CN102622198A - Method and system for visualizing data - Google Patents
Method and system for visualizing data Download PDFInfo
- Publication number
- CN102622198A CN102622198A CN2012100493758A CN201210049375A CN102622198A CN 102622198 A CN102622198 A CN 102622198A CN 2012100493758 A CN2012100493758 A CN 2012100493758A CN 201210049375 A CN201210049375 A CN 201210049375A CN 102622198 A CN102622198 A CN 102622198A
- Authority
- CN
- China
- Prior art keywords
- data
- visualization
- visual
- gridding
- texture
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 71
- 230000000007 visual effect Effects 0.000 claims abstract description 53
- 238000012800 visualization Methods 0.000 claims abstract description 41
- 238000005516 engineering process Methods 0.000 claims abstract description 36
- 238000004458 analytical method Methods 0.000 claims abstract description 15
- 238000007906 compression Methods 0.000 claims abstract description 12
- 230000006835 compression Effects 0.000 claims abstract description 12
- 239000013598 vector Substances 0.000 claims description 29
- 238000007794 visualization technique Methods 0.000 claims description 22
- 230000008569 process Effects 0.000 claims description 19
- 238000012545 processing Methods 0.000 claims description 19
- 230000008859 change Effects 0.000 claims description 18
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 6
- 238000013144 data compression Methods 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 4
- 239000012467 final product Substances 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 230000009467 reduction Effects 0.000 claims description 3
- 230000003252 repetitive effect Effects 0.000 abstract 1
- 238000013079 data visualisation Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000009877 rendering Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 238000005286 illumination Methods 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000005211 surface analysis Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 239000012530 fluid Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012067 mathematical method Methods 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
Images
Landscapes
- Image Generation (AREA)
Abstract
The embodiment of the invention discloses a method for visualizing data. The method comprises the following steps: meshing data through the method of superposition of trend surface and residual error; applying corresponding visualization technologies to the meshed data; removing repetitive data and leaving essential data through logic simplification, texture simplification and code compression; analyzing data format, assembling logic relation and generating visual data; and displaying the visual data. The embodiment of the invention further discloses a system for visualizing the data. According to the embodiment of the invention, the visualization technologies are combined so as to be suitable for different data fields very well; and on the basis that the basic visualization technology is combined with the visualization technology based on feature analysis, the visualization quality can be further improved, and the calculated amount is not large.
Description
Technical field
The present invention relates to the Computer Processing technical field, relate in particular to a kind of method and system of data being carried out visualization processing.
Background technology
The visualization in scientific computing technology has important application in scientific researches such as medical science, molecular chemistry and biology, computational fluid dynamics, finite element analysis, meteorology, geophysics and national economy field.Because visual is data and compute-intensive applications, in the early stage research of visualization technique, parallel and distribution technique just begins to obtain to use.General high-end parallel computer and the graphics workstation of using of early stage parallel volume rendering.Get into the middle and later periods nineties 20th century; PC popularizes gradually; The drawing ability of PC graphic hardware improves constantly, the network technology develop rapidly, and the PC cluster has progressively replaced traditional high-end parallel machine with networking PC (particularly PC cluster) and has become the ideal platform of parallel visualization application.And the rise of Internet makes visualization application utilize storage, calculating and the drafting resource of Internet scope, and in the Internet scope, provide the service become a kind of maybe with the demand of reality.Traditional network visualization research does not have to consider isomerism, interoperability, dynamic and the extensibility towards Internet.And from focus on dispersion, from be closely coupled to loose coupling, among a small circle to being parallel computer and development of internet technology trend on a large scale.A kind of prospective distributed computing towards the internet that gridding technique produces under this condition just, it is traditional parallel computation and the expansion of Distributed Calculation on the degree of depth and range.Though gridding technique is still among development; But functions such as the resource that it provided converges, autonomous coordination will make visualization application in wider scope, carry out data storage and calculating; Integrated with scientific program better, and let the user of wider scope use visualization application with long-range or cooperation mode through network.
The existing visual processing is structurally to be divided into calculation server that is used for data processing and the client that is used for the user-specified rule demonstration.Calculation server is mainly accomplished following steps successively: generate data network, science is calculated the data that relate to, and not only number of parameters itself two is big but also data volume that gathered also is a magnanimity; Find the solution physical equation, after handling, carry out finding the solution of physical equation through the gridding of data; Generate result data and data are deposited.In client, at first the user should the clear and definite data that oneself should see, thus set visual parameter earlier, such as viewpoint, size of result set or the like; Demand client according to the user reads out the corresponding results data load in internal memory from the hard disk of server; Carry out visual drafting and demonstration at last, its corresponding step is referring to shown in Fig. 1.
