Concrete surface defect detection method and system based on image recognition

Abstract

The invention discloses a concrete surface defect detection method and system based on image recognition, and belongs to the technical field of image data processing. A to-be-recognized concrete surface image is preprocessed, and the preprocessed to-be-recognized concrete surface image is obtained; and carrying out feature extraction on the preprocessed concrete surface image to be recognized by adopting a surface defect feature extraction algorithm to obtain features of the image to be recognized, and finally, calling a pre-deployed concrete defect recognition model to recognize the features of the image to be recognized. And the surface defect detection result corresponding to each to-be-identified concrete surface image is obtained, so that the concrete surface defects can be effectively detected, the defect detection precision can be effectively improved, and the concrete surface defect detection efficiency is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of image data processing, and in particular relates to a method and system for detecting concrete surface defects based on image recognition. Background Art

[0002] Concrete, also known as concrete for short, is an artificial stone made of cement, sand, gravel and water in a certain proportion. It has the advantages of high strength, good durability and strong plasticity, and is the main material for modern construction and infrastructure construction. Concrete can be made into various shapes and sizes according to needs, and is widely used in houses, bridges, roads, water conservancy projects and other projects. After hardening, it forms a solid whole that can withstand large loads and has good anti-seepage and anti-freeze properties. With the development of science and technology, the types of concrete are constantly enriched, such as high-performance concrete and green concrete, which have brought more possibilities to the construction industry. As the main material of modern construction, the surface quality of concrete directly affects the safety and service life of the building. However, concrete is prone to surface defects such as cracks, holes, and pockmarks during pouring and curing. Traditional manual detection methods are inefficient and subjective, and it is difficult to meet the needs of large-scale and high-precision detection. Summary of the invention

[0003] The present invention provides a concrete surface defect detection method and system based on image recognition, which are used to solve the problem of low efficiency in the existing concrete detection process.

[0004] In one aspect, the present invention provides a method for detecting concrete surface defects based on image recognition, comprising: Determine a defect area to be identified, control the drone to collect images of the defect area to be identified, and obtain at least one image of the concrete surface to be identified; For any concrete surface image to be identified, preprocessing the concrete surface image to be identified is performed to obtain the concrete surface image to be identified after the preprocessing; A surface defect feature extraction algorithm is used to extract features of the concrete surface image to be identified after preprocessing to obtain features of the image to be identified; The pre-deployed concrete defect recognition model is called to recognize the features of the image to be recognized, and the surface defect detection result corresponding to each concrete surface image to be recognized is obtained.

[0005] Furthermore, after obtaining the surface defect detection result corresponding to each concrete surface image to be identified, the method further includes: According to the surface defect detection result corresponding to each concrete surface image to be identified, the defective area is marked on the map corresponding to the defective area to be identified.

[0006] Further, preprocess the concrete surface image to be recognized to obtain the preprocessed concrete surface image to be recognized, including: Perform grayscale processing on the concrete surface image to be recognized to obtain the grayscale processed concrete surface image to be recognized; Perform grayscale non-linear transformation on the grayscale processed concrete surface image to be recognized to obtain the grayscale non-linearly transformed concrete surface image to be recognized; Perform image sharpening on the grayscale non-linearly transformed concrete surface image to be recognized to obtain the preprocessed concrete surface image to be recognized.

[0007] Further, performing grayscale processing on the concrete surface image to be recognized to obtain the grayscale processed concrete surface image to be recognized includes: Determine the R value, G value, and B value in the concrete surface image to be recognized to obtain the R value, G value, and B value corresponding to each pixel point; For any pixel point in the concrete surface image to be recognized, take the maximum value among the R value, G value, and B value as the grayscale value corresponding to the pixel point to obtain the grayscale processed concrete surface image to be recognized.

[0008] Further, performing grayscale non-linear transformation on the grayscale processed concrete surface image to be recognized to obtain the grayscale non-linearly transformed concrete surface image to be recognized includes: For any pixel point in the grayscale processed concrete surface image to be recognized, perform gamma transformation on the grayscale value of the pixel point to obtain the grayscale non-linearly transformed concrete surface image to be recognized.

