Template installation gap real-time detection method and system based on image three-dimensional reconstruction

Abstract

The invention provides a template installation gap real-time detection method and system based on image three-dimensional reconstruction, and relates to the technical field of image processing, and the method comprises the steps: obtaining image sequence data collected at multiple angles in real time in the installation process of a to-be-detected vertical wall template; performing image inversion according to the image sequence data to generate a three-dimensional moving point cloud field; performing curved surface reconstruction according to the three-dimensional moving point cloud field to obtain a fracture sensitive curved surface model; performing cross-scale image feature fusion of the installation gap according to the fracture sensitive curved surface model to generate a probability confidence map; performing geometric constraint relaxation processing on the gap boundary according to the probability confidence map to obtain a mounting gap vector boundary; and evaluating according to the installation gap vector boundary to obtain a quantization parameter set. According to the method, the defects that manual detection and the prior art are insufficient in adaptability to dynamic light change and low-texture surfaces and lack of engineering semantics are effectively overcome.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method and system for real-time detection of template installation gaps based on three-dimensional image reconstruction. Background Art

[0002] In the field of computer vision applications for concrete structure construction quality control, with the iterative advancement of image processing technology, gap detection for cast-in-place formwork installation has gradually evolved from manual judgment to automated detection. However, existing mainstream technical solutions still face fundamental bottlenecks. Existing methods based on multi-view stereo matching (such as sparse reconstruction based on SfM combined with edge detection algorithms) are limited in their adaptability to low-texture mirror materials when processing large-area steel formwork joints. This is due to the inability of the feature point matching mechanism to reliably capture sub-pixel displacement trajectories on highly reflective surfaces. Furthermore, under dynamic illumination interference, conventional edge detection operators (such as Canny and Sobel) respond slowly to subtle grayscale gradient changes at steel plate joints, resulting in distorted reconstruction of the temporal deformation field. More importantly, existing methods lack modeling of construction physics rules at the engineering semantic level. By manually setting empirical thresholds to segment gap regions, they fail to construct a cross-domain mapping model between image gradient distribution and concrete seepage mechanical parameters. This makes it difficult for detection results to meet the quantitative acceptance requirements for structural waterproofing in bridge projects, hindering the in-depth implementation of high-precision image processing technology in construction quality control scenarios.

[0003] Based on the above shortcomings of the prior art, there is an urgent need for a real-time detection method and system for template installation gaps based on three-dimensional image reconstruction. Summary of the Invention

[0004] The purpose of the present invention is to provide a real-time detection method for template installation gaps based on three-dimensional image reconstruction to improve the above-mentioned problem. To achieve the above-mentioned purpose, the technical solution adopted by the present invention is as follows: In a first aspect, the present application provides a method for real-time detection of template installation gaps based on three-dimensional image reconstruction, comprising: Acquire image sequence data collected in real time from multiple angles during the installation of the vertical wall formwork to be inspected, wherein the image sequence data includes local reflection characteristics and time sequence information of the template joint area under natural light or structured light illumination; Performing image inversion based on the image sequence data to generate a three-dimensional motion point cloud field; Performing surface reconstruction based on the three-dimensional motion point cloud field to obtain a fracture-sensitive surface model; Performing cross-scale image feature fusion of the installation gap based on the fracture-sensitive surface model, and generating a probability confidence map by combining the surface geometric mutation features with the texture fracture features of the seam area in the original image; Performing geometric constraint relaxation processing on the gap boundary according to the probability confidence map, guiding the initial contour to adaptively shrink along the confidence peak trajectory through the image gradient flow field, so that the boundary is locked at the physical fracture position of the template joint, and obtaining the installation gap vector boundary; An evaluation is performed based on the installation gap vector boundary to obtain a set of quantitative parameters characterizing the amount of installation misalignment, gap width, and leakage risk.

[0005] In a second aspect, the present application also provides a real-time detection system for template installation gaps based on three-dimensional image reconstruction, comprising: An acquisition module is used to acquire image sequence data collected in real time from multiple angles during the installation of the vertical wall formwork to be inspected, wherein the image sequence data includes local reflection characteristics and time sequence information of the template joint area under natural light or structured light illumination; an inversion module, configured to perform image inversion based on the image sequence data to generate a three-dimensional motion point cloud field; A reconstruction module, configured to perform surface reconstruction based on the three-dimensional motion point cloud field to obtain a fracture-sensitive surface model; a fusion module for fusing cross-scale image features of the installation gap according to the fracture-sensitive surface model, and generating a probability confidence map by combining the surface geometric mutation features with the texture fracture features of the seam area in the original image; a processing module, configured to perform geometric constraint relaxation processing on the gap boundary according to the probability confidence map, guide the initial contour to adaptively shrink along the confidence peak trajectory through the image gradient flow field, lock the boundary at the physical fracture position of the template joint, and obtain the installation gap vector boundary; An evaluation module is used to evaluate according to the installation gap vector boundary to obtain a quantitative parameter set representing the installation misalignment amount, gap width and leakage risk.