Because result of calculation is stored in the hard disk earlier; And then in the graphics workstation or common PC terminal is carried out visual to it; But project data is very huge sometimes; Particularly variable in time large-scale data field, terminal does not at this moment often have so big space to store raw data, does not have so big internal memory to carry out visual to it simultaneously yet.Like this, with the prompting that low memory often occurs, even can not be visual.This method is not optimized, is simplified result data in implementation process, visually wants to separate with numerical evaluation, and the size of data volume also directly has influence on traffic load end to end, as if unstable networks, with the running that seriously influences visual part.
Summary of the invention
The objective of the invention is to propose a kind of data to be carried out visualization method and system; Can accomplish the gridding of data and adopt various ways data visualization; Satisfy client's multiple demand, carry out effective load balance process, effectively accelerated processing speed and system stability.
Based on the problems referred to above, the embodiment of the invention provides a kind of data has been carried out the method for visualization processing,, comprise the steps:
Adopt trend surface and residual error method of superposition to realize the gridding of data;
To the corresponding visualization technique of the The data of gridding;
Adopt logic simplifying, texture simplification and encoding compression that data are gone heavily to stay essence;
To the parsing of data form, the assembling of logical relation and the generation of visualized data;
Visualized data is shown.
Said employing trend surface and residual error method of superposition realize that the gridding of data comprises:
Selected according to actual needs m value is used for repeatedly curved surface needs of back match;
Utilize the trend surface fitting process, simulate a m trend surface;
Make the residual error between data dot values and this trend surface;
With Trend value on the net point and the addition of residual error match value, as the grid point value.
The corresponding visualization technique of said The data to gridding comprises:
Parallel visual, visual based on signature analysis of parallel visual, the vector of scalar data field and tensor data fields.
Said employing logic simplifying, texture simplify and encoding compression goes heavily to stay spermatophore to draw together to data:
Logic simplifying adopts formal inference with calculation that the logical relation of the data in the module is clear and definite, through analyzing the data relationship from different pipelines, removes repeating data and the data that can directly generate;
It is a simplifications technology of carrying out to some texture images that texture is simplified, usually texture all be have rule repeat occur, at this moment utilize this attribute to keep a texture properties, only need directly extraction to get final product when needing to reproduce at every turn;
Data compression is at other data compression method of byte level, adopts Huffman encoding compressed encoding or other Variable Length Code algorithms to realize.
The assembling of said parsing to the data form, logical relation and the generation of visualized data comprise:
The parsing of data layout is responsible for accepting the data that transmit from different far-ends, to its decode, format analysis, reduction needs data presented;
The assembling of logical relation is a kind of foundation to internal logic between the parallel data, adopts the good logical term of predefine, and each data all is mapped to a logic of class type, also need state its required continuous logical term in front and back simultaneously;
According to the call format of different display devices, change into different data layouts to data.
Said visualized data is shown comprises:
Visual data of needs and specific are played up the preprocessing process that technology combines;
Adopt the OpenGL technology to realize the transplantability of data;
Hardware resource is abstracted into an object, and shielding top layer details makes the upper strata developer needn't know the various ins and outs of bottom, and provides a kind of unified DLL;
Data are transferred to hardware to be shown.
Accordingly, the embodiment of the invention also provides a kind of data has been carried out the system of visualization processing, comprising:
The data gridding module is used to adopt trend surface and residual error method of superposition to realize the gridding of data;
Visualization model is used for the corresponding visualization technique of the The data of gridding;
Simplify module, be used to adopt logic simplifying, texture simplification and encoding compression that data are gone heavily to stay essence;
Knockdown block is used for the parsing to the data form, the assembling of logical relation and the generation of visualized data;
The basic display module is used for visualized data is shown.
The visualization technique that said visualization model adopts comprises: parallel visual, visual based on signature analysis of parallel visual, the vector of scalar data field and tensor data fields.