[0009] Further, performing image sharpening on the grayscale non-linearly transformed concrete surface image to be recognized to obtain the preprocessed concrete surface image to be recognized includes: Using the Sobel operator to perform image sharpening on the grayscale non-linearly transformed concrete surface image to be recognized to obtain the preprocessed concrete surface image to be recognized.

[0010] Further, use the surface defect feature extraction algorithm to extract features from the preprocessed concrete surface image to be recognized to obtain the features of the image to be recognized, including: Binarize the preprocessed concrete surface image to be recognized to obtain the binarized concrete surface image to be recognized; For the binarized concrete surface image to be recognized, extract the boundary features of the concrete surface image to be recognized to obtain the features of the image to be recognized.

[0011] Further, the concrete defect recognition model is set as a CNN model or a YOLO model.

[0012] Further, a method for pre-deploying a concrete defect recognition model includes: deploying the concrete defect recognition model by using an improved pigeon flock optimization algorithm.

[0013] On the other hand, the present invention provides a concrete surface defect detection system based on image recognition, including: a concrete image acquisition module, an image preprocessing module, an image feature extraction module, and a concrete defect detection module; The concrete image acquisition module is used to determine a defect area to be recognized, control a drone to perform image acquisition on the defect area to be recognized, and obtain at least one concrete surface image to be recognized; The image preprocessing module is used to perform preprocessing on any concrete surface image to be recognized, and obtain a preprocessed concrete surface image to be recognized; The image feature extraction module is used to extract features from the preprocessed concrete surface image to be recognized by using a surface defect feature extraction algorithm, and obtain image features to be recognized; The concrete defect detection module is used to call a pre-deployed concrete defect recognition model to recognize the image features to be recognized, and obtain a surface defect detection result corresponding to each concrete surface image to be recognized.

[0014] A concrete surface defect detection method and system based on image recognition provided by the present invention preprocess a concrete surface image to be recognized, obtain a preprocessed concrete surface image to be recognized, extract features from the preprocessed concrete surface image to be recognized by using a surface defect feature extraction algorithm, obtain image features to be recognized, and finally call a pre-deployed concrete defect recognition model to recognize the image features to be recognized, and obtain a surface defect detection result corresponding to each concrete surface image to be recognized. It can not only effectively detect concrete surface defects, but also effectively improve the defect detection accuracy and the concrete surface defect detection efficiency. Description of the Drawings

[0015] The drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present invention, and are used together with the specification to explain the principles of the present invention.

[0016] Figure 1 It is a flowchart of a concrete surface defect detection method based on image recognition provided by an embodiment of the present invention.

[0017] Figure 2 It is a schematic structural diagram of a concrete surface defect detection system based on image recognition provided by an embodiment of the present invention.

[0018] Through the above-mentioned accompanying drawings, specific embodiments of the present invention have been shown, and more detailed descriptions will be given hereinafter. These drawings and textual descriptions are not intended to limit the scope of the inventive concept in any way, but to illustrate the concept of the present invention to those skilled in the art by reference to specific embodiments. Detailed Description of Specific Embodiments

[0019] Here, exemplary embodiments will be described in detail, and examples are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0020] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0021] As Figure 1 shown, an embodiment of the present invention provides a method for detecting concrete surface defects based on image recognition, including: S101. Determine the defect area to be recognized, control the drone to collect images of the defect area to be recognized, and obtain at least one image of the concrete surface to be recognized; The defect area to be recognized can be a target area input by a staff through human-computer interaction. This target area contains coordinates in a specific world coordinate system (i.e., the target area contains corresponding longitude and latitude) to facilitate planning the image collection route. Existing route planning algorithms can be used to plan the flight route of the drone to ensure that all positions of the defect area to be recognized are collected.

[0022] S102. For any image of the concrete surface to be recognized, preprocess the image of the concrete surface to be recognized to obtain the preprocessed image of the concrete surface to be recognized; By preprocessing the image of the concrete surface to be recognized in the embodiment of the present invention, the features related to the defects in the image can be effectively enhanced to improve the final defect recognition accuracy.

