Fast reconstruction method for hull segment planking based on feature splicing

Abstract

The invention belongs to the technical field of ship block construction. The invention provides a rapid reconstruction method for a ship body segment outer plate based on feature splicing, and the method comprises the following steps: S1, carrying out the coordinate conversion according to the coincident point data of original three-dimensional point cloud data and target three-dimensional point cloud data; s2, performing curved surface decomposition and characteristic value calculation and splitting on the three-dimensional point cloud data of the ship segmented outer plate; s3, according to the multiple split curved surfaces, ship body section outer plates are classified; s4, extracting local characteristic geometric parameters of the ship body section outer plate; s5, performing curved surface feature recognition on the curved plate according to the local feature geometric parameters; s6, feature points are extracted to represent local and overall curved surface information of the hull segment outer plate; and S7, according to a node interpolation algorithm, continuous splicing is carried out to form an outer plate segmented curved surface. According to the method, the out-of-tolerance position can be determined through comparison with a design model, the detection efficiency and precision of the ship body section outer plate are improved, and the method becomes a key link for promoting the zero-margin building digital transformation and upgrading of the ship.

Description

Technical Field

[0001] The invention relates to the technical field of ship segment construction, and in particular to a method for rapid reconstruction of ship segment outer plates based on feature splicing. Background Art

[0002] No-residue construction is an advanced shipbuilding process. Its core idea is to minimize or completely eliminate the margins between components or sections by using high-precision processing, assembly and measurement technology to compensate for processing and assembly errors. This process can significantly improve the construction accuracy and efficiency of ship sections and reduce construction costs. It is an important development direction of the modern shipbuilding industry.

[0003] In traditional block construction, due to the influence of factors such as processing allowance, material properties, and equipment errors, the actual size of the hull outer plate is often larger than the ideal design size. There are errors in the processes of marking, measuring, cutting, assembly, and hoisting. The accumulated errors caused by the coupling, transmission, and superposition of these processing steps lead to deviations in the accuracy of the segmentation. These errors will greatly affect the precision control of the segmented outer plate forming. The hull segmented outer plate usually has the characteristics of large scale and variable curvature. Its processing and forming quality is an important part of ship construction and is closely related to various performance indicators of the ship. Whether the error of the accuracy of the segmented outer plate can be effectively reduced and controlled will directly affect the accuracy of the segmented construction and is the key to ensuring the segmented construction without allowance.

[0004] With the advancement of digital manufacturing, when performing precision inspection on the completed segments, the incomplete and unstructured data of the measurement system makes it impossible to guarantee the assembly accuracy and the efficiency is low. There is still room for improvement in the reconstruction method of ship segments. Summary of the invention

[0005] In order to solve the problem that the incomplete and unstructured data of the measurement system in the above-mentioned prior art make it impossible to ensure the assembly accuracy and low efficiency, resulting in insufficient stability of the no-residue construction, the present invention proposes a method for rapid reconstruction of the hull segmented outer plate based on feature splicing, so as to realize the pre-modification of the curved plate, ensure the accuracy of the segmented outer plate, and improve the stability of the no-residue construction.

[0006] The present invention is achieved through the following technical solution: comprising the following steps:

[0007] S1. Scan the original three-dimensional point cloud data of the ship's section outer plate one by one based on the original coordinate system, take the midpoint of the target surface as the coordinate origin, the normal vector of the midpoint of the target surface as the Z' axis direction, the direction of the first-order fitting straight line of the shorter central curve of the surface as the X' axis direction, and the direction perpendicular to the X' axis and the Z' axis as the Y' axis direction, establish the target coordinate system, obtain the target three-dimensional point cloud data, obtain the conversion parameters, verify the conversion accuracy and implement the coordinate conversion according to the original three-dimensional point cloud data and the target three-dimensional point cloud data. The converted three-dimensional point cloud data of the ship's section outer plate;

[0008] S2, performing surface decomposition and eigenvalue calculation and splitting on the three-dimensional point cloud data of the ship segment outer plate converted in step 1 to obtain a number of split surfaces;

[0009] S3, classifying the hull segment outer plates according to the surface expression and the several split surfaces obtained in step 2;

