Remote non-contact concrete crack depth detection method

Abstract

The invention particularly relates to a remote non-contact concrete crack depth detection method, which comprises the following steps of: shooting a visible light image and an infrared image of a concrete crack in an unreachable area, and measuring environmental parameters during shooting; inputting the shot visible light image, the shot infrared image and the measured environment parameters into the trained deep learning model; the trained deep learning model predicts the concrete crack depth according to the quantitative relation between the concrete crack depth and the concrete surface temperature distribution and the environmental parameters and outputs the concrete crack depth. The invention further relates to a remote non-contact concrete crack depth detection device, computer equipment and a storage medium. According to the method, the infrared image and the visible light image of the surface of the concrete crack with the known depth under different environmental parameters are learned by artificial intelligence, so that the quantitative relationship among the concrete crack depth, the concrete surface temperature distribution and the environmental parameters is established, and remote non-contact detection of the concrete crack depth is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of remote non-contact detection of concrete crack depth, and in particular to a remote non-contact concrete crack depth detection method, device, computer equipment and storage medium. Background Art

[0002] Cracks are one of the most harmful defects in concrete structures. The presence of cracks greatly accelerates the intrusion of outside air and harmful ions, which in turn causes the internal steel bars to rust and swell. The width, length and depth of concrete cracks are important indicators for measuring the harmfulness of cracks. At present, the industry has developed a variety of detection technologies for the width and length of concrete cracks, but there is no remote non-contact detection technology for the depth of concrete cracks. For example, the patent with the publication number CN117781964A and the invention name "A Concrete Crack Depth Indicator Tracing Radar Visualization Detection Method" discloses a concrete crack depth detection technology combining an indicator and a radar method. This technology requires the use of a handheld radar and requires contact with the concrete crack surface to inject the indicator, which is not suitable for concrete crack depth detection in inaccessible areas. The patent with the publication number CN219869566U and the invention name "A Concrete Structure Crack Depth Detection Method for Ultrasonic Single-sided Flat Detection Method" discloses a concrete crack depth detection technology using an ultrasonic flat detection method. This technology requires that the ultrasonic transmitter and ultrasonic receiver be installed and fixed on the concrete surface, which is not suitable for concrete crack depth detection in inaccessible areas.

[0003] In summary, there is currently no reliable detection and assessment technology for the depth of concrete cracks in inaccessible areas. Summary of the invention

[0004] The purpose of the present invention is to provide a remote non-contact concrete crack depth detection method, device, computer equipment and storage medium, which adopt artificial intelligence to learn infrared images and visible light images of the surface of concrete cracks of known depth under different environmental parameters, so as to establish a quantitative relationship between the concrete crack depth and the concrete surface temperature distribution and environmental parameters, and realize remote non-contact detection of concrete crack depth.

[0005] In order to achieve the above object, the present invention provides a remote non-contact concrete crack depth detection method, comprising the following steps:

[0006] Take visible light and infrared images of concrete cracks in inaccessible areas and measure environmental parameters during the images;

[0007] Input the captured visible light images and infrared images and the measured environmental parameters into the trained deep learning model;

[0008] The trained deep learning model predicts and outputs the concrete crack depth based on the quantitative relationship between the concrete crack depth and the concrete surface temperature distribution and environmental parameters.

[0009] In the present invention, the trained deep learning model is obtained by the following steps:

[0010] Create concrete crack samples of known width and depth;

[0011] Take visible light images and infrared images of concrete crack samples under different environmental parameters, and record the environmental parameters during shooting; environmental parameters include ambient temperature, humidity and light intensity;

[0012] A deep learning dataset is established based on visible light images and infrared images of concrete crack samples under different environmental parameters; the deep learning dataset includes a deep learning training dataset and a deep learning verification dataset;

[0013] The deep learning model is trained using a deep learning training data set. The training effect of the deep learning model is verified using a deep learning verification data set. The model parameters are adjusted and the number of training times is increased until the training effect of the deep learning model meets the requirements, thereby obtaining a trained deep learning model.

[0014] In the present invention, visible light images and infrared images of concrete crack samples under different environmental parameters are taken, including:

[0015] The concrete crack samples were placed in a natural environment, and visible light images and infrared images of the concrete crack samples were taken at different times of the day.

[0016] Repeat the above steps every day until you have accumulated enough deep learning training materials.

[0017] In the present invention, the ambient temperature and humidity refer to the ambient air temperature and humidity, and the light intensity refers to the light intensity at the concrete crack sample.