The present invention proposes a kind of parallel visualization system of large-scale data field, comprise from the gridding of data and show a whole set of flow process to final terminal.In the gridding module of data, adopt the method for trend surface and residual error stack, taken into account the overall situation and the local characteristics of plain tool; Adopted multiple visualization technique to combine in visualization model, complementation is mutually, and has comprised the parallel method for visualizing of the data fields of scalar, vector and tensor; In order to reach better image quality, we have also adopted the technology of opportunity signature analysis, make visual quality more true to nature; Load balancing is an important factor in order of parallel system operational efficiency.The present invention adopts a kind of technology of task pool to can be good at load balancing in the resolution system.Then, simplify module and go heavily to deposit essence, required data toward next module are minimum simply, reduce volume of transmitted data; Knockdown block is responsible for assembling the parallel data from different far-ends, transfers to the basic display module at last and shows.
Combining through multiple visualization technique can be good at being fit to different data fields, comprises vector, tensor sum scalar; The visualization technique that on basic visualization technique, combines the opportunity signature analysis again can further promote visual quality, and calculated amount is also little simultaneously; In parallel framework, in order to reach the high-level efficiency of parallel performance, adopt the task pool technology that a big task is cut apart, come to finish the work respectively by the sub-thread of a group, can be good at load balancingization.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art; To do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art below; Obviously, the accompanying drawing in describing below only is some embodiments of the present invention, for those of ordinary skills; Under the prerequisite of not paying creative work, can also obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the process flow diagram that data visualization of the prior art is handled;
Fig. 2 carries out visual system architecture diagram to data in the embodiment of the invention;
Fig. 3 is that the method for superposition in the embodiment of the invention is done the data gridding flow process;
Fig. 4 is the visualization model structural representation in the embodiment of the invention;
Fig. 5 is the visible process figure of the tensor field in the embodiment of the invention;
Fig. 6 is the task pool dynamic task allocation flow figure in the embodiment of the invention;
Fig. 7 is the simplification modular structure synoptic diagram in the embodiment of the invention;
Fig. 8 is the Knockdown block structural representation in the embodiment of the invention;
Fig. 9 is the basic display modular structure synoptic diagram in the embodiment of the invention.
Embodiment
To combine the accompanying drawing in the embodiment of the invention below, the technical scheme in the embodiment of the invention is carried out clear, intactly description, obviously, described embodiment only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills are not making the every other embodiment that is obtained under the creative work prerequisite, all belong to the scope of the present invention's protection.
The objective of the invention is to propose a kind ofly can handle the numerical evaluation of large-scale data field and the system of data visualization by efficient parallel; System can accomplish the gridding of data and adopt various ways with data visualization; Satisfy client's multiple demand; Carry out effective load balance process, effectively accelerated processing speed and system stability.
What the present invention had implemented to provide in the example carries out the method for visualization processing to data, comprises the steps: to adopt trend surface and residual error method of superposition to realize the gridding of data; To the corresponding visualization technique of the The data of gridding; Adopt logic simplifying, texture simplification and encoding compression that data are gone heavily to stay essence; To the parsing of data form, the assembling of logical relation and the generation of visualized data; Visualized data is shown.
It is the computation host that performance is higher, concurrency is good that the parallel visualization system of a complete large-scale data field comprises one in two main frames at least, and another is the used general PC of user, is used to show needed result; Overwhelming majority's work of system all concentrates on the former.
As shown in Figure 2 data are carried out the system of visualization processing, this System Framework comprises data gridding module, visualization model, simplification module, Knockdown block and basic display module.It is accomplished on the high performance parallel computation machine and calculates and visualization process, but output visual pattern element, rather than result images, thereby at user side viewpoint, light source, illumination model or the like can be set, obtain visual image at last; On calculation server, user's method for visualizing that only definition is adopted in control documents is like profile drawing, contour surface drafting, streamline drafting etc.Because visualization process and computation process are all carried out on same computing machine simultaneously, will result of calculation not be kept on the dish, thereby avoided the restriction of storage space; The magnanimity internal memory completion that can make full use of calculation server simultaneously is visual, and last visualization result is much littler than the original calculation data volume usually.We are also according to user's the time requirement and the restriction of terminal computer internal memory and hard disk, and the quantity that the parallel simplification of usefulness module reduces the visual pattern element that outputs to user side is to reach real-time display requirement.
The gridding module of data of the present invention adopt one can response data local feature can react the method for global property, trend surface and residual error method of superposition (abbreviation method of superposition) again.Specifically referring to Fig. 3.