[0006] The beneficial effects of the present invention are: The present invention generates a three-dimensional motion point cloud field through image inversion to accurately reconstruct the dynamic deformation process. It generates a probability confidence map based on the fracture-sensitive surface model by fusing multi-scale images and geometric features. It combines the adaptive boundary optimization with physical constraints to lock the actual fracture position of the seam, and finally realizes the automatic output of engineering mechanics parameters, effectively overcoming the defects of manual detection and existing technologies in terms of insufficient adaptability to dynamic light changes and low-texture surfaces and the lack of engineering semantics. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0008] Figure 1 A schematic flow diagram of a method for real-time detection of a template installation gap based on three-dimensional image reconstruction according to an embodiment of the present invention; Figure 2 Schematic diagram of the structure of a real-time detection system for template installation gaps based on three-dimensional image reconstruction according to an embodiment of the present invention; Figure 3 Schematic diagram of the structure of a real-time detection device for template installation gaps based on three-dimensional image reconstruction according to an embodiment of the present invention.

[0009] Markings in the figure: 800, a real-time detection device for template installation gaps based on image three-dimensional reconstruction; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component; 901, acquisition module; 902, inversion module; 903, reconstruction module; 904, fusion module; 905, processing module; 906, evaluation module. DETAILED DESCRIPTION

[0010] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0011] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are used only to distinguish the description and should not be understood as indicating or implying relative importance.

[0012] Example 1: This embodiment provides a real-time detection method for template installation gaps based on three-dimensional image reconstruction.

[0013] See also Figure 1 , the figure shows that the method includes steps S100 to S600.

[0014] Step S100: Acquire image sequence data collected in real time from multiple angles during the installation of the vertical wall formwork to be inspected, the image sequence data including local reflection characteristics and time sequence information of the template joint area under natural light or structured light illumination; It can be understood that this step targets the complexity of the bridge vertical wall formwork construction environment, directly capturing multi-angle time-series image data from the dynamic installation site, focusing on the changes in the reflective characteristics of the joint area under natural or controllable light sources, and providing basic light field information with both spatial details and temporal evolution for subsequent processing.

[0015] Step S200: performing image inversion based on the image sequence data to generate a three-dimensional motion point cloud field; It should be noted that this step uses the spatiotemporal correlation of the image sequence to reversely map the pixel displacement trajectory to three-dimensional space. By constructing a physical coupling mechanism between pixel motion and structural deformation, the two-dimensional brightness gradient distribution is converted into a three-dimensional continuous motion field, breaking through the limitations of manual detection of non-visual domain displacement and solving the problem of inaccurate surface motion inversion of the mirror template.

[0016] Step S300: reconstructing a surface based on the three-dimensional motion point cloud field to obtain a fracture-sensitive surface model; It can be understood that this step is based on the spatial distribution characteristics of the point cloud of the three-dimensional motion field, introduces the topological modeling idea of ​​structural continuity constraints, reconstructs the continuous surface while retaining the local deformation gradient details of the steel plate, and at the same time strengthens the expression of the non-smooth geometric features of the joint area, eliminating the smooth annihilation of the installation misalignment features caused by conventional surface reconstruction.

[0017] Step S400: performing cross-scale image feature fusion of the installation gap based on the fracture-sensitive surface model, and generating a probability confidence map by combining the surface geometric mutation features with the texture fracture features of the seam area in the original image; It should be noted that this step combines the geometric mutation characteristics of the surface model with the multi-scale texture clues of the original image to establish a cross-domain correlation model driven by engineering physics rules. By fusing the physical evidence of structural fracture with the abnormal response of visual texture, the spatial probability distribution of the existence of gaps is quantified at the pixel level, avoiding the risk of missing slight installation deviations in single modality detection.

[0018] Step S500: performing geometric constraint relaxation processing on the gap boundary based on the probability confidence map, guiding the initial contour to adaptively shrink along the confidence peak trajectory through the image gradient flow field, so that the boundary is locked at the physical fracture position of the template joint, and obtaining the installation gap vector boundary; It can be understood that the probabilistic confidence map is converted into a constraint field to drive the autonomous evolution of the geometric boundary, and the coupling mechanism of the image gradient flow and the structural fracture energy is used to guide the contour to lock the real seam boundary, thereby achieving adaptive convergence from the probability density peak to the physical fracture position.

[0019] Step S600: Evaluate the installation gap vector boundary to obtain a set of quantitative parameters representing the amount of installation misalignment, gap width, and leakage risk.

[0020] Finally, based on the spatial geometric properties of the boundary curve, the correlation between the gap depth gradient and fluid behavior is analyzed within the framework of the structural functional failure model. By mapping the geometric parameters into seepage mechanics indicators, a quality quantification set that directly serves the construction acceptance is output, filling the gap in the traditional methods' lack of quantification of anti-seepage requirements at the engineering semantic level.

[0021] Furthermore, step S200 includes steps S210 to S230.

