A pipeline parameterized rendering method applied to a light-weight display platform

Abstract

The application discloses a pipeline parameterized rendering method applied to a light-weight display platform, relates to the technical field of pipeline parameterized rendering, and comprises the following steps: acquiring pipeline parameter data, performing contour boundary point selection by utilizing uniform angle sampling, obtaining initial contour boundary points, and optimizing and adjusting the initial contour boundary points by a step adjustment algorithm; and constructing vertex coordinates according to the optimized and adjusted contour boundary points; constructing vertex normal vectors by the pipeline parameter data, constructing vertex texture coordinates by utilizing a texture mapping method in combination with the vertex coordinates, and constructing a triangular mesh by utilizing a Delaunay triangulation algorithm; compressing the triangular mesh by utilizing a topological guide micro-mesh, obtaining compressed data, and rendering the compressed data by a visibility buffer sputtering to generate rendering data. The application can greatly reduce the data amount, improve the rendering efficiency, reduce platform resource consumption, and is more suitable for the application requirements of a light-weight graphics platform.

Description

Technical Field

[0001] This invention relates to the field of pipeline parametric rendering technology, and more specifically, to a pipeline parametric rendering method applied to a lightweight display platform. Background Technology

[0002] Currently, traditional CAD models are widely used in industrial design, equipment manufacturing, and shipbuilding engineering. However, due to their complex structure and large data volume, they typically employ high-precision boundary representation or parametric modeling methods, leading to inefficiencies and high resource consumption during storage, transmission, and rendering. This makes them difficult to directly apply to lightweight platforms such as web, mobile, or embedded systems. Therefore, it is necessary to convert these models into lightweight models and display them on appropriate platforms.

[0003] Pipeline models, as a common geometric structure in CAD modeling, are widely used in the design and simulation of various fluid systems. Currently, a known method for lightweighting pipeline models is to directly convert the solid data of the pipeline model into triangular facet data and display the conversion result on a lightweighting platform. The drawback of this method is that if the pipeline model is complex, the amount of converted triangular facet data is very large, failing to meet the requirements for lightweighting. It also increases the burden on data transmission and reduces the performance of the lightweighting platform.

[0004] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0005] To address the problems in related technologies, this invention proposes a pipeline parametric rendering method for lightweight display platforms to overcome the aforementioned technical issues in existing related technologies.

[0006] Therefore, the specific technical solution adopted by the present invention is as follows:

[0007] A pipeline parametric rendering method for lightweight display platforms, comprising the following steps:

[0008] S1. Obtain pipeline parameter data, select contour boundary points based on pipeline parameter data using uniform angle sampling to obtain initial contour boundary points, and optimize and adjust the initial contour boundary points using a step size adjustment algorithm, and construct vertex coordinates based on the optimized and adjusted contour boundary points;

[0009] S2. Construct vertex normal vectors using pipeline parameter data, and construct vertex texture coordinates using texture mapping combined with vertex coordinates. Construct triangular meshes based on vertex texture coordinates and vertex normal vectors using the Delaunay triangulation algorithm.

[0010] S3. The triangular mesh is compressed using a topology-guided micromesh to obtain compressed data, and the compressed data is rendered using visibility buffer sputtering to generate rendering data.

[0011] Furthermore, pipeline parameter data is acquired, and contour boundary points are selected based on the pipeline parameter data using uniform angle sampling to obtain initial contour boundary points. These initial contour boundary points are then optimized and adjusted using a step-size adjustment algorithm. Vertex coordinates are constructed based on the optimized contour boundary points, including:

[0012] S11. Obtain pipeline parameter data, and select contour boundary points based on pipeline parameter data by using uniform angle sampling combined with a preset distance deviation threshold to obtain initial contour boundary points;

[0013] S12. Initialize the step size adjustment algorithm parameters, and perform preliminary adjustment on the initial contour boundary points using the initialized step size adjustment algorithm to obtain the preliminary adjusted contour boundary points.

[0014] S13. Estimate the error of the preliminary adjusted contour boundary points, and dynamically adjust the step size adjustment algorithm parameters based on the error estimation results to obtain the adjusted step size adjustment algorithm.

[0015] S14. Obtain the optimized contour boundary points based on the adjusted step size adjustment algorithm, and construct vertex coordinates based on the optimized contour boundary points.

[0016] Furthermore, pipeline parameter data is acquired. Based on this data, contour boundary points are selected using uniform angle sampling combined with a preset distance deviation threshold. The initial contour boundary points include:

[0017] S111. Obtain pipeline parameter data. Based on the pipeline parameter data, use uniform angle sampling to sample the contour boundary points of the pipeline circular contour to obtain the adjacent points of the contour boundary.

[0018] S112. Calculate the deviation between the line segment where the adjacent point of the contour boundary is located and the circular contour line to obtain the deviation distance, and adjust the error between the deviation distance and the preset distance deviation threshold in combination with the preset error threshold.

[0019] S113. Based on the adjustment results, combined with the preset distance deviation threshold and pipeline parameter data, determine the maximum angle between adjacent points of the contour boundary. Based on the maximum angle of the adjacent points of the contour boundary, uniformly select the contour boundary points to obtain the initial contour boundary points.