[0023] S103. Use a surface defect feature extraction algorithm to extract features from the preprocessed image of the concrete surface to be recognized to obtain the features of the image to be recognized; After preprocessing the image of the concrete surface to be recognized, the features become easier to extract. Therefore, the features of the preprocessed image of the concrete surface to be recognized can be extracted to obtain the features of the image to be recognized, and deep learning technology can be used to recognize the features of the image to be recognized, thereby realizing defect recognition.

[0024] S104. Invoke the pre-deployed concrete defect recognition model to recognize the to-be-recognized image features, and obtain the surface defect detection results corresponding to each to-be-recognized concrete surface image.

[0025] The surface defect detection results can be that there are defects or no defects, or can be specific defect types (such as cracks, holes, pitted surfaces), which can be defined by the staff themselves, so as to achieve customized detection tasks.

[0026] In the embodiment of the present invention, after obtaining the surface defect detection results corresponding to each to-be-recognized concrete surface image, it further includes: According to the surface defect detection results corresponding to each to-be-recognized concrete surface image, mark the defective areas on the map corresponding to the to-be-recognized defect areas.

[0027] By marking the defective areas on the map corresponding to the to-be-recognized defect areas, a defect detection record can be formed, which is convenient for the staff to view, and at the same time, can be used as a basis to remind the staff, so as to achieve automatic abnormal detection and notification.

[0028] In the embodiment of the present invention, preprocess the to-be-recognized concrete surface image to obtain the preprocessed to-be-recognized concrete surface image, including: Perform grayscale processing on the to-be-recognized concrete surface image to obtain the grayscale-processed to-be-recognized concrete surface image; Perform grayscale non-linear transformation on the grayscale-processed to-be-recognized concrete surface image to obtain the grayscale non-linearly transformed to-be-recognized concrete surface image; Perform image sharpening on the grayscale non-linearly transformed to-be-recognized concrete surface image to obtain the preprocessed to-be-recognized concrete surface image.

[0029] In the embodiment of the present invention, by sequentially performing grayscale processing, grayscale non-linear transformation, and image sharpening on the to-be-recognized concrete surface image, the defect features can be effectively highlighted, the influence of other unnecessary features can be reduced, and the final recognition accuracy can be effectively improved.

[0030] In the embodiment of the present invention, perform grayscale processing on the to-be-recognized concrete surface image to obtain the grayscale-processed to-be-recognized concrete surface image, including: Determine the R value, G value, and B value in the to-be-recognized concrete surface image to obtain the R value, G value, and B value corresponding to each pixel point; For any pixel point in the to-be-recognized concrete surface image, take the maximum value among the R value, G value, and B value as the grayscale value corresponding to the pixel point to obtain the grayscale-processed to-be-recognized concrete surface image.

[0031] In an embodiment of the present invention, a gray-scale non-linear transformation is performed on the concrete surface image to be recognized after gray-scale processing, and the concrete surface image to be recognized after gray-scale non-linear transformation is obtained, including: for any pixel point in the concrete surface image to be recognized after gray-scale processing, a gamma transformation is used to transform the gray value of the pixel point, and the concrete surface image to be recognized after gray-scale non-linear transformation is obtained.

[0032] In an embodiment of the present invention, an image sharpening process is performed on the concrete surface image to be recognized after gray-scale non-linear transformation, and the concrete surface image to be recognized after preprocessing is obtained, including: using a Sobel operator to perform an image sharpening process on the concrete surface image to be recognized after gray-scale non-linear transformation, and the concrete surface image to be recognized after preprocessing is obtained.

[0033] In an embodiment of the present invention, a surface defect feature extraction algorithm is used to extract features from the concrete surface image to be recognized after preprocessing, and the features of the image to be recognized are obtained, including: Binarize the concrete surface image to be recognized after preprocessing to obtain the binarized concrete surface image to be recognized; For the binarized concrete surface image to be recognized, perform boundary feature extraction on the concrete surface image to be recognized to obtain the features of the image to be recognized.