[0010] S4, extracting local characteristic geometric parameters of the hull segment outer plate classified in step 3, wherein the local characteristic geometric parameters include the normal vector, Gaussian curvature and mean curvature of the curved plate surface;

[0011] S5, according to the local characteristic geometric parameters extracted in step 4, the curved plate is subjected to surface feature recognition to determine the type of points in the target three-dimensional point cloud data and the local surface shape;

[0012] S6, selecting characteristic points of the hull section outer plate, extracting characteristic points according to the curvature of each point on the surface, and then characterizing the local and overall surface information of the hull section outer plate;

[0013] S7, according to the node interpolation algorithm, G 1 Continuous splicing to form the segmented curved surface of the outer plate.

[0014] Furthermore, the specific steps of implementing the coordinate transformation in step S1 include:

[0015] S11, determining the conversion type according to the relative position, direction and ellipsoid parameters of the original coordinate system and the target coordinate system;

[0016] S12, collecting the coincidence point data of the original coordinate system and the target coordinate system;

[0017] S13, obtaining conversion parameters according to the coincidence point data collected in step S12 and the conversion type determined in step S11;

[0018] S14, randomly selecting a number of coincident point data to verify the conversion parameters, and obtaining the verified conversion parameters;

[0019] S15, performing coordinate transformation on the points in the original coordinate system according to the verified transformation parameters to obtain corresponding points in the target coordinate system;

[0020] S16: Check the conversion result. If there is an error, repeat steps S11 to S15.

[0021] Furthermore, the specific steps of the feature point extraction method in step S6 are as follows:

[0022] S61. Establish a point P i is the local coordinate system (u, v, w) of the center point, where the w axis is at the same angle as point P i The normal vector h i coincide with each other, and the u-axis and the v-axis are located at a point passing through P i and perpendicular to the normal vector h i on the plane;

[0023] S62, point P i All neighboring points Q around i Transform to this local coordinate system, and then use the least squares method to transform the transformed data points (u i ,v i ,w i ) is accurately fitted to construct a parabola C;

[0024] S63. Compare the parabola C at point P i The average curvature at and the preset threshold are used to filter out the candidate set of boundary points;

[0025] S64, if point P i If the minimum curvature M and the maximum curvature N of the parabola C at the point are the extreme values ​​in their respective main directions, it is determined as a boundary point;

[0026] S65, determine the feature point by finding the point where the normal vector changes greatly, and calculate P i The k different point sets {P1,P2,...,P k} as the center of gravity w of the area;

[0027] S66. Using the eigenvector of the matrix, obtain the normal vector of the local surface;

[0028] S67, calculate P i Point and its surrounding data points P i+1 The angle between the respective normal vectors and the two normal vectors is used to measure the direction difference according to the angle threshold.

[0029] Furthermore, in step S7, G 1 The specific steps of continuous splicing are as follows:

[0030] S71. Constructing mathematical space equations of two surfaces and Among them, u, v are coordinate parameters, p is the B-spline order, P i,j ,Q i,j is the characteristic mesh of the B-spline surface, is the B-spline basis function, and the corresponding node vector is set to U = V = {t0, t1, ..., t p ,...,t p+n+1}, with a repetition degree of p+1, determine the common boundary of the two surfaces β(v)=C(0,v)=S(0,v);

[0031] S72. Analyze the common boundary of two surfaces, including its shape, position and geometric constraints, and the following conditions must be met;

[0032]

[0033]

[0034] S73, adjusting the control vertices of the two surfaces to be fitted;

[0035] S74. Perform node interpolation according to the surface type and boundary continuity splicing conditions, and obtain the following formula:

[0036]

[0037] Furthermore, the plurality of split curved surfaces in step S2 include a transverse curved surface, a longitudinal curved surface and a twisted surface.

[0038] Furthermore, the characteristic value in step S2 includes the lateral bending height H x , Longitudinal bending height H y , twist height H t .

