Visual image three-dimensional modeling and simulation method for floating installation
By improving the feature point recognition algorithm and neural network model, the problem of poor feature point selection in floating installation was solved, the installation accuracy was improved, and the safety and quality of the installation were ensured.
Patent Information
- Application Number
- CN202511218802.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-28
- Publication Date
- 2026-01-27
AI Technical Summary
Existing 3D modeling and simulation methods suffer from poor feature point selection during the floating installation process, are easily affected by the ocean background, leading to simulation errors, affecting installation accuracy, and potentially causing safety accidents and economic losses.
An improved matching feature point recognition algorithm is adopted. Based on the corner detection method and the bilateral filtered image with Gaussian function filtering coefficient correction, the optimal feature point is identified as the origin of the three-dimensional model, and combined with the neural network model to simulate the dynamic operation of the installation method.
This improved the accuracy of the floatation installation, reduced the incidence of safety accidents and economic losses, and ensured the safety and quality of the installation.
Smart Images
Figure CN121413055A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The computer and auxiliary device of the present application belong to the technical field of visual image three-dimensional reconstruction simulation, and particularly relates to a visual image three-dimensional modeling and simulation method for float-over installation. BACKGROUND
[0002] Float-over installation refers to the installation technology that utilizes a barge to carry an upper block of a platform for offshore operation, relies on tide level, barge load adjustment and lifting mechanism to implement the lifting of the upper block, and simultaneously uses special connecting components to complete the docking operation of the upper block and the lower jacket platform. In the field of marine engineering, float-over installation is an important way for the installation of large offshore platforms, wind power facilities and the like. However, any slight mistake in the float-over installation process can lead to serious safety accidents and economic losses. In order to help engineers discover potential problems in advance and ensure the safety and quality of float-over installation, three-dimensional modeling and simulation of float-over installation need to be performed.
[0003] The existing three-dimensional modeling and simulation directly perform two-dimensional image data matching of feature points in combination with the calculation of the three-dimensional coordinates of each point in the scene based on the image parameters, generate point cloud data, and then construct a three-dimensional model to dynamically simulate the scene process that needs to be simulated and generate simulation animation and data results.
[0004] Through the above analysis, the problems and defects of the prior art are as follows:
[0005] At present, there are still great defects in the precise modeling and simulation of the complex float-over installation engineering operation process through digital and visual means. The three-dimensional float-over installation scene simulated has poor feature point selection and is extremely susceptible to graphic data disasters. In addition, the feature point matching selection used in three-dimensional modeling often uses a conventional corner point detection method, which cannot well adapt to the important feature points selected in the float-over installation with the ocean background and the huge amount of two-dimensional image data of the float-over installation scene. At the same time, in the three-dimensional modeling, the conventional feature corner point detection method has obvious defects, is very sensitive in terms of scale, does not have geometric scale invariance, and extracts pixel differences, which shows pixel-level data disaster problems for the float-over installation with the ocean as the background, makes the angular feature points clearer and less distorted, and thus forms simulation errors for the three-dimensional model construction of the installation scene, and further buries errors in details for the dynamic operation of the neural network model used in subsequent training to simulate the installation method, which further limits the accuracy of float-over installation in real marine engineering and finally leads to major safety accidents and economic losses. In order to better understand, plan and optimize the float-over installation operation and ensure the safety and quality of float-over installation, the present application proposes a visual image three-dimensional modeling and simulation method for float-over installation. SUMMARY
[0006] To solve the above technical problems, the present application provides a visual image three-dimensional modeling and simulation method for float-over installation.
[0007] In the first aspect of the present application, a visual image three-dimensional modeling and simulation method for float-over installation is provided, which comprises:
[0008] Collecting two-dimensional image data of the float-over installation scene, identifying feature points based on an improved matching feature point identification algorithm, selecting the identified feature points as the origin of the three-dimensional model to construct a three-dimensional simulation model;
[0009] Collecting point cloud data of the interfacing modules, installation site features, and environmental physical parameters in the three-dimensional simulation model, and processing to obtain float-over installation positioning features, while obtaining a first interfacing and installation simulation method input by the staff into the three-dimensional simulation model;
[0010] Constructing a simulation modeling interfacing model according to the float-over installation positioning features and the first interfacing and installation simulation method, obtaining a second interfacing and installation simulation method for float-over installation based on the simulation modeling interfacing model, and the staff planning float-over installation operations according to the second interfacing and installation simulation method.
