Bridge model establishment method based on multi-source point cloud data

Abstract

The application relates to a bridge model establishment method based on multi-source point cloud data, which comprises the following steps: acquiring land-based point cloud data and air-based point cloud data corrected through a global navigation satellite system; after point cloud preprocessing is performed on the land-based point cloud data and the air-based point cloud data, a bridge global point cloud model is constructed; feature points in the bridge global point cloud model are identified, and a first structure parameter of the bridge global point cloud model is calculated based on the feature points; a historical structure parameter of the bridge global point cloud model is acquired; if the deviation of the first structure parameter and the historical structure parameter exceeds a preset model enhancement threshold, enhanced point cloud data of the feature points is acquired, and the bridge global point cloud model is updated based on the enhanced point cloud data to obtain a bridge local enhanced point cloud model. By using the method, the precise fusion and intelligent analysis of multi-source point cloud data can be realized, the pain points of insufficient precision and excessively high cost in traditional bridge modeling can be solved, and an efficient technical path is provided for the digitization and intelligentization of bridge engineering.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of bridge modeling, and particularly relates to a bridge model establishing method based on multi-source point cloud data. BACKGROUND

[0002] In the field of bridge engineering, three-dimensional modeling and monitoring technology has always been the core means to ensure construction quality and safety. With the development of laser scanning technology, remote sensing technology and computer vision, point cloud data, as a carrier of high-precision three-dimensional spatial information, has gradually become an important tool for bridge design, construction and operation.

[0003] In the traditional technology, bridge modeling mainly relies on manual measurement, two-dimensional drawings and single sensor data. Manual measurement is not only inefficient, but also difficult to guarantee model accuracy, and has great limitations in complex terrain and large bridges. The two-dimensional drawing modeling method is difficult to intuitively reflect the three-dimensional structural characteristics of the bridge, and is relatively cumbersome in model updating and modification. The data collected by a single sensor can only provide limited perspective and information, which is easy to cause modeling blind area and difficult to fully reflect the complex structure of the bridge.

[0004] In addition, the existing fusion of multi-source point cloud data lacks a unified standard, and the data collected by different devices has differences in accuracy, density and coordinate system, which leads to redundancy and deviation of the spliced data, affecting the construction and analysis of the bridge point cloud model. SUMMARY

[0005] Therefore, it is necessary to provide a bridge model establishing method based on multi-source point cloud data, which can realize high-precision modeling, dynamic adaptability, intelligent structure analysis and risk warning, and efficient data model construction and model analysis.

[0006] The application provides a bridge model establishing method based on multi-source point cloud data, which comprises:

[0007] Obtaining land-based point cloud data and air-based point cloud data corrected by a global navigation satellite system;

[0008] After the land-based point cloud data and the air-based point cloud data are preprocessed, a global point cloud model of the bridge is constructed;

[0009] Identifying feature points in the global point cloud model of the bridge, and calculating a first structure parameter of the global point cloud model of the bridge based on the feature points;

[0010] Obtaining a historical structure parameter of the global point cloud model of the bridge, if a deviation between the first structure parameter and the historical structure parameter exceeds a preset model enhancement threshold, obtaining enhanced point cloud data of the feature points, and updating the global point cloud model of the bridge based on the enhanced point cloud data to obtain a local enhanced point cloud model of the bridge.

[0011] In one of the embodiments, the land-based point cloud data and the air-based point cloud data based on the geodetic coordinate system corrected by the global navigation satellite system are obtained, comprising:

[0012] A global navigation satellite system monitoring point is arranged at the bridge corresponding to the bridge model;

[0013] Global navigation satellite coordinate data of the global navigation satellite system monitoring point is obtained, and planar projection coordinate data of the global navigation satellite coordinate data is generated;

[0014] Land-based initial point cloud data and air-based initial point cloud data including the global navigation satellite system monitoring point are obtained, and monitoring point cloud coordinates of the global navigation satellite system monitoring point are extracted from the land-based initial point cloud data and the air-based initial point cloud data;

[0015] The coordinate data of the land-based initial point cloud data and the coordinate data of the air-based initial point cloud data are corrected based on the planar projection coordinate data as a reference and in combination with the monitoring point cloud coordinates, so as to obtain the land-based point cloud data and the air-based point cloud data.

[0016] In one of the embodiments, after the land-based point cloud data and the air-based point cloud data are subjected to point cloud preprocessing, a global bridge point cloud model is constructed, comprising:

[0017] The land-based point cloud data and the air-based point cloud data are subjected to point cloud filtering through a statistical outlier filtering algorithm;

[0018] Registration control points corresponding to the global navigation satellite system monitoring points are extracted from the filtered land-based point cloud data and the air-based point cloud data;

[0019] Based on the registration control points, a registration conversion model is solved by a least square method, and a preliminary registration bridge point cloud model is obtained in combination with the land-based point cloud data and the air-based point cloud data;

[0020] The preliminary registration bridge point cloud model is subjected to fine registration based on an iterative closest point algorithm, and a global bridge point cloud model is constructed.

[0021] In one of the embodiments, the preliminary registration bridge point cloud model is subjected to fine registration based on an iterative closest point algorithm, and a global bridge point cloud model is constructed, comprising:

[0022] Based on the preliminary registration bridge point cloud model, an initial source point cloud, a target point cloud, an initial rotation matrix and an initial translation vector of the iterative closest point algorithm are set;

[0023] A loss function is constructed according to distances of the registration control points in the iterative source point cloud and the target point cloud and distances of the paired point pairs;

[0024] For each point in the source point cloud of the iteration, find the nearest neighbor point in the target point cloud to construct a pair of points for this iteration;

[0025] Solve for the local optimal rotation matrix and the local optimal translation vector in this iteration, and update the source point cloud, cumulative transformation matrix and cumulative translation vector based on the local optimal rotation matrix and the local optimal translation vector;

[0026] If the difference between the loss function value of the current iteration and the loss function value of the previous iteration is less than the preset registration accuracy threshold, the iteration stops, and a global point cloud model of the bridge is constructed based on the initial source point cloud, the target point cloud, the cumulative transformation matrix of the current iteration, and the cumulative translation vector of the current iteration.

[0027] In one embodiment, the expression for the loss function is:

[0028] ;

[0029] In the formula, For the first The loss function for the next iteration. and The first The cumulative transformation matrix and cumulative translation vector of each iteration and These are the registration control point weighting coefficients and the paired point pair weighting coefficients, respectively. and These represent the total number of registration control points and the total number of paired point pairs, respectively. and The first The target point cloud in the next iteration The coordinates of the first registration control point and the first The coordinates of the nearest neighbor of the target point cloud and The first The first iteration of the source point cloud The coordinates of the first registration control point and the first The coordinates of the source point cloud points and The first The local optimal rotation matrix and local optimal translation vector for each iteration.

[0030] In one embodiment, enhanced point cloud data of feature points is acquired, and the global point cloud model of the bridge is updated based on the enhanced point cloud data to obtain a local enhanced point cloud model of the bridge, including:

[0031] The initial enhanced point cloud data including global navigation satellite system monitoring points of the feature point position is acquired, and the enhanced monitoring point cloud coordinates of the global navigation satellite system monitoring points are extracted from the enhanced point cloud data;

[0032] The enhanced point cloud coordinate data of the enhanced point cloud data is corrected based on the planar projection coordinate data as a reference and in combination with the enhanced monitoring point cloud coordinates, to obtain the enhanced point cloud data.

