Deformation area self-adaptive segmentation method for high-redundancy tunnel deformation point cloud
By combining three-dimensional Gaussian kernel voxel downsampling and empirical cumulative distribution function with seed point growth strategy, the adaptive segmentation problem of highly redundant tunnel deformation point cloud is solved, accurate segmentation of large deformation areas is achieved, and the accuracy and efficiency of tunnel deformation monitoring are improved.
Patent Information
- Application Number
- CN202510799706.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-16
- Publication Date
- 2025-10-17
AI Technical Summary
Traditional 3D laser scanning technology has difficulty effectively distinguishing discrete noise from real deformation when processing highly redundant tunnel deformation point clouds, resulting in misjudgment of the segmentation algorithm, lack of adaptability, and inability to meet the precise requirements of tunnel deformation monitoring.
A three-dimensional Gaussian kernel is used for voxel downsampling. Combined with the empirical cumulative distribution function and seed point growth strategy, adaptive segmentation is performed through deformation intensity guidance to suppress discrete noise interference and achieve accurate segmentation of large deformation areas.
While suppressing discrete noise interference, the geometric continuity of the segmented area is guaranteed, the accurate segmentation of the highly redundant tunnel deformation area is achieved, and the data depth mining capability of tunnel safety monitoring is improved.
Smart Images
Figure CN120807544A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application relates to the technical field of point cloud data processing, in particular to a deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds, a device, a storage medium and an electronic equipment. BACKGROUND
[0002] With the large-scale application of underground cavern projects in complex geological environments, their safety monitoring faces multiple technical challenges. Traditional monitoring methods are limited by single-point data acquisition mode and are difficult to meet the monitoring needs of the strong spatial heterogeneity of cavern groups, the complex interaction between surrounding rock and supporting structures and other engineering characteristics. Three-dimensional laser scanning technology can quickly obtain high-precision three-dimensional point cloud data of rock mass surfaces by emitting high-density laser beams, realizing the digital reconstruction of underground space forms. With a collection rate of millions of points per second, this technology can complete comprehensive scanning of large-section caverns within a few minutes, significantly improving the temporal and spatial resolution of deformation monitoring. Comprehensive tunnel deformation calculation results can be obtained through multi-period tunnel laser scanning data.
[0003] However, large-scale engineering volume and ultra-high-density laser scanning accuracy mean high-redundancy cavern point cloud calculation and analysis results. At the same time, the precise mixture of discrete noise and real deformation is disturbed by rock mass fractures, equipment errors, etc. A hard segmentation algorithm based on a fixed threshold is prone to misjudgment of discrete noise as large deformation points. The traditional threshold segmentation method has poor effect, parameter adjustment depends on experience and lacks adaptability. How to quickly identify the large deformation concentrated area from the high-redundancy tunnel deformation data is a key problem. SUMMARY
[0004] The embodiments of the application provide a deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds, a device, a storage medium and an electronic equipment, which can effectively suppress discrete noise interference while ensuring the geometric continuity of the segmented region, and realize accurate segmentation of the large deformation region.
[0005] The embodiments of the application provide a deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds, which comprises: Obtaining a tunnel point cloud file with deformation attributes, and constructing a tunnel point cloud original model with deformation attributes based on the tunnel point cloud file; Performing voxelization down-sampling on the tunnel point cloud original model based on a three-dimensional Gaussian kernel to obtain a tunnel deformation intensity field model; Performing empirical cumulative distribution function calculation on the data of the tunnel deformation intensity field model, and determining a deformation intensity segmentation threshold based on the empirical cumulative distribution function; Screening potential seed points in the tunnel point cloud original model with the deformation attribute based on the deformation intensity segmentation threshold, setting a seed point growth space neighborhood threshold, and iteratively growing the spatial topology of the seed points guided by the deformation intensity to obtain a tunnel concentrated deformation region segmentation result while suppressing discrete noise interference.
[0006] Further, the deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds described above, wherein after the step of obtaining a tunnel point cloud file with a deformation attribute and constructing a tunnel point cloud original model with a deformation attribute based on the tunnel point cloud file, the method comprises: Based on the data of the tunnel point cloud file, the axial bounding box extreme value coordinates of the tunnel point cloud model are calculated.
