset detection point of the brick sample. Then, the overall interference degree of the brick sample is determined by using the local interference degree of each first preset detection point of the brick sample. Then, according to the overall interference degree of the brick sample, the second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the brick sample is constructed. Thus, the target model is trained using the first ultrasonic image corresponding to each first preset detection point of the brick sample and the second ultrasonic image corresponding to each first preset detection point of the virtual sample to obtain a brick detection model. In this way, the embodiment of the present application fully considers the influence of the environmental noise of the construction on the ultrasonic data of the brick, and constructs the corresponding virtual sample according to the overall interference degree of the brick sample, so that the ultrasonic image of the virtual sample excludes the influence of the environmental noise as much as possible while meeting the characteristics of the ultrasonic data of the brick, and realizes the expansion of the training sample set of the neural network, which can improve the detection effect of the brick detection model, thereby improving the accuracy of brick quality detection.

The following describes a specific embodiment of a brick quality detection method and device for building construction provided by an embodiment of the present invention.

The following first introduces a brick quality detection method for building construction provided by an embodiment of the present invention.

FIG1 provides a flow chart of a brick quality detection method for construction. The brick quality detection method for construction can be applied to a server. The brick quality detection method for construction can include the following steps S101 to S105.

S101, obtaining first ultrasonic images corresponding to first preset detection points of a plurality of brick samples.

In this embodiment, the first preset detection point is a detection position point preset on the brick sample, and the position of the first preset detection point corresponding to each brick sample is the same. For example, as shown in FIG2, a schematic diagram of the first preset detection point of a brick sample is provided. Among them, seven first preset detection points 0, 1, 2, 3, 4, 5, and 6 are preset on each brick sample.

Among them, the first ultrasonic image is collected by an ultrasonic device. As shown in Figure 3, a structural schematic diagram of an ultrasonic device is provided. Specifically, for the acquisition of the first ultrasonic image corresponding to each first preset detection point of the brick sample, the terminal probe 303 of the ultrasonic device is first connected to the first probe connection 304 corresponding to the terminal probe, and the starting probe 302 of the ultrasonic device is connected to the second probe connection 305 corresponding to the starting probe. And the controller button 306 of the ultrasonic device and the controller knob 307 of the ultrasonic device are used to adjust the data parameters, such as the transmission frequency or the sampling rate.

Then, the starting probe 302 of the ultrasonic device and the terminal probe 303 of the ultrasonic device are placed on the brick to be detected 301, and appropriate coupling agents are used to ensure that the ultrasonic wave can be effectively transmitted and received. The starting probe 302 and the terminal probe 303 are slid to each first preset detection point on the brick to be detected 301 in sequence, so as to obtain the first ultrasonic image of each first preset detection point of the brick to be detected and display it on the display screen 308 of the ultrasonic device.

S102, using the first ultrasonic image corresponding to each first preset detection point of the brick sample, determine the local interference degree of each first preset detection point of the brick sample.

In this embodiment, the local interference degree is used to characterize the interference degree of the first preset detection point corresponding to the brick sample by the environmental noise. For example, the environmental noise may include at least one of the operation sound of mechanical equipment and the sound of human construction activities.

The greater the local interference degree is, the greater the interference degree of the first preset detection point corresponding to the brick sample is caused by the environmental noise.

As an example, the local interference degree may be a value input by the user after analyzing the first ultrasonic image displayed on the display screen 308 of the ultrasonic device, or may be a value obtained by the server after analyzing the interference degree of the first ultrasonic image generated by the ultrasonic device.

S103, using the local interference degree of each first preset detection point of the brick sample to determine the overall interference degree of the brick sample.

In this embodiment, the overall interference degree is used to represent the degree to which the brick sample is interfered with by the environmental noise as a whole. The greater the overall interference degree is, the greater the degree to which the brick sample is interfered with by the environmental noise as a whole.

As an example, the server receives the local interference degree of the brick sample input by the user corresponding to each first preset detection point, and then accumulates each local interference degree to obtain the overall interference degree of the brick sample.

S104, constructing a second ultrasonic image corresponding to each second preset detection point of a virtual sample corresponding to the brick sample according to the overall interference degree of each brick sample, wherein the second preset detection point matches the first preset detection point.

In this embodiment, there is a one-to-one matching relationship between the second preset detection points of the virtual sample and the first preset detection points of the brick sample.