[0020] Furthermore, the preset error threshold is the minimum value between the radius of the circular outline (a preset multiple) and the diagonal distance of the pipe enclosure (a preset multiple).

[0021] Furthermore, the step size adjustment algorithm parameters are initialized, and the initial contour boundary points are initially adjusted using the initialized step size adjustment algorithm to obtain the initially adjusted contour boundary points, including:

[0022] S121. Initialize the error weight parameters of the step size adjustment algorithm, and calculate the chord length angle error of the initial contour boundary points through the initialized step size adjustment algorithm;

[0023] S122. Based on the initialized error weight parameters, the chord length angle error result is discretized by weighted buckets to obtain the contour boundary points with bucket indices.

[0024] S123. Using the initialized step size adjustment algorithm combined with the initialized error weight parameters, perform quadtree hierarchical sampling on the contour boundary points with bucket index to obtain the preliminary adjusted contour boundary points.

[0025] Furthermore, the quadtree hierarchical sampling of the bucket-indexed contour boundary points is performed using the initialized step size adjustment algorithm combined with the initialized error weight parameters to obtain the preliminarily adjusted contour boundary points, including:

[0026] S1231. Use the initialized step size adjustment algorithm to perform four-quadrant partitioning of the contour boundary points with bucket index, and initialize the root node of the quadtree based on the four-quadrant partitioning results.

[0027] S1232. Based on the initialized quadtree root node and the initialized error weight parameters, perform quadtree parallel error detection on the quadrant corresponding to each non-empty bucket, and update the contour boundary points with bucket index according to the error detection results to obtain the preliminary adjusted contour boundary points.

[0028] Furthermore, error estimation is performed on the initially adjusted contour boundary points. Based on the error estimation results, the step size adjustment algorithm parameters are dynamically adjusted to obtain the adjusted step size adjustment algorithm, which includes:

[0029] S131. Calculate the viewing distance pixel error based on the preliminary adjusted contour boundary points and the preset viewing distance, and perform a weighted evaluation of the deviation degree of the chord length of the adjacent points of the preliminary adjusted contour boundary points according to the preset adjacent point arc length, so as to obtain the chord length offset weight.

[0030] S132. Use linear weighted combination to perform weighted fusion of view distance pixel error and chord length offset weight, and calculate the adjustment factor of step size adjustment algorithm parameters based on the weighted fusion result;

[0031] S133. The step size adjustment algorithm parameters are dynamically adjusted according to the adjustment factor of the step size adjustment algorithm parameters to obtain the adjusted step size adjustment algorithm.

[0032] Furthermore, vertex normals are constructed using pipeline parameter data, and vertex texture coordinates are constructed using texture mapping combined with vertex coordinates. Based on the vertex texture coordinates and vertex normals, a triangular mesh is constructed using the Delaunay triangulation algorithm, including:

[0033] S21. Construct vertex normal vectors using pipeline parameter data, and construct vertex texture coordinates using texture mapping combined with vertex coordinates;

[0034] S22. Based on vertex texture coordinates and vertex normal vectors, the Delaunay triangulation algorithm is used to select three non-collinear points, and the initial Delaunay triangulation is constructed according to the three-point selection results to obtain a set of triangles.

[0035] S23. Construct a spatial index structure based on vertex texture coordinates and vertex normal vectors, and use the spatial index structure to traverse the triangle set and insert vertices to obtain the reconstructed Delaunay triangulation.

[0036] S24. Perform a triangle circumcircle check on the insertion points of the reconstructed Delaunay triangulation. Based on the triangle circumcircle check results, combine the edge flip operation to adjust the triangles and obtain the triangular mesh.

[0037] Furthermore, the triangular mesh is compressed using a topology-guided micromesh to obtain compressed data, which is then rendered using visibility buffer sputtering to generate rendering data including:

[0038] S31. Perform topological analysis on the triangular mesh using a topology-guided micromesh, extract triangular feature edges and connectivity relationships based on the topological analysis results, and construct a topological connected graph based on the extraction results.

[0039] S32. Based on the topological connectivity graph, the triangular mesh is divided along the axis to obtain several triangular mesh blocks, and each triangular mesh block is compressed through micro-mesh encoding to obtain compressed data;

[0040] S33. Decode the compressed data by sputtering the visibility buffer to obtain decoded data, and perform shading calculations on the decoded data by combining multiple indirect drawing mechanisms with the lighting model to generate rendering data.

[0041] Furthermore, the compressed data is decoded through visibility buffer sputtering to obtain decoded data. Then, through a multi-stage indirect rendering mechanism combined with a lighting model, the decoded data is used to perform shading calculations, generating rendering data including:

[0042] S331. Decode the compressed data by sputtering the visibility buffer to obtain decoded data, and write the decoded data into the visibility buffer through a multiple indirect drawing mechanism to obtain a filled visibility buffer.

[0043] S332. Based on the filled visibility buffer and preset tiles, the pixels are classified to obtain the classification results. Based on the classification results and the lighting model, the shading calculation is performed to generate rendering data.

[0044] The beneficial effects of this invention are as follows:

[0045] 1. This invention achieves a significant reduction in data volume, improved rendering efficiency, and reduced platform resource consumption while ensuring visual quality and geometric features by selecting and optimizing contour boundary points, combining high-quality geometric representation with normal vectors and texture coordinates, and using a topology-guided micromesh compression and efficient rendering mechanism. This makes it more suitable for the application needs of lightweight graphics platforms such as Web and mobile devices.