[0007] Further, the deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds described above, wherein the voxelization down-sampling of the tunnel point cloud original model based on a three-dimensional Gaussian kernel to obtain a tunnel deformation intensity field model comprises: Setting a voxelization grid size and calculating a voxel number and a voxel center coordinate set; Taking a non-empty voxel center point coordinate as a search point, querying all points within a specified neighborhood range of the search point in the tunnel point cloud original model as local neighborhood points, calculating the spatial distance weight of the local neighborhood points using a three-dimensional Gaussian kernel function, and performing weight normalization, and calculating a local deformation intensity value based on the data in the neighborhood; Binding the local deformation intensity value with the point cloud coordinates after voxel down-sampling to obtain a tunnel deformation intensity field model.
[0008] Further, the deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds described above, wherein the voxel number and the voxel center coordinate set are calculated by the following formula:
[0009]
[0010] wherein the voxel number is: , and the voxel center coordinate set is: , is the maximum value coordinate and the minimum value coordinate of the axial bounding box information, is the voxelization grid size.
[0011] Further, the deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds described above, wherein the local deformation intensity value is calculated by the following formula:
[0012]
[0013]
[0014] in, It's on point v n The local deformation strength value at is the midpoint of the original point cloud The deformation property value at is the bandwidth parameter, which is used to control the smoothness of the kernel. Is the neighborhood range m × h The number of points found in the query, m is the scalar coefficient of the spatial retrieval distance, To take the absolute value operation, is the distance calculation function, for point The spatial convergence coefficient at for point The spatial pooling normalized weight at , is the spatial aggregation coefficient of each point in the neighborhood.
[0015] Furthermore, in the above-mentioned deformation region adaptive segmentation method for highly redundant tunnel deformation point clouds, the step of performing empirical cumulative distribution function calculation on the data of the tunnel deformation intensity field model and determining the deformation intensity segmentation threshold based on the empirical cumulative distribution function includes: The empirical cumulative distribution function is calculated based on the local deformation intensity value:
[0016] in, is the indicator function, when 1 if yes, 0 otherwise. Accumulate distribution functions for experience; Setting a probability threshold based on the empirical cumulative distribution function and , based on the probability threshold and Calculate the deformation intensity segmentation threshold and .
[0017] Furthermore, the above-mentioned deformed region adaptive segmentation method for highly redundant tunnel deformed point clouds, wherein potential seed points are screened in the original model of the tunnel point cloud with deformation attributes based on the deformation intensity segmentation threshold, a threshold for the spatial neighborhood of the seed point growth is set, and the spatial topology of the seed points is guided by the deformation intensity to iteratively grow, thereby obtaining a segmentation result of the concentrated deformation region of the tunnel while suppressing discrete noise interference, including: Segmentation threshold based on deformation strength and , traversing the original model of the tunnel point cloud, dividing the points in the original model of the tunnel point cloud into seed points, pending points and non-clusterable points, screening and creating a candidate point set including the seed points and the pending points, and setting the status of all points in the candidate point set to an unvisited state; Determine the neighborhood threshold of the seed point growth space , set if the seed point lie in of In the neighborhood, it is called Depend on Connected directly, all spatially connected seed points and neighboring pending points are grouped into one cluster. Unconnected points do not belong to any cluster, and clusters with a number of seed points less than the low clustering screening threshold are considered noise clusters. Traverse the unvisited seed points, starting from any unvisited seed point, and search the unvisited seed points by breadth-first search strategy. All candidate points in the neighborhood are included in the current cluster, and unvisited seed points are added to the queue until the queue is empty; the iterative clustering operation is repeated. When all seed points in the candidate point set are set to the visited state, the region growing operation is stopped, and the adaptive deformation clustering is completed; The noise clusters whose seed points in the deformation clusters are less than the low cluster screening threshold are removed to obtain the segmentation results of the concentrated deformation area of the tunnel.
[0018] The embodiment of the present application further provides a deformation region adaptive segmentation device for a highly redundant tunnel deformation point cloud, comprising: a tunnel point cloud deformation acquisition module, configured to acquire a tunnel point cloud file with deformation attributes, and construct a tunnel point cloud original model with deformation attributes based on the tunnel point cloud file; A Gaussian kernel-based voxel downsampling module is used to perform voxel downsampling on the original tunnel point cloud model based on a three-dimensional Gaussian kernel to obtain a tunnel deformation intensity field model; a deformation strength segmentation threshold adaptive determination module, configured to perform empirical cumulative distribution function calculation on the data of the tunnel deformation strength field model and determine the deformation strength segmentation threshold based on the empirical cumulative distribution function; The concentrated deformation area identification module is used to screen potential seed points in the original tunnel point cloud model with deformation attributes based on the deformation intensity segmentation threshold, set the seed point growth space neighborhood threshold, guide the spatial topology iterative growth of the seed points through the deformation intensity, and obtain the tunnel concentrated deformation area segmentation result while suppressing discrete noise interference.