[0022] Step S210: Perform dynamic pixel trajectory analysis based on the image sequence data, and obtain the sub-pixel displacement trajectory field of the template surface by extracting the spatiotemporal correlation information of the brightness gradient of the pixels in the seam area in consecutive frames; Step S220: Perform non-rigid motion 3D reconstruction based on the displacement trajectory field, and generate a 3D displacement vector set reflecting the actual deformation of the template by jointly solving the perspective projection geometric constraints and the motion continuity prior. Step S230: Perform physical deformation pattern association processing based on the three-dimensional displacement vector set, and perform physical consistency correction on the displacement vector by introducing the elastic potential energy model of the template material to obtain a three-dimensional motion point cloud field that characterizes the deformation characteristics during the installation process.

[0023] In the dynamic monitoring of the bridge vertical wall formwork installation process, step S210 first extracts the brightness gradient correlation features of adjacent frames in the joint area from a multi-angle time-series image sequence. This involves analyzing the brightness and darkness variation patterns and spatial distribution patterns of specific pixels at consecutive time points, and converting subtle displacements on the steel formwork surface into a trajectory field with sub-pixel accuracy. Step S220, based on the spatiotemporal evolution characteristics of the displacement trajectory field, integrates the camera perspective projection principle with the physical continuity constraints of structural motion (for example, the deformation transfer relationship between adjacent bolt nodes). By inversely solving the mapping model between pixel displacement and three-dimensional spatial deformation, the true three-dimensional displacement vector set of the steel plate under vibration load is reconstructed. Step S230 further introduces a material constitutive model and constructs an elastic potential energy field based on the elastic modulus and thickness parameters of the steel plate (simulating the bending stiffness characteristics of Q235 steel). Physical consistency correction is applied to the three-dimensional displacement vectors to suppress abnormal displacement components that exceed the material deformation limit, ultimately generating a three-dimensional motion point cloud field that conforms to engineering practice. The correction mechanism of the above process overcomes the non-physical deformation caused by light noise or matching errors in image inversion, so that the reconstruction results strictly follow the material response law of the bridge formwork under the action of bolt preload. The steps form a progressive optimization chain of "image trajectory capture-spatial deformation reconstruction-physical property rule calibration", which solves the problem of deformation reconstruction distortion caused by mirror reflection and vibration interference of low-texture steel plates in dynamic installation scenarios from the algorithm level.

[0024] Specifically, step S220 constructs a non-rigid motion reconstruction energy function based on the trajectory field. This function integrates the geometric constraints of the camera perspective projection matrix and the motion continuity prior (such as the displacement transfer attenuation characteristics of the bolt connection node). By minimizing the weighted combination of reprojection error and structural deformation energy, it solves the three-dimensional displacement vector set that meets the local rigid constraints of the steel plate. The non-rigid motion reconstruction energy function is: ; Where, Indicates the The three-dimensional displacement vector of the node; Indicates the Node initial space coordinates; Represents the camera projection function; represents the projection matrix; Indicates the image sequence Point trajectory coordinates; represents the coupling weight; Represents the set of structural connection edges; represents the connection stiffness coefficient; represents the stiffness penalty function; , Indicates the node's sequence number; Indicates the total number of nodes; Represents a set of three-dimensional displacement vectors; Step S230 further introduces an elastic potential energy correction model to calculate the connection stiffness based on the elastic modulus and thickness of Q235 steel. By suppressing displacement components that exceed the material's yield limit (such as abnormal stretching caused by welding thermal deformation), a physically consistent three-dimensional motion point cloud field is generated. The elastic potential energy correction model is expressed as: ; Where, Indicates the Three-dimensional displacement vector of the node after correction; Representation node to The connection vector of represents the node displacement difference; Indicates the installation normal deviation angle; represents the elastic modulus of the steel plate; Indicates the thickness of the template; Indicates the initial connection length; represents the creep relaxation factor; represents the stiffness attenuation coefficient; represents the Hadamard product; represents the tensor product; Represents the surface normal vector.

[0025] Furthermore, step S300 includes steps S310 to S330.

[0026] Step S310: performing spatial structure dependency modeling based on the three-dimensional motion point cloud field, and constructing a dynamic association network between point clouds based on elastic deformation constraints to obtain a spatial association topology map containing local geometric continuity features; Step S320: fusion of continuous and discontinuous features is performed based on the spatial correlation topology graph, and a hybrid geometric expression that simultaneously characterizes the smooth area of ​​the surface and the fracture zone is generated by decoupling the second-order differentiable function family of the continuous surface and the non-smooth operator of the joint fracture feature; Step S330 : performing fracture energy threshold compression according to the hybrid geometric expression, selectively strengthening the non-smooth operator region by introducing a fracture energy density function, and outputting a fracture sensitive surface model that is sensitive to the installation joint.