The second ultrasonic image is an ultrasonic image corresponding to a virtual sample constructed according to the first ultrasonic image of the brick sample.

As an example, the server presets an interference threshold, and then determines any two brick samples whose sum of overall interference is less than or equal to the interference threshold as a sample extension group. For example, if the interference threshold is 1, the overall interference of brick sample A is 0.8, the overall interference of brick sample B is 0.6, and the overall interference of brick sample C is 0.4, then brick sample B and brick sample C can be determined as a sample extension group.

Then, the difference between the ratio of the overall interference degree of each brick sample in the sample extension group to the sum of the overall interference degree in the sample extension group and 1 is used as the weight to construct the second ultrasonic image corresponding to the virtual sample. Specifically, the overall interference degree of brick sample B in the sample extension group is 0.6, and the overall interference degree of brick sample C is 0.4, then the weight of brick sample B is 0.4, and the weight of brick sample C is 0.6.

Then multiply the weight of brick sample B by the signal strength of the A detection point of brick sample B at time A1 to obtain the first strength, multiply the weight of brick sample C by the signal strength of the A detection point of brick sample C at time A1 to obtain the second strength, and then add the first strength to the second strength to obtain the signal strength of the A detection point of the virtual sample at time A.

In this way, after the signal strength of the virtual sample at the A detection point at each time is obtained, the second ultrasonic image of the virtual sample at the A detection point can be constructed.

S105, training the target model through the first ultrasonic images corresponding to the first preset detection points of the brick sample and the second ultrasonic images corresponding to the first preset detection points of the virtual sample to obtain a brick detection model, so that the brick detection model can be used to detect the quality of bricks in construction.

In this embodiment, the server combines the brick sample and the virtual sample as a training sample set, and uses the first ultrasonic image corresponding to each first preset detection point of the brick sample and the second ultrasonic image corresponding to each first preset detection point of the virtual sample as input information to train the target model. When the training cutoff condition is met, the brick detection model can be obtained. The target model is an RNN neural network model, and the loss function of the target model is a cross entropy loss function.

When the quality of bricks used in construction needs to be tested later, it is only necessary to input the ultrasonic images corresponding to the bricks used in construction into the brick detection model to obtain the test results of the quality of the bricks used in construction.

Through this embodiment, the first ultrasonic image corresponding to each first preset detection point of the brick sample is first used to determine the local interference degree of each first preset detection point of the brick sample. Then, the local interference degree of each first preset detection point of the brick sample is used to determine the overall interference degree of the brick sample. Then, based on the overall interference degree of the brick sample, a second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the brick sample is constructed. Thus, the target model is trained using the first ultrasonic image corresponding to each first preset detection point of the brick sample and the second ultrasonic image corresponding to each first preset detection point of the virtual sample to obtain a brick detection model. In this way, the embodiment of the present application fully considers the impact of the environmental noise of the construction on the ultrasonic data of the brick, and constructs the corresponding virtual sample according to the overall interference degree of the brick sample, so that the ultrasonic image of the virtual sample excludes the impact of the environmental noise as much as possible while meeting the characteristics of the ultrasonic data of the brick, thereby expanding the training sample set of the neural network, and can improve the detection effect of the brick detection model, thereby improving the accuracy of brick quality detection.

As an optional embodiment, S102 may specifically include:

Using the first ultrasonic image corresponding to the first preset detection point of the brick sample, determining the symmetry of each extreme point in the first ultrasonic image, where the symmetry is used to characterize whether the data distribution in the preset neighborhood of the extreme point is regular;

According to the symmetry of each extreme point, the data distribution irregularity between each extreme point and the corresponding next extreme point is obtained;

The local interference degree of the first preset detection point is determined by using the data distribution irregularity between each extreme point and the corresponding next extreme point.

In this embodiment, the first ultrasonic image is a relationship between the amplitude of the ultrasonic signal and time. The first ultrasonic image includes at least one extreme value point, and the extreme value point includes a maximum value point and a minimum value point.

The symmetry is used to characterize whether the data distribution in the preset neighborhood of the extreme point is regular, and the data distribution irregularity is used to characterize the degree of difference between the symmetry of two extreme points.