[0019] The embodiment of the present application also provides a computer readable storage medium, wherein a plurality of instructions are stored in the computer readable storage medium, and the instructions are suitable for being loaded by a processor to execute any one of the deformation region adaptive segmentation methods for high-redundancy tunnel deformation point clouds.
[0020] The embodiment of the present application also provides an electronic device, comprising a processor and a memory, wherein the processor is electrically connected with the memory, the memory is used for storing instructions and data, and the processor is used for the steps in any one of the deformation region adaptive segmentation methods for high-redundancy tunnel deformation point clouds.
[0021] The deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds, the device, the storage medium and the electronic device provided by the present application are used for performing deformation attribute voxelization down-sampling and empirical cumulative distribution function distribution fitting on a high-redundancy tunnel point cloud deformation calculation result, adaptively determining a segmentation threshold according to the distribution fitting situation, and then realizing accurate segmentation of a large deformation concentrated region based on a spatial connectivity feature, thereby providing strong data deep mining technical support for improving tunnel safety control, being capable of effectively suppressing discrete noise interference while guaranteeing the geometric continuity of a segmented region, and realizing accurate segmentation of a large deformation region. BRIEF DESCRIPTION OF DRAWINGS
[0022] The technical solutions and other beneficial effects of the present application will be apparent through the following detailed description of the specific embodiments of the present application in combination with the accompanying drawings.
[0023] Figure 1 A flowchart of the deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds provided by the embodiment of the present application.
[0024] Figure 2 A rendering diagram of a tunnel point cloud original model with deformation attributes provided by the embodiment of the present application.
[0025] Figure 3 A rendering diagram of a tunnel deformation intensity field model with deformation attributes provided by the embodiment of the present application.
[0026] Figure 4 A schematic diagram of an empirical cumulative distribution function curve and a tunnel deformation intensity frequency distribution provided by the embodiment of the present application.
[0027] Figure 5 A schematic diagram of dividing deformation clusters from seed points provided by the embodiment of the present application.
[0028] Figure 6 A schematic diagram of a tunnel concentrated deformation region segmentation result provided by the embodiment of the present application.
[0029] Figure 7A structural schematic diagram of a deformation region adaptive segmentation device for high-redundancy tunnel deformation point clouds provided by an embodiment of the present application.
[0030] Figure 8 A structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION
[0031] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person skilled in the art without creative work fall within the scope of protection of the present application.
[0032] The embodiments of the present application provide a deformation region adaptive segmentation method, device, storage medium and electronic device for high-redundancy tunnel deformation point clouds. The deformation region adaptive segmentation device for high-redundancy tunnel deformation point clouds provided by an embodiment of the present application can be integrated in an electronic device, which can be a terminal, a server or the like, wherein the terminal can include a tablet computer, a notebook computer, a personal computer (PC), a micro processing box or other devices, etc.
[0033] Please refer to Figure 1 , Figure 1 A flowchart of a deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds provided by an embodiment of the present application is applied in an electronic device, and the deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds includes the following steps: S1, a tunnel point cloud file with deformation attributes is acquired, and a tunnel point cloud original model with deformation attributes is constructed based on the tunnel point cloud file.
[0034] Specifically, the tunnel point cloud original model is , wherein represents spatial coordinate information of a tunnel geometric shape, is a deformation value bound with a spatial point, is the number of points contained in the tunnel point cloud. Figure 2 A rendering diagram of a tunnel point cloud original model with deformation attributes provided by an embodiment of the present application.
[0035] Further, after step S1, the following steps are further included: Based on the data of the tunnel point cloud file, axial bounding box extreme coordinate of the tunnel point cloud model is calculated .
[0036] Specifically, first, the axial bounding box extreme coordinates are initialized, then each data point in the tunnel point file is traversed, the coordinates thereof are compared with the size of the current initial axial bounding box extreme coordinates, and the minimum value and the maximum value of each axis are updated.
[0037] Further, an octree, a Kd tree or other spatial index structure can be established to improve the efficiency of neighborhood point retrieval in subsequent processing.
[0038] S2, voxelizing and down-sampling the tunnel point cloud original model based on a three-dimensional Gaussian kernel to obtain a tunnel deformation intensity field model.
