Tunnel crack identification method and system based on laser radar and unmanned aerial vehicle photographing, electronic equipment and storage medium

Abstract

The invention discloses a tunnel crack identification method and system based on laser radar and unmanned aerial vehicle photographing, an electronic device and a storage medium, and the method comprises the steps: obtaining space geometric information, photographing the surface image information of the interior of a tunnel, and constructing a three-dimensional texture point cloud model; carrying out feature extraction on geometric morphology features and surface texture features of the three-dimensional texture point cloud model to obtain a preliminary fracture feature set; classifying and preliminarily screening tunnel internal texture features based on the preliminary fracture feature set to obtain a preliminarily screened fracture region set; expanding the fissure regions based on the primarily screened fissure regions to generate an expanded fissure region set; analyzing the point cloud density in the expanded fracture region set through a Gaussian mixture model, judging the depth and width features of the fracture, and obtaining a fine screening fracture feature set; and for the fine screening fracture feature set, adopting a probability optimization algorithm based on a Markov random field to carry out joint probability distribution modeling on geometric and textural features of the fractures so as to obtain a fracture identification result.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel crack detection, and particularly to a tunnel crack identification method, system, electronic device and storage medium based on lidar and UAV photography. Background Art

[0002] The safety and durability of tunnel engineering are crucial research directions in the infrastructure field, which are directly related to the smoothness of transportation and the safety of public life and property. During the service life of a tunnel, cracks, as a common disease, not only affect the structural stability but may also induce larger-scale damage. Therefore, the timely identification and screening of cracks have become an indispensable technical link to ensure the long-term operation of the tunnel. Currently, the methods based on traditional manual inspections or single-sensor detections have obvious deficiencies in practical applications. For example, manual inspections are inefficient and easily interfered by subjective factors, while single sensors such as optical cameras or laser devices have limited data dimensions and are difficult to comprehensively capture the geometric and texture features of cracks, resulting in limited recognition accuracy or a high false positive rate.

[0003] Despite the continuous progress of technical means in recent years, tunnel crack identification still faces significant challenges. The most core technical factors focus on the effective fusion of multi-source data and the accuracy of the screening mechanism. Since lidar is good at obtaining spatial geometric information and the high-definition images taken by UAVs are more suitable for providing surface texture details, how to efficiently match and fuse these two heterogeneous data in the spatio-temporal dimension has become a key problem in improving the recognition ability. In addition, the diversity and complexity of crack features make it difficult for a single-threshold screening strategy to adapt to different scenarios and easily lead to missed reports or false alarms. How to design a hierarchical progressive screening mechanism to balance efficiency and accuracy further exacerbates the difficulty of technical implementation. These unsolved technical factors directly limit the reliability and practicality of crack identification, and a systematic solution is urgently needed.

[0004] Therefore, how to design a multi-level screening mechanism based on the fusion of the spatial data of lidar and the texture information captured by UAVs, and through the collaborative work of preliminary screening, refined screening and manual confirmation, gradually optimize the accuracy and robustness of crack identification has become the key problem to be solved in this study. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a tunnel crack identification method based on lidar and UAV photography, and the method includes:

[0006] Obtain spatial geometric information by using lidar, capture surface image information inside the tunnel by using a UAV, and construct a three-dimensional texture point cloud model based on the spatial geometric information and the surface image information;

[0007] Extract the geometric shape features and surface texture features of the three-dimensional texture point cloud model to obtain a preliminary crack feature set;

[0008] Classify the internal texture features of the tunnel based on the preliminary crack feature set to obtain a classified crack feature subset, and use the threshold segmentation method to preliminarily screen the high-risk crack areas in the classified crack feature subset to obtain a preliminary screening crack area set;

[0009] Obtain the curvature and texture contrast of the boundary features from the preliminary screening crack area set, and expand the crack area to generate an expanded crack area set;

[0010] If the curvature change trend of the expanded crack area set matches the preset crack shape template, analyze the point cloud density in the expanded crack area set through the Gaussian mixture model, judge the depth and width features of the crack, and obtain a refined screening crack feature set;

[0011] For the refined screening crack feature set, adopt a probability optimization algorithm based on the Markov random field to perform joint probability distribution modeling on the geometric and texture features of the crack to obtain a crack recognition result.