[0018] In the present invention, a deep learning data set is established based on visible light images and infrared images of concrete crack samples under different environmental parameters, including: taking the same concrete crack sample, visible light images, infrared images, environmental parameters, crack width and crack depth taken at the same time as a group of data.

[0019] In the present invention, the crack cross section of the concrete crack sample is "V" shaped, the crack shape is "I" shaped, "Y" shaped and "X" shaped, the crack width is (0.1-10) mm, and the crack depth is (10-50) mm.

[0020] In the present invention, the training effect evaluation indicators include the determination coefficient, accuracy and mean absolute percentage error of the deep learning model on the deep learning training data set and the deep learning verification data set; the accuracy is the proportion of samples whose errors between the predicted value and the actual value are within ±1 mm and ±2 mm to the total samples.

[0021] In the present invention, the deep learning model is an AdaBoost model; the prediction error of the trained AdaBoost model is less than 10%.

[0022] In order to achieve the above object, the present invention also provides a remote non-contact concrete crack depth detection device, comprising:

[0023] A data acquisition module, used to acquire visible light images, infrared images and environmental parameters of concrete cracks in inaccessible areas and input them into a depth prediction module;

[0024] The depth prediction module is equipped with a trained deep learning model, which is used to predict and output the depth of concrete cracks based on the quantitative relationship between the depth of concrete cracks and the temperature distribution on the concrete surface and environmental parameters.

[0025] In order to achieve the above object, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the above-mentioned remote non-contact concrete crack depth detection method are implemented.

[0026] In order to achieve the above object, the present invention also provides a computer-readable storage medium, on which computer-readable instructions are stored. When the computer-readable instructions are executed, the steps of the above-mentioned remote non-contact concrete crack depth detection method are implemented.

[0027] Beneficial technical effects of the present invention:

[0028] The present invention provides a remote non-contact concrete crack depth detection method, device, computer equipment and storage medium, which use artificial intelligence to learn infrared images and visible light images of the surface of concrete cracks of known depth under different environmental parameters, thereby establishing the potential law between the concrete crack depth and the concrete surface temperature distribution and environmental parameters, and realizing remote non-contact detection of the concrete crack depth; the concrete crack depth detection in the inaccessible area can be realized based on the visible light image and infrared image of the concrete crack acquired at a long distance and the environmental parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 A flow chart of an embodiment of a remote non-contact concrete crack depth detection method of the present invention;

[0030] Figure 2 It is a schematic diagram of the crack cross section of the concrete crack sample;

[0031] Figure 3 Schematic diagram of crack shape of concrete crack sample;

[0032] Figure 4 This is a schematic diagram of a "I"-shaped crack in a concrete crack sample;

[0033] Figure 5 This is a schematic diagram of a "Y"-shaped crack in a concrete crack sample;

[0034] Figure 6 This is a schematic diagram of a "cross" crack in a concrete crack sample;

[0035] Figure 7 It is a schematic structural view of an embodiment of a remote non-contact concrete crack depth detection device of the present invention;

[0036] Figure 8 The figure is a schematic structural view of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which this application belongs; the terms used in the specification of the application are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned figure descriptions are intended to cover non-exclusive inclusions.

[0038] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0039] The technical solution of the present invention is clearly and completely described below in conjunction with the accompanying drawings and specific embodiments.

[0040] refer to Figure 1 , shows a flow chart of an embodiment of a remote non-contact concrete crack depth detection method, the remote non-contact concrete crack depth detection method comprising the following steps:

[0041] Take visible light and infrared images of concrete cracks in inaccessible areas and measure environmental parameters during the images;

[0042] Input the captured visible light images and infrared images and the measured environmental parameters into the trained deep learning model;

[0043] The trained deep learning model predicts and outputs the concrete crack depth based on the quantitative relationship between the concrete crack depth and the concrete surface temperature distribution and environmental parameters.

[0044] In this embodiment, the trained deep learning model is obtained by the following steps:

[0045] Create concrete crack samples of known width and depth;

[0046] Take visible light images and infrared images of concrete crack samples under different environmental parameters, and record the environmental parameters during shooting; environmental parameters include ambient temperature, humidity and light intensity;

[0047] A deep learning dataset is established based on visible light images and infrared images of concrete crack samples under different environmental parameters; the deep learning dataset includes a deep learning training dataset and a deep learning verification dataset;

[0048] The deep learning model is trained using a deep learning training data set. The training effect of the deep learning model is verified using a deep learning verification data set. The model parameters are adjusted and the number of training times is increased until the training effect of the deep learning model meets the requirements, thereby obtaining a trained deep learning model.