As an example, the server obtains the extreme point in the first ultrasonic image according to the first preset detection point of the brick sample, and then determines the 12 data closest to the extreme point in the left and right neighborhoods of the extreme point. Then the first data in the left neighborhood of the extreme point is compared with the first data in the right neighborhood of the extreme point. When the signal strength difference between the two data is less than the preset strength difference threshold, the two data are considered to be symmetrical. By comparing them one by one, the number of groups of symmetrical data is determined, and the number of groups of symmetrical data is determined as the symmetry of the extreme point, thereby obtaining the symmetry of each extreme point in the first ultrasonic image.

Then, according to the time sequence in the first ultrasonic image, the symmetry of each extreme point is subtracted from the symmetry of the next extreme point to obtain the data distribution irregularity between each extreme point and the corresponding next extreme point. Then, the data distribution irregularities are accumulated and divided by the number of data with data distribution irregularities to obtain the local interference degree corresponding to the first preset detection point.

Through this embodiment, the symmetry of each extreme point in the first ultrasonic image is determined by using the first ultrasonic image. Then, the data distribution irregularity is determined according to the difference in symmetry between each extreme point and the corresponding next extreme point. Finally, the local interference degree of the first preset detection point is determined according to the data distribution irregularity. In this way, the degree of influence of the environmental noise on the first preset detection point can be accurately evaluated, providing a basis for eliminating the influence of environmental noise when the training sample set is subsequently expanded.

As an optional embodiment, using the first ultrasonic image corresponding to the first preset detection point of the brick sample to determine the symmetry of each extreme point in the first ultrasonic image may specifically include:

Acquire a first signal strength difference between two symmetrical data within a preset neighborhood of a target extreme point in a first ultrasonic image corresponding to a first preset detection point;

The reciprocals of the first signal strength differences are accumulated to obtain the symmetry of the target extreme point in the first ultrasonic image.

In this embodiment, the symmetry of the target extreme point can be determined by the following formula 1:

Formula 1

In the formula, represents the symmetry of the kth extreme point of the jth first preset detection point of the i-th brick sample, ∑ represents the cumulative summation operation, T represents the number of data that need to be compared and analyzed in the left and right neighborhoods of the kth extreme point, Represents the signal strength difference between the t-th data in the left and right neighborhoods of the k-th extreme point of the j-th first preset detection point of the ith brick sample.

in, It is the reciprocal of the signal strength difference between the t-th data in the left and right neighborhoods of the k-th extreme point of the j-th first preset detection point of the ith brick sample. The larger the signal strength difference, the smaller the reciprocal of the signal strength difference, indicating that the symmetry between the two data is weaker.

The symmetry of the kth extreme point of the jth first preset detection point of the i-th brick sample The larger the value is, the more regular the data distribution is in the left and right neighborhoods of the kth extreme point. The more regular the data distribution is, the less affected it is by environmental noise.

Through this embodiment, the symmetry of the target extreme point in the first ultrasonic image can be accurately calculated based on the first signal strength difference between two symmetrical data within a preset neighborhood of the target extreme point, which helps to subsequently use the symmetry to accurately evaluate the extent to which the first preset detection point is affected by environmental noise, and provides a basis for eliminating the influence of environmental noise when subsequently expanding the training sample set.

As an optional embodiment, according to the symmetry of each extreme point, obtaining the data distribution irregularity between each extreme point and the corresponding next extreme point may specifically include:

Obtaining a second signal strength difference between the target extreme value point and the corresponding next extreme value point and a symmetry difference between the target extreme value point and the corresponding next extreme value point;

The second signal strength difference is multiplied by the symmetry difference to obtain the data distribution irregularity between the target extreme value point and the corresponding next extreme value point.

In this embodiment, the data distribution irregularity can be determined by the following formula 2:

Formula 2

In the formula, Indicates the data distribution irregularity between the kth extreme point of the jth first preset detection point of the i-th brick sample and the corresponding k+1th extreme point, Represents the symmetry difference between the kth extreme point of the jth first preset detection point of the i-th brick sample and the corresponding k+1th extreme point, Represents the signal strength difference between the kth extreme point of the jth first preset detection point of the i-th brick sample and the corresponding k+1th extreme point.

in, It represents the product of the symmetry difference and the signal strength difference. The larger the symmetry difference, the less similar the data symmetry distribution between the kth extreme point and the corresponding k+1th extreme point is, which means the data distribution is more irregular; the larger the signal strength difference, the more obvious the change in signal strength between the kth extreme point and the corresponding k+1th extreme point is, which means the data distribution is more irregular.

