Concrete bridge efflorescence defect detection method, device and system and storage medium

Abstract

The invention discloses a concrete bridge efflorescence defect detection method, device and system, and a storage medium, and belongs to the technical field of bridge defect detection, and the method comprises the steps: dividing a concrete bridge efflorescence image into a training set and a test set; an ADSN-YOLO11 model is trained according to the training set; carrying out structured pruning on the trained ADSN-YOLO11 model, and carrying out precision fine tuning training on the model after the structured pruning, so as to obtain an ADSN-YOLO11 Pruned model; the ADSN-YOLO11Pruned model is subjected to model format conversion, and the ADSN-YOLO11Pruned And deploying the ADSN-YOLO11Pruned model subjected to model format conversion to an RKNN hardware platform, performing multi-thread real-time detection on a test set, and when concrete bridge efflorescence is detected, marking a defect position and displaying a detection confidence coefficient. By adopting the technical scheme provided by the invention, the rapid and accurate detection of the saltpetering defect of the bridge is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of bridge defect detection, and particularly relates to a method, device, system, and storage medium for detecting the efflorescence defect of concrete bridges. Background Art

[0002] Concrete bridges are long-term exposed to the natural environment and are easily affected by factors such as rain, temperature and humidity changes, and freeze-thaw cycles, resulting in the efflorescence phenomenon. If not repaired in time, the efflorescence will not only affect the appearance, but also cause the concrete surface to spall, reduce the compactness and durability of the concrete, and threaten the safety of the bridge structure.

[0003] However, there are many problems with existing detection methods. Especially for the efflorescence detection method of concrete bridges based on deep learning, the real-time performance is not high and the efficiency is low. Usually, data needs to be transmitted to a remote server for processing, and the detection results cannot be quickly fed back to guide the bridge maintenance work. Secondly, the generalization of the model is not high, and these methods cannot identify the efflorescence defects under different background features. The reason is that the constructed dataset has a single background, and the deep learning model cannot learn the efflorescence features of bridges with rich background features in reality. At the same time, there are obvious deficiencies in the feature extraction of the efflorescence of concrete bridges, and it is difficult to effectively identify the subtle features of the efflorescence of concrete bridges, and the accuracy and reliability of the detection results are limited. Finally, most of the existing models have complex structures and consume a large amount of computing resources, and it is difficult to meet the requirements of lightweight and low power consumption of detection equipment in this special outdoor environment of bridges. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method, device, system, and storage medium for detecting the efflorescence defect of concrete bridges, which can accurately extract the features of the efflorescence of concrete bridges to achieve rapid and accurate detection of the efflorescence defect of bridges.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A method for detecting the efflorescence defect of a concrete bridge, comprising: Step S1, preprocessing the obtained efflorescence image of the concrete bridge; Step S2, dividing the preprocessed efflorescence image of the concrete bridge into a training set and a test set; Step S3, training the ADSN-YOLO11 model according to the training set; wherein, in the ADSN-YOLO11 model, the convolutional block of the YOLO11 model is replaced by an Adown module; the ADSN-YOLO11 model adopts an OD2SN neck structure, and the OD2SN neck structure is based on the SlimNeck neural network neck, and two convolutional blocks of the input channels in the VoVGSCSP module are replaced by an ODConv module; Step S4: Perform structured pruning on the trained ADSN-YOLO11 model, and then perform fine-tuning training on the accuracy of the model after structured pruning to obtain the ADSN-YOLO11_Pruned model; Step S5: Perform model format conversion on the ADSN-YOLO11_Pruned model; Step S6: Deploy the ADSN-YOLO11_Pruned model after model format conversion to the RKNN hardware platform, start a thread pool to perform multi-threaded real-time detection on the test set. When alkali efflorescence on the concrete bridge is detected, mark the defect location and display the detection confidence.

[0006] Preferably, in step S4, the pruning operation on specific modules in the trained ADSN-YOLO11 model includes: pruning Bottleneck in the C3k2 module, pruning Adown, pruning GSConv in the OD2SN neck structure, specifying the convolutional kernels between different modules for pruning, and pruning the detection head; among them, for the pruning of the detection head, the pruned part includes the Reg branch and the Cls branch of the P3, P4, and P5 layers.