The concrete steps of gridding that method of superposition is done data are following:
Step1: the selected according to actual needs m value of user, be used for repeatedly curved surface needs of back match, change Step2;
Step2: utilize the trend surface fitting process, simulate a m trend surface, change Step3;
Step3: make the residual error between data dot values and this trend surface, change Step4;
Step4: with Trend value on the net point and the addition of residual error match value, as the grid point value.
The trend surface analysis that in method of superposition, relates to is field, a ground mathematical method commonly used, and it can change macrocyclic tendency variation and short-period locality separately, changes according to tendency and can sum up rule, changes and can note abnormalities according to locality.See that from the computerized mapping angle trend surface analysis is to come the complicated ground of match to learn curved surface with simple power series polynomial expression, and the effect of scabbling, filling and leading up actual curved surface is arranged.
Visualization model adopts different pieces of information is adopted different visualization techniques, mainly comprises: parallel visual, visual based on signature analysis of parallel visual, the vector of scalar data field and tensor data fields.Specifically as shown in Figure 4.
We adopt the profile drawing technology in scalar data field parallel visual.Profile drawing is the technology that the most often adopts in visual, because it is simple, efficient, each visual software all provides should technology.The present invention goes back integrating parallel contour surface, volume drawing, interval rendering technique.Because the contour surface method for drafting only can disclose the information of sub-fraction field, thus through adopting the parallel volume rendering method, directly from 3 d data field, draw out the distribution situation of various physical quantitys through the body illumination model, thus visual whole data fields.Further auxiliary mutually with interval volume drawing, more further improving again on than simple object plotting method efficient on the drawing efficiency.
We adopt streamline technique in the visualization model of vector field, and with this technological expansion in parallel environment.In order to disclose more three-dimensional information; Give radius value with each point on the streamline according to the size of vector; Form the stream pipe; Thereby can reflect more three-dimensional rotation information through the variation of illumination brightness, radius and color value can reflect two physical quantitys simultaneously, draw multilist than original streamline and have reached one-dimension information.Because streamline method is based upon on the basis of stream field discrete sampling, its visual quality depends critically upon choosing of seed points.If seed points is chosen too intensively, then possibly cause visual confusion, but, might miss structure important in the flow field and details again if choose too sparsely.Method based on texture then can address this problem preferably; Because it all carries out brightness calculation at each pixel place; Thereby can avoid the problem of sampling; By means of the distortion of texture, the direction of vector is visual to clearly reveal out simultaneously, and therefore the method based on texture is the most potential at present vector field visualization method.The present invention adopts this technology to realize as the technology of vector field module.
Tensor field is very general in engineering, but because it comprises 9 components, is difficult on the two-dimensional screen so many information representation is come out, so tensor field visual is to have a challenging research field in visual always.Ultra streamline (Hyperstream lines) method is owing to can disclose 9 components of tensor simultaneously on a continuous three-dimensional path.So adopt ultra streamline method to can be good at visual tensor field data.
The visual tensor field of ultra streamline method can be referring to Fig. 5, and concrete steps are following:
Step1: 9 components of tensor are decomposed into 3 proper vectors, are called main proper vector, sub-eigenvector, minimal characteristic vector successively, change Step2 by the size of its individual features value;
Step2: selected seed points, from the direction generation trajectory of seed points, change Step3 then according to main proper vector;
Step3: the additional one oval stream pipe that forms on each point of this trajectory, long axis of ellipse and short-axis direction have reflected the direction of sub-eigenvector and minimal characteristic vector respectively, its major axis and minor axis length then are the size of its corresponding vector, change Step4 then;
Step4: the size of main proper vector is mapped as the color that flows respective point on the pipe face, and so far, 3 proper vectors just show to come simultaneously.
In order to improve visual quality, system has further adopted the technology of signature analysis.As previously mentioned, during with streamline, the visual vector field of ultra streamline, tensor field, its quality depends critically upon choosing of seed points, should choose the seed points that can reflect the data fields characteristic more as far as possible.But the user is before visual, and the characteristic to the field is not to understand very much usually, so the characteristic that system obtains automatically just becomes extremely important.Native system adopts the topological property analytical technology of vector field.At first feature extraction is carried out in the field, find unique point, characteristic face, determine the density and the position of seed points then according to these unique points and face automatically.Make streamline and ultra streamline in the field, have rationally, effectively distribute.In order to improve parallel efficiency, be distributed on each processor choosing of seed points with also having considered to make their relative equilibriums simultaneously.Sometimes the user perhaps can be to unimportant regional interested; Therefore we also provide multiple mode easily to allow user oneself definition seed points simultaneously; Coordinate like input point; The overall situation coding of input grid cell (or grid node), the processor at input grid cell (or grid node) place number and local code thereof or the like.