[0012] Preferably, the method for constructing a three-dimensional texture point cloud model includes:

[0013] Use lidar to obtain point cloud data, and extract the spatial geometric features in the point cloud data through a stereoscopic geometric algorithm to obtain a three-dimensional point cloud coordinate set;

[0014] Use an unmanned aerial vehicle to capture two-dimensional texture data inside the tunnel, and separate the surface details through an image segmentation algorithm to obtain a texture feature set;

[0015] Use a preset coordinate system mapping algorithm to align the three-dimensional point cloud coordinate set and the texture feature set in the space-time dimension, and adjust the coordinate mapping through the iterative closest point algorithm to obtain an optimized matching data set;

[0016] Based on the optimized matching data set, fuse the aligned point cloud coordinate set and the texture feature set to generate a three-dimensional texture point cloud model.

[0017] Preferably, the method for obtaining the preliminary crack feature set includes:

[0018] Obtain crack geometric data based on the three-dimensional texture point cloud model to obtain an initial point cloud data set;

[0019] Use stereomicroscopy technology to extract crack geometry and surface texture from the initial point cloud data set to obtain a separated feature set;

[0020] Based on the coincidence degree of the crack combination and the morphological features judged from the separated feature set, determine the geometric morphological features;

[0021] Separate the surface texture from the geometric morphological features to obtain the surface texture features;

[0022] Construct the preliminary crack feature set based on the geometric morphological features and the surface texture features.

[0023] Preferably, the method for obtaining the preliminary screening crack region set includes:

[0024] If the number of geometric morphological feature points in the preliminary crack feature set exceeds the first preset threshold, use a deep convolutional neural network to classify the internal texture features of the tunnel, judge the potential risk level of the crack, and obtain the classified crack feature subset;

[0025] Use an adaptive threshold segmentation algorithm to process the crack features in the classified crack feature subset, obtain the preliminary screening result set of the high-risk area, and extract the boundary feature description set in the preliminary screening result set;

[0026] If the feature value in the boundary feature description set exceeds the second preset threshold, optimize the adaptive threshold through threshold adjustment technology to obtain the preliminary screening crack region set.

[0027] Preferably, the method for generating the extended crack region set includes:

[0028] Obtain the curvature and texture contrast of the boundary features from the preliminary screening crack region set, adopt an edge detection algorithm to extract the curvature features and texture features, and obtain the initial feature set;

[0029] Based on the initial feature set, analyze the boundary curvature through a region growing algorithm to determine the expansion direction of the crack region. If the boundary curvature exceeds the third preset threshold, expand the crack region along the curvature features to obtain a preliminary expansion region;

[0030] Judge the edges of the preliminary expansion region according to the texture contrast, filter non-crack regions using the texture features, and optimize the non-crack regions through a growing algorithm to obtain the connectivity of crack expansion and generate a connected expansion set;

[0031] Extract the boundary features of the expansion region from the connected expansion set to obtain the extended crack region set.

[0032] Preferably, the method for obtaining the refined screening crack feature set includes:

[0033] If the curvature change trend of the expanded crack area set matches the preset crack form template, significant points of curvature change within the area are extracted through curvature distribution calculation to obtain a curvature feature point set;

[0034] According to the curvature feature point set, the point cloud subset within the crack area is divided using point cloud segmentation technology to determine the point cloud density distribution, and clustering analysis is performed on the point cloud density distribution through a Gaussian mixture model to obtain preliminary features of the crack depth and width;

[0035] If the proportion of the high-density area of the preliminary features exceeds that of the low-density area, the point cloud data of the high-density area is processed through morphological filtering to obtain a smoothed point cloud subset, and the spatial distance between the crack boundary points is calculated according to the smoothed point cloud subset to determine the exact value of the crack width;

[0036] Based on the preliminary features and the exact value, a three-dimensional crack feature vector is constructed to obtain the refined screened crack feature set.

[0037] Preferably, the method for obtaining the crack recognition result includes:

[0038] Using the Markov random field algorithm, for the geometric features and texture features in the refined screened crack feature set, an initial joint probability distribution is constructed to obtain a preliminary probability model;

[0039] The preliminary probability model is adjusted using a stochastic optimization algorithm to obtain an optimized probability model, and the joint probability value of the crack features is calculated according to the optimized probability model to obtain a crack probability distribution;

[0040] Based on the crack probability distribution, a crack extraction result is generated, and a spatial clustering algorithm is used to group the crack areas in the crack extraction result to obtain the crack recognition result.