[0007] Preferably, in step S5, the ADSN-YOLO11_Pruned model is converted into an ONNX model through the first model conversion; the ONNX model is converted into an RKNN model through the second model conversion; among them, during the second model conversion, the ONNX model is transplanted to the Linux system, the model is quantized to convert floating-point storage into integer storage, and then the model is packaged into an RKNN model suitable for deployment to edge devices.

[0008] The present invention also provides a device for detecting alkali efflorescence defects on a concrete bridge, including: A first processing module, configured to preprocess the obtained alkali efflorescence image of the concrete bridge; A second processing module, configured to divide the preprocessed alkali efflorescence image of the concrete bridge into a training set and a test set; A third processing module, configured to train the ADSN-YOLO11 model according to the training set; among them, in the ADSN-YOLO11 model, the convolutional block of the YOLO11 model is replaced with an Adown module; the ADSN-YOLO11 model adopts an OD2SN neck structure, and the OD2SN neck structure is based on the SlimNeck neural network neck, and two convolutional blocks of the input channels in the VoVGSCSP module are replaced with ODConv modules; A fourth processing module, configured to perform structured pruning on the trained ADSN-YOLO11 model, and then perform fine-tuning training on the accuracy of the model after structured pruning to obtain the ADSN-YOLO11_Pruned model; The fifth processing module is used to convert the model format of the ADSN-YOLO11_Pruned model; The sixth processing module is used to deploy the ADSN-YOLO11_Pruned model after model format conversion to the RKNN hardware platform, start a thread pool to perform multi-threaded real-time detection on the test set, and when alkali efflorescence of the concrete bridge is detected, mark the defect location and display the detection confidence.

[0009] Preferably, the pruning operation of specific modules in the trained ADSN-YOLO11 model by the fourth processing module includes: pruning the Bottleneck in the C3k2 module, pruning Adown, pruning GSConv in the OD2SN neck structure, specifying the convolutional kernels between different modules for pruning, and pruning the detection head; among them, for the pruning of the detection head, the pruned part includes the Reg branch and the Cls branch of the P3, P4, and P5 layers.

[0010] Preferably, the fifth processing module converts the ADSN-YOLO11_Pruned model into an ONNX model through the first model conversion; and converts the ONNX model into an RKNN model through the second model conversion.

[0011] The present invention also provides a concrete bridge alkali efflorescence defect detection system, including: a memory and a processor, where a computer program run by the processor is stored on the memory, and the computer program executes the concrete bridge alkali efflorescence defect detection method when run by the processor.

[0012] The present invention also provides a storage medium, on which a computer program is stored, and the computer program executes the concrete bridge alkali efflorescence defect detection method when running.

[0013] Compared with the prior art, the present invention has the following technical effects: 1. Strong background adaptability: The present invention collects concrete bridge alkali efflorescence pictures under different backgrounds and uses various methods to perform data augmentation on the data, including diverse background information. During the model training process, the model comes into contact with a wider range of visual effects, enabling the model to learn more extensive background features. Compared with the prior art, the dataset background of the present invention is more rich and diverse, can better simulate the complex background conditions in the real scene, and significantly improves the robustness of the model to complex backgrounds.

[0014] 2. Precise feature extraction ability: The present invention optimizes the YOLO11 algorithm. By improving the feature extraction layer and network structure of the algorithm, it can more effectively capture the detailed features of the alkali efflorescence area, thereby significantly improving the accuracy and reliability of the detection results.

[0015] 3. Lightweight and High Efficiency: The present invention improves the YOLO11 model to obtain the lightweight ADSN-YOLO11 model, and on this basis, lightweight operations such as structured pruning and quantization are also carried out, so that while maintaining high detection performance, the complexity of the model and the consumption of computing resources are greatly reduced, making it more suitable for deployment and operation on edge devices.