[0039] Furthermore, the lateral bending height H x , Longitudinal bending height H y , twist height H t The calculation formula is as follows:

[0040]

[0041] Among them, z′ xmax ,z′ xmin They represent the maximum and minimum values ​​of the lateral decomposition surface in the Z′ axis direction, respectively. xmid Indicates the z′ value at the center point of the lateral decomposition surface;

[0042]

[0043] Among them, z′ ymax ,z′ ymin They represent the maximum and minimum values ​​of the lateral decomposition surface in the Z′ axis direction, respectively. ymid Indicates the z′ value at the center point of the lateral decomposition surface;

[0044] H t =max{|ΔH t1 |,|ΔH t2 |,|ΔH t3 |,|ΔH t4 |}

[0045] Where ΔH ti , i=1,2,3,4 represents the height difference obtained by subtracting the z′ value of the midpoint of the surface from the z′ value of the four vertices on the torsion decomposition surface. The distortion height is the maximum value of the absolute values ​​of the four height differences.

[0046] Furthermore, in step S3, the hull segment outer plates are classified according to the following method:

[0047] The flat plate has no twisting height, transverse bending height and longitudinal bending height;

[0048] The plate with a single curvature has a twisting height greater than the lateral bending height or the longitudinal bending height and has twisting in the same direction.

[0049] The transverse bending height has the same sign as the longitudinal bending height, and the twisting height is lower than both the transverse and longitudinal bending heights, which is a sail-shaped board.

[0050] A saddle plate is one in which the transverse bending height and the longitudinal bending height have different signs and the degree of distortion is lower than both the transverse bending height and the longitudinal bending height.

[0051] Furthermore, the three-dimensional point cloud data in step S1 is collected by measuring with a three-dimensional point cloud scanner. The three-dimensional laser scanner emits a large number of laser beams to the object to be measured and receives reflected signals, calculates the three-dimensional coordinates of the surface points of the object to be measured, and records information such as reflectivity and texture to obtain three-dimensional point cloud data of the ship's segmented outer plate.

[0052] Furthermore, in step S7, G 1 The continuous splicing uses B-spline curve.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] 1. Based on three-dimensional point cloud scanning technology, the present invention proposes a method for extracting the normal vector of the boundary points of the curved plate point cloud, and creates a curved plate point cloud splicing method based on a node interpolation algorithm, which can ensure that the outer plates of the hull can be accurately docked during assembly without a large amount of on-site trimming, which can effectively improve the construction accuracy and product quality of the hull and improve the efficiency of ship segment construction.

[0055] 2. The present invention ensures the accuracy of the segmented outer plate and improves the stability of the zero-residue construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The figure is a flow chart of the method of the present invention.

[0057] Figure 2 It is the reverse modeling flow chart of the present invention.

[0058] Figure 3 It is a schematic diagram of implementing coordinate transformation in the present invention.

[0059] Figure 4 This is a diagram of a type of partially curved panel according to the present invention.

[0060] Figure 5 It is a data processing flow chart of the present invention. DETAILED DESCRIPTION

[0061] Advantages and features of the present invention will be illustrated and explained by the following non-limiting description of preferred embodiments thereof, which are given by way of example only with reference to the accompanying drawings.

[0062] like Figures 1 to 3 As shown, the present invention provides a method for rapid reconstruction of hull segment outer plates based on feature splicing, comprising the following steps:

[0063] Step 1, scan the original three-dimensional point cloud data of the ship section outer plate based on the original coordinate system one by one, take the midpoint of the target surface as the coordinate origin, the normal vector of the midpoint of the target surface as the Z' axis direction, the direction of the first-order fitting straight line of the shorter central curve of the surface as the X' axis direction, and the direction perpendicular to the X' axis and Z' axis as the Y' axis direction, establish the target coordinate system, obtain the target three-dimensional point cloud data, and obtain the conversion parameters, verify the conversion accuracy and implement the coordinate conversion according to the original three-dimensional point cloud data and the target three-dimensional point cloud data. The converted three-dimensional point cloud data of the ship section outer plate is obtained.