[0046] 2. This invention selects initial contour boundary points by uniform angle sampling and optimizes them by combining an error-driven step size adjustment algorithm. This enables adaptive control of boundary point density, reducing redundant points while ensuring geometric accuracy, and providing a reasonable and compact geometric foundation for subsequent vertex construction.

[0047] 3. This invention constructs vertex normal vectors and texture coordinates, and generates triangular meshes based on Delaunay triangulation. This not only preserves the geometric features and topological structure of the original CAD model, but also improves the lighting performance and texture mapping quality during rendering, thereby enhancing the visual realism.

[0048] 4. This invention uses topology-guided micromesh compression technology to structurally divide and efficiently compress triangular meshes, significantly reducing data volume. Then, it uses visibility buffer sputtering technology for efficient rendering, performing lighting calculations only on visible pixels, reducing redundant rendering operations, and improving the rendering performance and resource utilization of the lightweight platform. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a flowchart of a pipeline parameterized rendering method applied to a lightweight display platform according to an embodiment of the present invention;

[0051] Figure 2 This is a triangular mesh generation diagram in practical application of a pipeline parameterized rendering method for a lightweight display platform according to an embodiment of the present invention.

[0052] Figure 3This is a schematic diagram of the contour boundary points of a pipeline parameterized rendering method applied to a lightweight display platform according to an embodiment of the present invention during practical application.

[0053] Figure 4 This is a schematic diagram of the outer rectangle of the bottom surface of a pipeline parameterized rendering method applied to a lightweight display platform according to an embodiment of the present invention in practical application.

[0054] Figure 5 This is a schematic diagram of the outer rectangle on the side of a pipeline parameterized rendering method applied to a lightweight display platform according to an embodiment of the present invention during practical application.

[0055] Figure 6 This is a schematic diagram of a cylindrical surface in practical application of a pipeline parametric rendering method for a lightweight display platform according to an embodiment of the present invention. Detailed Implementation

[0056] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention.

[0057] According to an embodiment of the present invention, a pipeline parametric rendering method is provided for use on a lightweight display platform.

[0058] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to an embodiment of the present invention, a pipeline parametric rendering method applied to a lightweight display platform includes the following steps:

[0059] S1. Obtain pipeline parameter data. Based on the pipeline parameter data, select contour boundary points using uniform angle sampling to obtain initial contour boundary points. Optimize and adjust the initial contour boundary points using a step size adjustment algorithm. Construct vertex coordinates based on the optimized contour boundary points.

[0060] Specifically, acquiring pipeline parameter data involves first extracting the 3D pipeline model from the CAD system into parametric data. In the lightweight platform, this parametric data is read and converted into rendering data in real time. Pipeline parameterization includes two description methods, as shown in Table 1 (straight pipe parameter list) and Table 2 (bend pipe parameter list).

[0061] Table 1 Straight Pipe Parameter List

[0062]

[0063] The parametric data of a straight pipe can be converted into rendering data in two parts. The straight section: Based on the annular cross-section formed by the outer and inner diameters of the pipe, a hollow pipe entity is formed by stretching along a straight line, and the surface of the entity is converted into triangular facet data. The bending section: After the bending lines are broken into polygons, multiple straight lines are formed, and these multiple straight lines are converted into triangular facet data, referring to the description of the straight section above.

[0064] Table 2 List of parameters for pipe bends

[0065]

[0066] The parametric data for bends is similar to that for straight pipes. The axis of the bend can be broken into a polygonal line, and a hollow pipe entity can be constructed using straight lines. Then, the surface of the entity can be converted into triangular face data.

[0067] Specifically, pipeline parameter data is acquired, and contour boundary points are selected based on this data using uniform angle sampling to obtain initial contour boundary points. These initial contour boundary points are then optimized and adjusted using a step-size adjustment algorithm. Vertex coordinates are constructed based on the optimized contour boundary points, including:

[0068] S11. Obtain pipeline parameter data. Based on the pipeline parameter data, use uniform angle sampling combined with a preset distance deviation threshold to select contour boundary points and obtain the initial contour boundary points.

[0069] Specifically, pipeline parameter data is acquired, and based on this data, contour boundary points are selected using uniform angle sampling combined with a preset distance deviation threshold. The initial contour boundary points include:

[0070] S111. Obtain pipeline parameter data. Based on the pipeline parameter data, use uniform angle sampling to sample the contour boundary points of the pipeline circular contour to obtain the adjacent points of the contour boundary.

[0071] S112. Calculate the deviation between the line segment where the adjacent point of the contour boundary is located and the circular contour line to obtain the deviation distance, and adjust the error between the deviation distance and the preset distance deviation threshold in combination with the preset error threshold.

[0072] Specifically, the preset error threshold is the minimum value between the radius of the circular outline (a preset multiple) and the diagonal distance of the pipe enclosure (a preset multiple).

[0073] S113. Based on the adjustment results, combined with the preset distance deviation threshold and pipeline parameter data, determine the maximum angle between adjacent points of the contour boundary. Based on the maximum angle of the adjacent points of the contour boundary, uniformly select the contour boundary points to obtain the initial contour boundary points.