[0011] Further, the improved matching feature point identification algorithm uses a feature corner detection method to determine the optimal feature point with the largest corner feature value, and constructs a three-dimensional simulation model with the optimal feature point as the origin of the three-dimensional model.
[0012] Further, the improved matching feature point identification algorithm uses a discriminant to determine the corner feature value, which is calculated based on a corner matrix M obtained by a method of modifying the pixel values of a bilateral filter image with a Gaussian function filter coefficient.
[0013] Further, the corner feature value is calculated based on the corner matrix M.
[0014] Further, the interfacing module point cloud data includes the three-dimensional coordinates of the upper module and the connecting components of the optimal three-dimensional simulation model constructed with the optimal feature point as the origin of the three-dimensional model, the installation site features include the force and direction of the installation site, and the environmental physical parameters include the tide level, barge loading, and wind force features, and the float-over installation positioning features are obtained by feature splicing.
[0015] Further, the simulation modeling interfacing model uses a neural network multi-classification model.
[0016] A visual image three-dimensional modeling and simulation system for float-over installation is also provided, which comprises an installation data collection module, a float-over installation scene simulation module, a simulation modeling interfacing model construction module, and an interactive display module.
[0017] The installation data acquisition module is used for acquiring two-dimensional image data of the float-over installation scene.
[0018] The float-over installation scene simulation module is used for identifying feature points based on the improved matching feature point identification algorithm, selecting the identified feature points as three-dimensional model origins to construct a three-dimensional simulation model.
[0019] The simulation modeling docking model construction module is used for acquiring docking module point cloud data, installation site features and environmental physical parameters in the three-dimensional simulation model, processing to obtain float-over installation positioning features, and obtaining a first docking installation simulation method input by a worker in the three-dimensional simulation model.
[0020] The simulation modeling docking model is constructed according to the float-over installation positioning features and the first docking installation simulation method, a second docking installation simulation method for float-over installation is obtained based on the simulation modeling docking model, and a worker plans a float-over installation operation according to the second docking installation simulation method.
[0021] The interactive display module is used for displaying three-dimensional modeling results and simulation animations, so that a user can interact with three-dimensional models and simulation scenes, such as rotating, scaling models, viewing details from different angles, and adjusting simulation parameters.
[0022] Further, the feature corner detection method is used to determine and select the optimal feature point with the maximum corner feature value as the three-dimensional model origin to construct the three-dimensional simulation model.
[0023] In the existing rehabilitation planning of burn plastic surgery, the generative adversarial network model based on regional part feature parameter generation correction coefficient is used to solve the problem that the model applicability is not enough due to the low data sample and patient privacy of the burn image. In addition, the patient and medical staff are considered at the same time, and the burn area is divided to generate images for specific parts, so that the generated data is more accurate, the automation design accuracy of burn plastic rehabilitation planning is improved, the actual deviation is reduced, and the rehabilitation method can be well adapted to the rehabilitation of nursing personnel.
[0024] The present application precisely models and simulates the complex engineering process of float-over installation through digital and visual means. The simulated three-dimensional float-over installation scene adopts a corner detection method based on multi-image filtering for feature point selection, which can well adapt to the presence of an ocean background and greatly reduce the use of two-dimensional image data of the float-over installation scene.
[0025] Meanwhile, after the three-dimensional model is constructed based on the selected specific feature point as the origin, the dynamic operation of the simulation installation method is given by using the trained neural network model, which facilitates the staff to select the simulation installation method, considers the influence of image detail parameters and related parameters of the float installation connection, such as connecting components, upper and lower assemblies and environmental physical parameters, inputs the model to obtain the simulation three-dimensional construction model of the float installation method, which improves the accuracy of the float installation, reduces the incidence of major safety accidents and economic losses, and makes contributions to better understanding, planning and optimizing the float installation operation, and ensures the safety and quality of the float installation.
