Pipe gallery complex algorithms and large mesh parameterization data online rendering

In recent years, the rapid development of network technology, mankind forward highly information-oriented society forward. 3DGIS + BIM in the domestic construction industry to form a boom, despite strong support from the government and society, BIM scene of large-scale Web application development side still faces many great challenges:
First, the web storage bottlenecks:
With BIM scene size increased complexity of the internal structure of the building space and improve the model body mass pipe rack structure is increasing, BIM data in the future may be up to city level (more than 100 TB), generally speaking, PC side of the browser can use the memory to only 1.5 2 GB, slightly larger BIM model could lead to the collapse of the browser.
Second, the calculated rendering bottlenecks:
the conventional loading mode, the IFC member is added one by one in the scene, the model has a number of pipe rack triangulated surface sheet, resulting in the slow initial load and low FPS scene, building models easily browse obvious Caton situation.
On the basis of the above problems, in particular the pipe rack members are often in such line in urban building model a certain proportion, for example water pipes, ventilation ducts, and the building interior pipe rack model array tend to have a certain complexity, optimization this parameter in charge of gallery model can effectively reduce the number of triangular facets of the scene, to help improve the loading rate model, to improve the above two bottlenecks has played a positive role.
Thus the above problems, this paper proposes a pipe rack mesh parameterization algorithm is complicated, firstly by semantic analysis and geometric analysis, extraction pipe gallery category scale mesh member IFC scene. And a generally cylindrical shaped pipe rack for a grid model, bend the pipe rack models and different gallery model three features a hollow tube, the end of its service parameters in the classification process, and returns the result parameter to the web page rendering end. In order to reduce the rendering load the page end, the pipe rack, while the auxiliary model is instantiated based on the adaptive rendering method progressive levels of detail LOD technique, the total data amount effective to reduce the number of triangular faces and pipe rack portion in the scene model, accelerate scene initial load time and improve the frame rate of the overall scene loaded for lightweight WebBIM large-scale online gallery tube model of the scene visualization provides support.
1 related research
1.1 Parametric principle gallery
original building model BIM After IFC THE ANALYSIS divided into several different categories, each category has hundreds of small mesh model information. These triangular mesh vertex information page after end rendering, architectural models make up the browser side. However, this analytic model generated by the pipe rack is usually irregular grid model, and a more complicated situation. If those irregular triangular mesh information extracted, converted into parameterized through the pipe line gallery well known characteristic information, then the information through pipe rack mesh vertex, radius information, the main characteristic parameters of the re-direction Gallery depicts a cylindrical mesh tube, and presents standard triangular mesh rendering.
FIG configuration generally straight tube gallery grid model shown below, specifically for the rendering: the straight tube gallery grid model abstract right circular cylinder, the center line abstract right cylinder axis, the center line of the line start and end points the bottom surface of the vertical cylinder is a positive center of the circle, the radius of the cylinder diameter.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
After the extraction tube to the gallery above the model parameters, when the browser may render the use Three. Js provided
THREE.Cylinder Geometry objects pipe rack batch model. It can be modeled as a cylinder receives THREE.Cylinder Geometry objects using a plurality of parameters, while in the actual modeling process mainly uses three parameters, namely: the radius of the top surface, a bottom surface of the height and radius of the cylinder, respectively, corresponding to three radius of pipelines and pipeline length. After generating the line geometry object model, the need for using a paste of suitable texture THREE.Material, through THREE.Mesh (geometry, material) produce a complete line segment model.
Gallery 1.2 plurality LOD (LOD)
Levels of Detail LOD technology that is short, which means multi-levels of detail. LOD means the technique in which the nodes of the model object in the display environment of the location and importance, resource allocation decisions rendered object, and reducing the number of surfaces of the non-essential details of the object, thereby obtaining high efficiency rendering operation. A typical internal ice storage room BIM schematic model line model shown below, (A) the structure of the ice storage room BIM models, (b) is a partially enlarged schematic detail, the pipe rack model (b) for the next in FIG. adjust the degree of precision less impact on the user's perspective, but with great significance to enhance the scene loaded fluency.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
(A) the ice storage room gallery inner tube model; (b) the model detail
LOD main idea is to reduce complexity, when the 3D model of the object away from the viewpoint, according to the human visual system, distant objects become blurred or smaller , then we can use the simplified version of the object. Varied to achieve the object of a simplified version, is typically accomplished by less data, or alternatively triangular geometry and texture geometric model. No matter how simplified version of the implementation, the final results of its strategy is the smaller one compared to the original model of the object to the GPU to handle pressure, reduce rendering complexity, increase the rate scene loaded.
For pipe rack grid model, LOD adjustment is mainly technical parameters of the pipe gallery after cutting the grid blocks, when the camera angle close, cutting the number of blocks increases, the finer the pipe rack model, and when the camera angle principle when possible to reduce the number of cutting block model, the model presented more rough pipe gallery.