[0033] After the enhanced point cloud data is preprocessed, the bridge global point cloud model is updated based on the enhanced point cloud data to obtain a bridge local enhanced point cloud model.

[0034] In one of the embodiments, the bridge corresponding to the bridge global point cloud model is a corrugated steel web composite beam bridge, the feature point position includes a first wave peak point position, a first wave trough point position, a first bottom flange weld toe point position, a first corrugated web weld toe point position, a second wave peak point position, a second wave trough point position, a second bottom flange weld toe point position and a second corrugated web weld toe point position, the first structure parameter includes a corrugated steel web linear parameter, a corrugated steel web curvature parameter and a corrugated steel web texture parameter, the feature point position in the bridge global point cloud model is identified, and the first structure parameter of the bridge global point cloud model is calculated based on the feature point position, including:

[0035] The corrugated steel web outer surface point cloud region and the corrugated steel web inner surface point cloud region of the bridge global point cloud model are fitted by a plane fitting algorithm and a triangular net interpolation algorithm;

[0036] The first wave peak point position, the first wave trough point position, the first bottom flange weld toe point position and the first corrugated web weld toe point position in the corrugated steel web outer surface point cloud region are identified, and the second bottom flange weld toe point position, the second corrugated web weld toe point position, the second wave peak point position corresponding to the first wave peak point position and the second wave trough point position corresponding to the first wave trough point position in the corrugated steel web inner surface point cloud region are identified;

[0037] The corrugated steel web linear parameter is calculated based on the coordinate data of the first wave peak point position, the first wave trough point position, the first bottom flange weld toe point position, the first corrugated web weld toe point position, the second wave peak point position, the second wave trough point position, the second bottom flange weld toe point position and the second corrugated web weld toe point position;

[0038] The corrugated steel web curvature parameter is calculated based on the field point cloud coordinate data of the first wave peak point position, the first wave trough point position, the second wave peak point position and the second wave trough point position by spherical fitting calculation;

[0039] The corrugated steel web texture parameter is calculated based on the point cloud texture features of the first bottom flange weld toe point position, the first corrugated web weld toe point position, the bottom flange weld toe point position and the second corrugated web weld toe point position.

[0040] In one of the embodiments, the bridge model establishment method based on multi-source point cloud data further comprises:

[0041] The feature point calculates the second structure parameter of the bridge local enhanced point cloud model based on the bridge local enhanced point cloud model;

[0042] According to the second structure parameter and the historical structure parameter, the local deterioration parameter of the feature point position is calculated;

[0043] If the local deterioration parameter exceeds the preset deterioration evaluation threshold, the structure deterioration trend score data of the bridge corresponding to the bridge local enhanced point cloud model is generated according to the local deterioration parameter.

[0044] In one of the embodiments, the local deterioration parameter includes a corrosion volume parameter, a crack volume parameter and a deformation volume parameter, the structure deterioration trend score data includes predicted life data, structure score data and damage evaluation data, and the structure deterioration trend score data of the bridge corresponding to the bridge local enhanced point cloud model is generated according to the local deterioration parameter, including:

[0045] Based on the corrosion volume parameter, the crack volume parameter and the deformation volume parameter, the damage evaluation data is calculated;

[0046] According to the damage evaluation data, the bridge structure state represented by the bridge local enhanced point cloud model is scored to obtain the structure score data;

[0047] Based on the structure score data, the predicted life data is predicted.

[0048] In one of the embodiments, the expression of the damage evaluation data is:

[0049] ;

[0050] In the formula, is the damage evaluation data, , and are corrosion weight coefficient, crack weight coefficient and deformation weight coefficient respectively, , and are corrosion volume parameter, crack volume parameter and deformation volume parameter respectively, , and are maximum allowed corrosion volume, maximum allowed crack volume and maximum allowed deformation volume respectively.

[0051] The bridge model establishment method based on multi-source point cloud data can realize efficient fusion of multi-source heterogeneous data by integrating land-based point cloud data corrected by a global navigation satellite system (GNSS) and air-based point cloud data, eliminate coordinate deviations caused by differences in sensor installation positions and scanning angles, and construct a bridge global point cloud model with high geometric fidelity, so that the spatial accuracy of the bridge model can be significantly improved, and reliable data foundation can be provided for long-term deformation analysis in the operation and maintenance stage. By calculating the structural parameters based on the feature points, and combining with historical data comparison and analysis, abnormal deviations can be automatically identified. If the deviation exceeds the preset threshold, the system triggers local enhanced scanning, and high-resolution point cloud data of the key area can be further accurately obtained, so that early warning of potential risks can be realized, and the initiative and accuracy of bridge health monitoring can be significantly improved. The multi-source point cloud data fusion technology can be adapted to different bridge types and complex environments, has strong universality, and can be flexibly applied to existing bridge maintenance and post-disaster rapid evaluation scenes. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the embodiments or the related art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0053] Figure 1 A flowchart of a bridge model establishment method based on multi-source point cloud data provided by an embodiment of the present application is shown in the figure.

[0054] Figure 2 A flowchart of another bridge model establishment method based on multi-source point cloud data provided by an embodiment of the present application is shown in the figure.

[0055] Figure 3 A flowchart of a method for constructing a bridge global point cloud model based on an iterative closest point algorithm provided by an embodiment of the present application is shown in the figure.

[0056] Figure 4 A flowchart of a method for evaluating bridge structure deterioration trend provided by an embodiment of the present application is shown in the figure.

[0057] Figure 5 A structural diagram of a bridge model establishment system based on multi-source point cloud data provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0058] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0059] In an exemplary embodiment of the present application, as shown in Figure 1 A bridge model establishment method based on multi-source point cloud data is provided, and the present embodiment is exemplified by the method applied to a bridge model construction platform. In the present embodiment, the method comprises the following steps:

[0060] In step S101, land-based point cloud data and air-based point cloud data corrected by a global navigation satellite system are obtained.

[0061] Specifically, the bridge model construction platform can set global navigation satellite system monitoring points on the model construction target bridge and obtain positioning data of the global navigation satellite system monitoring points. The bridge model construction platform can collect land-based point cloud data and air-based point cloud data including the global navigation satellite system monitoring points, and correct the land-based point cloud data and the air-based point cloud data by combining the point cloud coordinate data of the global navigation satellite system monitoring points in the land-based point cloud data and the air-based point cloud data with the positioning data of the global navigation satellite system monitoring points.

[0062] Optionally, the land-based point cloud data can include, but is not limited to, terrestrial laser scanner point cloud data, oblique photography point cloud data and vehicle-mounted mobile scanning point cloud data.

[0063] Optionally, the air-based point cloud data can include, but is not limited to, unmanned aerial vehicle laser radar point cloud data.

[0064] Optionally, the global navigation satellite system (GNSS) can include, but is not limited to, one or more of GPS, Beidou and GLONASS. The global navigation satellite system can provide high-precision three-dimensional coordinates to correct the spatial coordinates of the point cloud data and ensure the coordinate system consistency of point clouds from different sources.