[0041] The present invention also provides a tunnel crack recognition system based on lidar and drone photography. The recognition system applies the recognition method described in any one of the above, and includes: a point cloud model construction module, a feature extraction module, a preliminary screening module, a crack expansion module, a refined screening module, and a recognition module;

[0042] The point cloud model construction module uses lidar to obtain spatial geometric information, uses a drone to capture surface image information inside the tunnel, and constructs a three-dimensional texture point cloud model based on the spatial geometric information and the surface image information;

[0043] The feature extraction module is used to extract the geometric shape features and surface texture features of the three-dimensional texture point cloud model to obtain a preliminary crack feature set;

[0044] The initial screening module classifies the internal texture features of the tunnel based on the preliminary crack feature set, obtains the classified crack feature subset, and uses the threshold segmentation method to preliminarily screen the high-risk crack areas in the classified crack feature subset to obtain the initial screening crack area set;

[0045] The crack expansion module is used to obtain the curvature and texture contrast of the boundary features from the initial screening crack area set, and expand the crack area to generate the expanded crack area set;

[0046] If the curvature change trend of the expanded crack area set matches the preset crack form template, the fine screening module analyzes the point cloud density in the expanded crack area set through the Gaussian mixture model, judges the depth and width features of the crack, and obtains the fine screening crack feature set;

[0047] The recognition module is used to, for the fine screening crack feature set, adopt a probability optimization algorithm based on the Markov random field to perform joint probability distribution modeling on the geometric and texture features of the crack to obtain the crack recognition result.

[0048] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the tunnel crack recognition method based on lidar and drone photography as described above is implemented.

[0049] The present invention also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed, the tunnel crack recognition method based on lidar and drone photography as described above is implemented.

[0050] Compared with the prior art, the beneficial effects of the present invention are:

[0051] The present invention generates a three-dimensional texture point cloud model by fusing point cloud data and two-dimensional image data, extracts crack features using stereomicroscopy technology, and classifies the texture features using a deep convolutional neural network to judge the crack risk level; subsequently, an adaptive threshold segmentation and region growing algorithm are used to preliminarily screen and expand the high-risk crack areas, and then the point cloud density is analyzed through the Gaussian mixture model to judge the depth and width features of the crack; finally, a probability optimization algorithm based on the Markov random field is used to model the crack features to generate the final recognition result. The present invention realizes the accurate recognition and risk assessment of cracks in complex environments, providing strong support for geological disaster early warning and engineering safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] To more clearly illustrate the technical solution of the present invention, the accompanying drawings required for use in the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0053] Figure 1 Schematic diagram of the method flow of the embodiment of the present invention;

[0054] Figure 2 Schematic diagram of the structure of the electronic device of the embodiment of the present invention.

[0055] Explanation of reference numerals:

[0056] 1010, processor; 1020, memory; 1030, input / output interface; 1040, communication interface; 1050, bus. Specific implementation manners

[0057] 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 belong to the scope of protection of the present invention.

[0058] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present disclosure should have the ordinary meaning understood by those of ordinary skill in the art in the field to which the present disclosure belongs. The "first", "second" and similar terms used in the embodiments of the present disclosure do not represent any order, quantity or importance, but are only used to distinguish different components. Words such as "including" or "comprising" mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0059] Embodiment 1

[0060] In this embodiment, as Figure 1 shown, a tunnel crack identification method based on lidar and UAV photography, the method includes:

[0061] S1. Obtain spatial geometric information using lidar, capture surface image information inside the tunnel using a drone, and construct a three-dimensional texture point cloud model based on the spatial geometric information and surface image information.

[0062] The method for constructing a three-dimensional texture point cloud model includes: obtaining point cloud data using lidar, extracting spatial geometric features from the point cloud data through a stereovision geometry algorithm to obtain a three-dimensional point cloud coordinate set; capturing two-dimensional texture data inside the tunnel using a drone, separating surface details through an image segmentation algorithm to obtain a texture feature set; using a preset coordinate system mapping algorithm to align the three-dimensional point cloud coordinate set and the texture feature set in terms of spatial and temporal dimensions, and adjusting the coordinate mapping through the iterative closest point algorithm to obtain an optimized matching data set; based on the optimized matching data set, fusing the aligned three-dimensional point cloud coordinate set and the texture feature set to generate a three-dimensional texture point cloud model.

[0063] In this embodiment, a robot equipped with lidar is used to scan inside the tunnel to capture distance information of the surrounding environment, generating point cloud data containing millions of points. These point cloud data are recorded in three-dimensional coordinate form, representing positions in space. Then, spatial geometric features are extracted through a stereovision geometry algorithm to obtain a three-dimensional point cloud coordinate set. Subsequently, a drone equipped with a high-definition camera is used to take images inside the tunnel with an image resolution of 4000×3000 pixels, and an image segmentation algorithm is employed to separate surface details and extract the texture of the tunnel wall to form a texture feature set. Then, the coordinate system of the lidar is unified with that of the drone camera, and using the preset coordinate system mapping algorithm, the three-dimensional point cloud coordinate set and the texture feature set are aligned in terms of spatial and temporal dimensions. The global coordinate origin is set as the starting point of the street, and the point cloud coordinates and the pixel points in the texture data are aligned through rotation and translation transformations to generate a preliminary matching data set. If there is an offset, for example, the corner position in the point cloud deviates from the edge in the texture by 2 cm, then the distance between the closest points in the two data sets is calculated by adjusting the coordinate mapping through the iterative closest point algorithm, gradually optimizing the coordinate mapping until the deviation is reduced to the millimeter level to obtain an optimized matching data set; based on the optimized matching data set, fusing the aligned three-dimensional point cloud coordinate set and the texture feature set to generate a three-dimensional texture point cloud model.