[0074] Specifically, the generation of rendering data for pipeline entities involves transforming the surface of the hollow pipeline entity into triangular faces during the generation of rendering data for both types of pipeline parameterized data. For example... Figure 2 As shown, the hollow pipe contains two solid surfaces: the bottom surface of the pipe formed by two concentric circles and the cylindrical side surface of the pipe.

[0075] The conversion of the bottom face to a triangle is done in the same way as the conversion of the side face to a triangle, both involving the selection of contour boundary points and the construction of vertex coordinates.

[0076] The contour boundary points are selected by using uniform angle sampling combined with a preset distance deviation threshold. That is, based on the two circular contour boundary lines on the bottom surface, uniform contour boundary points are generated with a given precision.

[0077] like Figure 3 As shown, the radius of the preset circle is R The given accuracy value (i.e., the preset distance deviation threshold) is e The maximum deviation (i.e., deviation distance) between the line connecting two adjacent points on the contour boundary and the circle in which they lie. t Must be smaller than e To satisfy the above conditions, given... e and t The difference ∆ (i.e., the preset error threshold) is the minimum of 0.001 times the base radius and 0.0001 times the diagonal distance of the model bounding box, that is:

[0078] ;

[0079] in, Dist Represents the distance function between two points; Box min This represents the smallest point of the bounding box; Box max ∆ represents the maximum point of the bounding box; ∆ represents the radius of the preset circle. R .

[0080] Therefore, the maximum angle of the central angle formed by two adjacent points can be obtained. A max for:

[0081] .

[0082] To ensure that the central angle formed by two adjacent points is less than 1 / 2 inch, A max Take the number of points on the circular outline boundary. n for:

[0083] ;

[0084] in, FloorThis represents the floor function.

[0085] Finally, based on the parametric equation of the circle, take... n A uniform contour boundary point:

[0086] ;

[0087] in, i Values ​​[0, n -1]; P ( i () represents a point on the circle; O Indicates the origin; A i Indicates the first i The central angle (for example, if a circle is divided into 6 equal parts, then the first central angle) is... i The central angle is 60°. i ); Indicates the first direction vector; This represents the second direction vector.

[0088] S12. Initialize the step size adjustment algorithm parameters, and perform preliminary adjustment on the initial contour boundary points using the initialized step size adjustment algorithm to obtain the preliminary adjusted contour boundary points.

[0089] Specifically, the step size adjustment algorithm parameters are initialized, and the initial contour boundary points are initially adjusted using the initialized step size adjustment algorithm. The resulting initially adjusted contour boundary points include:

[0090] S121. Initialize the error weight parameters of the step size adjustment algorithm, and calculate the chord length angle error of the initial contour boundary points through the initialized step size adjustment algorithm;

[0091] S122. Based on the initialized error weight parameters, the chord length angle error result is discretized by weighted buckets to obtain the contour boundary points with bucket indices.

[0092] S123. Using the initialized step size adjustment algorithm combined with the initialized error weight parameters, perform quadtree hierarchical sampling on the contour boundary points with bucket index to obtain the preliminary adjusted contour boundary points.

[0093] Specifically, the error weight parameters of the initialization step size adjustment algorithm include the initialization of the chord length angle error threshold and the initialization of the quadtree parallel error threshold. Based on the number of initial contour boundary points and the circle radius, the initial angle step size is calculated. ,in δ To initialize the chord length angle error threshold; set the chord length error weight. α =0.5, Angle Change Error Weight β =0.3, the output contains {Δ θ ,α , β The parameter set of} is used for subsequent error assessment and sampling control.

[0094] Specifically, the chord length angle error weighted bucket discretization uses the output parameter set as input to calculate the chord length angle error segment by segment for the initial contour boundary point set. ε The calculation formula is: The ε value is linearly mapped to eight weighted buckets from 0 to 7, and the output is a set of points marked with bucket indices (i.e., contour boundary points with bucket indices), thus realizing error-driven hierarchical management.

[0095] Specifically, the quadtree hierarchical sampling of the bucket-indexed contour boundary points is performed using the initialized step size adjustment algorithm combined with the initialized error weight parameters to obtain the preliminarily adjusted contour boundary points, including:

[0096] S1231. Use the initialized step size adjustment algorithm to perform four-quadrant partitioning of the contour boundary points with bucket index, and initialize the root node of the quadtree based on the four-quadrant partitioning results.

[0097] S1232. Based on the initialized quadtree root node and the initialized error weight parameters, perform quadtree parallel error detection on the quadrant corresponding to each non-empty bucket, and update the contour boundary points with bucket index according to the error detection results to obtain the preliminary adjusted contour boundary points.

[0098] Specifically, quadtree hierarchical sampling uses outline boundary points with bucket indices as input to construct a quadtree structure within each non-empty bucket; and dynamically sets the error threshold based on the line-of-sight (i.e., initializes the quadtree parallel error threshold). τ If the parallel error of the current node's quadtree is greater than τ If the condition is met, a new contour boundary point is inserted at the midpoint of the corresponding arc segment to improve local accuracy; otherwise, the original number of contour boundary points is retained. The final output is a lightweight contour boundary point set with optimized structure and controllable error (i.e., the contour boundary points after preliminary adjustment), which is suitable for efficient rendering and interactive operation on lightweight platforms.