[0065] In step S102, after the land-based point cloud data and the air-based point cloud data are preprocessed, a bridge global point cloud model is constructed.

[0066] Specifically, the bridge model construction platform can construct a bridge global point cloud model after preprocessing the land-based point cloud data and the air-based point cloud data corrected by the global navigation satellite system. The point cloud preprocessing can include, but is not limited to, point cloud denoising processing, point cloud registration processing, point cloud enhancement processing and point cloud downsampling processing.

[0067] Optionally, the point cloud denoising processing can include, but is not limited to, one or more of statistical outlier filtering, radius filtering, conditional filtering, voxel filtering, Gaussian filtering, and bilateral filtering.

[0068] Optionally, the point cloud enhancement processing can include, but is not limited to, point cloud defect repair and point cloud data encoding.

[0069] In step S103, feature points in the global bridge point cloud model are identified, and a first structural parameter of the global bridge point cloud model is calculated based on the feature points.

[0070] Specifically, the bridge model construction platform can identify feature points in the global bridge point cloud model according to the spatial geometric features of the global bridge point cloud model, and calculate a first structural parameter of the global bridge point cloud model based on the feature points in the global bridge point cloud model.

[0071] Optionally, the feature points in the global bridge point cloud model can include, but are not limited to, stress concentration points, pile top points, cap corner points, pier column center points, bent cap center points, bearing pad top surface points, main beam end support points, expansion joint edge points, and truss nodes.

[0072] Optionally, the first structural parameter of the global bridge point cloud model can include, but is not limited to, mechanical structural parameters and texture structural parameters. The texture structural parameters can include, but are not limited to, texture roughness parameters and crack damage parameters.

[0073] For example, for a corrugated steel web composite beam bridge, the feature points can include first wave peak points, first wave valley points, first bottom flange weld toe points, first corrugated web weld toe points, second wave peak points, second wave valley points, second bottom flange weld toe points, and second corrugated web weld toe points, and the first structural parameters can include corrugated steel web linear parameters, corrugated steel web curvature parameters, and corrugated steel web texture parameters.

[0074] Optionally, for a corrugated steel web, the bottom flange weld toe point corresponding to the bottom flange weld toe point can be referred to as an S point. The bottom flange weld toe point is at the weld toe of the fillet weld connecting the corrugated steel web and the tensile flange on the flange side, specifically at the endpoint of the transition arc on one side of the inclined section of the web corrugated parallel section, which is one of the most significant locations of stress concentration in the corrugated steel web composite beam. The S point is a high-risk point for fatigue cracks. Once fatigue cracks appear at the S point, they will quickly spread along the flange, posing a serious threat to structural safety and significantly reducing the fatigue life of the structure.

[0075] Optionally, for the corrugated web, the corrugated web weld toe point corresponding to the corrugated web weld toe point position can be referred to as point B, the corrugated web weld toe point is located at the weld toe position of the fillet weld connecting the corrugated web and the flange on the web side, at the intersection of the web surface and the fillet weld, due to the welding process and the stress difference between the web and the flange, there is a certain stress concentration. When the bridge bears load, point B will generate additional stress due to the deformation of the web and the flange. The stress concentration at point B may cause fatigue cracks, the cracks will propagate along the web, weaken the reliability of the web and the flange connection, and further affect the overall stress performance of the structure.

[0076] In step S104, the historical structure parameters of the bridge global point cloud model are obtained. If the deviation of the first structure parameters and the historical structure parameters exceeds the preset model enhancement threshold, the enhanced point cloud data of the feature point position is obtained, and the bridge global point cloud model is updated based on the enhanced point cloud data to obtain the bridge local enhanced point cloud model.

[0077] Specifically, the bridge model construction platform can obtain the historical structure parameters of the bridge global point cloud model, and the database storing the historical structure parameters can be constructed based on the bridge model construction platform or mounted on other servers. If the deviation of the first structure parameters of the bridge global point cloud model calculated by the bridge model construction platform based on the feature point position in the bridge global point cloud model from the preset historical structure parameters exceeds the preset model enhancement threshold, the bridge model construction platform can obtain the enhanced point cloud data at the feature point position, and update the bridge global point cloud model based on the enhanced point cloud data at the feature point position to obtain the bridge local enhanced point cloud model.

[0078] Optionally, the bridge model construction platform can obtain the enhanced point cloud data at the feature point position by a structured light scanner and / or a high-precision scanner.

[0079] In the bridge model establishing method based on multi-source point cloud data, the land-based point cloud data and the air-based point cloud data are uniformly corrected by means of the global navigation satellite system, so as to construct a standardized and high-precision multi-source data fusion data base. The land-based point cloud data and the air-based point cloud data can be anchored in the same coordinate system by the correction capability of the global navigation satellite system, so as to eliminate the positioning errors and system deviations in the acquisition of different devices from the root, thereby ensuring the high precision of the multi-source data splicing, providing a complete, consistent and accurate data base for the bridge full-structure three-dimensional modeling, and further ensuring the reliability of subsequent structure parameter calculation and model updating. Through intelligent identification of feature points and dynamic tracking of structure parameters, the key feature points of the bridge can be intelligently extracted, and the first structure parameters can be calculated through three-dimensional coordinate fitting, so as to effectively solve the problems of low identification efficiency, large parameter calculation error and late structure change discovery in the traditional bridge monitoring, and provide accurate quantitative basis for early warning and maintenance decision making of the bridge disease, so as to significantly reduce the safety risk of the bridge caused by structure instability. Through directional scanning of the feature point area to obtain enhanced point cloud data, and embedding the enhanced data into the global model through the point cloud fusion algorithm, and locally reconstructing the deviation area, the updating efficiency of the bridge point cloud model can be greatly improved, the calculation time and resources can be saved, and the microstructure changes in the enhanced area can be more clearly captured, which can meet the global demand of macrostructure analysis and support the detail requirements of microdisease detection, thereby promoting the efficient implementation of bridge digital management.

[0080] In an optional embodiment of the present application, referring to Figure 1 and Figure 2 , step S101, obtaining the land-based point cloud data and the air-based point cloud data based on the geodetic coordinate system corrected by the global navigation satellite system, can include:

[0081] Step S201, setting a global navigation satellite system monitoring point at the bridge corresponding to the bridge model.

[0082] Optionally, the setting position of the global navigation satellite system monitoring point can include but is not limited to the pier, tower, anchor and main beam of the bridge corresponding to the bridge model.

[0083] Step S202, obtaining global navigation satellite coordinate data of the global navigation satellite system monitoring point, and generating plane projection coordinate data of the global navigation satellite coordinate data.

[0084] Optionally, the plane projection coordinate data of the global navigation satellite coordinate data can include but is not limited to the WGS84 coordinate system and the local engineering coordinate system.

[0085] Step S203, obtain the land-based initial point cloud data and the air-based initial point cloud data of the global navigation satellite system monitoring point, and extract the monitoring point cloud coordinates of the global navigation satellite system monitoring point from the land-based initial point cloud data and the air-based initial point cloud data.

[0086] Step S204, based on the planar projection coordinate data, correct the coordinate data of the land-based initial point cloud data and the coordinate data of the air-based initial point cloud data in combination with the monitoring point cloud coordinates, to obtain the land-based point cloud data and the air-based point cloud data.