[0064] First, scan the original 3D point cloud data of the outer plates that make up the ship sections one by one based on the original coordinate system. Take the midpoint of the target surface as the origin of the coordinates, the normal vector of the midpoint of the surface as the Z' axis direction, the direction of the first-order fitting straight line of the shorter center curve of the surface as the X' axis direction, and the direction perpendicular to the X' axis and Z' axis as the Y' axis direction, establish the target coordinate system, and obtain the target 3D point cloud data. The target coordinate system is as follows: Figure 3 As shown. Collect a certain amount of original 3D point cloud data and target 3D point cloud data overlap point data as the basis, calculate the conversion parameters, verify the conversion accuracy and implement coordinate conversion. The acquisition of 3D point cloud data mainly adopts 3D point cloud scanner measurement. The 3D laser scanner emits a large number of laser beams to the measured object and receives the reflected signals, calculates the 3D coordinates of the surface points of the measured object, and records the reflectivity and texture information, so as to obtain the 3D point cloud data of the ship's segmented outer plate.

[0065] The coordinate transformation of 3D point cloud data includes determining the transformation type, collecting coincident point data, calculating transformation parameters, verifying transformation accuracy, implementing coordinate transformation, and checking transformation results. The specific steps are as follows:

[0066] Step 11: Determine the conversion type according to the relative position, direction and ellipsoid parameters of the original coordinate system and the target coordinate system.

[0067] Clarify the characteristics of the original coordinate system and the target coordinate system to determine which conversion method to use. The choice of method depends on factors such as the relative position, direction and ellipsoid parameters of the two coordinate systems.

[0068] Step 12: Collect the coincidence point data of the original coordinate system and the target coordinate system.

[0069] In the original coordinate system and the target coordinate system, a certain number of coincident point data need to be collected as the basis for transformation. These coincident points should have known coordinates in both coordinate systems so that the transformation parameters can be calculated later.

[0070] Step 13: Based on the coincidence point data collected in step 12 and the conversion type determined in step 11, conversion parameters are obtained.

[0071] The collected coincidence point data are used to calculate the conversion parameters according to the selected conversion method.

[0072] Step 14: randomly select a number of coincident point data to verify the conversion parameters, and obtain the verified conversion parameters.

[0073] In order to ensure the accuracy and stability of the conversion, it is necessary to use a portion of coincident point data to verify the calculated conversion parameters. By comparing the difference between the converted coordinates of the verification points and the actual coordinates, the accuracy and reliability of the conversion parameters can be evaluated.

[0074] Step 15: Perform coordinate transformation on the points in the original coordinate system according to the verified transformation parameters to obtain corresponding points in the target coordinate system.

[0075] After verifying that the transformation parameters meet the requirements, these parameters are used to transform the points in the original coordinate system to obtain the corresponding points in the target coordinate system. This step usually involves mathematical transformations and calculations, such as translation, rotation, and scaling.

[0076] Step 16: Check the conversion result. If there is an error, repeat steps 11 to 15.

[0077] Step 2: Surface decomposition and eigenvalue calculation and splitting of the three-dimensional point cloud data of the ship's segmented outer plate converted in step 1 to obtain a number of split surfaces, specifically split into transverse curved surfaces, longitudinal curved surfaces and twisted surfaces. The eigenvalues ​​include the transverse bending height H x , Longitudinal bending height H y , twist height H t , the calculation formula is as follows:

[0078]

[0079] Among them, z′ xmax ,z′ xmin They represent the maximum and minimum values ​​of the lateral decomposition surface in the Z′ axis direction, respectively. xmid Indicates the z′ value at the center point of the lateral decomposition surface.

[0080]

[0081] Among them, z′ ymax ,z′ ymin They represent the maximum and minimum values ​​of the lateral decomposition surface in the Z′ axis direction, respectively. ymid Indicates the z′ value at the center point of the lateral decomposition surface.

[0082] H t =max{|ΔH t1 |,|ΔH t2 |,|ΔH t3 |,|ΔH t4 |}

[0083] Where ΔH ti , i=1,2,3,4 represents the height difference obtained by subtracting the z′ value of the midpoint of the surface from the z′ value of the four vertices on the torsion decomposition surface. The twist height is the maximum value of the absolute values ​​of the four height differences.

[0084] Through surface decomposition and eigenvalue calculation of various types of exterior panels acquired through 3D scanning, the complex curved panel surfaces can be accurately divided into three basic forms: transverse curved surfaces, longitudinal curved surfaces, and twisted surfaces.

