CN115359003A - Crack recognition method, system, medium and equipment for two-step tunnel grayscale image - Google Patents

Abstract

The invention belongs to the technical field of crack recognition of rock and soil structures, and provides a two-step crack recognition method, system, medium and equipment of tunnel grayscale images. Among them, the crack identification method of the two-step tunnel grayscale image includes performing edge extraction on batches of tunnel grayscale images to obtain the corresponding tunnel edge image, and then based on the comparison of the average grayscale of the edge image with the grayscale threshold, all possible cracks are screened out. There are suspected images of cracks; based on the suspected images and the fine segmentation network model, the crack segmentation images of the batch tunnel grayscale images are obtained, and then the crack identification results are obtained by statistics of the crack segmentation images; the crack identification results include images of actual cracks Number, number of cracks and characteristic parameters of each crack.

Description

Crack recognition method, system, medium and equipment for two-step tunnel grayscale image

technical field

The invention belongs to the technical field of crack recognition of rock and soil structures, and in particular relates to a crack recognition method, system, medium and equipment of a two-step tunnel grayscale image.

Background technique

The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

Cracks are common diseases in tunnel structures and are key factors affecting the safety of tunnel structures. The traditional detection method relies on manual judgment and auxiliary tool judgment; the detection efficiency is low, the work intensity is high, and the professional knowledge of personnel is high, so it is necessary to rely on rapid detection methods. At present, the commonly used method is to use image processing technology to identify diseases .

Commonly used methods for crack identification are mainly divided into two categories: traditional image processing methods and crack identification methods based on machine learning. Traditional image processing methods include image segmentation and skeleton extension, crack width transformation algorithm, multi-scale neighborhood information to realize automatic detection of pixel-level cracks, and use 3D scene reconstruction to evaluate the state of the structure. The crack recognition methods based on machine learning include A spatially tuned robust multi-feature classifier, using random forests for automatic detection of road cracks, and deep convolutional neural networks for crack classification and region extraction segmentation. Traditional image processing methods are poor for crack images with complex backgrounds, fractures, and rough walls. The machine learning method has high requirements on the number and types of crack samples, and has high requirements on training samples.

The inventors found that for the tunnel apparent grayscale image data collected automatically in large batches, the amount of data is relatively large, while the cracks are relatively small, and the amount of data including cracks is also small. Can spend a lot of time processing invalid data. Therefore, the existing machine learning methods are not suitable for the identification of tunnel structural cracks in large-scale data scenarios.

Contents of the invention

In order to solve the technical problems in the above-mentioned background technology, the present invention provides a two-step tunnel grayscale image crack recognition method, system, medium and equipment, which is suitable for tunnel structural crack recognition in a large-scale data scenario, and can greatly Reduce the measurement of data and greatly improve the efficiency of identification and detection.

In order to achieve the above object, the present invention adopts the following technical solutions:

A first aspect of the present invention provides a method for identifying cracks in a two-step tunnel grayscale image, which includes:

Edge extraction is performed on batches of tunnel grayscale images to obtain the corresponding tunnel edge images, and then based on the comparison of the average grayscale of the edge images with the grayscale threshold, all suspected images that may have cracks are screened out;

Based on the suspected image and the fine segmentation network model, the crack segmentation image of the batch tunnel grayscale image is obtained, and then the crack identification result is obtained by counting the crack segmentation image; the crack identification result includes the number of images of actual cracks, the number of cracks and each Characteristic parameters of cracks.

A second aspect of the present invention provides a two-step crack recognition system for tunnel grayscale images, which includes:

A rough screening module, which is used to perform edge extraction on batches of tunnel grayscale images to obtain corresponding tunnel edge images, and then screen out all suspected images that may have cracks based on the comparison between the average grayscale of the edge image and the grayscale threshold;

A fine identification module, which is used to obtain crack segmentation images of batches of tunnel grayscale images based on the suspected image and the fine segmentation network model, and then obtain crack identification results by statistical crack segmentation images; the crack identification results include actual cracks Number of images, number of cracks and characteristic parameters of each crack.

A third aspect of the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps in the above-mentioned two-step method for identifying cracks in tunnel grayscale images are realized .

A fourth aspect of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and operable on the processor. When the processor executes the program, the above-mentioned two Steps in the crack identification method for gray-scale images of tunnels.