1.3 rendering large-scale online WebBIM
provide a broader stage, "Internet +" era for the transfer and sharing of information on BIM, the integration of the "Internet + BIM", the development is unstoppable. Web browser is the most widely spoken on the mobile Internet information sharing platform, in a web browser on BIM visualization scene reproduced directly out becomes more attractive. Users only need to click on browse the web to access massive scenes interact with BIM. In particular, it is a popular HTMLS / WebGL provides a free plug-in installed Web3D development platform, which greatly convenient Internet BIM visualization shared browsing, so WebBIM online visualization will become the "Internet + BIM visualization" of mainstream trends.
However, due to the presence of large-scale real-time BIM scene is slow to respond, rendering weak, slow data transmission and other issues, WebBIM online visualization will suffer because of poor access to the web browser cache and lead to limited, especially compared with the pipe rack model BIM building more large-scale, large amount of data may lead to paralysis of the browser. Internet bandwidth is limited due to transport caused by the sluggish, affected by the restricted Web browser rendering capabilities caused by roaming delay, these effects will seriously hamper WebBIM scene of large-scale real-time visualization of online sharing. Therefore, the parametric method and line pipe gallery complex mesh models studied in this paper can be made to improve visualization techniques for rendering web pages online issue of great practical significance.
In summary, the innovation of this paper is mainly reflected in the proposed loading and rendering bottleneck for large-scale WebBIM model of a parametric algorithm based on the pipe rack grid model, can be based on different classification of the pipe rack grid model parameterization processing in the browser using progressive loading based LOD to ease the pressure data browser, improve rendering efficiency model for online analytical models with large-scale building pipe rack grid model rendering WebBIM It provides new optimization direction.
2 technical route
techniques employed herein follows the route shown in FIG. On the server side, the work need to complete the model including extraction pipe rack member semantic analysis and geometric analysis angle, and Parametric Algorithm cylindrical tube member gallery grid based algorithm to further expand the parametric model and the elbow pipe rack parametric algorithm hollow tube gallery complex models based on Boolean calculation.
Web work done at the end, including the use of progressive technology based on load weight WebBIM LOD enhance loading rates lightweight online rendering, and visualization applications to provide online support.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
3 Key Technologies
extract 3.1 gallery model member
IFC is a set of customized specifically for the construction industry open standard data format, in order to resolve the BIM model IFC format, as used herein, lightweight BIM large online data visualization system, after building model uploaded IFC format, the model-by layer analysis, from a project example, dismantling the building model comprising a plurality of all types of walls, doors and other building components and beams. The inner member can be classified according to a preliminary classification of semantic information model, we can see a scene in a variety of different constructs based presentation, as shown below, IFCCovering class diagram showing for the pipe rack housing model, there are more cylinder Gallery mesh tube shape, so this class IFCCovering member selected from the semantic wherein the angle of the target tube as a possible model gallery processed, carry out further geometrical analysis.
The entire pipe rack model extraction process as shown in FIG objects, extracted from the semantic analysis component member is a member from a plurality of different types IFC above, the extraction process may be a member of the cylinder-type model gallery pipe segment pipelines. From the perspective of semantic analysis, often there may be errors, such as the non-cylindrical body model that is stretched pipe rack model.
Based on this, analysis geometry is is a member analyzed morphologically pipe rack model, determining means whether the characteristic cylindrical member bounding box, and the upper and lower bottom surface to the central axis of the vertex information in the model can be calculated from and other parameters, and thus its induction type, the final extract a target model may be a class of the pipe rack mesh parameterization process.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
(A) Classification ice storage room model member; (b) IfcCovering class model
3.2 gallery mesh parameterization
3.2.1 cylindrical pipe rack grid model parameterization
BIM building model through a parsing step of generating the original cylindrical Gallery tube member is dat model format, the information consists of three parts, a first part of a three-dimensional coordinates of each vertex, starting from the number 0, the second portion is a triangular model information, each of three vertices of a triangle, in the scene examples of rendering, all the triangles in a scene depicted in a heavy cylindrical tube model gallery, the third portion is the second portion of each direction vector information corresponding to triangulated method.
It can be seen, for an original pipe gallery complex mesh models, which often have hundreds, even thousands of vertices, while a larger number of triangular mesh. Model for large buildings, the amount of data once the gallery pipe model is large, there is a big impact will render the initial model and loading efficiency, and therefore, cylindrical paper pipe rack mesh model parameters proposed manner redraw using ThreeJS instantiate the cylinder can be effectively simplified model data and to enhance the amount of scene rendering rate.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
(A) the original BIM model; (b) BIM parametric model; (c) details the original end Videos; (d) the details of the end surface parameterization