[0087] In the above bridge model establishment method based on multi-source point cloud data, the satellite coordinate data of the monitoring point is obtained and the planar projection coordinate is generated, and then the point cloud coordinates of the monitoring point are extracted from the initial point cloud data. The physical monitoring point can be used as an intermediate connection between the satellite coordinate system and the point cloud coordinate system, which can ensure the coordinate system consistency of the land-based initial point cloud data and the air-based initial point cloud data, avoid the error accumulation that may occur when the traditional method relies on virtual feature points for coordinate matching, and enable point cloud data from different sources to be calibrated based on the same set of reference coordinates. Therefore, the coordinate system difference of the multi-source point cloud acquisition equipment can be effectively overcome, the coordinate system uniformity for multi-source point cloud data fusion is laid, and the spatial position accuracy of the fused point cloud data is ensured.

[0088] In an optional embodiment of the present application, please refer to Figure 1 and Figure 2 , after the land-based point cloud data and the air-based point cloud data are preprocessed, the step S102 of constructing the bridge global point cloud model can include:

[0089] Step S205, perform point cloud filtering on the land-based point cloud data and the air-based point cloud data by using the statistical outlier filtering algorithm.

[0090] Optionally, the statistical outlier filtering algorithm is a filtering method based on the statistical characteristics of point cloud data, which can be used to detect and remove outliers in point cloud. The statistical outlier filtering algorithm assumes that most of the point cloud data conforms to a certain statistical distribution, and the outliers deviate from this distribution. By calculating the statistical quantity of each point cloud point and comparing it with the set threshold, it is determined whether the point is an outlier. If it is an outlier, it is removed.

[0091] Step S206, extract the registration control points corresponding to the global navigation satellite system monitoring points from the filtered land-based point cloud data and air-based point cloud data.

[0092] Step S207, based on the registration control points, solve the registration conversion model by using the least square method, and combine the land-based point cloud data and the air-based point cloud data to obtain the preliminary registration bridge point cloud model.

[0093] Optionally, the least square method can be used to solve the least square solution of linear or nonlinear equations. In point cloud registration, the least square method is used to solve the parameters of the registration conversion model, by minimizing the sum of squares of residuals between the observed values and the estimated values, so that the registered point cloud data is as close as possible to the reference point cloud data.

[0094] Step S208, fine registration of the preliminary registration bridge point cloud model based on the iterative closest point algorithm is performed to construct a global bridge point cloud model.

[0095] In the above bridge model building method based on multi-source point cloud data, the outlier filtering algorithm is used to filter the land-based point cloud data and the air-based point cloud data, accurately remove outliers deviating from the statistical rule, realize the noise reduction and purification of the point cloud data, retain effective point cloud information that can correctly reflect the bridge structure, reduce the interference of invalid data on subsequent processing, avoid false deletion of edge details, and ensure the contour integrity of the bridge structure, thereby ensuring the initial data quality of the model construction.

[0096] Further, in the above bridge model building method based on multi-source point cloud data, the registration control points corresponding to the global navigation satellite system monitoring points are extracted from the filtered point cloud data, and then the registration conversion model is solved based on these control points by the least square method, which can realize the preliminary registration and fusion of multi-source point cloud data, avoid macroscopic misplacement caused by different data sources, and reduce the blindness and complexity of parameter solving based on the conversion model of known control points, thereby improving the calculation efficiency and accuracy of the bridge model construction.

[0097] Further, in the above bridge model building method based on multi-source point cloud data, the iterative closest point algorithm is used for fine registration of the bridge point cloud model, which can continuously search for the nearest neighbor points of corresponding point pairs in the point cloud, calculate and optimize the conversion matrix, gradually reduce the distance error between the point clouds, and achieve the alignment of the point cloud model until the preset precision threshold is reached, so that the structures in the land-based and air-based point clouds can be accurately overlapped, the structural overlap or gap caused by the residual error of preliminary registration is avoided, and the global bridge point cloud model constructed can truly reflect the physical form of the bridge in the overall structure and local details, thereby providing a high-precision model basis for subsequent structure parameter extraction.

[0098] In an optional embodiment of the present application, as shown in Figure 3 the fine registration of the preliminary registration bridge point cloud model based on the iterative closest point algorithm can include:

[0099] Step S301, based on the preliminary registration bridge point cloud model, setting the initial source point cloud, target point cloud, initial rotation matrix and initial translation vector of the iterative closest point algorithm.

[0100] Step S302: Construct a loss function based on the distance between the registration control points in the source point cloud and the target point cloud and the distance between the paired point pairs.

[0101] Step S303: For each source point in the source point cloud, find the nearest neighbor point in the target point cloud to construct a pair of points for this iteration.

[0102] Step S304: Solve for the local optimal rotation matrix and the local optimal translation vector of this iteration, and update the iterative source point cloud, cumulative transformation matrix, and cumulative translation vector based on the local optimal rotation matrix and the local optimal translation vector.

[0103] Step S305: If the difference between the value of the loss function in this iteration and the value of the loss function in the previous iteration is less than the preset registration accuracy threshold, stop the iteration, and construct a global point cloud model of the bridge based on the initial source point cloud, the target point cloud, the cumulative transformation matrix of this iteration, and the cumulative translation vector of this iteration.

[0104] In the bridge model establishment method based on multi-source point cloud data described above, by constructing a loss function based on the distance between registration control points and the distance between paired point pairs, the alignment degree of key control points and the matching effect of the overall point cloud can be comprehensively considered. This avoids getting trapped in a local optimum solution where the overall structure is deviated due to good matching of local point pairs. Furthermore, the high-precision coordinates of the registration control points can be used as a global calibration benchmark to avoid the accumulation of local registration errors. This ensures that each iteration moves towards improving the overall registration accuracy, making the fine registration process more targeted and effective.

[0105] In an optional embodiment of this application, the expression for the loss function can be:

[0106] ;

[0107] In the formula, For the first The loss function for the next iteration. and The first The cumulative transformation matrix of the nth iteration and the nth iteration The cumulative translation vector of the next iteration and These are the registration control point weighting coefficients and the paired point pair weighting coefficients, respectively. and These represent the total number of registration control points and the total number of paired point pairs, respectively. and The first The target point cloud in the next iteration The coordinates of the first registration control point and the first The target point cloud in the next iteration The coordinates of the nearest neighbor of the target point cloud and The first The first iteration of the source point cloud The coordinates of the first registration control point and the first The first iteration of the source point cloud The coordinates of the source point cloud points and The first The local optimal rotation matrix of the nth iteration and the nth iteration The local optimal translation vector for the next iteration.

[0108] Optional, and The first The cumulative transformation matrix of the nth iteration and the nth iteration The cumulative translation vector of the next iteration It is the second norm.

[0109] Optional, registration control point weighting coefficients It can be greater than the weight coefficient of the paired point pair. .

[0110] In an optional embodiment of this application, please refer to Figure 2 The process involves acquiring enhanced point cloud data of feature points and updating the global point cloud model of the bridge based on this enhanced point cloud data to obtain a local enhanced point cloud model of the bridge. This can include:

[0111] Specifically, the bridge model building platform can acquire initial enhanced point cloud data of feature points, including global navigation satellite system monitoring points, and extract the enhanced monitoring point cloud coordinates of global navigation satellite system monitoring points from the enhanced point cloud data.