[0064] S2. Extract the geometric shape features and surface texture features of the three-dimensional texture point cloud model to obtain a preliminary crack feature set.

[0065] The method for obtaining the preliminary crack feature set includes: obtaining crack geometric data based on a three-dimensional texture point cloud model to obtain an initial point cloud data set; using stereomicroscopy technology to extract crack geometry and surface texture from the initial point cloud data set to obtain a separated feature set; judging the coincidence degree of crack combination and morphological features based on the separated feature set to determine geometric morphological features; separating surface texture from the geometric morphological features to obtain surface texture features; constructing a preliminary crack feature set based on the geometric morphological features and surface texture features.

[0066] In this embodiment, a point cloud model is generated through three-dimensional texture to obtain crack geometric data, and an initial point cloud data set is obtained. The initial point cloud data set is processed by stereomicroscopy technology to extract crack geometry and surface texture, and a separated feature set is obtained. For the separated feature set, the coincidence degree of crack geometry and morphological features is judged to determine a geometric morphological feature set. Surface texture is separated from the geometric morphological feature set to obtain a texture feature subset, and an independent texture description is obtained. If there is an intersection between the texture feature subset and the crack features, the intersection part is processed by a clustering algorithm to obtain a preliminary crack feature set.

[0067] S3. Classify the internal texture features of the tunnel based on the preliminary crack feature set to obtain a classified crack feature subset, and use the threshold segmentation method to preliminarily screen the high-risk crack areas in the classified crack feature subset to obtain a preliminary screening crack area set.

[0068] The method for obtaining the preliminary screening crack area set includes: if the number of geometric morphological feature points in the preliminary crack feature set exceeds the first preset threshold, use a deep convolutional neural network to classify the internal texture features of the tunnel, judge the potential risk level of the crack, and obtain a classified crack feature subset; use an adaptive threshold segmentation algorithm to process the crack features in the classified crack feature subset to obtain a preliminary screening result set of high-risk areas, and extract the boundary feature description set in the preliminary screening result set; if the feature values in the boundary feature description set exceed the second preset threshold, optimize the adaptive threshold through threshold adjustment technology to obtain a preliminary screening crack area set.

[0069] In this embodiment, if the number of geometric feature points in the initially screened fissure region set exceeds the first preset threshold, the distribution characteristics of the number of feature points are calculated through a statistical tool to obtain a preliminarily screened feature set. Texture feature data is obtained from the preliminarily screened feature set, and the texture features are classified through a deep convolutional neural network to obtain a classified texture feature set; according to the classified texture feature set, a clustering algorithm is used to group the fissure features to determine the distribution pattern of each group of fissures. If the distribution pattern of each group of fissures meets the preset conditions, the relationship between the texture features and the potential risks is analyzed through a logistic regression algorithm to judge the risk level of the fissures; high-risk fissure features are extracted from the judged risk levels to generate a classified fissure feature subset; an adaptive threshold segmentation algorithm is used to process the fissure features in the classified fissure feature subset to obtain a preliminary screening result set of high-risk regions; through a boundary feature extraction technique, regional boundary data is obtained from the preliminary screening result set to determine the initially screened region set; for the regional boundaries in the initially screened region set, a feature extraction method is used to obtain a boundary feature description set; if the feature values in the boundary feature description set exceed the preset threshold range, the adaptive threshold is optimized through a threshold adjustment technique to obtain an adjusted segmentation result set; according to the adjusted segmentation result set, an updated boundary of the high-risk region is obtained, the range of the risk region is judged, and the initially screened fissure region set is obtained.

[0070] S4. Obtain the curvature and texture contrast of the boundary features from the initially screened fissure region set, and expand the fissure region to generate an expanded fissure region set.