[0016] 4. Real-time and Convenience: Based on edge intelligence technology, the present invention adopts a multi-threaded inference mechanism to realize the real-time detection of efflorescence on concrete bridges by the lightweight model; it can be directly deployed on the bridge site without relying on a remote server, significantly shortening the detection cycle, and can quickly feedback the detection results to guide bridge maintenance work. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0018] Figure 1 It is a schematic flow chart of the method for detecting efflorescence defects on concrete bridges of the present invention; Figure 2 It is a network structure diagram of the unimproved YOLO11 model; Figure 3 Based on the present invention Figure 2 The designed network structure diagram of the ADSN-YOLO11 model; Figure 4 It is a network structure diagram of the Adown module; Figure 5 It is a network structure diagram of the VoVGSCSP module in the unimproved SlimNeck neck structure; Figure 6 Based on the present invention Figure 5 The designed improved VoVGSCSP module; Figure 7 It is a schematic diagram of the detection head structure; Figure 8 It is a schematic work flow diagram of multi-threaded inference on the RKNN hardware platform. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0020] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0021] Embodiment 1: As Figure 1 shown, the embodiment of the present invention provides a method for detecting the efflorescence defect of a concrete bridge, including: Step S1, preprocess the obtained efflorescence image of the concrete bridge; Step S2, divide the preprocessed efflorescence image of the concrete bridge into a training set and a test set; Step S3, train the ADSN-YOLO11 model according to the training set; Step S4, perform structured pruning on the trained ADSN-YOLO11 model, and then perform fine-tuning training on the accuracy of the structured pruned model to obtain the ADSN-YOLO11_Pruned model; Step S5, perform two model format conversions on the ADSN-YOLO11_Pruned model; Step S6, deploy the ADSN-YOLO11_Pruned model after model format conversion to the RKNN hardware platform, start a thread pool to perform multi-threaded real-time detection on the test set, and when the efflorescence of the concrete bridge is detected, mark the defect position and display the detection confidence.

[0022] As an implementation manner of the embodiment of the present invention, in step S1, the efflorescence image of the concrete bridge includes: efflorescence images of concrete bridges under various different backgrounds such as concrete river-crossing bridges, elevated highway bridges, and arch bridges.

[0023] As an implementation manner of the embodiment of the present invention, in step S1, the preprocessing of the obtained efflorescence image of the concrete bridge includes: Step S11, use the LabelImg tool to label the efflorescence targets in the efflorescence image of the concrete bridge to obtain a first bridge image dataset containing target categories and target positions; Step S12: Perform data augmentation on the first bridge image dataset to obtain a second bridge image dataset. Among them, the data augmentation enhances the first bridge image dataset by randomly adjusting brightness and contrast, randomly adding solar flares, randomly adding fog, randomly adding shadows, randomly blurring, randomly adjusting hue and saturation. As an implementation manner of the embodiment of the present invention, in step S2, the second bridge image dataset is divided into a training set and a test set.

