A construction quality monitoring method and system for 3D printed bridge prefabricated box girder

Abstract

The present application relates to the field of construction quality monitoring, and in particular to a construction quality monitoring method and system for 3D printed bridge prefabricated box girders, the method comprising: obtaining a grayscale image of the beam surface, calculating the ratio of the number of pixels of each grayscale level in the grayscale image to the number of all pixels, and obtaining multiple classification ranges by comparing the values; in each classification range, calculating the crack coefficient, and taking the mean of the crack coefficient of each classification range in each frame of the grayscale image as the correction value of the pixel in each classification range, and obtaining the correction weight according to the correction value; calculating the probability of each pixel being a crack pixel according to a preset calculation model, calculating the product of the probability and the correction weight, and obtaining an abnormal monitoring result of the presence of cracks in response to the product being greater than a preset threshold. The present application has the effect of reducing the error rate of quality recognition of bridge prefabricated box girders.

Description

Technical Field

[0001] The present application relates to the field of construction quality monitoring, and in particular to a construction quality monitoring method and system for a 3D printed bridge prefabricated box girder. Background Art

[0002] The application of 3D printing technology in the construction field, especially in the use of concrete and other cement-based materials, has shown great potential. For example, 3D printing can directly print out building components such as walls, columns, beams, and floor slabs. These components can be prefabricated and transported to the construction site for assembly. 3D printing can customize building components according to customer needs to meet personalized and differentiated market needs.

[0003] For example, a Chinese patent with the patent publication number CN116630282A discloses a 3D printed bridge prefabricated box girder construction quality monitoring system based on deep learning, and the method used includes the following steps: obtaining a defect image of the printing area during the construction process of the 3D printed bridge prefabricated box girder; annotating the defect type of the defect image after preprocessing, and measuring the actual size of the actual frame of the annotated defect target in reality and the pixel size in the picture; dividing the annotated defect image into a training set and a test set; constructing a convergence model for target detection and measurement, the convergence model includes a target detection algorithm and a conversion function module, the target detection algorithm is used to obtain the pixel size of the prediction frame in the defect image, and the conversion function module is used to obtain the predicted actual size of the defect according to the pixel size of the prediction frame using the actual size of the actual frame and the conversion ratio M of the pixel size; training the target detection algorithm with the training set, outputting the pixel size of the prediction frame, and obtaining the predicted actual size of the defect through the conversion function module; inputting the test set into the trained target detection algorithm to obtain the pixel width of the test set prediction frame, and using the calibration algorithm to correct the conversion ratio M in the conversion function module.

[0004] This method is not ideal for identifying damage to prefabricated box girders of bridges when the light intensity is low or high. In particular, the damage identification effect is poor or even impossible for overexposed or darker locations, resulting in an increase in the error rate of quality identification. Summary of the invention

[0005] In order to solve the problem of high error rate in quality identification of prefabricated box girders of bridges caused by lighting influence, the present application provides a construction quality monitoring method and system for 3D printed prefabricated box girders of bridges.

[0006] In the first aspect, the present application provides a method for monitoring the construction quality of a 3D printed bridge prefabricated box girder, which adopts the following technical solution:

[0007] A construction quality monitoring method for a prefabricated box girder of a 3D printed bridge comprises the following steps: obtaining a grayscale image of a beam body surface, calculating the ratio of the number of pixel points of each grayscale level in the grayscale image to the number of all pixel points, and obtaining a plurality of classification ranges by comparing the values; calculating a crack coefficient in each classification range, and taking the mean value of the crack coefficient of each classification range in each frame of the grayscale image as a correction value of the pixel points in each classification range, and obtaining a correction weight according to the correction value; calculating the probability that each pixel point is a crack pixel point according to a preset calculation model, calculating the product of the probability and the correction weight, and obtaining an abnormal monitoring result indicating the existence of cracks in response to the product being greater than a preset threshold.