[0112] Step S211: Using the planar projection coordinate data as a reference, and combining it with the enhanced monitoring point cloud coordinates, correct the enhanced point cloud coordinate data to obtain the enhanced point cloud data.

[0113] Step S212: After preprocessing the enhanced point cloud data, update the global point cloud model of the bridge based on the enhanced point cloud data to obtain the local enhanced point cloud model of the bridge.

[0114] In the bridge model establishment method based on multi-source point cloud data described above, by extracting the enhanced monitoring point cloud coordinates of the global navigation satellite system monitoring points from the enhanced point cloud data and correcting them based on the plane projection coordinate data, the spatial accuracy of the enhanced point cloud data can be significantly improved, and the coordinate offset in the enhanced point cloud can be corrected, thereby ensuring the geometric accuracy and coordinate system consistency of the enhanced point cloud data.

[0115] In an optional embodiment of the present application, the bridge corresponding to the bridge global point cloud model can be a corrugated steel web composite beam bridge, the feature point positions can include first wave crest point positions, first wave trough point positions, first bottom flange weld toe point positions, first corrugated web weld toe point positions, second wave crest point positions, second wave trough point positions, second bottom flange weld toe point positions, and second corrugated web weld toe point positions, and the first structure parameters can include corrugated steel web linear parameters, corrugated steel web curvature parameters, and corrugated steel web texture parameters. Please refer to Figure 1 and Figure 3 In step S103, feature point positions in the bridge global point cloud model are identified, and first structure parameters of the bridge global point cloud model are calculated based on the feature point positions. The step can include:

[0116] In step S301, the corrugated steel web outer surface point cloud region and the corrugated steel web inner surface point cloud region of the bridge global point cloud model are fitted by a plane fitting algorithm and a triangular mesh interpolation algorithm.

[0117] Optionally, the bridge model construction platform can first divide the initial point cloud into regions by a plane fitting algorithm to lock the approximate range of the corrugated steel web, and then construct a continuous curved surface model based on the discrete point cloud data by a triangular mesh interpolation algorithm to completely restore the undulating shapes of the outer surface and the inner surface.

[0118] Optionally, the plane fitting algorithm can include, but is not limited to, a least squares plane fitting algorithm, a principal component analysis plane fitting algorithm, a plane random sample consensus algorithm, a distance-based iterative closest point plane fitting algorithm, and a nearest neighbor iterative fitting algorithm.

[0119] Preferably, the plane fitting algorithm can use the plane random sample consensus algorithm.

[0120] In step S302, the first wave crest point positions, the first wave trough point positions, the first bottom flange weld toe point positions, and the first corrugated web weld toe point positions in the corrugated steel web outer surface point cloud region are identified, and the second bottom flange weld toe point positions, the second corrugated web weld toe point positions, the second wave crest point positions corresponding to the first wave crest point positions, and the second wave trough point positions corresponding to the first wave trough point positions in the corrugated steel web inner surface point cloud region are identified.

[0121] Optionally, the first wave crest point positions can be point positions constructed from the neighborhood point cloud region of the geometric center point of the wave crest region of the corrugated steel web outer surface point cloud region, the first wave trough point positions can be point positions constructed from the neighborhood point cloud region of the geometric center point of the wave trough region of the corrugated steel web outer surface point cloud region, the second wave crest point positions can be point positions constructed from the neighborhood point cloud region of the geometric center point of the wave crest region of the corrugated steel web inner surface point cloud region, and the second wave trough point positions can be point positions constructed from the neighborhood point cloud region of the geometric center point of the wave trough region of the corrugated steel web inner surface point cloud region.

[0122] Optionally, the first bottom flange weld toe point position can be a point position constructed from the point cloud region of the neighborhood of the flange side weld toe of the bottom flange and the corrugated web connecting weld of the outer surface point cloud region of the corrugated web, the first corrugated web weld toe point position can be a point position constructed from the point cloud region of the neighborhood of the web side weld toe of the bottom flange and the corrugated web connecting weld of the outer surface point cloud region of the corrugated web, the second bottom flange weld toe point position can be a point position constructed from the point cloud region of the neighborhood of the flange side weld toe of the bottom flange and the corrugated web connecting weld of the inner surface point cloud region of the corrugated web, and the second corrugated web weld toe point position can be a point position constructed from the point cloud region of the neighborhood of the web side weld toe of the bottom flange and the corrugated web connecting weld of the inner surface point cloud region of the corrugated web.

[0123] Step S303, based on the coordinate data of the first peak point position, the first valley point position, the first bottom flange weld toe point position, the first corrugated web weld toe point position, the second peak point position, the second valley point position, the second bottom flange weld toe point position and the second corrugated web weld toe point position, the corrugated web linear parameter is calculated.

[0124] Optionally, the corrugated web linear parameter can include but is not limited to wavelength parameter, wave height parameter, straight plate segment length parameter, inclined plate segment projection length parameter, inclined plate segment length parameter, web shape height parameter, weld toe spacing parameter, web thickness parameter, bending radius parameter and bending angle parameter.

[0125] Step S304, based on the field point cloud coordinate data of the first peak point position, the first valley point position, the second peak point position and the second valley point position, the spherical fitting calculation curvature calculation is performed to obtain the corrugated web curvature parameter.

[0126] Specifically, the bridge model construction platform can extract the field point cloud coordinate data of the field point cloud in the field point cloud region represented by the first peak point position, the first valley point position, the second peak point position and the second valley point position. The field point cloud coordinate data of the field point cloud in the field point cloud region represented by each of the first peak point position, the first valley point position, the second peak point position and the second valley point position is subjected to spherical fitting, and the curvature is calculated based on the spherical radius to obtain the corrugated web curvature parameter of each of the first peak point position, the first valley point position, the second peak point position and the second valley point position of the corrugated web.

[0127] Step S305, based on the point cloud texture feature of the first bottom flange weld toe point position, the first corrugated web weld toe point position, the bottom flange weld toe point position and the second corrugated web weld toe point position, the corrugated web texture parameter is calculated.

[0128] Optionally, the corrugated web texture parameter can include but is not limited to roughness parameter, reflectivity coefficient of variation parameter and texture entropy parameter.

[0129] In the bridge model establishment method based on multi-source point cloud data, the plane fitting algorithm and the triangular net interpolation algorithm are used to fit the point cloud areas of the outer surface and the inner surface of the corrugated steel web, so that the inner and outer surface boundaries of the corrugated steel web can be clearly defined, the point cloud features of different surfaces can be effectively distinguished, and a clear area division basis is provided for the accurate positioning of subsequent feature points; by identifying the first wave peak, the first wave valley and various types of weld toes, and combining the structural characteristics of the corrugated steel web to establish a structure parameter identification rule, full coverage positioning of multiple types of feature points can be realized, point omission caused by dense corrugation and subtle weld toes can be effectively avoided, and by using the coordinate data of these feature points, the linear parameters of the corrugated steel web can be further calculated to realize accurate quantification of the geometric shape of the corrugated steel web, and provide reliable data support for the geometric modeling and structural analysis of the bridge; by using the spherical fitting method to calculate the curvature of the key feature points, the local deformation of the corrugated steel web under complex stress conditions can be analyzed in depth, the local deformation degree of the corrugated steel web can be effectively evaluated, so that potential structural damage can be found in time, key data support can be provided for the health monitoring and safety evaluation of the bridge, and the safety and reliability of the bridge in the operation process can be ensured; by calculating the texture parameters of the corrugated steel web, surface defects or corrosion of the welding position can be found in time, so that corresponding maintenance measures can be taken to prolong the service life of the bridge; by constructing a dedicated parameter system including linear, curvature and texture according to the structural characteristics of the corrugated steel web, the actual needs of the corrugated steel web can be met, evaluation deviation caused by mismatch between parameters and structural characteristics can be avoided, and the calculation efficiency and accuracy of the bridge structure analysis can be improved.