[0024] As an implementation manner of the embodiment of the present invention, in step S3, the YOLO11 model is improved to obtain the ADSN-YOLO11 model, as Figure 2 For the YOLO11 model, the improvement method is to replace the convolutional block of the YOLO11 model with an Adown module for enhancing the ability to extract efflorescence features of concrete bridges and lightweight. The network structure of the Adown module is as Figure 4 , the feature map output from the C3k2 module first undergoes an AvgPool2d operation and then is Split into two parts, with the number of channels in each part halved. One part directly enters the first CBS module; the other part first undergoes a MaxPool2d operation and then enters the second CBS module. Finally, the outputs of the two parts are merged in a Concat operation to restore to the original number of channels. This structure combines average pooling and max pooling operations, enhances the diversity and robustness of feature extraction, realizes downsampling of the feature map, reduces the number of parameters and computational amount at the same time, and improves the detection accuracy of the model. The ADSN-YOLO11 model adopts an OD2SN neck structure. The OD2SN neck structure is based on the SlimNeck neural network neck, and replaces two convolutional blocks of the input channels in the VoVGSCSP module with ODConv blocks with stronger extraction capabilities to enhance the ability to extract efflorescence features. The VoVGSCSP network structure is as Figure 5 , the feature map obtained through Concat fusion enters the first convolutional layer for feature extraction, and the output feature map is input into the GS bottleneck module for further feature processing. At the same time, the fused feature map also enters the second convolutional layer for feature extraction through another path. The two feature maps output after being processed by the GS bottleneck module and the second convolutional layer are input into a Concat operation for feature splicing. The spliced feature map finally undergoes a third convolutional layer for feature fusion to obtain the final output feature map. The optimized VoVGSCSP network structure is as Figure 6, the feature map obtained through Concat feature fusion enters the first ODConv layer for feature extraction, and the output feature map is input into the GS bottleneck module for further feature processing. At the same time, the fused feature map also enters the second ODConv layer for feature extraction through another path. The two output feature maps after being processed by the GS bottleneck module and the second ODConv layer are input into the Concat operation for feature concatenation. The concatenated feature map finally undergoes feature fusion through the Conv layer to obtain the final output feature map. This structure enhances the diversity of feature extraction and the expressive power of the network through parallel ODConv paths and the GS bottleneck module. The Concat operation effectively fuses the feature information of different paths, which helps to improve the performance and accuracy of the model when processing complex images. Combining the two improvements, the ADSN-YOLO11 model is obtained, and its network model structure is as Figure 3 , the image to be detected enters the Backbone and is sequentially processed by multiple layers of Conv, C3k2, ADown, etc., and finally reaches the C2PSA and is input into the neck. In the OD2SN neck structure, features at different levels are fused and processed: some features are directly output at the 4th layer of the Backbone and fused with other processed features at the 15th layer Concat layer, and after being processed by VoVGSCSP, they are input into Head1 of the Detection head structure; the features output by C2PSA are upsampled through the 11th layer Upsample and fused with the features from the 6th layer of the Backbone at the 12th layer Concat, and after being processed by the 13th layer VoVGSCSP, they are fused with other features at the 18th layer Concat, and are connected to Head2 by the 19th layer VoVGSCSP. Another branch passes through a GSConv and is fused with another feature output by the Backbone at the 21st layer Concat, and is then connected to Head3 by the 22nd layer VoVGSCSP. This structure realizes multi-scale object detection through multiple feature fusions, upsamplings, and the processing of specific modules GSConv and VoVGSCSP. Each component works closely together to improve the detection accuracy and efficiency, ensuring the effective recognition and localization of objects of different sizes.

[0025] Furthermore, in the training of the ADSN-YOLO11 model, the data path is set to point to a configuration file containing bridge-related category information for specifying the training and validation datasets. During training, the size of the bridge images in the training set is uniformly 640×640 pixels, and a total of 3000 rounds of iteration are performed. It is set that if the accuracy does not improve within 100 rounds, the training will be automatically terminated, and 32 images are processed per round. The optimizer uses Stochastic Gradient Descent (SGD).

[0026] As an implementation manner of the embodiment of the present invention, in step S4, structured pruning is performed on the trained ADSN-YOLO11 model, aiming to reduce the complexity and the number of parameters of the model by removing unimportant weights or channels in the model, thereby improving the inference efficiency and running speed of the trained ADSN-YOLO11 model. The specific process is as follows.

[0027] First, load the trained ADSN-YOLO11 model, and determine the pruning threshold through the `get_threshold` method of the `PRUNE` class. This threshold is based on the set pruning rate. After experiments, it is set to be most suitable at 0.6. The pruning rate is used to determine which weights or channels can be removed. Then, perform pruning operations on specific modules in the trained ADSN-YOLO11 model. The pruning operations include: pruning the Bottleneck in the C3k2 module, pruning Adown, pruning the GSConv in the OD2SN neck structure, pruning the detection head, and then resetting the gradients of the model parameters for precision fine-tuning training to obtain the ADSN-YOLO11_Pruned model. Among them, for the pruning of the detection head, the pruned parts include the Reg branches and Cls branches of the P3, P4, and P5 layers. As Figure 7 shown, P3, P4, and P5 correspond to the feature maps of different levels of the three detection heads. Reg is the regression prediction output, and Cls is the classification prediction output. The convolutional modules of these two branches are pruned layer by layer, removing redundant channels with little impact on the results, and adjusting the input channels of the subsequent modules to ensure the coherence of the network structure. The whole process not only reduces the computational burden of the model, but also, through a fine pruning strategy, retains the key features and performance of the model as much as possible.