[0039] The three-dimensional Gaussian kernel is used to spatially converge the tunnel deformation features, and the voxelization and down-sampling of the high-redundancy tunnel deformation field can be completed during the spatial convergence of the tunnel deformation values, and the spatially converged and down-sampled tunnel deformation intensity field model is: wherein sam refers to a feature space convergence method based on a Gaussian kernel, , , represents the spatial coordinate information of the voxelized and down-sampled tunnel point cloud, is a deformation space convergence result based on a spatial kernel function, is the number of points contained in the tunnel deformation intensity field. Figure 3 is a rendering diagram of the tunnel deformation intensity field model with deformation attributes provided by the embodiment.
[0040] The spatial feature convergence of the tunnel deformation attributes and the voxelization and down-sampling effectively reduce the amount of tunnel point cloud data, and can improve the data processing efficiency.
[0041] In one embodiment, step S2 includes the following steps: S21, setting a voxelization grid size, and calculating a voxel number and a voxel center coordinate set.
[0042] The voxel number and the voxel center coordinate set are calculated by the following formula:
[0043]
[0044] wherein the voxel number is: , and the voxel center coordinate set is: , is the maximum value coordinate and the minimum value coordinate of the axial bounding box information, is the voxelization grid size.
[0045] In a specific embodiment, S=0.1m is set for the tunnel point cloud voxelization and the calculation of the voxel center coordinate set.
[0046] S22, taking the non-empty voxel center point coordinates as the search point, querying all points within the specified neighborhood range of the search point in the tunnel point cloud original model as the local neighborhood points, calculating the spatial distance weight of the local neighborhood points using a three-dimensional Gaussian kernel function, and performing weight normalization, and calculating the local deformation intensity value based on the data weighting in the neighborhood.
[0047] The calculation is as follows:
[0048]
[0049]
[0050] wherein, is the local deformation intensity value at the point v n , is the deformation attribute value of the point in the original point cloud, is a bandwidth parameter for controlling the smoothing degree of the kernel, is the number of points queried within the neighborhood range m × h , m is a scalar coefficient of the spatial search distance, is an absolute value operation, is a distance calculation function, is the spatial convergence coefficient at the point , is a shorthand for the function calculation result, is the spatial convergence normalization weight at the point , is the spatial convergence coefficient of each point in the neighborhood range.
[0051] In a specific embodiment, the tunnel is provided with h =0.1m, m =3.
[0052] S23, binding the local deformation intensity value with the point cloud coordinates after voxel down-sampling, to obtain a tunnel deformation intensity field model.
[0053] S3, performing empirical cumulative distribution function calculation on the data of the tunnel deformation intensity field model, and determining the deformation intensity segmentation threshold based on the empirical cumulative distribution function.
[0054] Taking into account The local deformation intensity value presents a non-negative unimodal characteristic (most deformation values are concentrated in small deformation), the data calculation experience cumulative probability distribution function in the tunnel deformation intensity field model is calculated, and the deformation intensity segmentation threshold is adaptively determined based on the distribution characteristics and . Figure 4 The experience accumulation distribution function curve and the schematic diagram of the tunnel deformation intensity frequency distribution provided by the embodiment of the application.
[0055] In an embodiment, step S3 comprises the following steps: S31, calculating an experience cumulative distribution function based on the local deformation intensity value:
[0056] wherein, is an indicator function, 1 when , otherwise 0, is the experience cumulative distribution function.
[0057] S32, setting a probability threshold based on the experience cumulative distribution function, and determining the deformation segmentation threshold based on the probability threshold.
[0058] According to the fitted experience cumulative probability distribution function, the probability threshold is set as and , and the threshold and of the corresponding concentrated deformation region growth segmentation algorithm is obtained, which is used as the threshold of the subsequent large deformation region segmentation algorithm. The large deformation intensity segmentation threshold is adaptively determined according to the tunnel deformation amplitude distribution, and the automatic processing level of the algorithm is improved.
[0059] In a specific embodiment, = 0.9 and = 0.85, the adaptive confirmation segmentation threshold = 0.0496m and = 0.0247m, S4, screening potential seed points in the tunnel point cloud original model with deformation attributes based on the deformation intensity segmentation threshold, setting a seed point growth space neighborhood threshold, and iteratively growing the spatial topology of the seed points guided by the deformation intensity, while suppressing discrete noise interference, to quickly obtain the tunnel concentrated deformation region segmentation result.