Compared with prior art, the beneficial effect of the present invention is:

The present invention has the characteristics of large difference in size ratio between cracks and images, and small crack size. Edge extraction is performed on batches of tunnel grayscale images to obtain corresponding tunnel edge images, and then based on the comparison of the average grayscale of the edge images with the grayscale threshold, Screen out all suspected images that may have cracks; then based on the suspected images and the fine segmentation network model, obtain the crack segmentation images of batches of tunnel grayscale images, and then obtain the crack recognition results by counting the crack segmentation images, which is suitable for large-scale data scenarios The identification of cracks in the tunnel structure greatly reduces the measurement of data and improves the accuracy and efficiency of identification and detection.

Advantages of additional aspects of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

Description of drawings

The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention, and the schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations to the present invention.

Fig. 1 is the schematic diagram of the crack identification of the two-step tunnel grayscale image of the embodiment of the present invention;

Fig. 2 is a flowchart of a method for identifying cracks in a two-step tunnel grayscale image according to an embodiment of the present invention;

3 is a schematic structural diagram of a finely segmented network model according to an embodiment of the present invention;

Fig. 4 is a schematic structural diagram of a two-step tunnel grayscale image crack identification system according to an embodiment of the present invention.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

It should be noted that the following detailed description is exemplary and intended to provide further explanation of the present invention. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

It should be noted that the terminology used here is only for describing specific embodiments, and is not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly dictates otherwise, the singular is intended to include the plural, and it should also be understood that when the terms "comprising" and/or "comprising" are used in this specification, they mean There are features, steps, operations, means, components and/or combinations thereof.

Embodiment one

As shown in Figures 1 and 2, this embodiment provides a two-step method for identifying cracks in tunnel grayscale images, which specifically includes the following steps:

S101: Perform edge extraction on batches of tunnel grayscale images to obtain corresponding tunnel edge images, and calculate an average grayscale for the data. Then, based on the comparison between the average gray level of the edge image and the gray level threshold, all suspected images that may have cracks are screened out.

Optionally, the gray threshold can generally be determined in two ways. One is the preset gray threshold, which can be obtained through simulation experiments and can be set to 0.001. Another method is to count the average gray level of all edge extraction results, and use the average gray level as the gray level threshold to compare with the edge extraction results of each image to determine whether it is a suspected crack image.

In step S101, the acquisition process of the batch tunnel grayscale images is as follows:

Based on the gradient detection method, the upper threshold and lower threshold of the original tunnel grayscale image are calculated in batches, and the upper threshold and lower threshold of the original tunnel grayscale image are adjusted to the set grayscale range. In this way, the performance of the tunnel grayscale image is basically the same.

Use the Canny edge detection algorithm to perform edge extraction on batches of tunnel grayscale images.

In a specific implementation, the original grayscale image of the tunnel is acquired using the automatic acquisition equipment of the tunnel.

S102: Based on the suspected image and the fine segmentation network model, obtain the crack segmentation images of the batch gray image of the tunnel, and then obtain the crack identification result by counting the crack segmentation images; the crack identification result includes the number of images with actual cracks and the number of cracks and the characteristic parameters of each crack.

In a specific implementation, before the suspected image is input to the fine segmentation network model, it also includes:

All the suspected images are normalized.

The normalization processing method is as follows:

Scale the input image size to 640*640 pixels, and convert the image data into RGB color mode. Set the image mean mean to [0.485, 0.456, 0.406], set the image variance std to [0.229, 0.224, 0.225], and calculate the normalized image according to the formula P _img = (P _img -mean)/std, where P _img is the color value of each pixel in the image. Load image data into Dataloader for data training.

As shown in FIG. 3 , the finely segmented network model is a U-shaped convolutional neural network including a residual network module. The network model is divided into two parts, an encoding module and a decoding module. The encoding module includes a 4-layer downsampling module, and the decoding module includes a 4-layer upsampling module. The residual module is used to connect the corresponding layers of the encoding module and the decoding module.

For the tunnel crack image samples in the training data set and verification data set of the finely segmented network model, the crack area is marked as white, and the non-crack area is uniformly marked as black.

The training process of the fine segmentation network model is as follows:

Mark the extraction data of several existing (for example: 300) tunnel apparent grayscale images, mark the cracked area as white, and uniformly mark the non-cracked area as black;

Construct a training set and a verification set based on labeled data. The ratio of the training set to the verification set is preset, such as 0.85:0.15. Use this data set to train the U-shaped neural network for several rounds, such as 30 rounds. The learning rate can be set to 0.03 to complete the training of the network.