[0028] As an implementation manner of an embodiment of the present invention, in step S5, the ADSN-YOLO11_Pruned model is subjected to two model conversions. The ADSN-YOLO11_Pruned model is converted into an ONNX model through the first model conversion; the ONNX model is converted into an RKNN model through the second model conversion; wherein, during the second model conversion, the ONNX model is transplanted to the Linux system, the model is quantized to convert floating-point storage into integer storage, and then the model is encapsulated into an RKNN model suitable for deployment on edge devices.

[0029] As an implementation manner of an embodiment of the present invention, in step S6, in order to implement the analysis of real-time captured concrete bridge images, multi-threading work is enabled. As Figure 8 shown, thread A is a thread for reading video frames; thread pool B is enabled to set multiple threads to load, infer, and post-process the ADSN-YOLO11_Pruned model after model format conversion. In order to avoid the performance overhead caused by frequent creation and destruction of threads, a certain number of threads are pre-created and placed in a pool for unified management. In this way, tasks can be scheduled efficiently, ensuring that each thread can quickly respond to new tasks when idle, improving the inference speed and response ability of the model when processing continuous video frames; finally, another thread C is enabled to push the inferred video frames and save the results to the corresponding files.

[0030] Embodiment 2: The embodiment of the present invention further provides a concrete bridge efflorescence defect detection device, including: A first processing module for preprocessing the acquired concrete bridge efflorescence image; A second processing module for dividing the preprocessed concrete bridge efflorescence image into a training set and a test set; A third processing module for training the ADSN-YOLO11 model according to the training set; wherein, in the ADSN-YOLO11 model, the convolutional block of the YOLO11 model is replaced by an Adown module; the ADSN-YOLO11 model adopts an OD2SN neck structure, and the OD2SN neck structure is based on the SlimNeck neural network neck, and two convolutional blocks of the input channels in the VoVGSCSP module are replaced by ODConv modules; A fourth processing module for performing structured pruning on the trained ADSN-YOLO11 model, and then performing fine-tuning training on the accuracy of the structured-pruned model to obtain the ADSN-YOLO11_Pruned model; A fifth processing module for performing model format conversion on the ADSN-YOLO11_Pruned model; The sixth processing module is used to deploy the ADSN-YOLO11_Pruned model after model format conversion to the RKNN hardware platform, start a thread pool to perform multi-threaded real-time detection on the test set, and when concrete bridge efflorescence is detected, mark the defect location and display the detection confidence level.

[0031] As an implementation manner of an embodiment of the present invention, the fourth processing module performs pruning operations on specific modules in the trained ADSN-YOLO11 model, including: pruning Bottlenecks in the C3k2 module, pruning Adown, pruning GSConv in the OD2SN neck structure, specifying the convolutional kernels between different modules for pruning, and pruning the detection head; among them, for the pruning of the detection head, the pruned parts include the Reg branches and Cls branches of the P3, P4, and P5 layers.

[0032] As an implementation manner of an embodiment of the present invention, the fifth processing module converts the ADSN-YOLO11_Pruned model into an ONNX model through the first model conversion; and converts the ONNX model into an RKNN model through the second model conversion.

[0033] Embodiment 3: The present invention also provides a concrete bridge efflorescence defect detection system, including: a memory and a processor, where a computer program run by the processor is stored on the memory, and the computer program, when run by the processor, executes the concrete bridge efflorescence defect detection method.

[0034] Embodiment 4: The present invention also provides a storage medium, where a computer program is stored on the storage medium, and the computer program, when running, executes the concrete bridge efflorescence defect detection method.