The data distribution irregularity between the kth extreme point of the jth first preset detection point of the i-th brick sample and the corresponding k+1th extreme point The larger it is, the more irregular the data distribution between the kth extreme point and the corresponding k+1th extreme point is, indicating that it is more affected by environmental noise.

Through this embodiment, based on the second signal strength difference between the target extreme point and the corresponding next extreme point and the symmetry difference between the target extreme point and the corresponding next extreme point, the data distribution irregularity between the target extreme point and the corresponding next extreme point in the first ultrasonic image can be accurately calculated, which will help to use the data distribution irregularity to accurately evaluate the extent to which the first preset detection point is affected by environmental noise, and provide a basis for eliminating the influence of environmental noise when subsequently expanding the training sample set.

As an optional embodiment, determining the local interference degree of the first preset detection point by using the data distribution irregularity between each extreme point and the corresponding next extreme point may specifically include:

The data distribution irregularity between each extreme point and the corresponding next extreme point is accumulated to obtain the data distribution irregularity accumulation value;

The accumulated value of the data distribution irregularity is divided by the number of data distribution irregularities to obtain the local interference degree of the first preset detection point.

In this embodiment, the local interference degree can be determined by the following formula 3:

Formula 3

In the formula, represents the local interference degree of the jth first preset detection point of the i-th brick sample, K represents the number of extreme value points, ∑ represents the cumulative summation operation, Represents the data distribution irregularity between the kth extreme point of the jth first preset detection point of the i-th brick sample and the corresponding k+1th extreme point.

in, It represents the cumulative value of the data distribution irregularity between each extreme point and the corresponding next extreme point in the first ultrasonic image corresponding to the first preset detection point. The larger the cumulative value of the data distribution irregularity is, the greater the local interference degree of the first preset detection point is.

The local interference degree of the jth first preset detection point of the i-th brick sample The larger the value is, the greater the impact of environmental noise is on the j-th first preset detection point of the i-th brick sample.

Through this embodiment, the local interference degree of the first preset detection point can be accurately calculated according to the data distribution irregularity between each extreme point and the corresponding next extreme point, which helps to subsequently use the local interference degree to accurately evaluate the extent to which the first preset detection point is affected by environmental noise, and provides a basis for eliminating the influence of environmental noise when subsequently expanding the training sample set.

As an optional embodiment, S103 may specifically include:

Calculating the standard deviation of the local interference degree of each first preset detection point of the brick sample to obtain the standard deviation of the local interference degree;

The inverse value of the standard deviation of the local interference degree is multiplied by the average value of the local interference degree of the brick sample to obtain the calculated value of the interference degree of the brick sample;

The calculated interference value is normalized to obtain the overall interference of the brick sample.

In this embodiment, the overall interference degree can be determined by the following formula 4:

Formula 4

In the formula, represents the overall interference degree of the i-th brick sample, norm represents the normalization function, represents the calculated standard deviation, represents the local interference degree of the jth first preset detection point of the i-th brick sample, J represents the total number of the first preset detection points in the i-th brick sample, Represents the average value of local interference in the i-th brick sample.

in, It is used to characterize the consistency of the local interference degree of each first preset detection point in the i-th brick sample. The greater the difference between the local interference degrees of each first preset detection point in the i-th brick sample, the greater the overall interference degree of the i-th brick sample.

The overall interference degree of the i-th brick sample The larger the value is, the more the i-th brick sample is affected by the environmental noise as a whole.

Through this embodiment, the overall interference degree of the brick sample can be accurately calculated based on the local interference degree of each first preset detection point in the brick sample, which helps to subsequently use the overall interference degree to accurately evaluate the extent to which the brick sample is affected by environmental noise, and provides a basis for eliminating the influence of environmental noise when subsequently expanding the training sample set.

As an optional embodiment, S104 may specifically include:

According to the first quality label of each brick sample, each brick sample is grouped to obtain at least one brick sample group, wherein the brick sample group includes at least one brick sample with the same first quality label;

The necessity of capacity expansion of each brick sample is determined by using the overall interference degree of each brick sample;

Sort each brick sample in the brick sample group in the order of the necessity of expansion to obtain a sample sorting result;

According to the sample sorting results, at least one expansion combination is obtained;

The sample is expanded using the expansion combination, and a second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the expansion combination is constructed.