[0099] S13. Estimate the error of the contour boundary points after the initial adjustment, and dynamically adjust the step size adjustment algorithm parameters based on the error estimation results to obtain the adjusted step size adjustment algorithm.

[0100] Specifically, error estimation is performed on the initially adjusted contour boundary points, and the step size adjustment algorithm parameters are dynamically adjusted based on the error estimation results to obtain the adjusted step size adjustment algorithm, which includes:

[0101] S131. Calculate the viewing distance pixel error based on the preliminary adjusted contour boundary points and the preset viewing distance, and perform a weighted evaluation of the deviation degree of the chord length of the adjacent points of the preliminary adjusted contour boundary points according to the preset adjacent point arc length, so as to obtain the chord length offset weight.

[0102] S132. Use linear weighted combination to perform weighted fusion of view distance pixel error and chord length offset weight, and calculate the adjustment factor of step size adjustment algorithm parameters based on the weighted fusion result;

[0103] S133. The step size adjustment algorithm parameters are dynamically adjusted according to the adjustment factor of the step size adjustment algorithm parameters to obtain the adjusted step size adjustment algorithm.

[0104] Specifically, based on the initially adjusted contour boundary points, the viewing distance pixel error is calculated in combination with the preset viewing distance. That is, based on the viewing distance information of the current virtual viewpoint and the pipeline section, the projection error of the contour points in the screen space is estimated to obtain the viewing distance-related pixel deviation value of each point, reflecting its visual importance at different viewing distances. At the same time, based on the preset adjacent point arc length, the deviation between the actual chord length and its theoretical arc length between the initially adjusted contour boundary points is quantified. A weighted evaluation function is introduced to comprehensively consider curvature changes and geometric complexity to calculate the chord length offset weight of each side, reflecting the degree of local geometric distortion.

[0105] Specifically, a linear weighted combination is used to weight and fuse the view distance pixel error and chord length offset weights, and a fusion coefficient is set. a and b (like a =0.6, b =0.4) Construct a comprehensive error index fused = a * pixel + b * chord ,in, pixel This is the view distance pixel error. chord Assigning weights to chord length offsets to achieve synergistic optimization of visual perception and geometric fidelity; based on this fusion result, design a nonlinear mapping function (such as a sigmoid or piecewise linear function) to convert it into adjustment factors for the step size adjustment algorithm parameters. γ (0< γ ≤1), used for dynamic control of sampling density.

[0106] Specifically, based on the adjustment factor γ Dynamic scaling is applied to the initial chord length angle error threshold and the initial quadtree to cover large error areas. γ Smaller step sizes lead to smaller step sizes, more sampling points, and smaller error regions. γBy maintaining sparse sampling close to 1, the final output is an adaptively optimized step size adjustment algorithm parameter, achieving intelligent contour reconstruction that balances visual quality and data efficiency.

[0107] S14. Obtain the optimized contour boundary points based on the adjusted step size adjustment algorithm, and construct vertex coordinates based on the optimized contour boundary points.

[0108] S2. Construct vertex normal vectors using pipeline parameter data, and construct vertex texture coordinates using texture mapping combined with vertex coordinates. Construct triangular meshes based on vertex texture coordinates and vertex normal vectors using the Delaunay triangulation algorithm.

[0109] Specifically, vertex normals are constructed using pipeline parameter data, and vertex texture coordinates are constructed using texture mapping combined with vertex coordinates. Based on the vertex texture coordinates and vertex normals, a triangular mesh is constructed using the Delaunay triangulation algorithm, including:

[0110] S21. Construct vertex normal vectors using pipeline parameter data, and construct vertex texture coordinates using texture mapping combined with vertex coordinates;

[0111] S22. Based on vertex texture coordinates and vertex normal vectors, the Delaunay triangulation algorithm is used to select three non-collinear points, and the initial Delaunay triangulation is constructed according to the three-point selection results to obtain a set of triangles.

[0112] S23. Construct a spatial index structure based on vertex texture coordinates and vertex normal vectors, and use the spatial index structure to traverse the triangle set and insert vertices to obtain the reconstructed Delaunay triangulation.

[0113] S24. Perform a triangle circumcircle check on the insertion points of the reconstructed Delaunay triangulation. Based on the triangle circumcircle check results, combine the edge flip operation to adjust the triangles and obtain the triangular mesh.

[0114] Specifically, the vertex normal vector is constructed by taking the normal data of the plane containing the base, and all vertices have the same normal data. The base can be represented by the center of the base circle as the origin, and the normal vector as the axis vector of the cylinder. For example... Figure 4 As shown, to construct vertex texture coordinates, first construct the bounding rectangle of the base, taking the bottom left corner as (0, 0) and the top right corner as (1, 1). Calculate the texture coordinate values ​​based on the relative positions of the vertex coordinates and the bottom left corner. Assume the vertex's coordinates relative to the center of the circle are P(…). x , y The radius of the cylinder is R Then the relative coordinates of the vertex with respect to the lower left corner are (( R + x ) / R ,( R+ y ) / R ).

