Method and device for measuring joint fissure connectivity rate of tunnel surrounding rock tunnel face and storage medium

Abstract

The invention belongs to the technical field of geological engineering construction, and discloses a method and device for measuring the joint fissure connectivity rate of a tunnel surrounding rock tunnel face and a storage medium. The problems that in the prior art, a joint fissure recognition technology based on image processing is low in joint fissure recognition accuracy and analysis efficiency, the fissure connectivity rate cannot be accurately calculated, and automatic and standardized processes are lacked are solved. According to the scheme, firstly, a tunnel face image is segmented, and a tunnel face area image is extracted and preprocessed; dividing the preprocessed tunnel face area image, extracting fractures, and splicing the extracted fractures into a complete tunnel face joint fracture image according to a topological relation; then, carrying out vector measurement on fractures in the tunnel face joint fissure graph, grouping the fractures according to inclination angles through a clustering algorithm, and removing interference fractures; and finally, by setting a virtual measuring line and adopting a shortest path method, calculating the fracture connectivity rate on the strip where the virtual measuring line is located under the condition of different strip widths.

Description

Technical Field

[0001] The invention belongs to the technical field of geological engineering construction, and in particular relates to a method, a device and a storage medium for measuring the connectivity rate of joints and fissures in a tunnel surrounding rock face. Background Art

[0002] The connectivity of the joints and fissures in the tunnel surrounding rock face has a great impact on the stability of the surrounding rock, thus affecting the construction and operation. For example, connected fissures may lead to the fragmentation of the surrounding rock, increase the risk of landslides and collapses, and force the construction party to strengthen support measures. In addition, connected fissures will also cause groundwater infiltration and gushing, increasing the difficulty of construction and safety hazards. Therefore, in the construction design process, it is very necessary to identify and analyze the connectivity of the joints and fissures in the tunnel face.

[0003] In traditional technology, the identification and connectivity analysis of joints and fissures in tunnel faces mainly rely on manual observation and empirical judgment. Specifically, field engineers observe the face with the naked eye, draw a fissure map manually, record the number, length, spacing, inclination and other parameters of the fissures, and make a preliminary judgment on their connectivity based on experience. Subsequently, based on these data, combined with empirical formulas or numerical simulation tools, the impact of fissure connectivity on surrounding rock stability is evaluated. This method has the following main shortcomings: 1. High subjectivity and lack of standardization: Manual observation results rely heavily on the experience and subjective judgment of the observer. Different engineers may come up with different results, resulting in low reliability and consistency of the results.

[0004] 2. Limited accuracy: Manually drawn fracture maps usually cannot accurately reflect the detailed structure of fractures, especially when the fractures are dense or small in size, which is prone to omission or misjudgment. In addition, for complex joint systems, manual analysis is difficult to accurately capture the spatial relationship and topological structure between fractures, affecting the accuracy of connectivity analysis.

[0005] 3. Low efficiency: The manual observation and data recording process takes a long time, especially in large tunnel projects, where the site conditions are complex and the workload is large. The manual method is inefficient and difficult to meet the requirements of the project progress.

[0006] 4. Unable to conduct real-time dynamic analysis: Traditional methods are mostly post-analysis, which cannot timely reflect the dynamic changes of the joints and cracks in the face during the construction process, resulting in delayed construction decisions and increased engineering risks.

[0007] With the development of computer vision and image processing technology, especially the widespread application of deep learning models, automated analysis methods based on image information have gradually been introduced into tunnel engineering. This type of method can automatically identify and measure the geometric characteristics and spatial distribution of joints and fissures by processing and analyzing photos or image data of the tunnel face, and then calculate their connectivity. The use of image processing technology can effectively improve the accuracy and efficiency of crack identification and reduce the subjective errors caused by manual operation. In addition, the introduction of deep learning models enables the system to automatically adapt to various lighting, noise and color changes in complex environments, thereby improving the recognition rate of joints and fissures.