[0071] The method for generating the expanded fissure region set includes: obtaining the curvature and texture contrast of the boundary features from the initially screened fissure region set, using an edge detection algorithm to extract the curvature features and texture features to obtain an initial feature set; based on the initial feature set, analyzing the boundary curvature through a region growing algorithm to determine the expansion direction of the fissure region. If the boundary curvature exceeds the third preset threshold, the fissure region is expanded along the curvature features to obtain a preliminary expanded region; judging the edges of the preliminary expanded region according to the texture contrast, filtering non-fissure regions using texture features, and optimizing the non-fissure regions through a growing algorithm to obtain the connectivity of the fissure expansion, and generating a connected expansion set; extracting the boundary features of the expanded regions from the connected expansion set to obtain the expanded fissure region set.

[0072] In this embodiment, the boundary curvature can be characterized by detecting the degree of bending of the crack edge. For example, in a 50-cm-long crack, if the curvature radius of a certain section is less than 5 cm, it can be considered that the bending of this area is significant and may be a high-risk point. When using an edge detection algorithm to extract curvature features, the change rate of boundary pixels can be scanned using image processing techniques. The texture contrast is determined by comparing the gray-scale difference between the crack area and the surrounding environment. Suppose the average gray scale of the crack area is 80 and that of the surrounding area is 120. A higher contrast indicates that the crack is more easily recognizable. When using an edge detection algorithm to extract texture features, local window analysis can be used. When the region growing algorithm analyzes the boundary curvature, it will expand based on seed points. Suppose a point with a curvature radius less than 3 cm is selected as the seed. If the curvature continuity of the adjacent region remains above 80%, it will expand in this direction. If the boundary curvature exceeds the preset threshold, the crack area will be expanded along the curvature feature to obtain a preliminary expanded area. The edges of the preliminary expanded area are judged according to the texture contrast. The non-crack areas are filtered using texture features, and the non-crack areas are optimized through the growing algorithm to obtain the connectivity of crack expansion and generate a connected expansion set. The boundary features of the expanded area are extracted from the connected expansion set to obtain the expanded crack area set.

[0073] S5. If the curvature change trend of the expanded crack area set matches the preset crack morphology template, the point cloud density within the expanded crack area set is analyzed through a Gaussian mixture model to judge the depth and width characteristics of the crack, and a refined screened crack feature set is obtained.

[0074] The method for obtaining the refined screened crack feature set includes: if the curvature change trend of the expanded crack area set matches the preset crack morphology template, the significant points of curvature change within the area are extracted through curvature distribution calculation to obtain a curvature feature point set; according to the curvature feature point set, the point cloud subset within the crack area is divided using point cloud segmentation technology to determine the point cloud density distribution, and clustering analysis is performed on the point cloud density distribution through a Gaussian mixture model to obtain the preliminary characteristics of the crack depth and width; if the proportion of the high-density area of the preliminary characteristics exceeds that of the low-density area, the point cloud data of the high-density area is processed through morphological filtering to obtain a smoothed point cloud subset, and the spatial distance between the crack boundary points is calculated according to the smoothed point cloud subset to determine the exact value of the crack width; based on the preliminary characteristics and the exact value, a three-dimensional crack feature vector is constructed to obtain the refined screened crack feature set.

[0075] In this embodiment, when the curvature change trend in the extended fissure region set matches the preset fissure morphology template, significant points can be extracted by analyzing the curvature distribution: Assume that the curvature value of a certain fissure region shows an obvious mutation at the boundary, jumping from a gentle 0.2 to 0.8, which indicates that this may be a key point at the fissure edge. By recording the positions of these significant points, a curvature feature point set is formed. According to the curvature feature point set, the point cloud subset within the fissure region is divided using point cloud segmentation technology to determine the point cloud density distribution. Through clustering analysis of the point cloud density distribution using the Gaussian mixture model, high-density and low-density regions can be distinguished. Assume that after clustering the point cloud data of a certain fissure region, 60% of the points belong to the high-density region and 40% belong to the low-density region. If the proportion of the high-density region in the preliminary features exceeds that of the low-density region, the point cloud data of the high-density region is processed by morphological filtering to obtain a smoothed point cloud subset. Based on the smoothed point cloud subset, the spatial distance between the fissure boundary points is calculated to determine the exact value of the fissure width; Based on the preliminary features and the exact value, a three-dimensional fissure feature vector is constructed to obtain a refined screening fissure feature set.

[0076] S6. For the refined screening fissure feature set, a probability optimization algorithm based on the Markov random field is used to model the joint probability distribution of the geometric and texture features of the fissure, and the fissure recognition result is obtained.