[0115] Based on the steps described above, the boundary points of the concentric circles are obtained. The Delaunay triangulation algorithm is then used to connect these boundary points to form the base triangles (i.e., a triangular mesh). The Delaunay triangulation algorithm includes the following steps: initialization, point insertion, and local optimization. Initialization involves selecting three non-collinear points as initial triangles, forming the initial Delaunay triangulation. Point insertion involves sequentially inserting the remaining points into the existing triangulations. For each insertion point, a triangle containing that point needs to be found (this can be accelerated by traversing the triangle set or using a spatial index structure). Local optimization is necessary because inserting points may violate the properties of the Delaunay triangulation. Specifically, this involves checking the circumcircle of the new triangle formed by the insertion point and the three vertices of the triangle containing it. If other points exist within the circumcircle, the triangle needs to be adjusted using operations such as edge flipping to restore the Delaunay properties. The above steps are repeated continuously, including point insertion and local optimization, until all points are inserted into the triangulation.

[0116] The vertex coordinates for converting the side face into a triangle can be obtained by referring to the same method as for the bottom face to obtain the boundary points of the circular contour of the side face. For example... Figure 6 As shown, the vertex normal vector is constructed, that is, the normal vector at the position of the vertex is calculated according to the parametric equation of the cylindrical surface. P = O + x *cos( u 1)* R + y *sin( u 1)* R + z * u 2, of which P Let be a point on the cylindrical surface. O With the origin as the point, x It is the X-direction vector. y It is the Y-axis vector. z It is the Z-direction vector. R (Radius) is the radius of the cylinder's base. u 1. u 2 represents two parameters of the cylindrical surface. u 1. Angle relative to the X-axis u 2 represents the height relative to the origin in the Z-axis direction.

[0117] Construct vertex texture coordinates from the side, i.e. Figure 5As shown, after unfolding the cylindrical surface along the generatrix, a rectangle with a length equal to the circumference of the cylinder's base and a width equal to the cylinder's height is obtained. The lower left corner is designated as (0, 0), and the upper right corner as (1, 1). Based on the relative positions of the vertex coordinates and the lower left corner, the texture coordinate values ​​are calculated. Assume the vertex is... P The parameters are u 1 and u 2. The coordinates relative to the lower left corner are ( R * u 1, u 2), the relative coordinate value is ( u 1 / 2 π , u 2 / H ),in R Let be the radius of the cylinder's base. H Let be the height of the cylinder. Connect the boundary points of the upper and lower circular outlines sequentially to form a lateral triangle.

[0118] S3. The triangular mesh is compressed using a topology-guided micromesh to obtain compressed data, and the compressed data is rendered using visibility buffer sputtering to generate rendering data.

[0119] Specifically, the triangular mesh is compressed using a topology-guided micromesh to obtain compressed data, and then rendered using visibility buffer sputtering to generate rendering data, including:

[0120] S31. Perform topological analysis on the triangular mesh using a topology-guided micromesh, extract triangular feature edges and connectivity relationships based on the topological analysis results, and construct a topological connected graph based on the extraction results.

[0121] S32. Based on the topological connectivity graph, the triangular mesh is divided along the axis to obtain several triangular mesh blocks, and each triangular mesh block is compressed through micro-mesh encoding to obtain compressed data;

[0122] S33. Decode the compressed data by sputtering the visibility buffer to obtain decoded data, and perform shading calculations on the decoded data by combining multiple indirect drawing mechanisms with the lighting model to generate rendering data.

[0123] Specifically, the compressed data is decoded through visibility buffer sputtering to obtain decoded data, and then the decoded data is used for shading calculations through a multi-stage indirect rendering mechanism combined with a lighting model to generate rendering data, including:

[0124] S331. Decode the compressed data by sputtering the visibility buffer to obtain decoded data, and write the decoded data into the visibility buffer through a multiple indirect drawing mechanism to obtain a filled visibility buffer.

[0125] S332. Based on the filled visibility buffer and preset tiles, the pixels are classified to obtain the classification results. Based on the classification results and the lighting model, the shading calculation is performed to generate rendering data.

[0126] Specifically, the pipeline parametric rendering method stores the 3D pipeline model as parametric data. During rendering, it uses topology-guided micromesh to compress the triangular mesh, obtaining compressed data. This compressed data is then rendered using visibility buffer sputtering to generate the final rendering data. Compared to the traditional method of directly storing triangular faces, this significantly reduces the amount of data stored, meeting the requirements for a lightweight model without compromising rendering efficiency.

[0127] Specifically, the parametric description of a straight pipe fully expresses its parametric information through its outer diameter, inner diameter, bend radius, starting coordinate system, starting length, bend point, ending length, and ending coordinate system. The rendering data conversion for straight pipe parameters involves decomposing the straight pipe into multiple standard hollow pipe entities and converting them into triangular facet data. Similarly, the parametric description of a bent pipe fully expresses its parametric information through its outer diameter, inner diameter, starting coordinate system, starting length, polyline point, ending length, and ending coordinate system. The rendering data conversion for bent pipe parameters involves decomposing the bent pipe into multiple standard hollow pipe entities and converting them into triangular facet data.