[0008] However, the existing joint and fissure identification technology based on image processing still has the following defects: 1. Limited by image quality: The tunnel construction environment is complex, and factors such as uneven light and noise interference will affect the image quality, thereby reducing the accuracy of crack identification.

[0009] 2. Failure to distinguish factors with weak effects on rock mass integrity: In existing joint and fissure extraction and analysis, joints and fissures are often not effectively distinguished from other factors with weak effects on rock mass integrity, such as bedding planes and excavation traces. Misidentification of these factors will interfere with rock mass stability analysis, resulting in reduced accuracy of analysis results and affecting the reliability of subsequent support design.

[0010] 3. The calculation of crack connectivity is complex: Although image processing technology can identify cracks, how to accurately calculate the connectivity of cracks based on identification remains a challenge.

[0011] 4. Lack of automated standardized processes: Although image processing technology can improve the automation level of crack identification, further standardization and automation are still needed in aspects such as crack vector measurement, connectivity calculation and result analysis so that they can be reused and promoted in different projects. Summary of the invention

[0012] The technical problem to be solved by the present invention is to propose a method, device and storage medium for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face, so as to solve the problems of low joint and fissure identification accuracy, low analysis efficiency, inability to accurately calculate the fissure connectivity rate and lack of automation and standardized processes in the joint and fissure identification technology based on image processing in the prior art.

[0013] The technical solution adopted by the present invention to solve the above technical problems is: In one aspect, the present invention provides a method for measuring the connectivity rate of joints and fissures in a tunnel surrounding rock face, comprising the following steps: S1, obtaining a tunnel face image of the tunnel surrounding rock, segmenting the tunnel face image, and extracting a tunnel face area image; S2, preprocessing the image of the palm face area; S3, dividing the preprocessed tunnel face area image and extracting cracks, and splicing the extracted cracks into a complete tunnel face joint and crack map according to the topological relationship; S4, performing vector measurement on the cracks in the face joint and crack map, grouping the cracks according to the inclination angles by using a clustering algorithm, and removing interfering cracks to obtain a processed face joint and crack map; S5. Based on the processed joint and fissure map of the tunnel face, a virtual survey line is set and the shortest path method is used to calculate the fracture connectivity on the strip where the virtual survey line is located under different strip width conditions. The strip width refers to the width of the area extending to both sides with the virtual survey line as the axis.

[0014] Furthermore, in step S1, the pre-trained Unet deep learning model is used to segment the tunnel face image and extract the tunnel face area image.

[0015] Furthermore, in step S2, the preprocessing method of the tunnel face area image includes: defogging, grayscale, smoothing denoising and histogram equalization processing.

[0016] Furthermore, in step S3, the method of dividing the preprocessed tunnel face region image and extracting cracks, and splicing the extracted cracks into a complete tunnel face joint and crack map according to the topological relationship includes: Cutting the preprocessed palm face area image into a plurality of sub-image units with overlapping areas; For each sub-image unit, the cracks in each sub-image unit are extracted by threshold segmentation, morphological processing and skeletonization processing respectively; According to the topological relationship, the cracks in each sub-image unit are spliced ​​into a complete joint and crack map of the tunnel face.

[0017] Furthermore, in step S4, the method of performing vector measurement on the cracks in the joint and crack map of the tunnel face includes: The Canny edge detection algorithm and chain code algorithm are used to extract and track the edge contours of the cracks in the face joint crack map; The edge contour of the extracted crack is simplified based on the Douglas-Peucker algorithm; Straight lines and polyline segments are used to represent edge contours, and the edge contours of cracks represented by chain codes are converted into vector form.

[0018] Furthermore, in step S4, the method of grouping the cracks according to the inclination angles by using a clustering algorithm and removing interfering cracks includes: Based on the inclination of the cracks, the DBSCAN algorithm is used to group the cracks. The inclination and spatial distribution information of the crack groups are displayed through visualization tools, and the interfering cracks that need to be removed are manually marked. Finally, the interfering cracks are eliminated based on the marked information.