[0077] The method for obtaining the fissure recognition result includes: using the Markov random field algorithm, constructing an initial joint probability for the geometric and texture features in the refined screening fissure feature set to obtain a preliminary probability model; using a random optimization algorithm to adjust the preliminary probability model to obtain an optimized probability model, and calculating the joint probability value of the fissure features according to the optimized probability model to obtain the fissure probability distribution; generating a fissure extraction result based on the fissure probability distribution, and using a spatial clustering algorithm to group the fissure regions in the fissure extraction result to obtain the fissure recognition result.

[0078] It should be noted that the method of this embodiment of the present disclosure can be executed by a single device, such as a computer or a server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In this distributed scenario, one of these multiple devices can only execute one or more steps of the method of this embodiment of the present disclosure, and these multiple devices will interact with each other to complete the described method.

[0079] It should be noted that some embodiments of the present disclosure have been described above. Other embodiments are within the scope of the appended claims. In some cases, it should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not imply the order of execution. The order of execution of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention. The actions or steps recited in the claims may be executed in a different order than in the above embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0080] Embodiment 2

[0081] In this embodiment, a tunnel crack identification system based on lidar and UAV photography includes: a point cloud model construction module, a feature extraction module, a preliminary screening module, a crack expansion module, a refined screening module, and an identification module.

[0082] The point cloud model construction module uses lidar to obtain spatial geometric information, uses a UAV to capture surface image information inside the tunnel, and constructs a three-dimensional texture point cloud model based on the spatial geometric information and surface image information.

[0083] The feature extraction module is used to extract the geometric shape features and surface texture features of the three-dimensional texture point cloud model to obtain a preliminary crack feature set.

[0084] The preliminary screening module classifies the internal texture features of the tunnel based on the preliminary crack feature set to obtain a classified crack feature subset, and uses the threshold segmentation method to preliminarily screen the high-risk crack areas in the classified crack feature subset to obtain a preliminary screening crack area set.

[0085] The crack expansion module is used to obtain the curvature and texture contrast of the boundary features from the preliminary screening crack area set, and expand the crack area to generate an expanded crack area set.

[0086] If the curvature change trend of the expanded crack area set matches the preset crack morphology template, the refined screening module analyzes the point cloud density within the expanded crack area set through a Gaussian mixture model to judge the depth and width features of the crack and obtain a refined screening crack feature set.

[0087] The identification module is used to, for the refined screening crack feature set, adopt a probability optimization algorithm based on a Markov random field to perform joint probability distribution modeling on the geometric and texture features of the crack to obtain a crack identification result.

[0088] The system of the above embodiments is used to implement the corresponding tunnel crack identification method based on lidar and UAV photography in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0089] It should be noted that the above tunnel crack identification system based on lidar and UAV photography is embodied in the form of functional units. The term "module" here can be implemented in software and / or hardware forms, and no specific limitation is made thereto.

[0090] For example, a "module" can be a software program, a hardware circuit, or a combination of both to implement the above functions. The hardware circuit may include an application specific integrated circuit (ASIC), an electronic circuit, a processor (such as a shared processor, a dedicated processor, or a group of processors, etc.) for executing one or more software or firmware programs, a memory, a merged logic circuit, and / or other suitable components to support the described functions.

[0091] Embodiment III

[0092] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the tunnel crack identification method based on lidar and UAV photography described in any of the above embodiments.

[0093] Figure 2 Fig. shows a more specific schematic diagram of the hardware structure of the electronic device provided in this embodiment. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. Among them, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are communicatively connected to each other inside the device through the bus 1050.

[0094] The processor 1010 can be implemented in a general-purpose CPU (Central Processing Unit), a microprocessor, an application specific integrated circuit (ASIC), or one or more integrated circuits, etc., and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0095] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage devices, dynamic storage devices, etc. The memory 1020 can store an operating system and other application programs. When implementing the technical solutions provided in the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 1020 and called and executed by the processor 1010.

[0096] The input / output interface 1030 is used to connect to the input / output module to achieve information input and output. The input / output module can be configured as a component in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Among them, the input device can include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device can include a display, a speaker, a vibrator, an indicator light, etc.

[0097] The communication interface 1040 is used to connect to a communication module (not shown in the figure) to achieve communication interaction between this device and other devices. Among them, the communication module can achieve communication through a wired method (such as USB (Universal Serial Bus), network cable, etc.) or through a wireless method (such as a mobile network, WIFI (Wireless Fidelity), Bluetooth, etc.).

[0098] The bus 1050 includes a path for transmitting information between various components of the device (such as the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0099] It should be noted that although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040, and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, those skilled in the art can understand that the above device may also only include the components necessary to implement the solution of the embodiments of this specification and does not necessarily include all the components shown in the figure.