[0034] For any pixel point, there are at least three adjacent pixel points around it, that is, an image corner point; and there are at least five pixel points on the image boundary; there are eight pixel points around the remaining pixel points; in order to improve the recognition effect of defects while reducing the influence of noise, in an embodiment of the present invention, for any target pixel point, if the number of adjacent pixel points with a value of 1 around the target pixel point is less than 4, the value corresponding to the target pixel point is set to 0, otherwise it is set to 1, so as to further optimize the defect features.

[0035] In an embodiment of the present invention, the concrete defect recognition model is set as a CNN (Convolutional Neural Network) model or a YOLO model. It should be noted that in addition to the above models, other image recognition models can also be used as the concrete defect recognition model.

[0036] In an embodiment of the present invention, the method for pre-deploying the concrete defect recognition model includes: using an improved pigeon flock optimization algorithm to deploy the concrete defect recognition model.

[0037] Optionally, the improved pigeon flock optimization algorithm may include: A1. Initialize the hyperparameters of the concrete defect recognition model to obtain a pigeon flock; among them, the pigeon flock includes multiple different pigeons, and each pigeon contains some or all of the parameters to be optimized of the concrete defect recognition model, which can be customized by the staff.

[0038] For example, when the concrete defect recognition model is set as a CNN model, the pigeons can include all or some of the weights of the CNN model, so as to customize the training plan. During the initialization of the hyperparameters, a random initialization method can be used for initialization.

[0039] A2. Obtain the loss function value corresponding to each pigeon by using the pre-acquired sample image features and the artificial defect labels corresponding to the sample image features; The hyperparameters included in any pigeon can be applied to the concrete defect recognition model, and then, using the sample image features as the input and the artificial defect labels corresponding to the sample image features as the expected output, obtain the loss function value corresponding to each pigeon. The loss function value can be determined by the root mean square loss function or the cross-entropy loss function.

[0040] A3. Improve the flight process of the pigeon flock by using an adaptive inertia cooperation search strategy, and obtain the pigeon flock after one flight as:

[0041]

[0042]

[0043] Among them, represents the t th pigeon in the i th training process, represents the i th pigeon after one flight, i = 1, 2, …, N , N represents the total number of pigeons in the pigeon flock, represents the t +1th training process, the i th pigeon's corresponding speed, represents the t th training process, the i th pigeon's corresponding speed, represents the natural constant, represents the inertia weight, R represents the positioning track planning and magnet, and R belongs to a random number between (0, 1), represents the first random number between (0, 1), represents the second random number between (0, 1), represents thet In the i th training process, the randomly paired search pigeon corresponding to the th pigeon; denotes the preset maximum value of the inertia weight, denotes the preset minimum value of the inertia weight, and T denotes the maximum number of training times; In the embodiment of the present invention, an adaptive inertia cooperation search strategy is adopted to improve the flight process of the pigeon flock, changing the original flight towards the optimal position to a cooperative search flight, which can effectively improve the local area search ability of the algorithm. At the same time, it will also fly towards the optimal position, effectively improving the local search ability of the algorithm. Meanwhile, the algorithm can have stronger global search ability and search speed in the early stage, and more refined search ability in the later stage of the algorithm.

[0044] A4. Improve the elite elimination strategy of the pigeon flock by adopting a hybrid information learning strategy, and obtain the pigeons after the second flight, including: Randomly pair the pigeons after the first flight in pairs, and for any two paired pigeons, regard the pigeon with a smaller loss function value as the winner and the pigeon with a larger loss function value as the loser; Update the winner to:

[0045] where denotes the t th winner in the th training process, m = 1, 2,..., N / 2, denotes the rd random number between (0, 1), denotes the Cauchy distribution factor, denotes the th random number between (0, 1), denotes the average position of all winners, denotes the th random number between (0, 1), and G denotes the Gaussian distribution factor;

[0046] where denotes the t th winner in the th training process, n = 1, 2,..., N / 2, denotes the first learning factor, Represents the second learning factor, Represents the sixth random number between (0, 1), Represents the seventh random number between (0, 1), Represents the loser The corresponding winner, Represents the loser The corresponding historical optimal value.