[0085] Step 3: Classify the hull segmented outer plates according to the surface expression and the several split surfaces obtained in step 2. The specific classification is as follows:

[0086] By decomposing the lateral curvature, longitudinal curvature and distortion of the surface, the curved plates are classified into four categories: flat plates, single curvature plates, sail-shaped plates and saddle-shaped plates.

[0087] The flat plate has no twisting height, transverse bending height and longitudinal bending height;

[0088] The plate with a single curvature has a twisting height greater than the lateral bending height or the longitudinal bending height and has twisting in the same direction.

[0089] The transverse bending height and the longitudinal bending height have the same sign, and the twisting height is lower than both the transverse and longitudinal bending heights, which is a sail-shaped board;

[0090] A saddle plate is one in which the transverse bending height and the longitudinal bending height have different signs and the degree of distortion is lower than both the transverse bending height and the longitudinal bending height.

[0091] Step 4: extracting local characteristic geometric parameters of the hull segment outer plate classified in step 3, wherein the local characteristic geometric parameters include the normal vector, Gaussian curvature and mean curvature of the curved plate surface.

[0092] The extracted geometric parameters are used to identify the geometric features of the curved plate surface and measure the concavity and convexity of the local surface, including the tangent vector in the u and v parameter directions of the curved plate surface, the unit normal vector, the maximum and minimum normal curvature, the Gaussian curvature and the mean curvature.

[0093] Step 5: Based on the local characteristic geometric parameters extracted in step 4, the surface features of the curved plate are recognized to determine the type of points in the target three-dimensional point cloud data and the local surface shape.

[0094] Surface feature recognition is to extract the required local surface features from the surface geometric model. The present invention recognizes the surface features based on the principal curvature detection method of the surface curvature characteristics.

[0095] 1) The values ​​of Gaussian curvature and mean curvature do not change with the rotation of the surface. Rotational invariance facilitates the shape characteristics of its independent observation points.

[0096] 2) Gaussian curvature is isometrically invariant, so it is an intrinsic property of the surface.

[0097] 3) The mean curvature is not isometrically invariant, so it is an extrinsic property of the surface.

[0098] 4) The sign of Gaussian curvature does not change with the direction of the surface normal vector, while the sign of mean curvature changes when the direction of the normal vector is reversed. This is because when the normal vector changes, the sign of the first type of basic formal parameters remains unchanged, but the sign of the second type of basic formal parameters changes.

[0099] 5) The sign of the Gaussian curvature and mean curvature at a point on the surface can determine the type of the point and the local surface shape.

[0100] 6) The calculation of Gaussian curvature and mean curvature does not depend on the global characteristics of the surface, but focuses on each point on the surface and its neighborhood. Gaussian curvature has local surface characteristics, and Gaussian curvature and mean curvature do not rely on the overall characteristics of the surface.

[0101] like Figure 4 As shown in the figure, in space, the two principal curvatures k1 and k2 of a surface at a specific point can reveal the local geometric features near the point, and the two curvature values ​​do not depend on the parameterization method of the surface. By analyzing the Gaussian curvature K = k1·k2 of the point, the following conclusions can be drawn: if the Gaussian curvature is less than 0, the point is a hyperbolic point; if it is equal to 0, it is a parabolic point; if it is greater than 0, it is an elliptic point. On the other hand, the sign of the mean curvature H = (k1+k2) / 2 can show the concave and convex nature of the surface at this point. Combining the sign combination of Gaussian curvature K and mean curvature H, the local shape near a point on the surface can be subdivided into eight types: plane, ridge surface, valley surface, minimal surface, peak surface, well surface, saddle ridge surface and saddle valley surface. That is, when K and H are both 0, the surface type is a plane; when K is 0 and H is greater than 0, the surface type is a ridge surface; when K is 0 and H is less than 0, the surface type is a valley surface; when K is greater than 0 and H is equal to 0, the surface does not exist; when K and H are both greater than 0, the surface type is a peak surface; when K is greater than 0 and H is less than 0, the surface type is a well surface; when K is less than 0 and H is 0, the surface type is a minimal surface; when K is less than 0 and H is greater than 0, the surface type is a saddle ridge surface; when K and H are both less than 0, the surface type is a saddle valley surface.