[0019] Furthermore, in step S5, the method of setting the virtual survey line includes: The minimum circumscribed rectangle method is used to determine the center position of the processed joint and fissure map of the tunnel face. Based on this center, a virtual survey line is set every 10° between 0° and 180°. The end point of the virtual survey line is set at the edge of the image so that the survey line covers the entire image and forms a radial distribution from the center to the edge.

[0020] Furthermore, in step S5, the method of using the shortest path method to calculate the fracture connectivity rate on the strip where the virtual measuring line is located under different strip width conditions includes: The Dijkstra shortest path algorithm is used to calculate the shortest connected path on the strip where the virtual survey line is located under different strip width conditions. Among them, the connected path with the most connected cracks and the shortest path length is taken as the potential damage path. The connectivity rate of the cracks on the strip where each virtual measuring line is located under the selected strip width condition is calculated according to the following formula: : ; in, is the total length of cracks on the potential damage path, that is, the sum of the lengths of all cracks in the direction of the virtual survey line on the potential damage path; is the potential damage path length, that is, the shortest connected path length from the starting point to the end point along the virtual survey line.

[0021] In a second aspect, the present invention also provides a device for measuring the connectivity rate of joints and fissures in a tunnel surrounding rock face, which includes a processor and a memory, wherein a computer program is stored in the memory, and the processor implements the steps of the above-mentioned method for measuring the connectivity rate of joints and fissures in a tunnel surrounding rock face by executing the computer program.

[0022] In a third aspect, the present invention further provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face.

[0023] The beneficial effects of the present invention are: (1) Ability to exclude interference from non-face areas and focus on crack feature extraction: By segmenting and processing the face image, the face area can be accurately identified and interference factors from non-face areas such as supporting surrounding rock and mechanical equipment can be excluded. This processing method that focuses on the face ensures the accuracy of subsequent crack feature extraction, making crack identification more reliable and reducing errors caused by environmental complexity.

[0024] (2) Improving the accuracy and efficiency of crack identification: Through a series of image preprocessing techniques such as defogging, grayscale, smoothing and denoising, and histogram equalization, the visibility of cracks in the tunnel face image is effectively enhanced, and the influence of factors such as ambient light and noise is reduced. The tunnel face image is divided into several small units for processing by a cutting combination method, and then the cracks are extracted by threshold segmentation, morphological processing and skeletonization technology. Finally, the small unit cracks are spliced ​​into a complete tunnel face joint crack map based on the topological relationship. This processing flow not only improves the accuracy of crack identification, but also improves the processing efficiency, especially in complex construction environments. It can still maintain efficient and accurate crack analysis.

[0025] (3) Automated connectivity measurement and analysis: The extracted fracture maps are processed by vector measurement, and the fractures are automatically grouped according to their inclination angles through a clustering algorithm, thereby achieving effective classification of fractures with different inclination angles. Next, the shortest path method is used to calculate the fracture connectivity on the strip where the virtual survey line is located under different strip width conditions. This automated connectivity calculation method can accurately reflect the connectivity of the joints and fractures on the tunnel face, provide objective data support for the stability of the surrounding rock, and reduce the errors in previous manual calculations.

[0026] (4) Strong applicability and practicality: Through a fully automated and standardized process, the real-time dynamic identification and quantitative analysis of the joints and fissures on the tunnel face are realized. This real-time nature enables the construction site to obtain the surrounding rock stability assessment results in a timely manner so that rapid construction adjustment suggestions can be made based on the assessment results. This significantly improves the safety and efficiency of construction and ensures the smooth progress of the project. It is highly practical. In addition, this automated and standardized process can be reused and promoted in different projects, and is highly applicable. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 The present invention is a flow chart of a method for measuring the connectivity rate of joints and fissures in a tunnel surrounding rock face in an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The present invention aims to provide a method, device and storage medium for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face, so as to solve the problems of low joint and fissure identification accuracy, low analysis efficiency, inability to accurately calculate the fissure connectivity rate and lack of automation and standardized processes in the joint and fissure identification technology based on image processing in the prior art.