[0130] In an optional embodiment of the present application, the bridge model establishment method based on multi-source point cloud data can further include:

[0131] Specifically, the bridge model construction platform can calculate the second structure parameters of the local enhanced point cloud model of the bridge based on the feature points of the local enhanced point cloud model of the bridge.

[0132] Specifically, the bridge model construction platform can calculate the local deterioration parameters of the feature points according to the second structure parameters and the historical structure parameters.

[0133] Specifically, if the local deterioration parameters exceed the preset deterioration evaluation threshold, the bridge model construction platform can generate the structure deterioration trend score data of the bridge corresponding to the local enhanced point cloud model of the bridge according to the local deterioration parameters.

[0134] In the bridge model establishment method based on multi-source point cloud data, the second structure parameter is calculated based on the feature points of the local enhanced point cloud model of the bridge, which can further refine the quantitative analysis of the local structure of the bridge, focus on the geometric features and structure state of the specific part of the bridge, so that the second structure parameter can better reflect the true situation of the local structure, help to find the small changes or potential damage that may be ignored in the overall model, and ensure the structural safety of the key parts; the local deterioration parameter of the feature point position is calculated according to the second structure parameter and the historical structure parameter, which can realize the accurate quantification of the local structure deterioration degree of the bridge, master the change trend of the local structure of the bridge in the operation process, quantify the deterioration rate and degree of the local structure, help to establish a more scientific and reasonable bridge maintenance plan, so as to improve the bridge maintenance efficiency, reduce the bridge maintenance cost, and enhance the safety and reliability of the bridge.

[0135] In an optional embodiment of the present application, the local deterioration parameter can include a corrosion volume parameter, a crack volume parameter and a deformation volume parameter, the structure deterioration trend score data can include predicted life data, structure score data and damage assessment data, and the structure deterioration trend score data corresponding to the bridge of the local enhanced point cloud model of the bridge can include:

[0136] Specifically, the bridge model construction platform can calculate the damage assessment data based on the corrosion volume parameter, the crack volume parameter and the deformation volume parameter.

[0137] Specifically, the bridge model construction platform can score the bridge structure state represented by the local enhanced point cloud model of the bridge according to the damage assessment data to obtain the structure score data.

[0138] Specifically, the bridge model construction platform can predict the predicted life data based on the structure score data.

[0139] In an optional embodiment of the present application, the expression of the damage assessment data can be:

[0140] ;

[0141] In the formula, is the damage assessment data, , and are corrosion weight coefficients, crack weight coefficients and deformation weight coefficients, respectively, , and are respectively a corrosion volume parameter, a crack volume parameter and a deformation volume parameter, , and are respectively a maximum allowed corrosion volume, a maximum allowed crack volume and a maximum allowed deformation volume.

[0142] Optionally, for a bridge built far from the sea, the crack weight coefficient may be greater than the deformation weight coefficient , and the deformation weight coefficient may be greater than the corrosion weight coefficient . For a bridge built near the sea, the corrosion weight coefficient can be greater than the crack weight coefficient , and the crack weight coefficient may be greater than the deformation weight coefficient .

[0143] In an exemplary embodiment of the present application, as shown in Figure 2 , a bridge model establishment method based on multi-source point cloud data is provided, comprising:

[0144] Step S201, setting a global navigation satellite system monitoring point at the bridge corresponding to the bridge model.

[0145] Step S202, acquiring global navigation satellite coordinate data of the global navigation satellite system monitoring point, and generating planar projection coordinate data of the global navigation satellite coordinate data.

[0146] Step S203, acquiring land-based initial point cloud data and air-based initial point cloud data including the global navigation satellite system monitoring point, and extracting monitoring point cloud coordinates of the global navigation satellite system monitoring point from the land-based initial point cloud data and the air-based initial point cloud data.

[0147] Step S204, correcting coordinate data of the land-based initial point cloud data and coordinate data of the air-based initial point cloud data based on the planar projection coordinate data and in combination with the monitoring point cloud coordinates, to obtain land-based point cloud data and air-based point cloud data.

[0148] Step S205, performing point cloud filtering on the land-based point cloud data and the air-based point cloud data through a statistical outlier filtering algorithm.

[0149] Step S206, extracting a registration control point corresponding to the global navigation satellite system monitoring point from the filtered land-based point cloud data and the air-based point cloud data.

[0150] In step S207, based on the registration control points, a registration conversion model is solved by a least square method, the land-based point cloud data and the air-based point cloud data are combined, and a preliminary registration bridge point cloud model is obtained.

[0151] In step S208, the preliminary registration bridge point cloud model is precisely registered based on an iterative closest point algorithm, and a global bridge point cloud model is constructed.

[0152] In step S209, feature points in the global bridge point cloud model are identified, and a first structure parameter of the global bridge point cloud model is calculated based on the feature points.

[0153] In step S210, a historical structure parameter of the global bridge point cloud model is obtained, if a deviation between the first structure parameter and the historical structure parameter exceeds a preset model enhancement threshold, initial enhanced point cloud data of the feature points including global navigation satellite system monitoring points are obtained, and enhanced monitoring point cloud coordinates of the global navigation satellite system monitoring points are extracted from the enhanced point cloud data.

[0154] In step S211, taking the plane projection coordinate data as a reference, the enhanced point cloud data is corrected based on the enhanced monitoring point cloud coordinates, and the enhanced point cloud data is obtained.

[0155] In step S212, after the enhanced point cloud data is preprocessed, the global bridge point cloud model is updated based on the enhanced point cloud data, and a local enhanced bridge point cloud model is obtained.

[0156] In the above bridge model establishment method based on multi-source point cloud data, by setting the global navigation satellite system monitoring points and correcting the land-based point cloud data and the air-based point cloud data, a high-precision and reliable bridge model is constructed through steps such as filtering, registration, parameter calculation and model updating, which can timely reflect the structural changes of the bridge, improve the bridge management efficiency and the decision-making scientificity, and promote the digital transformation of the bridge engineering.

[0157] It should be understood that, although each step in the flowchart involved in each embodiment as described above is displayed in sequence according to the arrow, these steps are not necessarily executed in sequence according to the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other sequences. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include multiple steps or stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or steps or stages in other steps.

[0158] Based on the same inventive concept, the embodiments of the present application also provide a multi-source point cloud data based bridge model establishment system for implementing the multi-source point cloud data based bridge model establishment method described above. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme described in the above method, so the specific limitations in one or more multi-source point cloud data based bridge model establishment system embodiments provided below can be referred to the limitations of the multi-source point cloud data based bridge model establishment method described above, which will not be repeated here.