[0027] It should be noted that in the surface reconstruction process, step S310 targets the structural characteristics of the steel formwork bolt connection by establishing an elastic deformation constraint network between point cloud nodes (for example, simulating the deformation transfer relationship between adjacent bolt holes under vibration loads), converting the physical connection rules into a spatial topological association graph, thereby addressing the problem that traditional neighborhood search fails to model structural continuity. Step S320 implements feature decoupling based on this topological graph, using a family of surface differentiable functions to describe the smooth deformation of the steel plate body (preferably, such as bicubic spline surface fitting), while constructing a non-smooth operator to characterize the geometric mutation characteristics at the joint, achieving seamless integration of the structural continuity region and the fracture zone at the mathematical level. Step S330 further introduces a fracture energy density model (calculating the critical energy threshold based on the material yield strength and thickness), gradient amplification and noise suppression of the non-smooth region in the hybrid expression, and enhancing the fracture characteristic response when the local deformation energy exceeds the fatigue limit of the Q235 steel formwork. Ultimately, a fracture surface sensitive to installation misalignment is generated, for example, making the submillimeter crack at the edge of the weld seam appear as a spike in the surface curvature distribution. This progressive mechanism enables the visualization of fractures in high-strength connection structures unique to bridge engineering under vibration conditions.

[0028] Furthermore, step S400 includes steps S410 to S430.

[0029] Step S410: performing geometric-texture feature decomposition based on the fracture-sensitive surface model, and obtaining a geometric fracture feature vector and an image texture feature matrix by respectively extracting the geometric discontinuity magnitude of the gradient mutation region in the surface and the pixel gradient directional anomaly in the seam region in the original image; Step S420: perform dual-domain association rule modeling based on the geometric fracture feature vector and the image texture feature matrix, and generate pixel-level cross-domain association intensity distribution by constructing a linear response function between the geometric discontinuity magnitude and pixel gradient anomaly; Step S430: Perform leakage risk probability mapping based on the cross-domain correlation strength distribution, and obtain a probability confidence map representing the probability of joint sealing failure by integrating the correlation strength and the structural sealing failure threshold in energy space.

[0030] In the concrete structure construction quality inspection link, step S410 targets the installation characteristics of the steel formwork of the vertical wall of the bridge, and first implements multimodal feature decoupling of the fracture-sensitive surface: based on the sudden change amplitude of the reconstructed surface spatial curvature, the geometric discontinuity of the joint area is quantified (for example, the difference between the local curvature peak and valley exceeds the design threshold, indicating installation misalignment), and the gradient direction aggregation deviation of the joint edge pixels in the original image is simultaneously analyzed (that is, the abnormal fracture angle of the reflected light band), thereby separating the geometric feature vector representing the physical deformation of the structure and the texture feature matrix reflecting the surface optical response, eliminating the interference of pseudo-edge noise in dynamic light environment; step S420 constructs a dual-domain correlation response machine Specifically, a linear weighting function is used to establish a mapping relationship between the magnitude of geometric discontinuity (such as the rate of change of curvature per unit length) and the pixel gradient direction deviation (such as the energy ratio in the orthogonal direction), and a pixel-level cross-domain correlation intensity thermal field is generated through adaptive weight distribution to achieve collaborative verification of the physical deformation and optical appearance of the steel plate; step S430 introduces an engineering leakage failure physical model, and performs energy space integration on the cross-domain correlation intensity according to the structural sealing specification threshold (such as the minimum allowable joint width of anti-seepage concrete), and converts the spatially distributed thermal value into the sealing failure probability of the joint area, forming a probability confidence map that simultaneously encodes the geometric deformation risk and optical anomaly evidence.

[0031] Furthermore, step S500 includes steps S510 to S530.

[0032] Step S510: constructing a boundary motion constraint field based on the probability confidence map, and obtaining a dynamic constraint force distribution that guides contour motion by mapping the probability confidence gradient into a curvature-dependent potential energy field in pixel space; Step S520: Performing physical driven contour evolution according to the dynamic constraint force distribution, and generating a transient boundary trajectory that is topologically adaptively contracted along the joint fracture zone by simulating the minimum energy deformation process of the elastic boundary line in the constraint force field; Step S530 : performing fracture stability convergence processing according to the transient boundary trajectory, and outputting the installation gap vector boundary that locks the physical fracture position by detecting the motion steady state of the trajectory curvature under the joint energy release rate threshold.

[0033] Specifically, step S510 converts the confidence gradient representing the risk of sealing failure in the probability confidence map into a mechanical constraint field in pixel space based on the physical characteristics of the steel template joint. Preferably, the potential energy of the high curvature area (corresponding to the stress concentration point of the steel plate fracture zone) is enhanced through a curvature modulation mechanism, forming a non-uniform constraint force distribution that drives the contour movement (this design conforms to the energy release law of brittle fracture of steel). Step S520 innovatively constructs a physical driving model based on this constraint field, treating the boundary line as an elastic material (simulating the ductility of the steel plate). According to the principle of minimum deformation energy, it continuously evolves in the potential energy field, so that the topological structure of the contour along the joint fracture zone adaptively shrinks. Step S530 further introduces the structural fracture mechanics criterion to detect the local curvature change rate of the boundary trajectory in real time (reflecting the energy release intensity at the crack tip). When the curvature fluctuation is lower than the critical threshold (indicating that the energy release is tending to be stable), the boundary is determined to have converged, and finally the sub-pixel precision vector boundary is output at the actual physical fracture location of the template joint.