A generally cylindrical shaped pipe rack parametric model used herein is extracted from the valid vertices and triangular faces mesh information of the original pipe rack vertically in its bottom surface and the radius of the center circle, and the longitudinal axis, and wherein the determined direction, and the process can be re-drawn member in the scene. Standardization cylinder fitted with a source of data, such as the view (a) a non-standardized data source side surface triangulation and (b) after a standardized fitting triangulated side in FIG. The use of a parametric model of the embodiment of the cylinder, the cylinder need to redraw the data thus greatly reduced.
Due to the way the original pipe rack triangulation irregular grid model, as shown above (c) source data - non-standardized bottom surface triangulation and (d) after a standardized fitting bottom surface triangulation, the parameters of the process is a tube Gallery mesh standardization process, provide a basis for subsequent application online gallery cylindrical tube member. The number of divided blocks vertically bottom surface determines the degree of fine cylinder model in the scene.
3.2.2 elbow pipe gallery mesh model parameterized
in addition to general cylindrical pipe gallery model, building model there is a big portion of the bend pipe gallery model, as shown in FIG. This is typically part of the model at the interface of the drain pipes and ventilation ducts, occupy a certain proportion of the pipe rack grid model, to make a distinction between models are generally cylindrical pipe rack in the parameter of the linear process. As used herein, quadratic Bezier curve fit elbow pipe rack member, subjects is shown below an elbow on a circular line model.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
The first cylinder model parameters determined two end faces of the center, as the center two control points of the Bezier curve PO and P2, is provided which controls the point P1, the central axis of the fitted model should satisfy the following secondary baser Er curve formula:
B (T) = (1-T) 2P0 + 2T (T-1) + Pl t2P2, t ∈ [0, 1] (1)
By changing the parameters of the control point, if the distance of the axis of the elbow pipe fitting point on the shortest gallery model can be considered a best fit curve, plotted when performing parameterization, due to the symmetric pipe rack Model sex, on orders from the central axis of the elbow is the radius to the model, to one end of a circle, the radius of the circle, and then stretched along the Bezier curve fitting to obtain the bend model fitting the model parameters of the tensile member. And since the tensile properties of the body, the number of divided blocks of the fitted model can be adjusted dynamically, effectively reducing the number of triangular mesh model.
3.2.3 hollow tube gallery complex mesh model parameterization
based on the standard of research cylinder-shaped tube Gallery parametric approach, the study of complex parametric algorithms pipe gallery component-based Boolean operations are very valuable in the field of engineering and construction, as follows Gallery tube model shown in FIG. (a) a hollow, are based on a complex pipe rack member extending cylinder-type pipe rack extension member, an end surface and a side schematic model shown below (b) and (c), with which unique, it is necessary to explore the model feature targeted manner.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
(A) a hollow tube gallery; (b) end surface mesh model; (c) a grid model side

Aiming parameterization the hollow tube gallery complex model can be divided into the following steps:
A) due to the symmetry of the pipe rack model, all vertices located extracted mesh information and the triangular end surface of the same;
B) Alternatively repeated vertices, as a plurality of vertex number, minimum number Alternatively, reorganizing triangular grid information;
C) split into three triangular mesh information side information to extract information three sides <small number, a large number> form, and record frequency of use;
D) extracts all edges of the frequency of use of 1, redraw the scene, the complete extraction of edge boundaries;
E) according to the closed edge, vertex packet, and the outer shape into a hollow interior portion;
F) packets after the parameters of the fitted model Boolean operations, in the scene depicted in a new tube gallery heavy mesh.
Below for (a), the hollow interior pipe gallery stretching body model, the solutions described above, first, reads the entire model, which model is then separated by external and internal parameters of the geometric features of the model, its edges the extraction results as shown (b)), after the edge extraction result obtained, the model is divided into a large solid cylinder model, and the pipe rack pipe rack 7 of the hollow cylinder type, large solid cylindrical Gallery tube model Boolean operation, the hollow tube after redraw gallery model parameterized below (c) shown in FIG.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
(A) a hollow tube Gallery original model; (b) model edge extraction; (c) after parametric model