[0047] In the embodiment of the present invention, a hybrid information learning strategy is adopted to improve the elite elimination strategy of the pigeon flock. A speed and position update strategy based on a competition mechanism is used to perform two different update strategies on the pigeons in the population. The loser is updated using the information of the winner, thereby improving the convergence performance of the algorithm. The winner uses the Cauchy distribution, Levy distribution, and Gaussian distribution to ensure that the movement is not limited to a small local area, which can increase the diversity of the population and expand the search range. As the algorithm progresses, the pigeon flock gradually converges, enabling the algorithm to maintain the global search ability while also improving the convergence accuracy of the algorithm and ensuring the normal execution of the algorithm.

[0048] The winners after the second flight and the losers after the second flight are jointly used as the pigeons after the second flight.

[0049] A5. Determine whether the current training times are greater than or equal to the preset maximum training times. If so, re-determine the optimal pigeon according to the pigeons after the second flight, and use the hyperparameters included in the optimal pigeon as the final hyperparameters of the concrete defect recognition model to implement the deployment of the concrete defect recognition model. Otherwise, return to step A2.

[0050] As Figure 2 shown, the embodiment of the present invention provides a concrete surface defect detection system based on image recognition, including: a concrete image acquisition module 201, an image preprocessing module 202, an image feature extraction module 203, and a concrete defect detection module 204; The concrete image acquisition module 201 is used to determine the defect area to be recognized, control the drone to perform image acquisition on the defect area to be recognized, and obtain at least one concrete surface image to be recognized; The image preprocessing module 202 is used to perform preprocessing on any concrete surface image to be recognized to obtain the preprocessed concrete surface image to be recognized; The image feature extraction module 203 is used to extract features from the preprocessed concrete surface image to be recognized by using a surface defect feature extraction algorithm to obtain the image features to be recognized; The concrete defect detection module 204 is configured to call a pre-deployed concrete defect recognition model to recognize the image features to be recognized, and obtain the surface defect detection results corresponding to each concrete surface image to be recognized.

[0051] The concrete surface defect detection system based on image recognition provided by the embodiments of the present invention can execute the above method technical solutions, and its principle and beneficial effects are similar, so details are not described herein again.

[0052] The method and system for detecting concrete surface defects based on image recognition provided by the embodiments of the present invention preprocess the concrete surface images to be recognized to obtain the preprocessed concrete surface images to be recognized, extract features from the preprocessed concrete surface images to be recognized by using a surface defect feature extraction algorithm to obtain image features to be recognized, and finally call a pre-deployed concrete defect recognition model to recognize the image features to be recognized, and obtain the surface defect detection results corresponding to each concrete surface image to be recognized, which can not only effectively detect the concrete surface defects, but also effectively improve the defect detection accuracy and the concrete surface defect detection efficiency.

[0053] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0054] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for realizing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0055] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one process Figure 1 or more processes and / or blocks Figure 1 or more blocks.

[0056] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are performed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process Figure 1 or more processes and / or blocks Figure 1 or more blocks.

[0057] Those of ordinary skill in the art can understand that all or part of the steps in implementing the above facts and methods can be completed by instructing relevant hardware through a program. The program involved or the said program can be stored in a computer-readable storage medium. When the program is executed, it includes the following steps: At this time, the corresponding method steps are introduced. The storage medium can be ROM / RAM, magnetic disk, optical disc, etc.

[0058] The specific embodiments described above further elaborate on the object, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for detecting concrete surface defects based on image recognition, characterized in that: include: Determine a defect area to be identified, control the drone to collect images of the defect area to be identified, and obtain at least one image of the concrete surface to be identified; For any concrete surface image to be identified, preprocessing the concrete surface image to be identified is performed to obtain the concrete surface image to be identified after the preprocessing; A surface defect feature extraction algorithm is used to extract features of the concrete surface image to be identified after preprocessing to obtain features of the image to be identified; The pre-deployed concrete defect recognition model is called to recognize the features of the image to be recognized, and the surface defect detection result corresponding to each concrete surface image to be recognized is obtained.