[0034] Furthermore, step S600 includes steps S610 to S630.

[0035] Step S610: Perform three-dimensional structural functional failure modeling based on the installation gap vector boundary, project boundary points along the surface normal and analyze the curvature spatial gradient distribution to generate a local structural abnormality feature tensor that characterizes the three-dimensional direction of the gap; Step S620: Perform dislocation-leakage coupling analysis based on the local structural abnormality characteristic tensor, and output the installation dislocation amount and leakage risk coefficient by establishing a linear response model of the gap depth gradient and the concrete flow potential energy; Step S630: Perform dimension conversion of the project acceptance parameters according to the misalignment amount and the leakage risk coefficient, and generate a quantitative parameter set that meets the construction acceptance specifications based on the dual criteria of the bridge joint sealing standard and the structural deformation threshold.

[0036] In the final stage of the project acceptance assessment, step S610 projects the gap boundary points along the normal direction of the reconstructed surface (aligned with the stress and deformation direction of the steel plate) based on the structural characteristics of the steel formwork. By analyzing the distribution pattern of the curvature gradient in three-dimensional space (the rate of change in the main curvature direction reflects the stress concentration effect at the bolt connection), a structural anomaly feature tensor that integrates depth discontinuity information is generated. This tensor is used to accurately encode the three-dimensional deformation characteristics of the joint; step S620 constructs the concrete fluid mechanics response mechanism based on this tensor, converting the gap depth gradient (reflecting the degree of formwork misalignment) into concrete flow potential energy (simulating the viscous resistance of cement paste between rough joint walls), and synchronously calculates the installation through a linear response model. The amount of installation misalignment (such as the height difference between adjacent steel formwork panels) and the leakage risk coefficient (potential energy integral value mapped to leakage probability) are used to physically quantify the bridge's anti-seepage requirements. Step S630, based on project acceptance standards (such as the joint width limit and concrete impermeability grade specified in the JTGF80 specification), adopts a dual-criteria fusion strategy: the misalignment is filtered through a structural deformation threshold (exceeding the limit is considered unqualified), and the leakage coefficient is mapped to a sealing failure level. Ultimately, an acceptance parameter set consisting of [misalignment level, equivalent joint width, leakage index] is generated. This process addresses the problem of insufficient quantification of three-dimensional defects in manual assessment by automatically converting engineering semantics, and directly outputs a digital quality report that can be executed by construction supervisors.

[0037] Furthermore, step S620 includes steps S621 to S623.

[0038] Step S621: performing gap geometric invariant extraction processing based on the local structural abnormality feature tensor, extracting the depth gradient discontinuity distribution in the main direction of curvature by tensor eigenvalue decomposition, and obtaining the gap three-dimensional deformation gradient field; Step S622: Modeling the concrete fluid dynamics response based on the three-dimensional deformation gradient field of the gap, and generating an equivalent potential energy distribution of the concrete flow by establishing the gradient distribution and the boundary layer equation of Newtonian fluid under rough boundary conditions; Step S623: Structural risk coupling quantification is performed based on the equivalent potential energy distribution. By performing a convolution analysis on the potential energy peak and the construction load spectrum, the installation misalignment amount and the leakage risk coefficient related to the pouring pressure are output.

[0039] Specifically, step S621 performs spectral decomposition on the structural abnormal characteristic tensor based on the mechanical response characteristics of the steel structure (extracting the principal curvature depth gradient in the direction of the maximum eigenvalue). This constructs a three-dimensional deformation gradient field that reflects the stress concentration effect at the bolted joint. This 3D deformation gradient field quantifies the non-uniform deformation of the joint region along the thickness of the steel plate, overcoming the shortcomings of traditional methods in modeling the elastic rebound effect of the formwork. Step S622 introduces a concrete rheology model to establish a physical mapping between the deformation gradient field and the behavior of Newtonian fluids. The core of this method is to simulate the viscous resistance of cement paste within the crack using a rough boundary layer equation, converting the gradient distribution into an equivalent potential energy distribution for concrete flow (the potential energy peak corresponds to the high-risk leakage area). The adaptive adjustment mechanism of the boundary layer thickness is specifically designed to address the flow resistance characteristics of the corroded surface of Q235 steel plates. Step S623 further integrates construction dynamic loads (such as the frequency spectrum of the vibrating rod) and performs a time-domain convolution operation on the potential energy distribution. By analyzing the coupling strength between the potential energy peak and the pouring pressure pulsation, the installation misalignment amount (the integral value of the deformation gradient) and the leakage risk factor under dynamic working conditions are simultaneously output. These three steps realize the closed loop of "steel structure deformation analysis - concrete fluid modeling - construction load coupling", solving the industry problem that the static evaluation model is insensitive to the dynamic risks of the bridge casting process.