[0026] More embodiments and improvement effects of the present application will be further introduced in combination with the drawings and specific embodiments. BRIEF DESCRIPTION OF DRAWINGS
[0027] Figure 1 is a visual image three-dimensional modeling and simulation method flow chart of the float installation of the present application;
[0028] Figure 2 is a visual image three-dimensional modeling and simulation method system schematic diagram of the float installation of the present application;
[0029] Figure 3 is a corner point positioning diagram in the embodiment of the present application;
[0030] Figure 4 is a float installation part simulation diagram in the embodiment of the present application;
[0031] Figure 5 is an electronic device structure schematic diagram for realizing the method of the present application in the embodiment of the present application. DETAILED DESCRIPTION
[0032] In the following, the application will be further described in combination with the drawings and specific embodiments, and the image acquisition in the present application is a color image.
[0033] As Figure 2 shown, the system of the present application belongs to visual image three-dimensional reconstruction simulation, and thus belongs to the next generation of information network technology.
[0034] In the first aspect of the present application, a visual image three-dimensional modeling and simulation method of float installation is provided, the method comprising:
[0035] Collecting two-dimensional image data of the float installation scene, and identifying feature points based on an improved matching feature point identification algorithm, selecting the identified feature points as the origin of the three-dimensional model to construct a three-dimensional simulation model;
[0036] Collect point cloud data of the docking module, installation site features and environmental physical parameters in the three-dimensional simulation model, and process to obtain the float installation positioning feature, and obtain the first docking installation simulation method input by the staff in the three-dimensional simulation model;
[0037] According to the float installation positioning feature and the first docking installation simulation method, a simulation modeling docking model is constructed, a second docking installation simulation method for float installation is obtained based on the simulation modeling docking model, and the staff plans the float installation operation according to the second docking installation simulation method.
[0038] Further, the improved matching feature point recognition algorithm adopts a feature corner detection method to determine the optimal feature point with the maximum corner feature value, and constructs a three-dimensional simulation model with the optimal feature point as the three-dimensional model origin.
[0039] Further, the improved matching feature point recognition algorithm determines the feature points by discriminant, which is calculated based on the corner point matrix M. In three-dimensional modeling, the conventional feature corner detection method has obvious shortcomings, which is very sensitive in scale and does not have geometric scale invariance. At the same time, the extracted corner pixel difference is used as the background of the float installation in the sea, which shows the pixel-level data disaster problem, so that the corner feature point is clearer and not distorted. Therefore, the method of correcting the bilateral filter image with the Gaussian function filter coefficient is adopted for corner detection, and the corner point matrix M based on the Gaussian function filter coefficient correction of the bilateral filter image is:
[0040]
[0041] In the formula, M is the corner point matrix, i and j are the offset gray values, G(x, y) is the gray value after Gaussian function filtering, I(x, y) is the gray value after bilateral filtering, x and y represent the coordinate values of the pixel in the X axis and Y axis, g X is the difference of the image gray function g in the X direction, g Y is the difference of the image gray function g in the Y direction, g XY is the difference of the image gray function g in the X direction and the Y direction.
[0042] Further, the corner feature value is calculated based on the corner point matrix M, and the corner feature value D is:
[0043]
[0044] In the formula, det(M) represents the determinant of the corner point matrix M, T(M) is the trace of the corner point matrix M, the coefficient k is a discriminant constant, and the value is 0.04, g X is the difference of the image gray function g in the X direction, g YFor the difference of the image gray function g in Y direction, when the corner point feature value D at the image pixel point is greater than 200, it is determined as a corner point, and the corner point feature value D at the image pixel point is taken as the maximum value as the optimal feature point.
[0045] Further, the G(x, y) is a gray value after Gaussian function filtering, and the Gaussian function filtering expression is:
[0046]
[0047] In the formula, x and y represent the coordinate values of the pixels in the X axis and the Y axis, and sigma is the standard deviation of the pixel image.