[0100] The system of the above embodiment is used to implement the corresponding tunnel crack identification method based on lidar and UAV photography in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0101] Embodiment 4

[0102] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present invention also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the tunnel crack identification method based on lidar and drone photography as described in any of the above embodiments.

[0103] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible by a computing device.

[0104] The computer instructions stored in the storage medium of the above embodiment are used to cause the computer to execute the tunnel crack identification method based on lidar and drone photography as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0105] Those of ordinary skill in the art should understand that the discussion of any of the above embodiments is only exemplary and is not intended to imply that the scope of the present disclosure (including the claims) is limited to these examples; under the concept of the present disclosure, the technical features in the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations in different aspects of the embodiments of the present disclosure as described above, which are not provided in detail for the sake of brevity.

[0106] In addition, for simplicity of explanation and discussion, and so as not to make the embodiments of the present disclosure difficult to understand, well-known power / ground connections of integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Further, the devices may be shown in block diagram form so as not to make the embodiments of the present disclosure difficult to understand, and this also takes into account the fact that details of the implementation of these block diagram devices are highly dependent on the platform on which the embodiments of the present disclosure are to be implemented (i.e., these details should be entirely within the understanding of those skilled in the art). In cases where specific details (such as circuits) are set forth to describe exemplary embodiments of the present disclosure, it will be apparent to those skilled in the art that the embodiments of the present disclosure may be practiced without these specific details or with variations of these specific details. Accordingly, these descriptions should be regarded as illustrative rather than restrictive.

[0107] Although the present disclosure has been described in connection with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those of ordinary skill in the art based on the foregoing description. For example, other memory architectures (such as dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0108] Therefore, the units of the examples described in the embodiments of the present application can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present application.

[0109] The embodiments of the present disclosure are intended to cover all such alternatives, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent replacements, improvements, etc., made within the spirit and principles of the embodiments of the present disclosure shall be included within the protection scope of the present disclosure.

Claims

1. A tunnel crack identification method based on lidar and drone photography, characterized in that, The method includes: Obtaining spatial geometric information using a lidar, taking surface image information of the interior of the tunnel using a drone, and constructing a three-dimensional texture point cloud model based on the spatial geometric information and the surface image information; Extracting the geometric shape features and surface texture features of the three-dimensional texture point cloud model to obtain a preliminary crack feature set; Classifying the interior texture features of the tunnel based on the preliminary crack feature set to obtain a classified crack feature subset, and using a threshold segmentation method to preliminarily screen the high-risk crack areas in the classified crack feature subset to obtain a preliminary screening crack area set; Obtaining the curvature and texture contrast of the boundary features from the preliminary screening crack area set, and expanding the crack area to generate an expanded crack area set; If the curvature change trend of the expanded crack area set matches a preset crack shape template, then analyze the point cloud density within the expanded crack area set through a Gaussian mixture model to judge the depth and width features of the crack, and obtain a refined screening crack feature set; For the refined screening crack feature set, adopt a probability optimization algorithm based on a Markov random field to perform joint probability distribution modeling on the geometric and texture features of the crack to obtain a crack recognition result.

2. The tunnel crack identification method based on lidar and drone photography according to claim 1, wherein The method for constructing a three-dimensional texture point cloud model includes: Obtaining point cloud data using a lidar, extracting spatial geometric features from the point cloud data through a stereoscopic geometry algorithm to obtain a three-dimensional point cloud coordinate set; Taking two-dimensional texture data of the interior of the tunnel using a drone, separating surface details through an image segmentation algorithm to obtain a texture feature set; Using a preset coordinate system mapping algorithm to align the three-dimensional point cloud coordinate set and the texture feature set in the spatial and temporal dimensions, and adjusting the coordinate mapping through an iterative closest point algorithm to obtain an optimized matching data set; Based on the optimized matching data set, fuse the aligned point cloud coordinate set and the texture feature set to generate a three-dimensional texture point cloud model.

3. The tunnel crack identification method based on lidar and UAV photography according to claim 1, wherein The method for obtaining the preliminary crack feature set includes: Obtaining crack geometric data based on the three-dimensional texture point cloud model to obtain an initial point cloud data set; Using stereomicroscopy technology to extract crack geometry and surface texture from the initial point cloud data set to obtain a separated feature set; Judging the coincidence degree of the crack combination and morphological features based on the separated feature set to determine the geometric shape features; Separating the surface texture from the geometric shape features to obtain the surface texture features; Constructing the preliminary crack feature set based on the geometric shape features and the surface texture features.