[0040] Example 2: like Figure 2 As shown, this embodiment provides a real-time detection system for template installation gaps based on three-dimensional image reconstruction, the system comprising: An acquisition module 901 is configured to acquire image sequence data collected in real time from multiple angles during the installation of the vertical wall formwork to be inspected. The image sequence data includes local reflection characteristics and time sequence information of the template joint area under natural light or structured light illumination; Inversion module 902, used to perform image inversion based on image sequence data to generate a three-dimensional motion point cloud field; The reconstruction module 903 is used to perform surface reconstruction based on the three-dimensional motion point cloud field to obtain a fracture-sensitive surface model; A fusion module 904 is configured to fuse cross-scale image features of the installation gap based on the fracture-sensitive surface model, and generate a probability confidence map by combining the surface geometric mutation features with the texture fracture features of the seam area in the original image; Processing module 905 is used to relax the geometric constraints of the gap boundary based on the probability confidence map, guide the initial contour to adaptively shrink along the confidence peak trajectory through the image gradient flow field, so that the boundary is locked at the physical fracture position of the template joint, and obtain the installation gap vector boundary; The evaluation module 906 is configured to perform evaluation based on the installation gap vector boundary to obtain a set of quantitative parameters representing the amount of installation misalignment, gap width, and leakage risk.

[0041] In a specific embodiment of the present application, the inversion module 902 includes: The first inversion unit is used to perform dynamic pixel trajectory analysis based on the image sequence data, and obtain the sub-pixel displacement trajectory field of the template surface by extracting the spatiotemporal correlation information of the brightness gradient of the pixels in the seam area in consecutive frames; The second inversion unit is used to perform non-rigid motion 3D reconstruction based on the displacement trajectory field. By jointly solving the perspective projection geometric constraints and the motion continuity prior, a 3D displacement vector set reflecting the actual deformation of the template is generated. The third inversion unit is used to perform physical deformation pattern association processing based on the three-dimensional displacement vector set, and to perform physical consistency correction on the displacement vector by introducing the elastic potential energy model of the template material to obtain a three-dimensional motion point cloud field that characterizes the deformation characteristics during the installation process.

[0042] In a specific embodiment of the present application, the reconstruction module 903 includes: The first reconstruction unit is used to perform spatial structure dependency modeling based on the three-dimensional motion point cloud field, and obtain a spatial correlation topology map containing local geometric continuity features by constructing a dynamic correlation network between point clouds based on elastic deformation constraints; The second reconstruction unit is used to fuse continuous and discontinuous features based on the spatial correlation topology. By decoupling the second-order differentiable function family of the continuous surface and the non-smooth operator of the joint fracture characteristics, a hybrid geometric expression is generated that simultaneously characterizes the smooth area of ​​the surface and the fracture zone. The third reconstruction unit is used to perform fracture energy threshold compression according to the hybrid geometric expression, selectively strengthen the non-smooth operator area by introducing the fracture energy density function, and output a fracture sensitive surface model that is sensitive to the installation joint.

[0043] Example 3: Corresponding to the above method embodiment, this embodiment also provides a real-time detection device for template installation gaps based on image three-dimensional reconstruction. The real-time detection device for template installation gaps based on image three-dimensional reconstruction described below and the real-time detection method for template installation gaps based on image three-dimensional reconstruction described above can be referenced to each other.

[0044] Figure 3 FIG. 8 is a block diagram of a real-time detection device 800 for template installation gaps based on three-dimensional image reconstruction according to an exemplary embodiment. Figure 3 As shown, the real-time detection device 800 for template installation gaps based on three-dimensional image reconstruction may include: a processor 801 and a memory 802. The real-time detection device 800 for template installation gaps based on three-dimensional image reconstruction may also include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0045] The processor 801 is used to control the overall operation of the real-time detection device 800 for template installation gaps based on three-dimensional image reconstruction to complete all or part of the steps of the above-mentioned real-time detection method for template installation gaps based on three-dimensional image reconstruction. The memory 802 is used to store various types of data to support the operation of the real-time detection device 800 for template installation gaps based on three-dimensional image reconstruction. This data may include, for example, instructions for any application or method operating on the real-time detection device 800 for template installation gaps based on three-dimensional image reconstruction, as well as application-related data, such as contact information, sent and received messages, pictures, audio, video, etc. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may be further stored in the memory 802 or transmitted via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above-mentioned other interface modules can be a keyboard, a mouse, buttons, etc. These buttons can be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the template installation gap real-time detection device 800 based on image three-dimensional reconstruction and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 805 may include: Wi-Fi module, Bluetooth module, NFC module.

[0046] In an exemplary embodiment, a real-time detection device 800 for template installation gaps based on three-dimensional image reconstruction can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above-mentioned real-time detection method for template installation gaps based on three-dimensional image reconstruction.

[0047] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided. When executed by a processor, the program instructions implement the steps of the above-described method for real-time detection of template installation gaps based on three-dimensional image reconstruction. For example, the computer-readable storage medium may be the aforementioned memory 802 including the program instructions. The program instructions may be executed by the processor 801 of the device 800 for real-time detection of template installation gaps based on three-dimensional image reconstruction to implement the above-described method for real-time detection of template installation gaps based on three-dimensional image reconstruction.