[0029] The core idea is: (1) For the tunnel face image, the pre-trained Unet deep learning model is used to segment the image, accurately extract the tunnel face area, and exclude interference factors such as supporting surrounding rock and mechanical equipment, ensuring that subsequent processing focuses on the tunnel face itself. Subsequently, a series of image preprocessing techniques are used to enhance the visibility of tunnel face cracks, and the image is extracted by cutting combination. On this basis, combined with threshold segmentation, morphological processing and skeletonization techniques, the crack features are accurately identified, and a complete joint and crack map is formed through topological splicing; this method of comprehensive use of deep learning and image processing technology significantly improves the accuracy and efficiency of tunnel face crack identification, ensuring efficient and accurate crack feature extraction in complex construction environments. (2) A clustering algorithm is used to automatically group the cracks according to their inclination angles to ensure that the cracks in the same group have similar inclination characteristics. After clustering is completed, the interactive tool is used to manually check and remove unnecessary cracks (such as bedding planes or other interfering cracks) to further optimize the accuracy and rationality of crack grouping. This process, which combines automated clustering with manual interaction optimization, not only improves the accuracy of fracture grouping, but also enhances the reliability of the analysis results, making it suitable for fine fracture analysis under complex geological conditions. (3) The extracted fracture maps are vectorized, and the fracture connectivity under different conditions is automatically calculated using the shortest path method. This automated process of fracture vectorization and connectivity calculation can significantly reduce manual intervention and errors, and improve the reliability of the evaluation results.

[0030] The scheme of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0031] Example

[0032] This embodiment provides a method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face. Figure 1 , including the following implementation steps: S1. Segmentation of the tunnel face image to extract the tunnel face area image: In this step, the tunnel face image of the tunnel surrounding rock is first obtained, and then the tunnel face image is segmented to extract the tunnel face area image. The tunnel face image of the tunnel surrounding rock is obtained by using a high-definition camera or other image acquisition equipment to shoot the tunnel face at the tunnel construction site, ensuring that the photo covers the entire tunnel face and has sufficient resolution to capture details. Low-light photography technology is used to improve the image quality in low-light environments, while optimizing the shooting position and angle, reducing distortion and uneven lighting, and ensuring that the image can accurately reflect the actual situation of the tunnel face.

[0033] Then, based on the pre-trained Unet deep learning model, the palm face area is automatically extracted. The Unet model is a convolutional neural network widely used in image segmentation. It is particularly suitable for the palm face area extraction task in this embodiment because of its high-resolution detail retention, precise contour boundary extraction, multi-scale feature processing, efficient training and reasoning, and good generalization ability. Specifically, the Unet model first downsamples the input palm face image to extract multi-level features, and then gradually restores the spatial information of the image through upsampling, and combines the feature maps at all levels to achieve accurate segmentation of the palm face area. This method not only effectively eliminates the interference of non-palm face areas, but also provides a reliable foundation for subsequent crack feature extraction.

[0034] S2. Preprocessing of the image of the face area: In this step, the tunnel face area image is preprocessed through a series of image preprocessing techniques such as defogging, grayscale, smoothing denoising and histogram equalization processing, aiming to enhance the visibility of the tunnel face cracks and reduce the interference of environmental factors such as light and noise on crack identification. The specific preprocessing methods are as follows: (1) Defogging treatment: This process can process low-contrast, blurred images caused by dust, fog or smoke in the tunnel based on the Dehazenet network. DehazeNet represents an effective deep learning method in the field of image dehazing. It uses the powerful feature extraction ability of convolutional neural networks to automatically learn and remove haze from images, thereby significantly improving the quality of images. Compared with traditional dehazing algorithms, it has the advantages of automation, high efficiency, and strong feature extraction capabilities, and its calculation principle relies on the feature learning and end-to-end training process of deep convolutional networks. In this way, DehazeNet can achieve better dehazing effects under complex haze conditions.