[0060] Figure 5 The schematic diagram of the deformation cluster division from the seed point provided by the embodiment of the application. In an embodiment, step S4 comprises the following steps: S41, according to the deformation intensity segmentation threshold and , traverse the tunnel point cloud original model, divide the points in the tunnel point cloud original model into seed points, undetermined points and unclusterable points, screen and create a candidate point set containing the seed points and the undetermined points, and set the state of all points in the candidate point set to an unvisited state.
[0061] The division principle of the seed points is The division principle of the undetermined points is The division principle of the unclusterable points is .
[0062] S42, determine a seed point growth space neighborhood threshold value If the seed point is located in the neighborhood of , it is called connected directly by . All seed points and neighborhood undetermined points that are spatially connected are classified into a cluster, points that fail to be connected do not belong to any cluster, and a cluster with a seed point number less than a low clustering screening threshold value
[0063] S43, traverse the unvisited seed points, and from any unvisited seed point, all candidate points in the neighborhood of the unvisited seed point are included in the current cluster by using a breadth-first search strategy, and unvisited seed points in the candidate points are added to a queue until the queue is empty; the clustering operation is repeated and iterated, and when all seed points in the candidate point set are set to the visited state, the region growing operation is stopped, and adaptive deformation cluster clustering is completed.
[0064] S44, remove noise clusters with a seed point number less than a low clustering screening threshold value , to obtain a tunnel set deformation region segmentation result. Figure 6 The tunnel set deformation region segmentation result provided by the embodiment of the application is shown in the following schematic diagram.
[0065] At this time, all seed points and attached undetermined points that are spatially connected are classified into a cluster, the seed point iterative expansion mechanism guided by the deformation strength completes the fine clustering of the large deformation region, and effectively suppresses the discrete noise interference.
[0066] In a specific embodiment, the seed point growth space neighborhood threshold value = 0.05, = 1000.
[0067] According to the method described in the above embodiment, this embodiment will be further described from the perspective of a deformation region adaptive segmentation device for a high-redundancy tunnel deformation point cloud. The deformation region adaptive segmentation device for the high-redundancy tunnel deformation point cloud can be implemented as an independent entity or integrated in an electronic device, which can be a terminal, a server, or the like. The terminal can include a tablet computer, a notebook computer, a personal computer (PC), a micro processing box, or other devices.
[0068] Please refer to Figure 7 , Figure 7 The deformation region adaptive segmentation device for the high-redundancy tunnel deformation point cloud provided by the embodiments of the present application is specifically described. The device is applied in an electronic device and mainly includes the following modules. A tunnel point cloud deformation acquisition module is configured to acquire a tunnel point cloud file with deformation attributes, and construct a tunnel point cloud original model with deformation attributes based on the tunnel point cloud file. A voxelization down-sampling module based on a Gaussian kernel is configured to perform voxelization down-sampling on the tunnel point cloud original model based on a three-dimensional Gaussian kernel to obtain a tunnel deformation intensity field model. A deformation intensity segmentation threshold adaptive determination module is configured to perform empirical cumulative distribution function calculation on data of the tunnel deformation intensity field model, and determine a deformation intensity segmentation threshold based on the empirical cumulative distribution function. A concentrated deformation region identification module is configured to filter potential seed points in the tunnel point cloud original model with deformation attributes based on the deformation intensity segmentation threshold, set a seed point growth space neighborhood threshold, and perform spatial topology iterative growth of the seed points guided by the deformation intensity to obtain a tunnel concentrated deformation region segmentation result while suppressing discrete noise interference.
[0069] In specific implementation, each of the above modules and / or units can be implemented as an independent entity, or combined as the same or several entities. The specific implementation of each of the above modules and / or units can be referred to the method embodiments above, and the beneficial effects achieved by the specific implementation can be referred to the beneficial effects of the method embodiments above, which will not be described here again.
[0070] In addition, the electronic device provided by the embodiments of the present application can be a computer, a tablet computer, or the like. The electronic device can implement the steps in any of the deformation region adaptive segmentation methods for the high-redundancy tunnel deformation point cloud provided by the embodiments of the present application. Therefore, the electronic device can achieve the beneficial effects achieved by any of the deformation region adaptive segmentation methods for the high-redundancy tunnel deformation point cloud provided by the embodiments of the present application. Details are described in the above embodiments, which will not be described here again.
[0071] Figure 8 A specific structural block diagram of an electronic device provided by an embodiment of the present application is shown, which can be used to implement the deformation region adaptive segmentation method for high-redundancy tunnel deformation point cloud provided by the above-mentioned embodiments. The electronic device 500 can be a terminal, a server, or the like, wherein the terminal can include a tablet computer, a notebook computer, a personal computer (PC), a micro processing box, or other devices, etc.