[0008] Optionally, the method for calculating the crack coefficient includes the steps of: extracting a connected domain consisting of all pixel points whose grayscale values ​​are not equal to 255 within any classification range; obtaining the position coordinates of the midpoints of the two short sides of the minimum outer rectangle of the connected domain, and calculating the offset distance according to the position coordinates; the calculation formula of the offset distance is: ,in, For the The deviation distance of pixels, is the horizontal coordinate value of the first midpoint, is the ordinate value of the first midpoint, is the horizontal coordinate value of the second midpoint, is the ordinate value of the second midpoint, For the The vertical coordinate value of the pixel point, For the The horizontal coordinate value of the pixel point; the crack coefficient is calculated based on the offset distance.

[0009] Optionally, in calculating the crack coefficient according to the offset distance, the calculation formula of the crack coefficient is: ,in, is the crack coefficient, is the set of deviation distances of all pixels in the connected domain, For the The deviation distance of pixels.

[0010] Optionally, in calculating the crack coefficient according to the offset distance, the calculation formula of the crack coefficient is: ,in, is the crack coefficient, is the set of deviation distances of all pixels in the connected domain, For the The deviation distance of pixels, is the hyperbolic tangent function.

[0011] Optionally, a correction weight is obtained according to the correction value, and the calculation formula of the correction weight is:

[0012] ,in, Indicates The correction weight corresponding to the correction value is Indicates Correction value, Indicates Correction value, Indicates the total number of correction values.

[0013] Optionally, in calculating the probability that each pixel is a crack pixel according to a preset calculation model, the calculation model is a fully connected neural network model.

[0014] Optionally, obtaining a grayscale image of the beam surface includes the steps of: obtaining a captured video of the beam, extracting frames from the video to obtain a frame image; and gray-processing the frame image to obtain a grayscale image.

[0015] In the second aspect, the present application provides a construction quality monitoring system for a 3D printed bridge prefabricated box girder, which adopts the following technical solution:

[0016] A construction quality monitoring system for a 3D printed bridge prefabricated box girder comprises: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the construction quality monitoring method for the 3D printed bridge prefabricated box girder is implemented.

[0017] This application has the following beneficial effects:

[0018] The crack characteristics are calculated for different classification ranges, and the neural network is used to identify defects, which improves the recognition accuracy. The accuracy and reliability of the monitoring results are improved by quantitatively calculating the crack coefficient and applying the correction weight. The correction weight is set according to the actual application scenario, which adapts to beams with different lighting conditions and different surface textures, and improves the versatility and adaptability of the monitoring method. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present application will become easily understood. In the accompanying drawings, several embodiments of the present application are shown in an exemplary and non-restrictive manner, and the same or corresponding numbers represent the same or corresponding parts.

[0020] Figure 1 It is a method flow chart of a method for monitoring the construction quality of a 3D printed bridge prefabricated box girder according to an embodiment of the present application.

[0021] Figure 2 It is a flow chart of a method for calculating a crack coefficient in a method for monitoring the construction quality of a 3D printed bridge prefabricated box girder in an embodiment of the present application. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0023] It should be understood that when the terms "first", "second", etc. are used in the claims, specification and drawings of the present application, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the specification and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections.

[0024] The present application embodiment discloses a method for monitoring the construction quality of a 3D printed bridge prefabricated box girder, referring to Figure 1 , including steps S1 to S3, which are specifically as follows:

[0025] S1: Obtain a grayscale image of the beam surface, calculate the ratio of the number of pixels of each grayscale level in the grayscale image to the number of all pixels, and classify the comparison values ​​to obtain multiple classification ranges.

[0026] A captured video of the beam body is obtained, and frame extraction is performed on the video to obtain a frame image; and grayscale processing is performed on the frame image to obtain a grayscale image.