[0048] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered by the scope of protection of the present invention.

Claims

1. A real-time detection method for template installation gaps based on three-dimensional image reconstruction, characterized in that: include: Acquire image sequence data collected in real time from multiple angles during the installation of the vertical wall formwork to be inspected, wherein the image sequence data includes local reflection characteristics and time sequence information of the template joint area under natural light or structured light illumination; Performing image inversion based on the image sequence data to generate a three-dimensional motion point cloud field; Performing surface reconstruction based on the three-dimensional motion point cloud field to obtain a fracture-sensitive surface model; Performing cross-scale image feature fusion of the installation gap based on the fracture-sensitive surface model, and generating a probability confidence map by combining the surface geometric mutation features with the texture fracture features of the seam area in the original image; Performing geometric constraint relaxation processing on the gap boundary according to the probability confidence map, guiding the initial contour to adaptively shrink along the confidence peak trajectory through the image gradient flow field, so that the boundary is locked at the physical fracture position of the template joint, and obtaining the installation gap vector boundary; An evaluation is performed based on the installation gap vector boundary to obtain a set of quantitative parameters characterizing the amount of installation misalignment, gap width, and leakage risk.

2. The method for real-time detection of template installation gaps based on three-dimensional image reconstruction according to claim 1, characterized in that: Performing image inversion according to the image sequence data to generate a three-dimensional motion point cloud field includes: Performing dynamic pixel trajectory analysis on the image sequence data to obtain a sub-pixel displacement trajectory field on the template surface by extracting spatiotemporal correlation information of brightness gradients of pixels in the seam area in consecutive frames; Performing non-rigid motion 3D reconstruction based on the displacement trajectory field, generating a 3D displacement vector set reflecting the actual deformation of the template by jointly solving perspective projection geometric constraints and motion continuity priors; Physical deformation pattern association processing is performed based on the three-dimensional displacement vector set, and the displacement vector is corrected for physical consistency by introducing the elastic potential energy model of the template material to obtain a three-dimensional motion point cloud field that characterizes the deformation characteristics during the installation process.

3. The method for real-time detection of template installation gaps based on three-dimensional image reconstruction according to claim 1, characterized in that: Performing surface reconstruction based on the three-dimensional motion point cloud field to obtain a fracture-sensitive surface model includes: Performing spatial structure dependency modeling based on the three-dimensional motion point cloud field, and constructing a dynamic association network between point clouds based on elastic deformation constraints to obtain a spatial association topology map containing local geometric continuity features; Continuous-discontinuous feature fusion is performed based on the spatial correlation topology, and a hybrid geometric expression that simultaneously represents the smooth area of ​​the surface and the fracture zone is generated by decoupling the second-order differentiable function family of the continuous surface and the non-smooth operator of the joint fracture feature; Fracture energy threshold compression is performed according to the hybrid geometric expression, and the non-smooth operator region is selectively enhanced by introducing a fracture energy density function, and a fracture sensitive surface model that is sensitive to the installation joint is output.

4. The method for real-time detection of template installation gaps based on three-dimensional image reconstruction according to claim 1, characterized in that: Cross-scale image feature fusion of the installation gap is performed based on the fracture-sensitive surface model. By combining the surface geometric mutation features with the texture fracture features of the seam area in the original image, a probability confidence map is generated, including: Performing geometric-texture feature decomposition based on the fracture-sensitive surface model, and extracting the geometric discontinuity magnitude of the gradient mutation area in the surface and the pixel gradient directional anomaly in the seam area in the original image, respectively, to obtain a geometric fracture feature vector and an image texture feature matrix; Performing dual-domain association rule modeling based on the geometric fracture feature vector and the image texture feature matrix, and generating pixel-level cross-domain association intensity distribution by constructing a linear response function between the geometric discontinuity magnitude and pixel gradient anomaly; Leakage risk probability mapping is performed based on the cross-domain correlation strength distribution, and a probability confidence map representing the probability of joint sealing failure is obtained by integrating the correlation strength and the structural sealing failure threshold in energy space.

5. The method for real-time detection of template installation gaps based on three-dimensional image reconstruction according to claim 1, characterized in that: The geometric constraint relaxation process of the gap boundary is performed according to the probability confidence map, and the initial contour is guided to adaptively shrink along the confidence peak trajectory through the image gradient flow field, so that the boundary is locked at the physical fracture position of the template joint, and the installation gap vector boundary is obtained, including: Constructing a boundary motion constraint field based on the probability confidence map, and obtaining a dynamic constraint force distribution that guides contour motion by mapping the probability confidence gradient into a curvature-dependent potential energy field in pixel space; Performing physical driven contour evolution according to the dynamic constraint force distribution, and generating a transient boundary trajectory with topological adaptive contraction along the joint fracture zone by simulating the minimum energy deformation process of the elastic boundary line in the constraint force field; Fracture stability convergence processing is performed according to the transient boundary trajectory, and the motion steady state of the trajectory curvature under the joint energy release rate threshold is detected to output the installation gap vector boundary that locks the physical fracture position.