[0035] (2) Grayscale processing: This process converts the colored face area image into a grayscale image to reduce the interference of color information on crack identification and highlight the brightness and shape characteristics of the image. Common grayscale processing methods include the average method, weighted average method, maximum method, minimum method and single channel method. Among them, the weighted average method is selected as the grayscale processing method of this embodiment because it takes into account the sensitivity of the human eye to different colors and can more accurately reflect the brightness of the image. The specific expression is: ; in, , , are the light intensity values ​​of the red, green, and blue channels respectively.

[0036] (3) Smoothing and denoising processing: In this process, the image can be denoised using a bilateral filtering algorithm. Compared with traditional mean filtering, bilateral filtering takes into account both spatial distance and pixel value differences for weighted averaging, making noise smoothing more targeted and avoiding edge blurring while retaining key structural information in the image, such as crack edges, thereby removing noise while maintaining image clarity and detail integrity.

[0037] (4) Histogram equalization processing: This process performs histogram equalization on the image to enhance image contrast and balance brightness distribution, thereby improving the problem of difficulty in crack identification caused by uneven lighting.

[0038] S3. Extract the cracks and splice them into the joint and crack map of the tunnel face: In this step, the preprocessed tunnel face area image is divided and cracks are extracted, and the extracted cracks are spliced ​​into a complete tunnel face joint and crack map according to the topological relationship. The specific process is as follows: (1) Image cutting: This process cuts the preprocessed face area image into several sub-image units with overlapping areas, ensuring that the subsequent crack extraction algorithm can be processed in a smaller area, thereby improving the extraction accuracy. Considering the overlapping area during cutting helps reduce boundary effects and avoids the discontinuity problem that may cause cracks during splicing.

[0039] (2) Crack extraction: This process extracts the cracks in each sub-image unit obtained by image cutting by performing threshold segmentation, morphological processing and skeletonization processing. The process of crack extraction is as follows: ①Threshold segmentation: The process calculates the grayscale histogram of the image and applies Otsu's method (maximum inter-class variance method) to calculate the global threshold. The global threshold is used to binarize the image and initially separate the crack part from the background.

[0040] Otsu's method has significant advantages in face crack extraction. It automatically determines a global threshold by analyzing the grayscale histogram of the image, and divides the image into cracks (foreground part) and background part, which can effectively improve the accuracy and automation of crack extraction. The core is to calculate and maximize the inter-class variance by traversing all possible thresholds. , the formula is: ; in, and They are the proportion of foreground and background pixels in the entire image; and are the mean brightness of foreground and background respectively; is the mean brightness of the entire image.

[0041] ②Morphological processing: The process first uses an erosion operation to process the binarized image to remove small noise points; then, a dilation operation is performed to repair the broken parts of the cracks and fill the small holes in the cracks; further, a closing operation is applied to fill the small holes in the cracks and connect the broken parts to make the crack area more complete; finally, an opening operation is applied to remove pseudo-cracks and smooth the edges of the cracks to optimize the coherence and integrity of the cracks.

[0042] ③Skeletonization processing: This process skeletonizes the crack image after morphological processing, and can apply the Zhang-Suen algorithm to extract the crack centerline, so that the crack appears as a thin line and retains the topological structure of the original crack. The Zhang-Suen algorithm is a classic image skeletonization algorithm. The algorithm gradually removes non-edge pixels through repeated iterations, retains the connectivity and topological structure of the image, and extracts the crack as a skeleton line with a single pixel width. Its advantage is that it maintains the topological structure and connectivity of the crack, effectively avoiding crack breakage or accidental deletion.

[0043] (3) Crack splicing: The process is to splice the sub-image units after crack extraction through overlapping area fusion method to ensure the continuity of cracks. After splicing, the overall crack map is subjected to global morphological processing (closing and opening operations) to fill holes, repair fractures and remove pseudo cracks. Finally, through connected component analysis and local adjustment, the topological structure of the cracks is ensured to be complete, and a joint crack map of the tunnel face with satisfactory accuracy and integrity is generated.