[0049] In this embodiment, visible light images and infrared images of concrete crack samples under different environmental parameters are captured, including:

[0050] The concrete crack samples were placed in a natural environment, and visible light images and infrared images of the concrete crack samples were taken at different times of the day.

[0051] Repeat the above steps every day until you have accumulated enough deep learning training materials.

[0052] In this embodiment, the ambient temperature and humidity refer to the ambient air temperature and humidity, and the light intensity refers to the light intensity at the concrete crack sample.

[0053] In this embodiment, a deep learning data set is established based on visible light images and infrared images of concrete crack samples under different environmental parameters, including: taking the same concrete crack sample, visible light images, infrared images, environmental parameters, crack width and crack depth taken at the same time as a set of data.

[0054] See also Figure 2-6In this embodiment, the crack cross section of the concrete crack sample is "V"-shaped, the crack shapes are "I"-shaped, "Y"-shaped and "X"-shaped, the crack width is (0.1-10) mm, and the crack depth is (10-50) mm.

[0055] In this embodiment, the training effect evaluation indicators include the determination coefficient, accuracy and mean absolute percentage error of the deep learning model on the deep learning training data set and the deep learning verification data set; the accuracy is the proportion of samples whose errors between the predicted value and the actual value are within ±1 mm and ±2 mm to the total samples.

[0056] In this embodiment, the deep learning model is an AdaBoost model; the prediction error of the trained AdaBoost model is less than 10%.

[0057] The AdaBoost model is an integrated learning model that combines multiple weak learners. It optimizes the model by assigning a weight to each weak learner and dynamically adjusting the weight of each weak learner according to its performance.

[0058] In this embodiment, the determination coefficient R 2 The calculation formula is as follows:

[0059]

[0060] Among them, SS R is the regression sum of squares, SS T is the total sum of squares, y i is the true depth of the crack, t i To predict crack depth, is the average true crack depth, is the average predicted crack depth, and n is the number of training times;

[0061] The accuracy acc calculation formula is as follows:

[0062]

[0063] Among them, N true is the number of data with prediction results within the error range, and n is the number of training times;

[0064] The mean absolute percentage error (MAPE) is calculated as follows:

[0065]

[0066] Among them, y i is the true depth of the crack, t i is the predicted crack depth, and n is the number of training times.

[0067] refer to Figure 7As an implementation of the above method, the present invention provides an embodiment of a remote non-contact concrete crack depth detection device. The device embodiment corresponds to the embodiment of the above remote non-contact concrete crack depth detection method. The device can be specifically applied to various electronic devices.

[0068] The remote non-contact concrete crack depth detection device described in this embodiment includes:

[0069] A data acquisition module, used to acquire visible light images, infrared images and environmental parameters of concrete cracks in inaccessible areas and input them into a depth prediction module;

[0070] The depth prediction module is equipped with a trained deep learning model, which is used to predict and output the depth of concrete cracks based on the quantitative relationship between the depth of concrete cracks and the temperature distribution on the concrete surface and environmental parameters.

[0071] See also Figure 8 As an implementation of the above method, the present invention provides an embodiment of a computer device, which corresponds to the embodiment of the above remote non-contact concrete crack depth detection method.

[0072] The computer device described in this embodiment includes a memory, a processor and a network interface that are interconnected and communicated through a system bus. It should be noted that the figure only shows a computer device with component memory, a processor and a network interface, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, it can be understood by those skilled in the art that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application-specific integrated circuits, programmable gate arrays, digital processors, embedded devices, etc.

[0073] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0074] The memory includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory, random access memory, static random access memory, read-only memory, electrically erasable programmable read-only memory, programmable read-only memory, magnetic memory, disk, optical disk, etc. In some embodiments, the memory can be an internal storage unit of the computer device, such as a hard disk or memory of the computer device. In other embodiments, the memory can also be an external storage device of the computer device, such as a plug-in hard disk, smart memory card, secure digital card, flash memory card, etc. equipped on the computer device. Of course, the memory can also include both the internal storage unit of the computer device and its external storage device. In this embodiment, the memory is generally used to store the operating system and various application software installed on the computer device, such as the computer-readable instructions of the above-mentioned remote non-contact concrete crack depth detection method. In addition, the memory can also be used to temporarily store various types of data that have been output or are to be output.