[0102] Step 6: Select the feature points of the hull section outer plate, extract the feature points according to the curvature of each point on the surface, and then characterize the local and overall surface information of the hull section outer plate.

[0103] The points where the geometric features of the outer panel surface change relatively greatly are taken as the feature points of the outer panel itself. The feature points are extracted according to the curvature of each point on the surface. The local and overall surface information of the three-dimensional outer panel is characterized by extracting some feature points.

[0104] The specific steps of the feature point extraction method are as follows:

[0105] Step 61: Create a point P i is the local coordinate system (u, v, w) of the center point, where the w axis is at the same angle as point P i The normal vector h i coincide with each other, and the u-axis and the v-axis are located at a point passing through P i and perpendicular to the normal vector h i on the plane;

[0106] Step 62: Point P i All neighboring points Q around i Transform to this local coordinate system, and then use the least squares method to transform the transformed data points (u i ,v i ,w i ) is accurately fitted to construct a parabola C;

[0107] Step 63: Compare the parabola C at point P i The average curvature at and the preset threshold are used to filter out the candidate set of boundary points;

[0108] Set a threshold about the mean curvature. By comparing the parabola C at point P i The average curvature at P and the preset threshold can filter out possible boundary point candidate sets. If the average curvature exceeds the set threshold, then point P i will be considered a potential boundary point.

[0109] Step 64: If point P i If the minimum curvature M and the maximum curvature N of the parabola C at the point are the extreme values ​​in their respective main directions, it is determined as a boundary point;

[0110] Further screening is required in the candidate boundary point set. That is, by checking point P i Whether the minimum curvature M and maximum curvature N of the parabola C at the point reach the extreme value in its corresponding main direction. If this condition is met, then it can be confirmed that point P i It’s a real border point.

[0111] Step 65: Determine the feature points by finding the points where the normal vector changes greatly and calculate P i The k different point sets {P1,P2,...,P k} as the center of gravity w of the area;

[0112] Step 66: Using the eigenvector of the matrix, obtain the normal vector of the local surface;

[0113] Convert the problem in step 65 into the solution of the minimum value of the expression. Then the solution process of the minimization problem is transformed into the problem of finding the minimum eigenvector of a specific matrix, where the eigenvector corresponding to the minimum eigenvalue of the matrix just represents the P of interest. i The normal vector of the local surface of the point. The specific matrix is ​​as follows:

[0114]

[0115] Among them, x i ,y i ,z i is the coordinate value of the adjacent point, w x ,w y ,w z are the centroid coordinates of adjacent points.

[0116] Step 67: Calculate P i Point and its surrounding data points P i+1 The angle between the respective normal vectors and the two normal vectors measures the direction difference according to the angle threshold.

[0117] This step requires defining a reasonable angle threshold as a screening condition.

[0118] Step 7: According to the node interpolation algorithm, perform G 1 Continuous splicing to form the segmented curved surface of the outer plate.

[0119] G 1 The continuous splicing adopts B-spline curve, which is the result of further development on the basis of Bézier curve. It introduces the extended conditions of continuous splicing of Bézier surface, thus realizing the natural extension and expansion of the continuous splicing conditions of B-spline surface.

[0120] Based on G 1 The continuous splicing algorithm specifically includes the following steps:

[0121] Step 71: Construct mathematical space equations of two surfaces and Among them, u, v are coordinate parameters, p is the B-spline order, P i,j ,Q i,j is the characteristic mesh of the B-spline surface, is the B-spline basis function, and the corresponding node vector is set to U = V = {t0, t1, ..., t p ,...,t p+n+1}, with a repetition degree of p+1, determine the common boundary of the two surfaces β(v)=C(0,v)=S(0,v);

[0122] Step 72: Analyze the common boundary of the two surfaces, including its shape, position and geometric constraints, and the following conditions must be met;

[0123]

[0124] Step 73, adjusting the control vertices of the two surfaces to be fitted;

[0125] By adjusting control vertices near a common boundary, you can reshape a surface to match another surface.