In this embodiment, the first quality label is used to mark the quality of the brick sample. The first quality label may include "qualified product" and "defective product".

For example, a destructive test may be performed on a brick sample, and a technician determines the quality of the brick sample by observing the internal structure after destruction, thereby generating a first quality label for the brick sample.

As an example, the server divides each brick sample into a first brick sample group corresponding to the first quality label of "qualified products" or a second brick sample group corresponding to the first quality label of "defective products" according to the first quality label of each brick sample.

Then, for each brick sample group, the inverse of the overall interference degree of each brick sample in the brick sample group is determined as the necessity of capacity expansion of the brick sample. The greater the overall interference degree of the brick sample, the greater the impact of the environmental noise on the brick sample, and the weaker the corresponding necessity of capacity expansion.

Then, the brick samples in the brick sample group are sorted in the order of the necessity of expansion to obtain the sample sorting result. According to the sample sorting result, the two brick samples with the greatest and least necessity of expansion are combined into one expansion combination, and the two brick samples with the second greatest and second least necessity of expansion are combined into one expansion combination. And so on, multiple expansion combinations are obtained.

Finally, the sample is expanded according to each expansion combination to construct a second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the expansion combination. The second quality label of the virtual sample is consistent with the first quality labels of the two brick samples in the expansion combination.

Through this embodiment, a second ultrasonic image of a virtual sample corresponding to a brick sample is constructed according to the overall interference degree of each brick sample. By analyzing bricks with the same quality label together, the problem of confusion in the quality labels of virtual samples when different quality labels are expanded can be avoided. At the same time, brick samples with high interference degree and low interference degree are combined, so that a relatively balanced interference degree can be maintained while expanding the virtual sample. In this way, a virtual sample with better quality can be constructed, thereby improving the training effect of the brick detection model.

As an optional embodiment, using the expansion combination to expand the sample, and constructing the second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the expansion combination may specifically include:

Multiply the necessity of capacity expansion of each brick sample in the capacity expansion combination by the signal strength of the first preset detection point of the corresponding brick sample at the target time to obtain the signal strength proportion value of each brick sample at the target time;

Accumulate the signal strength proportion values of each brick sample at the target time to obtain the signal strength of the second preset detection point of the virtual sample corresponding to the expansion combination at the target time;

According to the signal strength of the second preset detection point of the virtual sample at each time, a second ultrasonic image corresponding to the second preset detection point of the virtual sample is obtained.

In this embodiment, the signal strength of the second preset detection point of the virtual sample at the target time can be determined by the following formula 5:

Formula 5

In the formula, represents the signal strength of the jth second preset detection point of the virtual sample constructed by the mth expansion combination at the bth time, Indicates the necessity of expanding the first brick sample in the mth expansion combination, It represents the signal strength of the jth first preset detection point of the first brick sample in the mth expansion combination at the bth time. Indicates the necessity of expanding the second brick sample in the mth expansion combination, Represents the signal strength of the jth first preset detection point of the second brick sample in the mth expansion combination at the bth time.

in, It represents the signal strength ratio of the jth first preset detection point of the first brick sample in the mth expansion combination at the bth time. It represents the signal strength proportion of the jth first preset detection point of the second brick sample in the mth expansion combination at the bth time. The signal strength of the jth second preset detection point of the virtual sample constructed in the mth expansion combination at the bth time can be determined by adding the corresponding signal strength proportions.

By combining the signal strengths of the jth second preset detection point of the virtual sample constructed by the mth expansion combination at various moments, the second ultrasonic image corresponding to the jth second preset detection point of the virtual sample constructed by the mth expansion combination can be obtained.

Through this embodiment, sample expansion is performed according to the necessity of expansion of each brick sample in the expansion combination and the signal strength of the first preset detection point of the corresponding brick sample. The virtual sample finally obtained by the expansion can maintain a relatively balanced interference degree, so that the ultrasonic data characteristics of the virtual sample are consistent with the quality of the corresponding brick and will not be affected by excessive environmental noise, thereby improving the training effect of the brick detection model.