In addition; We also are used to the topological characteristic analytical technology of vector field to quicken the generation of vector field texture image; According to the characteristic of vector field, the importance values of each point in the calculated field is determined the fine degree that texture calculates and is chosen the various textures granularity by its importance values; Thereby the generation of accelerogram picture, and can give prominence to the key character of vector field.Because the intensive of texture, will expand to based on the vector field visualization method of texture and usually be difficult to obtain effect preferably when said three-dimensional body is drawn, the transfer function that characteristic control volume that can use is drawn improves the quality of volume rendered images.
Load balancing is the important channel of improving the volume drawing parallel efficiency.The present invention adopts a kind of technology of task pool to can be good at solving the problem of load balancing, and is as shown in Figure 6.Concrete steps are following:
Step1: platform initialization, start main thread, and do initial work such as Memory Allocation, change Step2;
Step2: initial beggar's thread crowd, comprise the needed memory headroom of each thread, resource etc., change Step3;
Step3: task is divided into some subtasks, then each subtasks is distributed to each sub-thread, change Step4;
Step4: wait for the result of subtask, judge that whether task pool is empty, if do not dally Step5, otherwise changes Step6;
Step5: distribute a subtasks to give the idle sub-thread that returns, change Step4;
Step6: kill the sub-thread that returns, judge whether that all sub-threads are all dead,, otherwise change Step4 if then change Step7;
Step7: assembling also shows, finishes.
Simplify module: simplifying module is that data are gone heavily to stay smart module, and main applied logic is simplified, texture is simplified and encoding compression three aspect simplification technology realize.Logic simplifying adopts formal inference with calculation that the logical relation of the data in the module is clear and definite, through analyzing the data relationship from different pipelines, removes repeating data and the data that can directly generate.It is a simplification technology of carrying out to some texture images that texture is simplified.Usually texture all be have rule repeat occur, at this moment can utilize this attribute to keep a texture properties, only need directly extraction to get final product when needing to reproduce at every turn.Data compression is at other data compression method of byte level, can adopt the compression algorithm of existing maturation, and for example Huffman encoding compressed encoding or other Variable Length Code algorithms are realized.Its structure can be referring to shown in Figure 7.
Knockdown block, its formation is as shown in Figure 8.Knockdown block is realized the assembling with parallel data.The large-scale data normal open is often accomplished visual work from assembling between the data of difference end, and assembly working is exactly the process of data fusion.Knockdown block comprises the parsing of data layout, the assembling of logical relation and generation three parts of visualized data.The parsing of data layout is responsible for accepting the data that transmit from different far-ends, to its decode, format analysis, reduction needs data presented.The encoding compression module that decoding need and be simplified in the module is thought correspondence; If the Huffman encoding that the encoding compression module adopts; Also need adopt Huffman decoding so here; If what adopt is other encryption algorithms, then need its corresponding decoding algorithm, our default situation adopts Huffman encoding to get final product here.The assembling of logical relation is a kind of foundation to internal logic between the parallel data, adopts the good logical term of predefine, and each data all is mapped to a logic of class type, also need state its required continuous logical term in front and back simultaneously.In addition, in assembling process, also have logic strategy except logical term, logic strategy can carry out different logical optimization according to the deflection of strategy when multiple assembling mode is all feasible.At last parallel data is assembled the data that generate after assembling, data need be sent to next basic display module, according to the call format of different display modules to data, need change into different data layouts, and so far, the work of Knockdown block finishes.