[0128] Specifically, the original triangular mesh is used as input. First, a topological analysis is performed on the triangular mesh based on the micro-mesh structure to identify the shared edges and vertex relationships between each triangle and its neighboring triangles. Then, key geometric features are extracted, including feature edges (such as boundary edges, sharp edges, and high curvature edges) and the adjacency connections between triangles. Based on the extracted topological connectivity information, a topological connectivity graph is constructed to express the structural connectivity and local geometric characteristics of the entire triangular network. The output topological connectivity graph serves as the basis for subsequent partitioning and compression.

[0129] Specifically, the topological connectivity graph is used as input. Combined with geometric spatial distribution information, the triangular network is divided along the main axis to ensure that each partitioned region is topologically locally connected and geometrically compact. The partitioning result is a set of multiple local triangles, called triangular blocks, with each triangular block containing approximately 128 triangles. Subsequently, each triangular block is compressed using a microgrid encoder, requiring only 32 bytes per triangle. The final output is structured compressed data suitable for efficient transmission and fast decoding on the GPU.

[0130] Specifically, the filled visibility buffer serves as input. Pixels are categorized according to their tile size based on screen space, with each tile typically being 16×16 or 32×32 pixels. The output is a list of visible triangles within each tile and their corresponding geometric attribute indices, used for subsequent fragment shading. Using the tile classification results as input, the fragment shader interpolates the geometric attributes of the current pixel from the global vertex buffer. Simultaneously, it prepares the input parameters needed for lighting calculations by combining viewpoint direction, directional light, and ambient light. Using geometric attributes and lighting parameters as input, it performs lighting calculations using physically based rendering models (such as PBR) or classic lighting models (such as Phong, Blinn-Phong), sequentially calculating diffuse, specular, and ambient occlusion terms. Material properties (such as albedo, roughness, and metallicity) are then combined to shade each pixel. The lighting calculation results are written to the frame buffer, outputting a final rendered image (i.e., rendering data) with realistic lighting effects, completing the full conversion from geometric data to a visual image.

[0131] In summary, by utilizing the above-mentioned technical solutions of this invention, the present invention achieves significant reduction in data volume, improved rendering efficiency, and reduced platform resource consumption while ensuring visual quality and geometric features, through contour boundary point selection and optimization, high-quality geometric representation combining normal vectors and texture coordinates, and a topology-guided micromesh compression and efficient rendering mechanism. This makes it more suitable for lightweight graphics platforms such as web and mobile devices. Furthermore, the present invention selects initial contour boundary points through uniform angle sampling and optimizes them using an error-driven step size adjustment algorithm, which adaptively controls the boundary point density, reducing redundant points while maintaining geometric accuracy. This invention provides a structurally sound and data-compact geometric foundation for subsequent vertex construction. By constructing vertex normals and texture coordinates and generating triangular meshes based on Delaunay triangulation, it not only preserves the geometric features and topological structure of the original CAD model but also improves the lighting performance and texture mapping quality during rendering, enhancing visual realism. This invention uses topology-guided micromesh compression technology to structurally divide and efficiently compress the triangular mesh, significantly reducing data volume. Then, it uses visibility buffer sputtering technology for efficient rendering, performing lighting calculations only on visible pixels, reducing redundant rendering operations, and improving the rendering performance and resource utilization of the lightweight platform.

[0132] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for pipeline parameterized rendering applied to a lightweight display platform, characterized in that, The method comprises the following steps: S1, acquiring pipeline parameter data, using uniform angle sampling based on the pipeline parameter data to select contour boundary points to obtain initial contour boundary points, and optimizing and adjusting the initial contour boundary points through a step adjustment algorithm, and constructing vertex coordinates according to the optimized and adjusted contour boundary points; The S1 comprises: S11, acquiring pipeline parameter data, using uniform angle sampling combined with a preset distance deviation threshold based on the pipeline parameter data to select contour boundary points to obtain initial contour boundary points; S12, initializing step adjustment algorithm parameters, and preliminarily adjusting the initial contour boundary points through the initialized step adjustment algorithm to obtain preliminarily adjusted contour boundary points; S13, error estimation is performed on the preliminarily adjusted contour boundary points, the step adjustment algorithm parameters are dynamically adjusted based on the error estimation result to obtain adjusted step adjustment algorithm; S14, obtaining optimized and adjusted contour boundary points based on the adjusted step adjustment algorithm, and constructing vertex coordinates according to the optimized and adjusted contour boundary points; The S12 comprises: S121, initializing the error weight parameter of the step adjustment algorithm, and calculating the chord length angle error of the initial contour boundary points through the initialized step adjustment algorithm; S122, weight bucket discretization is performed on the chord length angle error result based on the initialized error weight parameter to obtain contour boundary points with bucket indexes; S123, using the initialized step adjustment algorithm combined with the initialized error weight parameter to perform quadtree hierarchical sampling on the contour boundary points with bucket indexes to obtain preliminarily adjusted contour boundary points; S2, constructing vertex normal vectors through the pipeline parameter data, and constructing vertex texture coordinates using texture mapping combined with the vertex coordinates, and constructing triangular meshes using the Delaunay triangulation algorithm based on the vertex texture coordinates and the vertex normal vectors; S3, using topology-guided micro-mesh to compress the triangular meshes to obtain compressed data, and rendering the compressed data through a visibility buffer to generate rendering data.