[0035] The above-described embodiments are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for detecting the efflorescence defect of a concrete bridge, characterized in that, Including: Step S1: Preprocess the obtained concrete bridge efflorescence image; Step S2: Divide the preprocessed concrete bridge efflorescence image into a training set and a test set; Step S3: Train the ADSN-YOLO11 model according to the training set; Among them, in the ADSN-YOLO11 model, replace the convolutional block of the YOLO11 model with the Adown module; The ADSN-YOLO11 model adopts the OD2SN neck structure, and the OD2SN neck structure is based on the SlimNeck neural network neck, and replaces the two convolutional blocks of the input channels in the VoVGSCSP module with the ODConv module; Step S4: Perform structured pruning on the trained ADSN-YOLO11 model, and then perform fine-tuning training on the accuracy of the structured pruned model to obtain the ADSN-YOLO11_Pruned model; Step S5: Convert the model format of the ADSN-YOLO11_Pruned model; Step S6: Deploy the ADSN-YOLO11_Pruned model with the converted model format to the RKNN hardware platform, start a thread pool to perform multi-threaded real-time detection on the test set, and when concrete bridge efflorescence is detected, mark the defect location and display the detection confidence.

2. The concrete bridge efflorescence defect detection method according to claim 1, characterized in that In step S4, the pruning operation on specific modules in the trained ADSN-YOLO11 model includes: pruning the Bottleneck in the C3k2 module, pruning Adown, pruning the GSConv in the OD2SN neck structure, and pruning the detection head; Among them, for the pruning of the detection head, the pruned part includes the Reg branch and the Cls branch of the P3, P4, and P5 layers.

3. The method for detecting the efflorescence defect of a concrete bridge according to claim 2, characterized in that, In step S5, the ADSN-YOLO11_Pruned model is converted into an ONNX model through the first model conversion; The ONNX model is converted into an RKNN model through the second model conversion; Among them, in the second model conversion, the ONNX model is transplanted to the Linux system, the model is quantized to convert the floating-point storage into integer storage, and then the model is encapsulated into an RKNN model suitable for deployment on edge devices.

4. A device for detecting the efflorescence defect of a concrete bridge, characterized in that, Including: The first processing module is used to preprocess the obtained concrete bridge efflorescence image; The second processing module is used to divide the preprocessed concrete bridge efflorescence image into a training set and a test set; The third processing module is used to train the ADSN-YOLO11 model according to the training set; Among them, in the ADSN-YOLO11 model, replace the convolutional block of the YOLO11 model with the Adown module; The ADSN-YOLO11 model adopts the OD2SN neck structure, and the OD2SN neck structure is based on the SlimNeck neural network neck, and replaces the two convolutional blocks of the input channels in the VoVGSCSP module with the ODConv module; The fourth processing module is used to perform structured pruning on the trained ADSN-YOLO11 model, and then perform fine-tuning training on the accuracy of the model after structured pruning to obtain the ADSN-YOLO11_Pruned model; The fifth processing module is used to perform model format conversion on the ADSN-YOLO11_Pruned model; The sixth processing module is used to deploy the ADSN-YOLO11_Pruned model after model format conversion to the RKNN hardware platform, start a thread pool to perform multi-threaded real-time detection on the test set, and when alkali efflorescence of the concrete bridge is detected, mark the defect location and display the detection confidence.

5. The concrete bridge efflorescence defect detection device according to claim 4, wherein, The pruning operations performed by the fourth processing module on specific modules in the trained ADSN-YOLO11 model include: pruning the Bottleneck in the C3k2 module, pruning Adown, pruning the GSConv in the OD2SN neck structure, and pruning the detection head; among them, for the pruning of the detection head, the pruned part includes the Reg branch and the Cls branch of the P3, P4, and P5 layers.

6. The concrete bridge efflorescence defect detection device according to claim 5, characterized in that, The fifth processing module converts the ADSN-YOLO11_Pruned model into an ONNX model through the first model conversion; and converts the ONNX model into an RKNN model through the second model conversion.

7. A detection system for the efflorescence defect of a concrete bridge, characterized in that, Including: A memory and a processor, where a computer program run by the processor is stored on the memory, and the computer program, when run by the processor, executes the concrete bridge alkali efflorescence defect detection method according to any one of claims 1-3.

8. A storage medium, characterized in that, A computer program is stored on the storage medium, and the computer program, when running, executes the concrete bridge alkali efflorescence defect detection method according to any one of claims 1-3.

Images