[0027] In one embodiment, a drone can be used to shoot around the beam body, and key frames can be extracted from the video or a frame can be extracted at regular intervals to reduce the amount of calculation and retain sufficient information. This can be achieved using video processing software or programming libraries (such as OpenCV). The extracted frame image is converted into a grayscale image, where the frame image is a color image, and each pixel in the frame image is converted into a single grayscale value. The grayscale value is between 0 (black) and 255 (white), and a grayscale image is obtained. Frame extraction and grayscale processing are prior arts and will not be described in detail here.

[0028] In the grayscale image, calculate the number of pixels of each grayscale level (grayscale value), that is, traverse each pixel in the grayscale image and count the number of pixels corresponding to each grayscale value.

[0029] Calculate the ratio of the number of pixels of each gray level in the grayscale image to the number of all pixels, and get the grayscale image in The ratio of the number of pixels of all gray levels between to the number of all pixels, with a total of 256 ratio values.

[0030] The proportion values ​​are classified, and the classification process satisfies the following conditional function: the grayscale levels are divided into Category range, exemplary, ,in, For the Category range, For the The first proportion value in the classification range, For the In the category range A ratio value, For the Category range, For the The first proportion value in the classification range, For the In the category range A ratio value.

[0031] Calculate the classification conditions of each classification range, take the classification range with the smallest classification condition as the final classification range, and take all the pixels in the classification range with the largest cumulative proportion value in the final classification range as the concrete pixel points. The classification condition expression is:

[0032] ,in, Indicates the classification conditions. represents the standard deviation, Table average.

[0033] The purpose of the above classification step is to eliminate the influence of reflection caused by light shining on concrete on the concrete area in the image, that is, to eliminate noise.

[0034] S2: In each classification range, the crack coefficient is calculated, and the mean value of the crack coefficient of each classification range in each frame of the grayscale image is used as the correction value of the pixel point in each classification range, and the correction weight is obtained according to the correction value.

[0035] Reference Figure 2 The method for calculating the crack coefficient includes steps S20 to S22, which are as follows:

[0036] S20: In any classification range, extract a connected domain consisting of all pixels whose grayscale values ​​are not equal to 255.

[0037] S21: Obtain the position coordinates of the midpoints of the two short sides of the minimum outer rectangle of the connected domain, and calculate the offset distance according to the position coordinates.

[0038] The calculation formula for the offset distance is: ,in, For the The deviation distance of pixels, is the horizontal coordinate value of the first midpoint, is the ordinate value of the first midpoint, is the horizontal coordinate value of the second midpoint, is the ordinate value of the second midpoint, For the The vertical coordinate value of the pixel point, For the The horizontal coordinate value of the pixel point.

[0039] The above formula is the distance formula from a point to a straight line, that is, to pass and The perpendicular distance of the straight line to express the point The degree of deviation from this straight line.

[0040] S22: Calculate the crack coefficient based on the offset distance.

[0041] In one embodiment, the calculation formula of the crack coefficient is: ,in, is the crack coefficient, is the set of deviation distances of all pixels in the connected domain, For the The deviation distance of pixels. is the base of natural logarithms, It is for Perform normalization.

[0042] In one embodiment, the calculation formula of the crack coefficient is: ,in, is the crack coefficient, is the set of deviation distances of all pixels in the connected domain, For the The deviation distance of pixels, is the hyperbolic tangent function, which is used to Normalize.

[0043] The correction weight is obtained according to the correction value. The calculation formula of the correction weight is:

[0044] ,in, Indicates The correction weight corresponding to the correction value is Indicates Correction value, Indicates Correction value, Indicates the total number of correction values.

[0045] S3: Calculate the probability that each pixel is a crack pixel according to a preset calculation model, calculate the product of the probability and the correction weight, and in response to the product being greater than a preset threshold, obtain an abnormal monitoring result indicating the presence of a crack.

[0046] The construction and training of the computational model includes the following steps:

[0047] A fully connected neural network is set, such as a BP (Backpropagation) model and a CNN (Convolutional Neural Network) model. The BP model and the CNN model are existing technologies and will not be described in detail here.