[0048] Further, the I(x, y) is a gray value after bilateral filtering, and the bilateral filtering expression is:
[0049]
[0050] In the formula, x and y represent the coordinate values of the pixels in the X axis and the Y axis, Omega represents the neighborhood of the filter, which is usually a square or circular window, m and m represent the pixel coordinate values of the neighborhood window, which are well known to those skilled in the art, A(x,y) represents the current pixel gray value, N L is a normalization factor for ensuring that the weight sum is 1, R(||A(x,y)-A(m,n)||) represents the gray value domain weight, represents the gray similarity weight between the current pixel and the neighborhood pixel, W(m,n) is a spatial domain weight, and represents the distance weight between the current pixel and the neighborhood pixel.
[0051] Further, the docking block point cloud data includes the upper block three-dimensional coordinates of the optimal three-dimensional simulation model constructed with the optimal feature point as the three-dimensional model origin and the three-dimensional coordinates of the connecting component, the installation site feature includes the force and direction of the installation site, the environmental physical parameters include the tide level, barge loading adjustment and wind force feature, and the floating installation positioning feature is obtained through feature splicing, and is specifically represented as (upper block three-dimensional coordinates, three-dimensional coordinates of the connecting component, force, direction, tide level, barge loading adjustment, wind force feature).
[0052] Further, the simulation modeling docking model adopts a neural network multi-classification model.
[0053] Further, the activation function of the neural network multi-classification model is represented as:
[0054]
[0055] Wherein F(v) is the activation function value, and v is the input floating installation positioning feature.
[0056] Also provided is a float-over installation visual image three-dimensional modeling and simulation system, comprising an installation data acquisition module, a float-over installation scene simulation module, a simulation modeling docking model construction module, and an interactive display module.
[0057] The installation data acquisition module is configured to acquire two-dimensional image data of a float-over installation scene.
[0058] The float-over installation scene simulation module is configured to identify feature points based on an improved matching feature point identification algorithm, select the identified feature points as three-dimensional model origins, and construct a three-dimensional simulation model.
[0059] The simulation modeling docking model construction module is configured to acquire docking module point cloud data, installation site features, and environmental physical parameters in the three-dimensional simulation model, process the float-over installation positioning features, and obtain a first docking installation simulation method input by a worker into the three-dimensional simulation model.
[0060] A simulation modeling docking model is constructed according to the float-over installation positioning features and the first docking installation simulation method, a second docking installation simulation method for float-over installation is obtained based on the simulation modeling docking model, and a worker plans a float-over installation operation according to the second docking installation simulation method.
[0061] The interactive display module is configured to display three-dimensional modeling results and simulation animations, so that a user can interact with the three-dimensional model and the simulation scene, such as rotating and scaling the model, viewing details from different angles, and adjusting simulation parameters.
[0062] Further, the feature point identification based on the improved matching feature point identification algorithm adopts a feature corner point detection method to determine optimal feature points with the largest corner feature values, and constructs a three-dimensional simulation model with the optimal feature points as three-dimensional model origins.
[0063] The present application precisely models and simulates the complex float-over installation engineering operation process through digitalization and visualization, and the simulated three-dimensional float-over installation scene adopts a corner point detection method based on multi-image filtering for feature point selection, which can well adapt to the existence of an ocean background and greatly reduce the use of float-over installation scene two-dimensional image data.
[0064] Meanwhile, after the three-dimensional model is constructed based on selecting a specific feature point as an origin, the dynamic operation of the simulation installation method is given by using the trained neural network model, so as to facilitate the staff to select the simulation installation method in a reminding mode, consider the influence of image detail parameters and related parameters of the float-over installation connection, such as connecting components, upper and lower assemblies and environmental physical parameters, input the model to obtain the float-over installation method in the simulation three-dimensional construction model, which improves the accuracy of the float-over installation, reduces the incidence of major safety accidents and economic losses, and contributes to better understanding, planning and optimization of the float-over installation operation, and ensures the safety and quality of the float-over installation.
[0065] Of course, it can be understood that each embodiment of the present application can realize one of the effects, and the combination of multiple embodiments of the present application can realize all the effects described above, but it is not required that each embodiment of the present application realizes all the advantages and effects described above, because each embodiment of the present application can constitute a separate technical solution and make one or more contributions to the prior art.