4. The tunnel crack identification method based on lidar and UAV photography according to claim 1, characterized in that, The method for obtaining the preliminary screening crack area set includes: If the number of geometric shape feature points in the preliminary crack feature set exceeds a first preset threshold, then use a deep convolutional neural network to classify the interior texture features of the tunnel, judge the potential risk level of the crack, and obtain the classified crack feature subset; Using an adaptive threshold segmentation algorithm to process the crack features in the classified crack feature subset to obtain a preliminary screening result set for high-risk areas, and extracting the boundary feature description set in the preliminary screening result set; If the eigenvalue in the boundary feature description set exceeds the second preset threshold, the adaptive threshold is optimized by threshold adjustment technology to obtain the initial screened fissure area set.

5. The tunnel crack identification method based on lidar and UAV photography according to claim 1, wherein The method for generating the expanded fissure area set includes: Obtain the curvature and texture contrast of the boundary feature from the initial screened fissure area set, and use an edge detection algorithm to extract the curvature feature and texture feature to obtain an initial feature set; Based on the initial feature set, analyze the boundary curvature through a region growing algorithm to determine the expansion direction of the fissure area. If the boundary curvature exceeds the third preset threshold, expand the fissure area along the curvature feature to obtain a preliminary expansion area; Judge the edges of the preliminary expansion area according to the texture contrast, filter non-fissure areas using the texture feature, and optimize the non-fissure areas through a growing algorithm to obtain the connectivity of fissure expansion and generate a connected expansion set; Extract the boundary feature of the expansion area from the connected expansion set to obtain the expanded fissure area set.

6. The tunnel crack identification method based on lidar and UAV photography according to claim 1, characterized in that, The method for obtaining the refined screened fissure feature set includes: If the curvature change trend of the expanded fissure area set matches the preset fissure morphology template, calculate through curvature distribution to extract significant points of curvature change in the area to obtain a curvature feature point set; According to the curvature feature point set, use point cloud segmentation technology to divide the point cloud subset in the fissure area, determine the point cloud density distribution, and perform clustering analysis on the point cloud density distribution through a Gaussian mixture model to obtain preliminary features of the fissure depth and width; If the proportion of the high-density area of the preliminary feature exceeds that of the low-density area, perform morphological filtering on the point cloud data of the high-density area to obtain a smoothed point cloud subset, calculate the spatial distance of the fissure boundary points according to the smoothed point cloud subset, and determine the exact value of the fissure width; Based on the preliminary feature and the exact value, construct a three-dimensional fissure feature vector to obtain the refined screened fissure feature set.

7. The tunnel crack identification method based on lidar and UAV photography according to claim 1, characterized in that The method for obtaining the fissure recognition result includes: Adopt a Markov random field algorithm to construct an initial joint probability distribution for the geometric features and texture features in the refined screened fissure feature set to obtain a preliminary probability model; Use a stochastic optimization algorithm to adjust the preliminary probability model to obtain an optimized probability model, and calculate the joint probability value of the fissure feature according to the optimized probability model to obtain a fissure probability distribution; Generate a fissure extraction result based on the fissure probability distribution, and use a spatial clustering algorithm to group the fissure areas in the fissure extraction result to obtain the fissure recognition result.

8. Tunnel crack identification system based on lidar and drone photography, the identification system applying the identification method according to any one of claims 1-7, characterized in that, Including: A point cloud model construction module, a feature extraction module, an initial screening module, a fissure expansion module, a refined screening module, and an identification module; The point cloud model construction module uses lidar to obtain spatial geometric information, uses an unmanned aerial vehicle to capture surface image information inside the tunnel, and constructs a three-dimensional texture point cloud model based on the spatial geometric information and the surface image information; The feature extraction module is used to extract the geometric shape features and surface texture features of the three-dimensional texture point cloud model to obtain a preliminary fissure feature set; The initial screening module classifies the internal texture features of the tunnel based on the preliminary crack feature set, obtains a subset of classified crack features, and uses the threshold segmentation method to initially screen the high-risk crack areas in the subset of classified crack features to obtain an initial screening crack area set; The crack expansion module is used to obtain the curvature and texture contrast of the boundary features from the initial screening crack area set, and expand the crack area to generate an expanded crack area set; If the curvature change trend of the expanded crack area set matches the preset crack morphology template, the fine screening module analyzes the point cloud density in the expanded crack area set through a Gaussian mixture model to judge the depth and width features of the crack, and obtains a fine screening crack feature set; The recognition module is used to perform a joint probability distribution modeling on the geometric and texture features of the crack by using a probability optimization algorithm based on a Markov random field for the fine screening crack feature set, and obtain a crack recognition result.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the tunnel crack recognition method based on lidar and UAV photography according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed, it implements the tunnel crack recognition method based on lidar and UAV photography according to any one of claims 1 to 7.

Images