2. The method for detecting concrete surface defects based on image recognition according to claim 1, characterized in that: After obtaining the surface defect detection result corresponding to each concrete surface image to be identified, it also includes: According to the surface defect detection result corresponding to each concrete surface image to be identified, the defective area is marked on the map corresponding to the defective area to be identified.

3. The method for detecting concrete surface defects based on image recognition according to claim 1, characterized in that: Preprocessing the concrete surface image to be identified to obtain the preprocessed concrete surface image to be identified, including: graying the concrete surface image to be identified to obtain a grayed concrete surface image to be identified; Performing grayscale nonlinear transformation on the concrete surface image to be identified after grayscale processing to obtain the concrete surface image to be identified after grayscale nonlinear transformation; The image of the concrete surface to be identified after the grayscale nonlinear transformation is subjected to image sharpening processing to obtain the concrete surface image to be identified after preprocessing.

4. The method for detecting concrete surface defects based on image recognition according to claim 3 is characterized in that: Graying the concrete surface image to be identified to obtain the grayed concrete surface image to be identified includes: Determine the R value, G value and B value in the concrete surface image to be identified, and obtain the R value, G value and B value corresponding to each pixel; For any pixel point in the concrete surface image to be identified, the maximum value among the R value, G value and B value is taken as the grayscale value corresponding to the pixel point to obtain the concrete surface image to be identified after grayscale processing.

5. The method for detecting concrete surface defects based on image recognition according to claim 4 is characterized in that: The grayscale nonlinear transformation is performed on the concrete surface image to be identified after the grayscale processing to obtain the concrete surface image to be identified after the grayscale nonlinear transformation, including: for any pixel point in the concrete surface image to be identified after the grayscale processing, the grayscale value of the pixel point is transformed by gamma transformation to obtain the concrete surface image to be identified after the grayscale nonlinear transformation.

6. The method for detecting concrete surface defects based on image recognition according to claim 5, characterized in that: The image of the concrete surface to be identified after the grayscale nonlinear transformation is subjected to image sharpening processing to obtain the concrete surface image to be identified after preprocessing, including: using a Sobel operator to perform image sharpening processing on the concrete surface image to be identified after the grayscale nonlinear transformation to obtain the concrete surface image to be identified after preprocessing.

7. The method for detecting concrete surface defects based on image recognition according to claim 1, characterized in that: The surface defect feature extraction algorithm is used to extract features from the preprocessed concrete surface image to be identified, and the features of the image to be identified are obtained, including: Binarizing the preprocessed concrete surface image to be identified to obtain a binarized concrete surface image to be identified; For the concrete surface image to be identified after binarization, boundary features are extracted to obtain features of the image to be identified.

8. The method for detecting concrete surface defects based on image recognition according to claim 1, characterized in that: The concrete defect recognition model is set as a CNN model or a YOLO model.

9. The method for detecting concrete surface defects based on image recognition according to claim 8, characterized in that: The pre-deployment method of the concrete defect recognition model includes: deploying the concrete defect recognition model using an improved pigeon flock optimization algorithm.

10. A concrete surface defect detection system based on image recognition, characterized in that: include: Concrete image acquisition module, image preprocessing module, image feature extraction module and concrete defect detection module; The concrete image acquisition module is used to determine the defect area to be identified, control the drone to collect images of the defect area to be identified, and obtain at least one concrete surface image to be identified; The image preprocessing module is used to preprocess any concrete surface image to be identified, and obtain the concrete surface image to be identified after preprocessing; The image feature extraction module is used to extract features from the pre-processed concrete surface image to be identified by using a surface defect feature extraction algorithm to obtain features of the image to be identified; The concrete defect detection module is used to call a pre-deployed concrete defect recognition model to recognize the features of the image to be recognized, and obtain a surface defect detection result corresponding to each concrete surface image to be recognized.

Images