[0066] The part of the module structure of the present application which is not particularly clear is subject to the content recorded in the prior art. The prior art mentioned in the foregoing background section and the specific embodiment section of the present application can be used as a part of the present application to understand the meaning of some technical features or parameters. The protection scope of the present application is subject to the content actually recorded in the claims.
Claims
1. A method for three-dimensional modeling and simulation of visual images in floating installation, characterized in that, The method includes: Two-dimensional image data of the floating installation scene were collected, and feature points were identified based on an improved matching feature point recognition algorithm. The identified feature points were selected as the origin of the three-dimensional model to construct a three-dimensional simulation model. Collect point cloud data of docking blocks, installation location features, and environmental physical parameters from the 3D simulation model, process them to obtain floating installation positioning features, and simultaneously obtain the first docking installation simulation method input by the staff into the 3D simulation model; Based on the floating installation positioning characteristics and the first docking installation simulation method, a simulation modeling docking model is constructed. Based on the simulation modeling docking model, a second docking installation simulation method for the floating installation is obtained. The staff plans the floating installation operation according to the second docking installation simulation method.
2. The method for three-dimensional modeling and simulation of visual images for floating installation as described in claim 1, characterized in that: The improved matching feature point recognition algorithm identifies feature points by using a feature corner detection method, and selects the feature point with the largest corner feature value as the optimal feature point. The optimal feature point is then used as the origin of the three-dimensional model to construct a three-dimensional simulation model.
3. The method for three-dimensional modeling and simulation of visual images for floating installation as described in claim 2, characterized in that: The improved matching feature point recognition algorithm determines corner feature values through a discriminant, which is calculated based on the corner matrix M. The corner matrix M is obtained by correcting the pixel values of the bilaterally filtered image using Gaussian function filtering coefficients.
4. A method for three-dimensional modeling and simulation of visual images for floating installation as described in claim 2 or 3, characterized in that: The corner feature values are calculated based on the corner matrix M.
5. The method for three-dimensional modeling and simulation of visual images for floating installation as described in claim 4, characterized in that: The docking block point cloud data includes the three-dimensional coordinates of the upper block of the optimal three-dimensional simulation model constructed with the optimal feature point as the origin of the three-dimensional model, as well as the three-dimensional coordinates of the connecting components. The installation part features include the force and direction of the installation part. The environmental physical parameters include tide level, barge loading and wind characteristics. The floating installation positioning features are obtained through feature splicing.
6. A method for three-dimensional modeling and simulation of visual images for floating installation as described in claim 1 or 5, characterized in that: The simulation modeling docking model adopts a neural network multi-classification model.
7. A three-dimensional modeling and simulation system for visual images of floating installation, the system implementing the method of claim 1, comprising an installation data acquisition module, a floating installation scene simulation module, a simulation modeling docking model construction module, and an interactive display module, characterized in that: The installation data acquisition module is used to acquire two-dimensional image data of the floating installation scene; The floating installation scenario simulation module identifies feature points based on an improved matching feature point recognition algorithm, and selects the identified feature points as the origin of the three-dimensional model to construct a three-dimensional simulation model. The simulation modeling docking model construction module: collects point cloud data of docking blocks, installation location features and environmental physical parameters in the three-dimensional simulation model, processes them to obtain floating installation positioning features, and simultaneously obtains the first docking installation simulation method input by the staff into the three-dimensional simulation model; Based on the floating installation positioning characteristics and the first docking installation simulation method, a simulation modeling docking model is constructed. Based on the simulation modeling docking model, a second docking installation simulation method for the floating installation is obtained. The staff plans the floating installation operation according to the second docking installation simulation method. The interactive display module is used to display the 3D modeling results and simulation animations, enabling users to interact with the 3D model and simulation scene, such as rotating and scaling the model, viewing details from different angles, and adjusting simulation parameters.
8. The three-dimensional modeling and simulation system for floating installation visual images as described in claim 7, characterized in that: The improved matching feature point recognition algorithm identifies feature points by using a feature corner detection method, and selects the feature point with the largest corner feature value as the optimal feature point. The optimal feature point is then used as the origin of the three-dimensional model to construct a three-dimensional simulation model.