6. The method for real-time detection of template installation gaps based on three-dimensional image reconstruction according to claim 1, characterized in that: An evaluation is performed based on the installation gap vector boundary to obtain a set of quantitative parameters characterizing the amount of installation misalignment, gap width, and leakage risk, including: Performing three-dimensional structural functional failure modeling based on the installation gap vector boundary, generating a local structural abnormality feature tensor representing the three-dimensional direction of the gap by projecting boundary points along the surface normal and analyzing the spatial gradient distribution of the curvature; Performing a dislocation-leakage coupling analysis based on the local structural abnormality characteristic tensor, and outputting the installation dislocation amount and leakage risk coefficient by establishing a linear response model of the gap depth gradient and the concrete flow potential energy; The engineering acceptance parameter dimensionality conversion is performed according to the misalignment amount and the leakage risk coefficient, and a quantitative parameter set that meets the construction acceptance specifications is generated based on the dual criteria fusion of the bridge joint sealing standard and the structural deformation threshold.

7. The method for real-time detection of template installation gaps based on three-dimensional image reconstruction according to claim 6, characterized in that: Based on the local structural abnormality characteristic tensor, a dislocation-leakage coupling analysis is performed. By establishing a linear response model of the gap depth gradient and the concrete flow potential energy, the installation dislocation amount and leakage risk coefficient are output, including: According to the local structural abnormality feature tensor, a gap geometric invariant extraction process is performed, and the depth gradient discontinuity distribution in the main direction of curvature is extracted by tensor eigenvalue decomposition to obtain the gap three-dimensional deformation gradient field; Concrete fluid dynamics response modeling is performed based on the three-dimensional deformation gradient field of the gap, and an equivalent potential energy distribution of concrete flow is generated by establishing the gradient distribution and the boundary layer equation of Newtonian fluid under rough boundary conditions; Structural risk coupling quantification is performed based on the equivalent potential energy distribution, and the installation misalignment amount and the leakage risk coefficient related to the pouring pressure are output through convolution analysis of the potential energy peak and the construction load spectrum.

8. A real-time detection system for template installation gaps based on three-dimensional image reconstruction, characterized in that: include: An acquisition module is used to acquire image sequence data collected in real time from multiple angles during the installation of the vertical wall formwork to be inspected, wherein the image sequence data includes local reflection characteristics and time sequence information of the template joint area under natural light or structured light illumination; an inversion module, configured to perform image inversion based on the image sequence data to generate a three-dimensional motion point cloud field; A reconstruction module, configured to perform surface reconstruction based on the three-dimensional motion point cloud field to obtain a fracture-sensitive surface model; a fusion module for fusing cross-scale image features of the installation gap according to the fracture-sensitive surface model, and generating a probability confidence map by combining the surface geometric mutation features with the texture fracture features of the seam area in the original image; a processing module, configured to perform geometric constraint relaxation processing on the gap boundary according to the probability confidence map, guide the initial contour to adaptively shrink along the confidence peak trajectory through the image gradient flow field, lock the boundary at the physical fracture position of the template joint, and obtain the installation gap vector boundary; An evaluation module is used to evaluate according to the installation gap vector boundary to obtain a quantitative parameter set representing the installation misalignment amount, gap width and leakage risk.

9. The template installation gap real-time detection system based on three-dimensional image reconstruction according to claim 8 is characterized in that: The inversion module includes: A first inversion unit is configured to perform dynamic pixel trajectory analysis based on the image sequence data, and obtain a sub-pixel displacement trajectory field of the template surface by extracting spatiotemporal correlation information of brightness gradients of pixels in the seam area in consecutive frames; A second inversion unit is used to perform non-rigid motion three-dimensional reconstruction based on the displacement trajectory field, and generate a three-dimensional displacement vector set reflecting the actual deformation of the template by jointly solving the perspective projection geometric constraint and the motion continuity prior; The third inversion unit is used to perform physical deformation mode association processing based on the three-dimensional displacement vector set, and to perform physical consistency correction on the displacement vector by introducing the elastic potential energy model of the template material to obtain a three-dimensional motion point cloud field that characterizes the deformation characteristics during the installation process.

10. The template installation gap real-time detection system based on three-dimensional image reconstruction according to claim 8, characterized in that: The reconstruction module includes: A first reconstruction unit is configured to perform spatial structure dependency modeling based on the three-dimensional motion point cloud field, and obtain a spatial correlation topology map containing local geometric continuity features by constructing a dynamic correlation network between point clouds based on elastic deformation constraints; A second reconstruction unit is configured to perform continuous-discontinuous feature fusion based on the spatial correlation topology, and generate a hybrid geometric expression that simultaneously characterizes the smooth area of ​​the surface and the fracture zone by decoupling the second-order differentiable function family of the continuous surface and the non-smooth operator of the joint fracture feature; The third reconstruction unit is used to perform fracture energy threshold compression according to the hybrid geometric expression, selectively strengthen the non-smooth operator area by introducing a fracture energy density function, and output a fracture sensitive surface model that is sensitive to the installation joint.

Images