[0044] S4, crack vector measurement and clustering, removal of interfering cracks: In this step, the cracks in the face joint and crack map are vector-measured, and the cracks are grouped according to the inclination angle through the clustering algorithm, and the interfering cracks are removed to obtain the processed face joint and crack map. The specific process is as follows: (1) Crack vector measurement: In this process, the Canny edge detection algorithm and the chain code algorithm can be used to extract and track the edge contours of the cracks in the face joint crack map. At the same time, in order to reduce the complexity of the crack contour, the contour can be simplified based on the Douglas-Peucker algorithm to reduce the number of vertices while maintaining the main geometric features; next, the crack contour is represented by straight lines and polylines, and the crack contour represented by the chain code is converted into a vector form. This conversion process converts the crack from a raster image to a continuous vector representation, which facilitates the accurate statistical analysis of the crack geometric features (such as length, width, inclination, etc.).

[0045] (2) Crack grouping and interference crack removal: This process is based on the inclination of the cracks. The cracks are grouped through a clustering algorithm, and the cracks with similar inclination angles are grouped into the same group. In terms of the selection of clustering algorithms, the DBSCAN algorithm can be used. The DBSCAN algorithm can adaptively identify and group cracks according to the spatial density of the cracks, and can effectively handle crack groups of different densities, avoiding the limitation of the preset number of groups. Secondly, the algorithm is highly robust to noise and isolated points, and can effectively remove interfering cracks and improve the accuracy of grouping.

[0046] After clustering is completed, the inclination and spatial distribution information of the fracture groupings are displayed through visualization tools, and the bedding planes or other interfering fractures that need to be removed are marked; then, unnecessary fractures are eliminated through the marked information to ensure that the final grouped data only contains valid fractures; finally, the grouped data is updated to ensure that subsequent connectivity analysis is based on accurate purified data.

[0047] S5. Calculate the fracture connectivity: In this step, based on the processed face joint fissure map, by setting a virtual survey line and using the shortest path method, the fracture connectivity rate on the strip where the virtual survey line is located under different strip width conditions is calculated. The strip here is the area extending to both sides with the virtual survey line as the axis, and the strip width is the width of the area. Different strip width conditions refer to the situation of multiple width values ​​selected according to the actual size of the face. The specific process of calculating the fracture connectivity rate is as follows: (1) Virtual measurement line settings: In this process, the minimum circumscribed rectangle method can be used to determine the center position of the processed face joint and fissure map. Taking this as the center point, a virtual survey line is set every 10° between 0° and 180°, and the end point of the virtual survey line is set at the edge of the image, so that the survey line covers the entire image and forms a radial distribution from the center to the edge.

[0048] (2) Calculation of fracture connectivity: The process can use Dijkstra's shortest path algorithm to calculate the shortest connected path on the strip where the virtual survey line is located under different strip width conditions, among which the connected path with the most cracks and the shortest path length is taken as the potential damage path; The connectivity rate of the cracks on the strip where each virtual measuring line is located under the selected strip width condition is calculated according to the following formula: : ; in, is the total length of cracks on the potential damage path, that is, the sum of the lengths of all cracks in the direction of the virtual survey line on the potential damage path; is the potential damage path length, that is, the shortest connected path length from the starting point to the end point along the virtual survey line.

[0049] Based on the above implementation process, the connectivity data of the cracks on the strips where each virtual measuring line is located under different strip width conditions can be obtained. Based on the statistics of these data, it can provide a reference for engineers to evaluate the stability of the surrounding rock and guide subsequent construction.

[0050] Finally, it should be noted that the above embodiments are only preferred implementations and are not intended to limit the present invention. It should be pointed out that for those skilled in the art, several modifications, equivalent replacements, improvements, etc. can be made without departing from the scope of the present invention and the scope of protection of the claims, and all of these should be included in the protection scope of the present invention.