Among them, the fracture identification results include but are not limited to the number of fracture images, the proportion of fracture images in the collected images, the number of fractures, the length, width, shape and other characteristic parameters of each fracture, and the fine segmentation results of each fracture.

Embodiment two

As shown in Figure 4, the present embodiment provides a two-step tunnel grayscale image crack identification system, which specifically includes the following modules:

Rough screening module 201, which is used to perform edge extraction on batches of tunnel grayscale images to obtain corresponding tunnel edge images, and then screen out all suspected images that may have cracks based on the comparison between the average grayscale of the edge images and the grayscale threshold;

Fine identification module 202, which is used to obtain crack segmentation images of batches of tunnel grayscale images based on the suspected image and fine segmentation network model, and then obtain crack identification results by statistical crack segmentation images; the crack identification results include actual cracks The number of images, the number of cracks and the characteristic parameters of each crack.

It should be noted here that each module in this embodiment corresponds to each step in Embodiment 1, and the specific implementation process is the same, so it will not be repeated here.

Embodiment Three

This embodiment provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps in the two-step method for identifying cracks in tunnel grayscale images as described above are implemented.

Embodiment four

This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and operable on the processor. When the processor executes the program, the two-step tunnel described above is realized Steps in the crack identification method for grayscale images.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (10)

1. A crack recognition method of a two-step tunnel grayscale image, characterized in that it comprises: Edge extraction is performed on batches of tunnel grayscale images to obtain the corresponding tunnel edge images, and then based on the comparison of the average grayscale of the edge images with the grayscale threshold, all suspected images that may have cracks are screened out; Based on the suspected image and the fine segmentation network model, the crack segmentation image of the batch tunnel grayscale image is obtained, and then the crack identification result is obtained by counting the crack segmentation image; the crack identification result includes the number of images of actual cracks, the number of cracks and each Characteristic parameters of cracks. 2. the crack identification method of two-step tunnel grayscale image as claimed in claim 1, is characterized in that, the acquisition process of described batch tunnel grayscale image is: Based on the gradient detection method, the upper threshold and lower threshold of the original tunnel grayscale image are calculated in batches, and the upper threshold and lower threshold of the original tunnel grayscale image are adjusted to the set grayscale range. 3. the crack identification method of two-step tunnel gray-scale image as claimed in claim 1, is characterized in that, use Canny edge detection algorithm to carry out edge extraction respectively to batch tunnel gray-scale image. 4. the crack recognition method of two-step tunnel grayscale image as claimed in claim 1, is characterized in that, before described suspicious image is input to fine segmentation network model, also comprises: All the suspected images are normalized. 5. the crack recognition method of two-step tunnel grayscale image as claimed in claim 1, is characterized in that, described fine segmentation network model is the U-shaped convolutional neural network that comprises residual network module. 6. the crack recognition method of two-step tunnel grayscale image as claimed in claim 1, is characterized in that, for the training data set of finely segmented network model and the tunnel crack image sample in verification data set, the crack region is marked as white , and the non-crack regions are uniformly marked in black. 7. A crack recognition system of a two-step tunnel grayscale image, characterized in that it comprises: A rough screening module, which is used to perform edge extraction on batches of tunnel grayscale images to obtain corresponding tunnel edge images, and then screen out all suspected images that may have cracks based on the comparison between the average grayscale of the edge image and the grayscale threshold; A fine identification module, which is used to obtain crack segmentation images of batches of tunnel grayscale images based on the suspected image and the fine segmentation network model, and then obtain crack identification results by statistical crack segmentation images; the crack identification results include actual cracks Number of images, number of cracks and characteristic parameters of each crack. 8. the crack identification system of two-step tunnel grayscale image as claimed in claim 7, is characterized in that, in described rough screening module, the acquisition process of described batch tunnel grayscale image is: Based on the gradient detection method, the upper threshold and lower threshold of the original tunnel grayscale image are calculated in batches, and the upper threshold and lower threshold of the original tunnel grayscale image are adjusted to the set grayscale range. 9. A computer-readable storage medium on which a computer program is stored, wherein when the program is executed by a processor, the two-step tunnel grayscale image according to any one of claims 1-6 is realized. Steps in the fracture identification method. 10. A computer device, comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, characterized in that, when the processor executes the program, it realizes any of claims 1-6. A step in the two-step crack recognition method for tunnel grayscale images.

Images