3.3 Progressive LOD online real-time rendering
after the adoption of model building with large-scale pipe gallery grid pipe gallery members were analytical extraction and model parameterization, how to efficiently render smooth BIM model is one of the main points of this research page end. LOD herein based adaptive incremental loading techniques, different levels of adjustment according to different classification mesh tube gallery, a schematic model is shown below.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
After obtaining the first two steps pipe gallery model parameterized result, this paper based LOD progressive loading of the distance from the camera, and the camera model LOD0 farther distance from, and to LOD2, the camera model are closer LOD1 stage, model small number of divided blocks dynamically adjusting the number of divided blocks pipe rack grid model, i.e., the more the model number from the divided blocks near the camera, the more detailed model, the distance the camera, model simplification. Gallery for standard cylindrical tube member, in accordance with the distance from the camera, a dynamic adjustment of the number of divisions.
As shown above, for the pipe rack is generally cylindrical mesh model, adaptive adjustment is the number of divided blocks of the end surface, from 3 LOD0 of the LOD1 to 8, then the 16 LOD2. For elbow pipe rack model, adaptive adjustment of the axis is the number of sub-blocks, and 5 blocks from LOD0 to LOD1 of 10, then 20 LOD2 can be seen more dense grid model. And for the hollow tube gallery grid model, and in LOD0 LOD1 stage, i.e. far from the camera, rendering the tube hollow performed only in accordance with the model gallery cylindrical pipe rack rendering, Boolean operation is not performed, the distance see is generally cylindrical pipe rack mesh model, only reached within LOD2 distance, rendered hollow pipe rack in mesh close to the camera model, this step can be optimized rendering efficiency, improve the FPS scene, reduce the load on the browser.
Gallery 3.4 based mesh parameterization tube cutting applications
as pipe racks model cylindrical body member gradually increasing the amount of BIM building model, associated also made more interaction requirements, such as the more common aliquot cutting cylinder Wait.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
After the article after the cylindrical tube model parameterization gallery based on the central axis is divided, as shown in FIG renderings for the parametric model redraw the scene, through the redraw parametric model itself member nonstandard after standard pipe rack into a cylindrical mesh model, through the effect of the cut-half times as shown in FIG. According to the actual demand, the application can also be extended to the n-th cutting and self-regulation and other means to facilitate the practical application engineering.
Experiment 4 Analysis
proposed in accordance with the lightweight pipe rack WebBIM mesh parameterization of the three way performance evaluation results for performance evaluation: the total amount of data pipe rack model; initial loading time of the scene (s); browser frames per second FPS.
4.1 Test Case
The static scene data used in the test is the basis of the format of the ice storage room BIM data model ifc, as shown (a), the ice storage room, a total of 6575 mesh, occupies a large 176 MB, the IFC experiment after analytical model, called the extracted IFCCovering class, as shown (b), the total of such BIM pipe rack 2184 mesh model stored in file xubinOutput -Vis.txt total percentage space 24.4MB.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
(A) the ice storage room original model; (b) IfcCovering class model

4.2 test environment
for this experiment using a desktop browser for testing. Server operating system used is WinServer2008, memory is 4G; CPU is Intel Xeon. The client operating system Windows7 memory to 8 G, COU is Intel i7-4700MQ. Used browser Google Chrome, network bandwidth is 100 Mbps.
4.3 Test procedure
a) were tested using the total amount of data before and after the model parameterization lightweight WebBIM pipe gallery.
b) were tested before and after the scene using lightweight pipe gallery WebBIM parametric initial load time (s).
c) were tested using the browser of the front and rear frames per second FPS WebBIM lightweight pipe gallery parameterization.
4.4 Test Results
The results of this comparative experiment shown in the following table, the amount of data can be seen that the use of such obviously after the pipe rack mesh parameterization algorithm, down from the original 20.5 MB to 4.2 MB, such data representing original the amount of about one-fifth. And the initial load time greatly enhance the scene, and the whole scene has improved transmission frames, proved the usefulness and value of the algorithm. After the pipe rack model parameterized scene in the schematic shown below.
Pipe gallery complex algorithms and large mesh parameterization data online rendering
Pipe gallery complex algorithms and large mesh parameterization data online rendering

5 Conclusions
To reduce the overall amount of data WebBIM line model of the scene and enhancing rendering efficiency, this paper proposes a pipe rack complex grid model parameterization method, the model has a lot of BIM pipe rack mesh by first IFC parsing, semantic analysis after and means for geometric analysis, the pipe rack can be extracted model parameterized classes. And a generally cylindrical shaped pipe rack for a model, and model complex elbow pipe gallery gallery hollow tube model parameterization scheme are proposed and verify its validity. Finally, the web client uses a progressive reduction in weight based LOD rendering mode, to cut the auxiliary model is applied based on the results of the parametric pipe rack.
Experiments show that the method has a huge amount of data to enhance the efficiency and rendering model, BIM engineering practice for a multi-tube model gallery presents important optimization direction. Future research include expansion of the cylindrical extension from the parametric model parameterization method in general to stretch the body, and the cylindrical model based on the parameters of collision detection parameters, for example, based on application of the automatic line arrangement geospatial layout.

Original from:http://www.ztmapinfo.com/blog/index.php/article/23.html

Guess you like

Origin blog.51cto.com/14758767/2483514