Claims

1. A method for measuring the connectivity rate of joints and fissures in tunnel surrounding rock faces, characterized in that: The following steps are involved: S1, obtaining a tunnel face image of the tunnel surrounding rock, segmenting the tunnel face image, and extracting a tunnel face area image; S2, preprocessing the image of the palm face area; S3, dividing the preprocessed tunnel face area image and extracting cracks, and splicing the extracted cracks into a complete tunnel face joint and crack map according to the topological relationship; S4, performing vector measurement on the cracks in the face joint and crack map, grouping the cracks according to the inclination angles by using a clustering algorithm, and removing interfering cracks to obtain a processed face joint and crack map; S5. Based on the processed joint and fissure map of the tunnel face, a virtual survey line is set and the shortest path method is used to calculate the fracture connectivity on the strip where the virtual survey line is located under different strip width conditions. The strip width refers to the width of the area extending to both sides with the virtual survey line as the axis.

2. The method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face according to claim 1, characterized in that: In step S1, the pre-trained Unet deep learning model is used to segment the tunnel face image and extract the tunnel face area image.

3. The method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face according to claim 1, characterized in that: In step S2, the preprocessing method of the tunnel face area image includes: defogging, grayscale, smoothing denoising and histogram equalization processing.

4. The method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face according to claim 1, characterized in that: In step S3, the method of dividing the preprocessed tunnel face region image and extracting cracks, and splicing the extracted cracks into a complete tunnel face joint and crack map according to the topological relationship includes: Cutting the preprocessed palm face area image into a plurality of sub-image units with overlapping areas; For each sub-image unit, the cracks in each sub-image unit are extracted by threshold segmentation, morphological processing and skeletonization processing respectively; According to the topological relationship, the cracks in each sub-image unit are spliced ​​into a complete joint and crack map of the tunnel face.

5. The method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face according to claim 1, characterized in that: In step S4, the method of performing vector measurement on the cracks in the joint and crack map of the tunnel face includes: The Canny edge detection algorithm and chain code algorithm are used to extract and track the edge contours of the cracks in the face joint crack map; The edge contour of the extracted crack is simplified based on the Douglas-Peucker algorithm; Straight lines and polyline segments are used to represent edge contours, and the edge contours of cracks represented by chain codes are converted into vector form.

6. The method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face according to claim 5, characterized in that: In step S4, the method of grouping the cracks according to the inclination angles by using a clustering algorithm and removing interfering cracks includes: Based on the inclination of the cracks, the DBSCAN algorithm is used to group the cracks. The inclination and spatial distribution information of the crack groups are displayed through visualization tools, and the interfering cracks that need to be removed are manually marked. Finally, the interfering cracks are eliminated based on the marked information.

7. The method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face according to claim 1, characterized in that: In step S5, the method of setting the virtual measurement line includes: The minimum circumscribed rectangle method is used to determine the center position of the processed joint and fissure map of the tunnel face. A virtual survey line is set every 10° between 0° and 180° with this center point as the center point. The end point of the virtual survey line is set at the edge of the image so that the survey line covers the entire image and forms a radial distribution from the center to the edge.

8. The method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face according to claim 7, characterized in that: In step S5, the method of using the shortest path method to calculate the fracture connectivity rate on the strip where the virtual measuring line is located under different strip width conditions includes: The Dijkstra shortest path algorithm is used to calculate the shortest connected path on the strip where the virtual survey line is located under different strip width conditions. Among them, the connected path with the most connected cracks and the shortest path length is taken as the potential damage path. The connectivity rate of the cracks on the strip where each virtual measuring line is located under the selected strip width condition is calculated according to the following formula: : ； in, is the total length of cracks on the potential damage path, that is, the sum of the lengths of all cracks in the direction of the virtual survey line on the potential damage path; is the potential damage path length, that is, the shortest connected path length from the starting point to the end point along the virtual survey line.

9. A device for measuring the connectivity rate of joints and fissures in a tunnel surrounding rock face, comprising a processor and a memory, wherein a computer program is stored in the memory, and wherein: The processor executes the computer program to implement the steps of the method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face as described in any one of claims 1 to 8.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for measuring the connectivity rate of joints and fissures in the tunnel surrounding rock face as described in any one of claims 1 to 8 are implemented.

Images