[0075] In some embodiments, the processor may be a central processing unit, a controller, a microcontroller, a microprocessor, or other data processing chip. The processor is generally used to control the overall operation of the computer device. In this embodiment, the processor is used to run the computer-readable instructions stored in the memory or process data, such as running the computer-readable instructions of the above-mentioned remote non-contact concrete crack depth detection method.

[0076] The network interface may include a wireless network interface or a wired network interface, which is generally used to establish a communication connection between the computer device and other electronic devices.

[0077] As an implementation of the above method, the present invention provides an embodiment of a computer-readable storage medium, and the embodiment of the computer-readable storage medium corresponds to the embodiment of the above remote non-contact concrete crack depth detection method.

[0078] The computer-readable storage medium described in this embodiment stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the above-mentioned remote non-contact concrete crack depth detection method.

[0079] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A remote non-contact concrete crack depth detection method, characterized in that: The method comprises the following steps: taking visible light images and infrared images of concrete cracks in the inaccessible area, and measuring environmental parameters during the taking of images; Input the captured visible light images and infrared images and the measured environmental parameters into the trained deep learning model; The trained deep learning model predicts and outputs the concrete crack depth based on the quantitative relationship between the concrete crack depth and the concrete surface temperature distribution and environmental parameters.

2. The remote non-contact concrete crack depth detection method according to claim 1 is characterized in that: The trained deep learning model is obtained through the following steps: Create concrete crack samples of known width and depth; Take visible light images and infrared images of concrete crack samples under different environmental parameters, and record the environmental parameters during shooting; environmental parameters include ambient temperature, humidity and light intensity; A deep learning dataset is established based on visible light images and infrared images of concrete crack samples under different environmental parameters; The deep learning dataset includes the deep learning training dataset and the deep learning verification dataset; The deep learning model is trained using a deep learning training data set. The training effect of the deep learning model is verified using a deep learning verification data set. The model parameters are adjusted and the number of training times is increased until the training effect of the deep learning model meets the requirements, thereby obtaining a trained deep learning model.

3. The remote non-contact concrete crack depth detection method according to claim 2 is characterized in that: Take visible light and infrared images of concrete crack samples under different environmental parameters, including: The concrete crack samples were placed in a natural environment, and visible light images and infrared images of the concrete crack samples were taken at different times of the day. Repeat the above steps every day until you have accumulated enough deep learning training materials.

4. The remote non-contact concrete crack depth detection method according to claim 2 is characterized in that: A deep learning dataset is established based on visible light images and infrared images of concrete crack samples under different environmental parameters, including: the same concrete crack sample, visible light images, infrared images, environmental parameters, crack width and crack depth taken at the same time as a set of data.

5. The remote non-contact concrete crack depth detection method according to claim 2, characterized in that: The crack cross section of the concrete crack sample is "V" shaped, the crack shapes are "I" shaped, "Y" shaped and "X" shaped, the crack width is (0.1~10) mm, and the crack depth is (10~50) mm.

6. The remote non-contact concrete crack depth detection method according to claim 2, characterized in that: The training effect evaluation indicators include the determination coefficient, accuracy and mean absolute percentage error of the deep learning model on the deep learning training data set and the deep learning verification data set; the accuracy is the proportion of samples with an error between the predicted value and the actual value within ±1 mm and ±2 mm to the total samples.

7. The remote non-contact concrete crack depth detection method according to claim 1, characterized in that: The deep learning model is the AdaBoost model; the prediction error of the trained AdaBoost model is less than 10%.

8. A remote non-contact concrete crack depth detection device, characterized in that: include: A data acquisition module, used to acquire visible light images, infrared images and environmental parameters of concrete cracks in inaccessible areas and input them into a depth prediction module; The depth prediction module is equipped with a trained deep learning model, which is used to predict and output the depth of concrete cracks based on the quantitative relationship between the depth of concrete cracks and the temperature distribution on the concrete surface and environmental parameters.

9. A computer device comprising a memory and a processor, wherein the memory stores computer-readable instructions, characterized in that: When the processor executes the computer-readable instructions, the steps of the remote non-contact concrete crack depth detection method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having computer-readable instructions stored thereon, characterized in that: When the computer-readable instructions are executed, the steps of the remote non-contact concrete crack depth detection method according to any one of claims 1 to 7 are implemented.

Images