[0072] The RF circuit 510 is configured to receive and send electromagnetic waves, and to convert the electromagnetic waves and electrical signals to each other, so as to communicate with a communication network or other devices. The RF circuit 510 can include various existing circuit elements for performing these functions, such as an antenna, a radio frequency transceiver, a digital signal processor, an encryption / decryption chip, a subscriber identity module (SIM) card, a memory, and the like. The RF circuit 510 can communicate with various networks, such as the Internet, an intranet, a wireless network, or communicate with other devices through the wireless network. The wireless network can include a cellular telephone network, a wireless local area network or metropolitan area network. The wireless network can use various communication standards, protocols and technologies, including but not limited to Global System for Mobile Communication (GSM), Enhanced Data GSM Environment (EDGE), Wideband Code Division Multiple Access (WCDMA), Code Division Access (CDMA), Time Division Multiple Access (TDMA), Wireless Fidelity (Wi-Fi) (such as Institute of Electrical and Electronics Engineers (IEEE) 802.11a, 802.11b, 802.11g and / or 802.11n), Voice over Internet Protocol (VoIP), Worldwide Interoperability for Microwave Access (Wi-Max), other protocols for email, instant messaging and short message service, and any other suitable communication protocol, even including those not yet developed.
[0073] The memory 520 can be used to store software programs and modules, such as the corresponding program instructions / modules in the above-described embodiments, and the processor 580 can execute various functions and data processing by running the software programs and modules stored in the memory 520, i.e., realize functions such as front camera shooting, processing of the shot image, and switching of display color of the display content on the display screen. The memory 520 can include a high-speed random access memory, and can further include a non-volatile memory such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 520 can further include memories disposed remotely with respect to the processor 580, which can be connected to the electronic device 500 through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.
[0074] The input unit 530 can be used to receive inputted digital or character information, and generate a keyboard, a mouse, and the like related to user settings and function control.
[0075] The display unit 540 can be used to display information inputted by the user or provided to the user, and various graphical user interfaces which can be constituted by graphics, text, icons, video, and any combination thereof. The display unit 540 can include a display panel 541, which can be configured in the form of an LCD (Liquid Crystal Display), an OLED (Organic Light-Emitting Diode), or the like.
[0076] The audio circuit 560, the speaker 561, and the microphone 562 can provide an audio interface between the user and the electronic device 500. The audio circuit 560 can convert received audio data into an electrical signal, transmit the electrical signal to the speaker 561, and convert the electrical signal into a sound signal outputted by the speaker 561; on the other hand, the microphone 562 can convert a sound signal collected into an electrical signal, and the audio circuit 560 can convert the electrical signal into audio data, output the audio data to the processor 580 for processing, and then transmit the audio data to another terminal through the RF circuit 510, or output the audio data to the memory 520 for further processing. The audio circuit 560 can further include an earphone jack to provide communication of an external earphone with the electronic device 500.
[0077] The electronic device 500 can help the user to receive requests, transmit information, and the like through the transmission module 570 (e.g., a Wi-Fi module), which provides the user with wireless broadband Internet access. Although the transmission module 570 is shown, it can be understood that it does not belong to the essential components of the electronic device 500, and can be omitted as needed without changing the essence of the application.
[0078] The processor 580 is a control center of the electronic device 500 that uses various interfaces and lines to connect the various parts of the entire mobile phone, performs various functions of the electronic device 500 and processes data by running or executing software programs and / or modules stored in the memory 520 and calling data stored in the memory 520, thereby overall monitoring the electronic device. Optionally, the processor 580 can include one or more processing cores; in some embodiments, the processor 580 can integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communication. Understandably, the above-mentioned modem processor can also not be integrated into the processor 580.
[0079] The electronic device 500 further includes a power supply 590 (such as a battery) for supplying power to various components, and in some embodiments, the power supply can be logically connected to the processor 580 through a power management system, so that the power management system can realize functions such as management of charging, discharging, and power consumption management. The power supply 590 can also include one or more direct or alternating power supplies, recharging systems, power failure detection circuits, power converters or inverters, power state indicators, and any other components.