As an optional embodiment, S105 may specifically include:

Acquire at least one training sample, the training sample comprising at least one of a brick sample and a virtual sample, the brick sample comprising a first ultrasonic image and a corresponding first quality label, and the virtual sample comprising a second ultrasonic image and a corresponding second quality label;

The target model is trained using each training sample until the target training stop condition is met to obtain a brick detection model.

In this embodiment, the target model is a preset model for training, and the brick detection model is a trained model for detecting brick quality.

The target training stop condition may be that the number of training times reaches a first number threshold or the function loss value is less than a first loss threshold.

As an example, when a brick detection model needs to be trained, the ultrasonic image of each training sample is first input into the target model as an input value to obtain the quality recognition result of each training sample. Then, the function loss value is calculated based on the quality recognition result and the quality label corresponding to the training sample. The function loss value is used to further train the target model until the target training stop condition is met, and the brick detection model can be obtained.

Through this embodiment, the target model is trained using training samples until the target training stop condition is met, and a brick detection model can be obtained. This helps to subsequently call the brick detection model to detect the quality of bricks, without the need for manual brick quality detection, thereby improving the detection efficiency and accuracy of bricks.

Based on the brick quality detection method for building construction, accordingly, the present application also provides a specific embodiment of a brick quality detection device for building construction.

FIG4 shows a schematic diagram of the hardware structure of a brick quality inspection device for construction provided in an embodiment of the present application.

The brick quality inspection device for building construction may include a processor 401 and a memory 402 storing computer program instructions.

Specifically, the processor 401 may include a central processing unit, or a specific integrated circuit, or may be configured to implement one or more integrated circuits of the embodiments of the present application.

Memory 402 may include a large capacity memory for data or instructions. By way of example and not limitation, memory 402 may include a hard disk drive, a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus drive, or a combination of two or more of these. Where appropriate, memory 402 may include a removable or non-removable (or fixed) medium. Where appropriate, memory 402 may be inside or outside the integrated gateway disaster recovery device. In a particular embodiment, memory 402 is a non-volatile solid-state memory.

The memory 402 may include read-only memory, random access memory, magnetic disk storage media devices, optical storage media devices, flash memory devices, electrical, optical or other physical/tangible memory storage devices. Thus, typically, the memory 402 includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present disclosure.

The processor 401 reads and executes the computer program instructions stored in the memory 402 to implement any one of the brick quality detection methods for building construction in the above embodiments.

In one example, the case processing device may further include a communication interface 403 and a bus 410. As shown in Fig. 4, the processor 401, the memory 402, and the communication interface 403 are connected via the bus 410 and communicate with each other.

The communication interface 403 is mainly used to implement communication between each module, device, unit and/or equipment in the embodiment of the present application.

Bus 410 includes hardware, software or both, and the parts of case handling equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port or other graphics bus, enhanced industrial standard architecture bus, front-end bus, hypertransport interconnection, industrial standard architecture bus, infinite bandwidth interconnection, low pin count bus, memory bus, micro channel architecture bus, peripheral component interconnection bus, serial advanced technology attachment bus, video electronics standard association local bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 410 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the present application considers any suitable bus or interconnection.

It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

It should also be noted that the exemplary embodiments mentioned in the present invention describe some methods or systems based on a series of steps or devices. However, the present invention is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiments, or in a different order from the embodiments, or several steps can be performed simultaneously.

The above is only a specific implementation of the present invention. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the system, module and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be covered within the protection scope of the present invention.

Claims (3)

1. A method for quality inspection of a brick for construction, the method comprising:

Acquiring first ultrasonic images corresponding to first preset detection points of a plurality of brick samples;

determining the local interference degree of each first preset detection point of the brick sample by using a first ultrasonic image corresponding to each first preset detection point of the brick sample;

determining the overall interference degree of the brick sample by utilizing the local interference degree of each first preset detection point of the brick sample;

Constructing a second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to each brick sample according to the overall interference degree of each brick sample, wherein the second preset detection points are matched with the first preset detection points;

Training a target model through a first ultrasonic image corresponding to each first preset detection point of the brick sample and a second ultrasonic image corresponding to each second preset detection point of the virtual sample to obtain a brick detection model, so that the brick quality of building construction is detected by using the brick detection model;