[0126] Step 74: Perform node interpolation according to the surface type and boundary continuity splicing conditions, and obtain the following formula:

[0127]

[0128]

[0129] When a surface boundary is spliced ​​with multiple surfaces at the same time, redundant constraints may appear. These redundant constraints need to be handled properly to avoid affecting the splicing results.

[0130] like Figure 5 As shown in the figure, the acquired ship outer plate point cloud and its original coordinates, geometric parameters and other data are combined with the B-spline surface stitching algorithm and the corresponding accuracy requirements to quickly reconstruct the segmented outer plate, and the model of the entire surface segment after the splicing of each curved plate can be obtained, so as to analyze and refine the processing and structural parameters and improve the processing technology, thereby improving the manufacturing quality and efficiency.

[0131] In addition to the above embodiments, the present invention may also have other implementation modes. Any technical solutions formed by equivalent replacement or equivalent transformation shall fall within the protection scope required by the present invention.

Claims

1. A method for rapid reconstruction of hull segment outer plates based on feature splicing, characterized by: The steps include: S1. Scan the original three-dimensional point cloud data of the ship's section outer plate one by one based on the original coordinate system, take the midpoint of the target surface as the coordinate origin, the normal vector of the midpoint of the target surface as the Z' axis direction, the direction of the first-order fitting straight line of the shorter central curve of the surface as the X' axis direction, and the direction perpendicular to the X' axis and the Z' axis as the Y' axis direction, establish the target coordinate system, obtain the target three-dimensional point cloud data, obtain the conversion parameters, verify the conversion accuracy and implement the coordinate conversion according to the original three-dimensional point cloud data and the target three-dimensional point cloud data. The converted three-dimensional point cloud data of the ship's section outer plate; S2, performing surface decomposition and eigenvalue calculation and splitting on the three-dimensional point cloud data of the ship segment outer plate converted in step 1 to obtain a number of split surfaces; S3, classifying the hull segment outer plates according to the surface expression and the several split surfaces obtained in step 2; S4, extracting local characteristic geometric parameters of the hull segment outer plate classified in step 3, wherein the local characteristic geometric parameters include the normal vector, Gaussian curvature and mean curvature of the curved plate surface; S5, according to the local characteristic geometric parameters extracted in step 4, the curved plate is subjected to surface feature recognition to determine the type of points in the target three-dimensional point cloud data and the local surface shape; S6, selecting feature points of the hull section outer plate, extracting feature points according to the curvature of each point on the surface, and then characterizing the local and overall surface information of the hull section outer plate; S7, according to the node interpolation algorithm, G 1 Continuous splicing to form the segmented curved surface of the outer plate.

2. The method for rapid reconstruction of hull segment outer plate based on feature splicing according to claim 1 is characterized in that: The specific steps of implementing the coordinate conversion in step S1 include: S11, determining the conversion type according to the relative position, direction and ellipsoid parameters of the original coordinate system and the target coordinate system; S12, collecting the coincidence point data of the original coordinate system and the target coordinate system; S13, obtaining conversion parameters according to the coincidence point data collected in step S12 and the conversion type determined in step S11; S14, randomly selecting a number of coincident point data to verify the conversion parameters, and obtaining the verified conversion parameters; S15, performing coordinate transformation on the points in the original coordinate system according to the verified transformation parameters to obtain corresponding points in the target coordinate system; S16. Check the conversion result. If there is an error, repeat steps S11 to S15.

3. The method for rapid reconstruction of hull segment outer plate based on feature splicing according to claim 2 is characterized in that: The specific steps of the feature point extraction method in step S6 are as follows: S61. Establish a point P i is the local coordinate system (u, v, w) of the center point, where the w axis is at the same angle as point P i The normal vector h i coincide with each other, and the u-axis and the v-axis are located at a point passing through P i and perpendicular to the normal vector h i on the plane; S62, point P i All neighboring points Q around i Transform to this local coordinate system, and then use the least squares method to transform the transformed data points (u i ,v i ,w i ) to perform accurate fitting to construct a parabola C; S63. Compare the parabola C at point P i The average curvature at and the preset threshold are used to filter out the candidate set of boundary points; S64, if point P i If the minimum curvature M and the maximum curvature N of the parabola C at the point are the extreme values ​​in their respective main directions, it is determined as a boundary point; S65, determine the feature point by finding the point where the normal vector changes greatly, and calculate P i The k different point sets {P1,P2,...,P k } as the center of gravity w of the area; S66. Using the eigenvector of the matrix, obtain the normal vector of the local surface; S67, calculate P i Point and its surrounding data points P i+1 The angle between the respective normal vectors and the two normal vectors is used to measure the direction difference according to the angle threshold.