[0080] Although not shown, the electronic device 500 also includes a camera (such as a front camera or a rear camera), a Bluetooth module, and the like, which are not described here in detail. Specifically, in the present embodiment, the display unit of the electronic device is a touch screen display, and the mobile terminal further includes a memory and one or more programs, wherein one or more programs are stored in the memory and configured to be executed by one or more processors, and the one or more programs include instructions for: obtain a tunnel point cloud file with deformation attribute, construct a tunnel point cloud original model with deformation attribute based on the tunnel point cloud file; perform voxelization down-sampling on the tunnel point cloud original model based on a three-dimensional Gaussian kernel to obtain a tunnel deformation intensity field model; perform empirical cumulative distribution function calculation on data of the tunnel deformation intensity field model, and determine a deformation intensity segmentation threshold based on the empirical cumulative distribution function; screen potential seed points in the tunnel point cloud original model with deformation attribute based on the deformation intensity segmentation threshold, set a seed point growth space neighborhood threshold, and perform spatial topology iterative growth of the seed points guided by the deformation intensity to obtain a tunnel concentrated deformation region segmentation result while suppressing discrete noise interference.
[0081] In practice, the above various modules can be implemented as independent entities, or combined as the same or several entities, and the specific implementation of the above various modules can refer to the method embodiments above, which will not be repeated here.
[0082] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions, or by related hardware controlled by the instructions, which can be stored in a computer readable storage medium and loaded and executed by a processor. Therefore, the embodiments of the present application provide a storage medium, which stores a plurality of instructions capable of being loaded by a processor to execute the steps of any one of the deformation region adaptive segmentation methods for high-redundancy tunnel deformation point clouds provided by the embodiments of the present application.
[0083] The computer readable storage medium can include a read only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.
[0084] Since the instructions stored in the storage medium can execute the steps in any one of the deformation region adaptive segmentation methods for high-redundancy tunnel deformation point clouds provided by the embodiments of the present application, the beneficial effects of any one of the deformation region adaptive segmentation methods for high-redundancy tunnel deformation point clouds provided by the embodiments of the present application can be achieved, which will be described in detail in the foregoing embodiments, and will not be repeated here.
[0085] The above provides a detailed description of the deformation region adaptive segmentation method for high-redundancy tunnel deformation point clouds, the device, the storage medium and the electronic equipment provided by the embodiments of the present application. The principle and implementation manner of the present application are described by applying specific examples in this paper, and the above embodiment description is only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation manner and application range will be changed, and the above description should not be understood as the limitation of the present application.
Claims
1. A deformation region adaptive segmentation method for highly redundant tunnel deformation point clouds, characterized by: The method comprises: Acquire a tunnel point cloud file with deformation attributes, and construct a tunnel point cloud original model with deformation attributes based on the tunnel point cloud file; Performing voxel downsampling on the original tunnel point cloud model based on a three-dimensional Gaussian kernel to obtain a tunnel deformation intensity field model; performing empirical cumulative distribution function calculation on the data of the tunnel deformation intensity field model, and determining a deformation intensity segmentation threshold based on the empirical cumulative distribution function; Based on the deformation intensity segmentation threshold, potential seed points are screened in the original model of the tunnel point cloud with deformation attributes, and a threshold for the spatial neighborhood of the seed point growth is set. The spatial topology of the seed points is guided to iteratively grow through the deformation intensity, and the segmentation result of the concentrated deformation area of the tunnel is obtained while suppressing discrete noise interference.
2. The method for adaptive segmentation of deformation regions for highly redundant tunnel deformation point clouds according to claim 1, characterized in that: After the step of obtaining the tunnel point cloud file with deformation attributes and constructing the tunnel point cloud original model with deformation attributes based on the tunnel point cloud file, the method includes: Based on the data of the tunnel point cloud file, the extreme coordinates of the axial bounding box of the tunnel point cloud model are calculated.
3. The method for adaptively segmenting deformation regions of highly redundant tunnel deformation point clouds according to claim 2, characterized in that: The voxel downsampling of the original tunnel point cloud model based on the three-dimensional Gaussian kernel to obtain the tunnel deformation intensity field model includes: Set the voxel grid size and calculate the number of voxels and the voxel center coordinates; Taking the coordinates of the center point of a non-empty voxel as a retrieval point, querying all points within a specified neighborhood of the retrieval point in the original tunnel point cloud model as local neighborhood points, calculating the spatial distance weights of the local neighborhood points using a three-dimensional Gaussian kernel function, performing weight normalization, and calculating the local deformation intensity value based on the weighted data within the neighborhood; The local deformation intensity value is bound to the point cloud coordinates after voxel downsampling to obtain a tunnel deformation intensity field model.