The method for determining the local interference degree of each first preset detection point of the brick sample by using the first ultrasonic image corresponding to each first preset detection point of the brick sample comprises the following steps: determining symmetry degree of each extreme point in the first ultrasonic image by using a first ultrasonic image corresponding to a first preset detection point of the brick sample, wherein the symmetry degree is used for representing whether data distribution in a preset neighborhood of the extreme point is regular or not; according to the symmetry degree of each extreme point, obtaining the data distribution irregularity between each extreme point and the corresponding next extreme point; determining the local interference degree of the first preset detection point by utilizing the data distribution irregularity between each extreme point and the corresponding next extreme point;

The determining the symmetry degree of each extreme point in the first ultrasonic image by using the first ultrasonic image corresponding to the first preset detection point of the brick sample comprises the following steps: acquiring a first signal intensity difference between two symmetrical data in a preset neighborhood of a target extreme point in a first ultrasonic image corresponding to the first preset detection point; accumulating the reciprocal of each first signal intensity difference to obtain the symmetry degree of the target extreme point in the first ultrasonic image;

The obtaining the data distribution irregularity between each extreme point and the corresponding next extreme point according to the symmetry degree of each extreme point comprises the following steps: acquiring a second signal intensity difference between a target extreme point and a corresponding next extreme point and a symmetry difference between the target extreme point and the corresponding next extreme point; multiplying the second signal intensity difference by the symmetry difference to obtain data distribution irregularity between the target extreme point and the corresponding next extreme point;

wherein determining the local disturbance degree of the first preset detection point by using the data distribution irregularity between each extreme point and the corresponding next extreme point includes: accumulating the data distribution irregularity between each extreme point and the corresponding next extreme point to obtain a data distribution irregularity accumulated value; dividing the accumulated value of the data distribution irregularity by the number of the data distribution irregularity to obtain the local interfered degree of the first preset detection point;

Wherein, the determining the overall interference degree of the brick sample by using the local interference degree of each first preset detection point of the brick sample includes: calculating the standard deviation of the local disturbance degree of each first preset detection point of the brick sample to obtain the standard deviation of the local disturbance degree; multiplying the reciprocal value of the local disturbance rejection standard deviation by the local disturbance rejection average value of the brick sample to obtain a disturbance rejection calculation value of the brick sample; normalizing the calculated interference degree value to obtain the overall interference degree of the brick sample;

The constructing a second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the brick sample according to the overall interference degree of each brick sample comprises the following steps: grouping each brick sample according to a first quality label of each brick sample to obtain at least one brick sample group, wherein the brick sample group comprises at least one brick sample with the same first quality label; determining the expansion necessity of each brick sample by utilizing the overall interference degree of each brick sample; sequencing each brick sample in the brick sample group according to the size sequence of the capacity expansion necessity to obtain a sample sequencing result; according to the sample sequencing result, at least one capacity expansion combination is obtained; performing sample expansion by using the expansion combination, and constructing a second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the expansion combination;

The performing sample expansion by using the expansion combination, and constructing a second ultrasonic image corresponding to each second preset detection point of the virtual sample corresponding to the expansion combination, including: multiplying the capacity expansion necessity of each brick sample in the capacity expansion combination with the signal intensity of the corresponding first preset detection point of the brick sample at the target moment to obtain the signal intensity ratio of each brick sample at the target moment; accumulating the signal intensity ratio of each brick sample at the target moment to obtain the signal intensity of a second preset detection point of the virtual sample corresponding to the expansion combination at the target moment; and obtaining a second ultrasonic image corresponding to the second preset detection point of the virtual sample according to the signal intensity of the second preset detection point of the virtual sample at each moment.

2. The method for detecting quality of bricks for construction according to claim 1, wherein training the target model by the first ultrasonic image corresponding to each first preset detection point of the brick sample and the second ultrasonic image corresponding to each second preset detection point of the virtual sample to obtain a brick detection model comprises:

obtaining at least one training sample, wherein the training sample comprises at least one of the brick sample and the virtual sample, the brick sample comprises a first ultrasonic image and a corresponding first quality label, and the virtual sample comprises a second ultrasonic image and a corresponding second quality label;

And training the target model by utilizing each training sample until the target training stopping condition is met, so as to obtain the brick detection model.

3. A brick quality detection apparatus for use in construction, the apparatus comprising: a processor and a memory storing computer program instructions;

The processor, when executing the computer program instructions, implements the brick quality detection method for construction of a building as claimed in claim 1.