[0159] In one exemplary embodiment, as shown in Figure 5 A multi-source point cloud data based bridge model establishment system 500 is provided, comprising:

[0160] The bridge point cloud data acquisition module 501 can be used to acquire land-based point cloud data and air-based point cloud data corrected by a global navigation satellite system.

[0161] The global point cloud model construction module 502 can be used to construct a bridge global point cloud model after point cloud preprocessing of the land-based point cloud data and the air-based point cloud data.

[0162] The bridge structure parameter generation module 503 can be used to identify feature points in the bridge global point cloud model, and calculate a first structure parameter of the bridge global point cloud model based on the feature points.

[0163] The bridge local point cloud enhancement module 504 can be used to acquire a historical structure parameter of the bridge global point cloud model, and if a deviation between the first structure parameter and the historical structure parameter exceeds a preset model enhancement threshold, acquire enhanced point cloud data of the feature points, and update the bridge global point cloud model based on the enhanced point cloud data to obtain a bridge local enhanced point cloud model.

[0164] In one optional embodiment of the present application, the bridge point cloud data acquisition module 501 can also be used to:

[0165] A global navigation satellite system monitoring point is arranged at the bridge corresponding to the bridge model.

[0166] Global navigation satellite coordinate data of the global navigation satellite system monitoring point is acquired, and planar projection coordinate data of the global navigation satellite coordinate data is generated.

[0167] Land-based initial point cloud data and air-based initial point cloud data including the global navigation satellite system monitoring point are acquired, and monitoring point cloud coordinates of the global navigation satellite system monitoring point are extracted from the land-based initial point cloud data and the air-based initial point cloud data.

[0168] The coordinate data of the initial land-based point cloud data and the coordinate data of the initial aerial-based point cloud data are corrected based on the planar projection coordinate data and in combination with the monitoring point cloud coordinates, to obtain land-based point cloud data and aerial-based point cloud data.

[0169] In an optional embodiment of the present application, the global point cloud model construction module 502 can also be used for:

[0170] The land-based point cloud data and the aerial-based point cloud data are subjected to point cloud filtering through a statistical outlier filtering algorithm.

[0171] The global navigation satellite system monitoring points corresponding to the registration control points are extracted from the filtered land-based point cloud data and the filtered aerial-based point cloud data.

[0172] Based on the registration control points, a registration conversion model is solved by a least square method, and in combination with the land-based point cloud data and the aerial-based point cloud data, a preliminary registration bridge point cloud model is obtained.

[0173] The preliminary registration bridge point cloud model is subjected to fine registration based on an iterative closest point algorithm, and a global bridge point cloud model is constructed.

[0174] In an optional embodiment of the present application, the global point cloud model construction module 502 can also be used for:

[0175] Based on the preliminary registration bridge point cloud model, an initial source point cloud, a target point cloud, an initial rotation matrix and an initial translation vector of the iterative closest point algorithm are set.

[0176] A loss function is constructed according to the distances of the registration control points in the iterative source point cloud and the target point cloud and the distances of the paired point pairs.

[0177] For each source point cloud point in the iterative source point cloud, a target point cloud nearest neighbor point in the target point cloud is found, and a paired point pair for this iteration is constructed.

[0178] A local optimal rotation matrix for this iteration and a local optimal translation vector for this iteration are solved, and the iterative source point cloud, the cumulative transformation matrix and the cumulative translation vector are updated based on the local optimal rotation matrix and the local optimal translation vector.

[0179] If the difference between the value of the loss function for this iteration and the value of the loss function for the last iteration is less than a preset registration accuracy threshold, the iteration is stopped, and a global bridge point cloud model is constructed based on the initial source point cloud, the target point cloud, the cumulative transformation matrix for this iteration and the cumulative translation vector for this iteration.

[0180] In an optional embodiment of the present application, the bridge local point cloud enhancement module 504 can also be used for:

[0181] The initial enhanced point cloud data including global navigation satellite system monitoring points is acquired, and the enhanced monitoring point cloud coordinates of the global navigation satellite system monitoring points are extracted from the enhanced point cloud data.

[0182] The enhanced point cloud coordinate data of the enhanced point cloud data is corrected based on the planar projection coordinate data and in combination with the enhanced monitoring point cloud coordinates, to obtain the enhanced point cloud data.

[0183] After the enhanced point cloud data is preprocessed, the bridge global point cloud model is updated based on the enhanced point cloud data, to obtain a bridge local enhanced point cloud model.

[0184] In an optional embodiment of the present application, the bridge structure parameter generation module 503 can also be used for:

[0185] The wave-shaped steel web outer surface point cloud region and the wave-shaped steel web inner surface point cloud region of the bridge global point cloud model are fitted through a plane fitting algorithm and a triangular net interpolation algorithm.

[0186] The first wave crest point, the first wave trough point, the first bottom flange weld toe point and the first corrugated web weld toe point in the wave-shaped steel web outer surface point cloud region are identified, and the second bottom flange weld toe point, the second corrugated web weld toe point, the second wave crest point corresponding to the first wave crest point and the second wave trough point corresponding to the first wave trough point in the wave-shaped steel web inner surface point cloud region are identified.

[0187] The wave-shaped steel web linear parameters are calculated based on the coordinate data of the first wave crest point, the first wave trough point, the first bottom flange weld toe point, the first corrugated web weld toe point, the second wave crest point, the second wave trough point, the second bottom flange weld toe point and the second corrugated web weld toe point.

[0188] The wave-shaped steel web curvature parameters are calculated based on the spherical surface fitting of the field point cloud coordinate data of the first wave crest point, the first wave trough point, the second wave crest point and the second wave trough point.

[0189] The wave-shaped steel web texture parameters are calculated based on the point cloud texture features of the first bottom flange weld toe point, the first corrugated web weld toe point, the bottom flange weld toe point and the second corrugated web weld toe point.

[0190] In an optional embodiment of the present application, the bridge model establishment system 500 based on multi-source point cloud data can also be used for:

[0191] The second structure parameters of the bridge local enhanced point cloud model are calculated based on the feature points of the bridge local enhanced point cloud model.

[0192] The local deterioration parameters of the feature points are calculated according to the second structure parameters and the historical structure parameters.

[0193] If the local deterioration parameter exceeds the preset deterioration evaluation threshold, a structure deterioration trend score data of the bridge corresponding to the bridge local reinforcement point cloud model is generated according to the local deterioration parameter.

[0194] In an optional embodiment of the present application, the bridge model establishment system 500 based on multi-source point cloud data can also be used for:

[0195] The damage evaluation data is calculated based on the corrosion volume parameter, the crack volume parameter and the deformation volume parameter.

[0196] The structure score data is obtained by scoring the bridge structure state represented by the bridge local reinforcement point cloud model according to the damage evaluation data.

[0197] The predicted service life data is predicted based on the structure score data.

[0198] In an embodiment, a computer device is provided, comprising a memory and a processor, the memory stores a computer program, and the processor implements the steps of the bridge model establishment method based on multi-source point cloud data as described above when executing the computer program.

[0199] In an embodiment, a computer readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the steps of the above method embodiments.