[0048] Collect multiple pictures of prefabricated box girders of 3D printed bridges, mark the crack pixels in the pictures, and put the marked pictures into a fully connected neural network for training. Use the cross entropy loss function and stop when the loss function is less than a preset value (for example, 0.001) or reaches a preset number of training times (for example, 1000 times) to obtain a calculation model.

[0049] The currently acquired frame image is input into the calculation model to obtain the probability that each pixel is a crack pixel. For a pixel, the product of the probability and the correction weight of the pixel is calculated. When the product is greater than a preset threshold, the pixel is determined to be a crack pixel, an abnormal monitoring result of the existence of cracks is generated, and an alarm is issued. For example, the area formed by the pixel determined to be a crack is highlighted in the image.

[0050] An embodiment of the present application also discloses a construction quality monitoring system for a 3D printed bridge prefabricated box girder, comprising a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, a construction quality monitoring method for a 3D printed bridge prefabricated box girder according to the present application is implemented.

[0051] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0052] In the present application, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high bandwidth memory HBM (High Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0053] Although this specification has shown and described a plurality of embodiments of the present application, it is obvious to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will think of many changes, modifications and alternatives without departing from the thought and spirit of the present application. It should be understood that in the process of practicing the present application, various alternatives to the embodiments of the present application described herein may be adopted.

[0054] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for monitoring the construction quality of a 3D printed bridge prefabricated box girder, characterized in that: The method comprises the steps of: obtaining a grayscale image of the beam surface, calculating the ratio of the number of pixels of each grayscale level in the grayscale image to the number of all pixels, and classifying the comparison values ​​to obtain multiple classification ranges; In each classification range, the crack coefficient is calculated, and the mean value of the crack coefficient of each classification range in each frame of the grayscale image is used as the correction value of the pixel point in each classification range, and the correction weight is obtained according to the correction value; The probability that each pixel point is a crack pixel point is calculated according to a preset calculation model, and the product of the probability and the correction weight is calculated. In response to the product being greater than a preset threshold, an abnormal monitoring result indicating the presence of a crack is obtained; The calculation method of the crack coefficient includes the following steps: In any classification range, extract the connected domain composed of all pixels whose gray value is not equal to 255; Get the position coordinates of the midpoints of the two short sides of the minimum external rectangle of the connected domain, and calculate the offset distance based on the position coordinates; the calculation formula for the offset distance is: ,in, For the The deviation distance of pixels, is the horizontal coordinate value of the first midpoint, is the ordinate value of the first midpoint, is the horizontal coordinate value of the second midpoint, is the ordinate value of the second midpoint, For the The horizontal coordinate value of the pixel point. For the The vertical coordinate value of each pixel point; the crack coefficient is calculated based on the offset distance; The calculation formula of the crack coefficient is: ,in, is the crack coefficient, is the set of deviation distances of all pixels in the connected domain, For the The deviation distance of pixels; The calculation formula for the modified weight is: ,in, Indicates The correction weight corresponding to the correction value is Indicates Correction value, Indicates Correction value, Indicates the total number of correction values.

2. The construction quality monitoring method of the 3D printed bridge prefabricated box girder according to claim 1 is characterized in that: The probability that each pixel is a crack pixel is calculated according to a preset calculation model, and the calculation model is a fully connected neural network model.

3. The construction quality monitoring method of the 3D printed bridge prefabricated box girder according to claim 1 is characterized in that: Obtaining the grayscale image of the beam surface includes the following steps: Obtain a video of the beam body, and extract frames from the video to obtain frame images; The frame image is gray-scaled to obtain a gray-scale image.

4. A construction quality monitoring system for 3D printed bridge prefabricated box girders, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the construction quality monitoring method of the 3D printed bridge prefabricated box girder according to any one of claims 1 to 3 is implemented.

Images