2. The method of claim 1, wherein, The step of acquiring pipeline parameter data, using uniform angle sampling combined with a preset distance deviation threshold based on the pipeline parameter data to select contour boundary points to obtain initial contour boundary points comprises: S111, acquiring pipeline parameter data, using uniform angle sampling based on the pipeline parameter data to sample contour boundary points on a pipeline circular contour to obtain contour boundary adjacent points; S112, calculating the deviation between the line segment on which the contour boundary adjacent points are located and the circular contour line to obtain a deviation distance, and adjusting the error between the deviation distance and the preset distance deviation threshold combined with a preset error threshold; S113, determining the maximum angle between the contour boundary adjacent points according to the adjustment result combined with the preset distance deviation threshold and the pipeline parameter data, and uniformly selecting contour boundary points based on the maximum angle of the contour boundary adjacent points to obtain initial contour boundary points.

3. The method of claim 2, wherein, The preset error threshold is the minimum value between the radius of the circular contour line multiplied by a preset multiple and the diagonal distance of the pipeline bounding box multiplied by a preset multiple.

4. The method of claim 1, wherein, The step length adjustment algorithm after initialization is combined with the initialized error weight parameter to perform quadtree hierarchical sampling on the profile boundary points with barrel indexing, to obtain the profile boundary points after preliminary adjustment, including: S1231, using the step length adjustment algorithm after initialization to perform four-quadrant division on the profile boundary points with barrel indexing, and initializing the quadtree root node based on the four-quadrant division result; S1232, based on the initialized quadtree root node, combining the initialized error weight parameter to perform quadtree parallel error detection on the corresponding quadrants of each non-empty barrel, and performing profile boundary point updating on the profile boundary points with barrel indexing according to the error detection result, to obtain the profile boundary points after preliminary adjustment.

5. The method of claim 1, wherein, The step length adjustment algorithm after initialization is combined with the initialized error weight parameter to perform quadtree hierarchical sampling on the profile boundary points with barrel indexing, to obtain the profile boundary points after preliminary adjustment, including: S131, based on the profile boundary points after preliminary adjustment, combining the preset visual distance to calculate the visual distance pixel error, and according to the preset adjacent point arc length, performing weighted evaluation on the deviation degree of the adjacent point chord length of the profile boundary points after preliminary adjustment, to obtain the chord length deviation weight; S132, using linear weighted combination to weight and fuse the visual distance pixel error and the chord length deviation weight, and calculating the adjustment factor of the step length adjustment algorithm parameter according to the weighted fusion result; S133, dynamically adjusting the step length adjustment algorithm parameter according to the adjustment factor of the step length adjustment algorithm parameter, to obtain the adjusted step length adjustment algorithm.

6. The method of claim 1, wherein, The vertex normal vector is constructed through the pipeline parameter data, and the vertex texture coordinate is constructed by using the texture mapping method combined with the vertex coordinate, and the triangular mesh is constructed by using the Delaunay triangulation algorithm based on the vertex texture coordinate and the vertex normal vector, including: S21, constructing the vertex normal vector through the pipeline parameter data, and constructing the vertex texture coordinate by using the texture mapping method combined with the vertex coordinate; S22, using the Delaunay triangulation algorithm to select three points that are not collinear based on the vertex texture coordinate and the vertex normal vector, and constructing the initial Delaunay triangulation according to the three-point selection result, to obtain a triangular set; S23, constructing a space index structure according to the vertex texture coordinate and the vertex normal vector, and using the space index structure to traverse the triangular set and insert the vertex, to obtain the reconstructed Delaunay triangulation; S24, performing triangle circumscribed circle checking on the inserted points of the reconstructed Delaunay triangulation, and performing triangle adjustment based on the triangle circumscribed circle checking result combined with edge flip operation, to obtain the triangular mesh.

7. The method of claim 1, wherein, The triangular mesh is compressed by using the topologically guided micro-mesh, to obtain the compressed data, and the compressed data is rendered by using the visibility buffer sputtering, to generate the rendering data, including: S31, performing topological analysis on the triangular mesh by using the topologically guided micro-mesh, and extracting the triangular characteristic edge and the connection relationship based on the topological analysis result, and constructing a topological connected graph according to the extraction result; S32, dividing the triangular mesh along the axial direction based on the topological connected graph, to obtain a plurality of triangular mesh blocks, and compressing each triangular mesh block by using micro-mesh coding, to obtain the compressed data; S33, decode the compressed data through the visibility buffer splatting to obtain decoded data, and perform shading calculation on the decoded data through the multi-indirect rendering mechanism combined with the illumination model to generate rendering data.

8. The method of claim 7, wherein, The decoding of the compressed data through the visibility buffer splatting to obtain decoded data, and the shading calculation on the decoded data through the multi-indirect rendering mechanism combined with the illumination model to generate rendering data comprises: S331, decode the compressed data through the visibility buffer splatting to obtain decoded data, and write the decoded data into the visibility buffer through the multi-indirect rendering mechanism to obtain a filled visibility buffer; S332, classify pixels based on the filled visibility buffer combined with a preset tile to obtain a classification result, and perform shading calculation based on the classification result combined with the illumination model to generate rendering data.

Images