4. The method for adaptive segmentation of deformation regions for highly redundant tunnel deformation point clouds according to claim 3, characterized in that: The number of voxels and the voxel center coordinate set are calculated by the following formula: The number of voxels is: , the voxel center coordinate set is: , are the maximum and minimum coordinates of the axial bounding box information, is the voxelization grid size.
5. The method for adaptively segmenting deformation regions of highly redundant tunnel deformation point clouds according to claim 3, characterized in that: The local deformation strength value is calculated by the following formula: in, It's on point v n The local deformation strength value at is the midpoint of the original point cloud The deformation property value at is the bandwidth parameter, which is used to control the smoothness of the kernel. Is the neighborhood range m × h The number of points found in the query, m is the scalar coefficient of the spatial retrieval distance, To take the absolute value operation, is the distance calculation function, for point The spatial convergence coefficient at for point The spatial pooling normalized weight at , is the spatial aggregation coefficient of each point in the neighborhood.
6. The method for adaptive segmentation of deformation regions for highly redundant tunnel deformation point clouds according to claim 5, characterized in that: The step of performing empirical cumulative distribution function calculation on the data of the tunnel deformation intensity field model and determining the deformation intensity segmentation threshold based on the empirical cumulative distribution function includes: The empirical cumulative distribution function is calculated based on the local deformation intensity value: in, is the indicator function, when 1 if yes, 0 otherwise. Accumulate distribution functions for experience; Setting a probability threshold based on the empirical cumulative distribution function and , based on the probability threshold and Calculate the deformation intensity segmentation threshold and .
7. The method for adaptively segmenting deformation regions of highly redundant tunnel deformation point clouds according to claim 1, characterized in that: Based on the deformation intensity segmentation threshold, potential seed points are screened in the original tunnel point cloud model with deformation attributes, a threshold for the seed point growth space neighborhood is set, and the spatial topology of the seed points is guided by the deformation intensity to iteratively grow. The segmentation result of the concentrated deformation area of the tunnel is obtained while suppressing discrete noise interference, including: Segmentation threshold based on deformation strength and , traversing the original model of the tunnel point cloud, dividing the points in the original model of the tunnel point cloud into seed points, pending points and non-clusterable points, screening and creating a candidate point set including the seed points and the pending points, and setting the status of all points in the candidate point set to an unvisited state; Determine the neighborhood threshold of the seed point growth space , set if the seed point lie in of In the neighborhood, it is called Depend on Connected directly, all spatially connected seed points and neighboring pending points are grouped into one cluster. Unconnected points do not belong to any cluster, and clusters with a number of seed points less than the low clustering screening threshold are considered noise clusters. Traverse the unvisited seed points, starting from any unvisited seed point, and search the unvisited seed points by breadth-first search strategy. All candidate points in the neighborhood are included in the current cluster, and unvisited seed points are added to the queue until the queue is empty; the iterative clustering operation is repeated. When all seed points in the candidate point set are set to the visited state, the region growing operation is stopped, and the adaptive deformation clustering is completed; The noise clusters whose seed points in the deformation clusters are less than the low cluster screening threshold are removed to obtain the segmentation results of the concentrated deformation area of the tunnel.
8. A deformation region adaptive segmentation device for highly redundant tunnel deformation point clouds, characterized in that: include: a tunnel point cloud deformation acquisition module, configured to acquire a tunnel point cloud file with deformation attributes, and construct a tunnel point cloud original model with deformation attributes based on the tunnel point cloud file; A Gaussian kernel-based voxel downsampling module is used to perform voxel downsampling on the original tunnel point cloud model based on a three-dimensional Gaussian kernel to obtain a tunnel deformation intensity field model; a deformation strength segmentation threshold adaptive determination module, configured to perform empirical cumulative distribution function calculation on the data of the tunnel deformation strength field model and determine the deformation strength segmentation threshold based on the empirical cumulative distribution function; The concentrated deformation area identification module is used to screen potential seed points in the original tunnel point cloud model with deformation attributes based on the deformation intensity segmentation threshold, set the seed point growth space neighborhood threshold, guide the spatial topology iterative growth of the seed points through the deformation intensity, and obtain the tunnel concentrated deformation area segmentation result while suppressing discrete noise interference.
9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a plurality of instructions, which are suitable for being loaded by a processor to execute the deformation region adaptive segmentation method for highly redundant tunnel deformation point clouds according to any one of claims 1 to 7.
10. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used to execute the steps of the method for adaptive segmentation of deformation areas for highly redundant tunnel deformation point clouds as described in any one of claims 1 to 7.