4. The method for rapid reconstruction of hull segment outer plate based on feature splicing according to claim 3 is characterized in that: In step S7, 1 The specific steps of continuous splicing are as follows: S71. Constructing mathematical space equations of two surfaces and Among them, u, v are coordinate parameters, p is the B-spline order, P i,j ,Q i,j is the characteristic mesh of the B-spline surface, is the B-spline basis function, and the corresponding node vector is set to U = V = {t0, t1, ..., t p ,...,t p+n+1 }, with a repetition degree of p+1, determine the common boundary of the two surfaces β(v)=C(0,v)=S(0,v); S72. Analyze the common boundary of two surfaces, including its shape, position and geometric constraints, and the following conditions must be met; S73, adjusting the control vertices of the two surfaces to be fitted; S74. Perform node interpolation according to the surface type and boundary continuity splicing conditions, and obtain the following formula:

5. The method for rapid reconstruction of hull segment outer plate based on feature splicing according to claim 4 is characterized in that: The plurality of split curved surfaces in step S2 include a transverse curved surface, a longitudinal curved surface and a twisted surface.

6. The method for rapid reconstruction of hull segment outer plate based on feature splicing according to claim 4 is characterized in that: The characteristic value in step S2 includes the lateral bending height H x , Longitudinal bending height H y , twist height H t .

7. The method for rapid reconstruction of hull segment outer plate based on feature splicing according to claim 6 is characterized in that: The transverse bending height H x , Longitudinal bending height H y , twist height H t The calculation formula is as follows: Among them, z x ' max ,z′ xmin They represent the maximum and minimum values ​​of the lateral decomposition surface in the Z′ axis direction, respectively. x ' mid Indicates the z′ value at the center point of the lateral decomposition surface; Among them, z y ' max ,z y ' min They represent the maximum and minimum values ​​of the lateral decomposition surface in the Z′ axis direction, respectively. y ' mid Indicates the z′ value at the center point of the lateral decomposition surface; H t =max{|ΔH t1 |,|ΔH t2 |,ΔH t3 |,|ΔH t4 |} Where ΔH ti , i=1,2,3,4 represents the height difference obtained by subtracting the z′ value of the midpoint of the surface from the z′ value of the four vertices on the torsion decomposition surface. The twist height is the maximum value of the absolute values ​​of the four height differences.

8. The method for rapid reconstruction of hull segment outer plate based on feature splicing according to claim 4 is characterized in that: The step S3 classifies the hull segment outer plates according to the following method: The flat plate has no twisting height, transverse bending height and longitudinal bending height; The plate with a single curvature has a twist height greater than the lateral bending height or the longitudinal bending height and has twists in the same direction. The transverse bending height and the longitudinal bending height have the same sign, and the twisting height is lower than both the transverse and longitudinal bending heights, which is a sail-shaped board; A saddle plate is one in which the transverse bending height and the longitudinal bending height have different signs and the degree of distortion is lower than both the transverse bending height and the longitudinal bending height.

9. The method for rapid reconstruction of hull segment outer plate based on feature splicing according to claim 4 is characterized in that: The three-dimensional point cloud data in step S1 is collected by measuring with a three-dimensional point cloud scanner. The three-dimensional laser scanner emits a large number of laser beams to the object to be measured and receives reflected signals, calculates the three-dimensional coordinates of the surface points of the object to be measured, and records information such as reflectivity and texture to obtain three-dimensional point cloud data of the ship's segmented outer plate.

10. The method for rapid reconstruction of hull segment outer plate based on feature splicing according to claim 4 is characterized in that: In step S7, 1 The continuous splicing uses B-spline curve.

Images