[0200] For the device embodiment, since it basically corresponds to the method embodiment, the related parts are described in the part of the method embodiment. The device embodiments described above are only schematic, and the components described as separate components can or can not be physically separate, and the components displayed as a unit can or can not be a physical unit, i.e. they can be located in one place, or distributed on multiple network units. According to the actual needs, some or all of the modules can be selected to achieve the purpose of the present disclosure. Those skilled in the art can understand and implement it without creative labor.

[0201] The above described embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the patent scope of the present application. It should be pointed out that for those skilled in the art, without departing from the concept of the present application, some modifications and improvements can be made, which are all within the protection scope of the present application.

Claims

1. A method for bridge model establishment based on multi-source point cloud data, characterized in that, The method comprises: obtaining land-based point cloud data and air-based point cloud data corrected by a global navigation satellite system; after point cloud preprocessing of the land-based point cloud data and the air-based point cloud data, a global bridge point cloud model is constructed; feature points in the global bridge point cloud model are identified, and a first structure parameter of the global bridge point cloud model is calculated based on the feature points; if a deviation between the first structure parameter and a historical structure parameter exceeds a preset model enhancement threshold, enhanced point cloud data of the feature points is obtained, and the global bridge point cloud model is updated based on the enhanced point cloud data to obtain a local enhanced bridge point cloud model; wherein the bridge corresponding to the global bridge point cloud model is a corrugated steel web composite beam bridge, the feature points include a first peak point, a first valley point, a first bottom flange weld toe point, a first corrugated web weld toe point, a second peak point, a second valley point, a second bottom flange weld toe point, and a second corrugated web weld toe point, the first structure parameter includes a corrugated steel web linear parameter, a corrugated steel web curvature parameter, and a corrugated steel web texture parameter, and the identification of the feature points in the global bridge point cloud model and the calculation of the first structure parameter of the global bridge point cloud model based on the feature points comprise: the corrugated steel web outer surface point cloud region and the corrugated steel web inner surface point cloud region of the global bridge point cloud model are fitted by a plane fitting algorithm and a triangular net interpolation algorithm; the first peak point, the first valley point, the first bottom flange weld toe point, and the first corrugated web weld toe point in the corrugated steel web outer surface point cloud region are identified, and the second bottom flange weld toe point, the second corrugated web weld toe point, the second peak point corresponding to the first peak point, and the second valley point corresponding to the first valley point in the corrugated steel web inner surface point cloud region are identified; the corrugated steel web linear parameter is calculated based on the coordinate data of the first peak point, the first valley point, the first bottom flange weld toe point, the first corrugated web weld toe point, the second peak point, the second valley point, the second bottom flange weld toe point, and the second corrugated web weld toe point; the corrugated steel web curvature parameter is calculated by spherical fitting based on the field point cloud coordinate data of the first peak point, the first valley point, the second peak point, and the second valley point; the corrugated steel web texture parameter is calculated based on the point cloud texture features of the first bottom flange weld toe point, the first corrugated web weld toe point, the bottom flange weld toe point, and the second corrugated web weld toe point.

2. The method of claim 1, wherein, The land-based point cloud data and the air-based point cloud data corrected by the global navigation satellite system based on the geodetic coordinate system are obtained, comprising: global navigation satellite system monitoring points are arranged at the bridge corresponding to the bridge model; Obtaining global navigation satellite coordinate data of the global navigation satellite system monitoring point, and generating planar projection coordinate data of the global navigation satellite coordinate data; Obtaining land-based initial point cloud data and air-based initial point cloud data including the global navigation satellite system monitoring point, and extracting monitoring point cloud coordinates of the global navigation satellite system monitoring point from the land-based initial point cloud data and the air-based initial point cloud data; Taking the planar projection coordinate data as a reference, combining the monitoring point cloud coordinates, correcting the coordinate data of the land-based initial point cloud data and the coordinate data of the air-based initial point cloud data, and obtaining the land-based point cloud data and the air-based point cloud data.

3. The method of claim 2, wherein, After the land-based point cloud data and the air-based point cloud data are preprocessed, a global bridge point cloud model is constructed, including: Performing point cloud filtering on the land-based point cloud data and the air-based point cloud data through a statistical outlier filtering algorithm; Extracting registration control points corresponding to the global navigation satellite system monitoring point from the filtered land-based point cloud data and air-based point cloud data; Based on the registration control points, a registration conversion model is solved by a least squares method, and a preliminary registration bridge point cloud model is obtained by combining the land-based point cloud data and the air-based point cloud data. The preliminary registration bridge point cloud model is precisely registered based on an iterative closest point algorithm to construct the global bridge point cloud model.

4. The method of claim 3, wherein, The preliminary registration bridge point cloud model is precisely registered based on an iterative closest point algorithm to construct the global bridge point cloud model, including: Based on the preliminary registration bridge point cloud model, an initial source point cloud, a target point cloud, an initial rotation matrix and an initial translation vector of the iterative closest point algorithm are set; According to the distances of the registration control points in the iterative source point cloud and the target point cloud and the distances of the paired point pairs, a loss function is constructed; For each source point cloud point in the iterative source point cloud, a target point cloud nearest neighbor point is found in the target point cloud, and the paired point pairs of this iteration are constructed; The local optimal rotation matrix of this iteration and the local optimal translation vector of this iteration are solved, and the iterative source point cloud, the cumulative transformation matrix and the cumulative translation vector are updated based on the local optimal rotation matrix and the local optimal translation vector; If the difference between the value of the loss function of this iteration and the value of the loss function of the last iteration is less than a preset registration accuracy threshold, the iteration is stopped, and the global bridge point cloud model is constructed based on the initial source point cloud, the target point cloud, the cumulative transformation matrix of this iteration and the cumulative translation vector of this iteration.

5. The method of claim 4, wherein, The expression of the loss function is: ； ； ； wherein, is the loss function for the th iteration, and are the accumulated transformation matrix and the accumulated translation vector for the th iteration, respectively, and are the registration control point weight coefficient and the pair point pair weight coefficient, respectively, and are the total number of registration control points and the total number of pair point pairs, respectively, and are the coordinates of the th registration control point in the target point cloud and the coordinates of the th nearest neighbor point of the target point cloud for the th iteration, respectively, and are the coordinates of the th registration control point in the source point cloud and the coordinates of the th source point cloud point for the th iteration, respectively, and are the local optimal rotation matrix and the local optimal translation vector for the th iteration, respectively.

6. The method of claim 2, wherein, The global navigation satellite system monitoring point is obtained, and the global navigation satellite system monitoring point is obtained. Obtaining initial enhanced point cloud data of the feature point position including the global navigation satellite system monitoring point, and extracting enhanced monitoring point cloud coordinates of the global navigation satellite system monitoring point from the enhanced point cloud data; The method further comprises: The method further comprises:

7. The method according to any one of claims 1 to 6, characterized in that, The method further comprises: The method further comprises: The method further comprises: The method further comprises:

8. The method of claim 7, wherein, The method further comprises: The method further comprises: The method further comprises: The method further comprises:

9. The method of claim 8, wherein, The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The method further comprises: The ； wherein is the damage assessment data, , and are a corrosion weight factor, a crack weight factor and a deformation weight factor, respectively, , and are the corrosion volume parameter, the crack volume parameter and the deformation volume parameter, respectively, , and are a maximum allowed corrosion volume, a maximum allowed crack volume and a maximum allowed deformation volume, respectively.

Images