The basic display module is responsible for being permitted visual data and is shown, and is as shown in Figure 9.It can adopt DirectX technology or OpenGL technology, and in order to realize better transplantability, we adopt the OpenGL technology, and OpenGL also has much the trend that becomes academic and industrial standard.The formation of basic display module is as shown in Figure 8.Data preprocessing module is that visual data of needs and specific are played up the preprocessing process that technology combines; Here the OpenGL technology that adopts, so we need carry out the form with the OpenGL coupling with data, for example the data-switching with vertex type becomes the Vertex type; So if adopted vertex buffer and buffer zone to combine; Then need generate corresponding index according to original vertex list, and change vertex list, each summit all is unique in the feasible tabulation.The OpenGL storehouse is the graph rendering storehouse, is the graphic presentation storehouse of academia and industry member specialty, and we can customize it as required, are transplanted to our platform.OpenGL does not directly come into contacts with hardware, but gets in touch with hardware abstraction layer.Hardware abstraction layer is abstracted into an object with hardware resource, and shielding top layer details makes the upper strata developer needn't know the various ins and outs of bottom, and provides a kind of unified DLL.Hardware abstraction layer is normally named difference to some extent according to the difference of system platform, but elementary tactics all is the same with framework, and purpose all is in order to shield the otherness of bottom.Driver is direct and hardware is come into contacts with, and it is provided by hardware manufacturer, and it follows the rule of hardware abstraction layer, and service upwards is provided.Last display module is transferred to hardware with data and is shown.
Generally speaking, the present invention proposes a kind of parallel visualization system of large-scale data field.Framework comprises from the gridding of data and shows a whole set of flow process to final terminal.In the gridding module of data, adopt the method for trend surface and residual error stack, taken into account the overall situation and the local characteristics of plain tool; Adopted multiple visualization technique to combine in visualization model, complementation is mutually, and has comprised the parallel method for visualizing of the data fields of scalar, vector and tensor; In order to reach better image quality, we have also adopted the technology of opportunity signature analysis, make visual quality more true to nature; Load balancing is an important factor in order of parallel system operational efficiency.The present invention adopts a kind of technology of task pool to can be good at load balancing in the resolution system.Then, simplify module and go heavily to deposit essence, required data toward next module are minimum simply, reduce volume of transmitted data; Knockdown block is responsible for assembling the parallel data from different far-ends, transfers to the basic display module at last and shows.
Combining through multiple visualization technique can be good at being fit to different data fields, comprises vector, tensor sum scalar; The visualization technique that on basic visualization technique, combines the opportunity signature analysis again can further promote visual quality, and calculated amount is also little simultaneously; In parallel framework, in order to reach the high-level efficiency of parallel performance, adopt the task pool technology that a big task is cut apart, come to finish the work respectively by the sub-thread of a group, can be good at load balancingization.
More than the method and system that the embodiment of the invention provided carries out visualization processing to data have been carried out detailed introduction; Used concrete example among this paper principle of the present invention and embodiment are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that on embodiment and range of application, all can change, in sum, this description should not be construed as limitation of the present invention.
Claims (8)
1. one kind is carried out the method for visualization processing to data, it is characterized in that, comprises the steps:
Adopt trend surface and residual error method of superposition to realize the gridding of data;
To the corresponding visualization technique of the The data of gridding;
Adopt logic simplifying, texture simplification and encoding compression that data are gone heavily to stay essence;
To the parsing of data form, the assembling of logical relation and the generation of visualized data;
Visualized data is shown.
2. as claimed in claim 1 data are carried out the method for visualization processing, it is characterized in that said employing trend surface and residual error method of superposition realize that the gridding of data comprises:
Selected according to actual needs m value is used for repeatedly curved surface needs of back match;
Utilize the trend surface fitting process, simulate a m trend surface;
Make the residual error between data dot values and this trend surface;
With Trend value on the net point and the addition of residual error match value, as the grid point value.
3. as claimed in claim 2 data are carried out the method for visualization processing, it is characterized in that the corresponding visualization technique of said The data to gridding comprises:
Parallel visual, visual based on signature analysis of parallel visual, the vector of scalar data field and tensor data fields.
4. as claimed in claim 3 data are carried out the method for visualization processing, it is characterized in that said employing logic simplifying, texture simplify and encoding compression goes heavily to stay spermatophore to draw together to data:
Logic simplifying adopts formal inference with calculation that the logical relation of the data in the module is clear and definite, through analyzing the data relationship from different pipelines, removes repeating data and the data that can directly generate;
It is a simplifications technology of carrying out to some texture images that texture is simplified, usually texture all be have rule repeat occur, at this moment utilize this attribute to keep a texture properties, only need directly extraction to get final product when needing to reproduce at every turn;
Data compression is at other data compression method of byte level, adopts Huffman encoding compressed encoding or other Variable Length Code algorithms to realize.
5. as claimed in claim 4 data are carried out the method for visualization processing, it is characterized in that the assembling of said parsing to the data form, logical relation and the generation of visualized data comprise:
The parsing of data layout is responsible for accepting the data that transmit from different far-ends, to its decode, format analysis, reduction needs data presented;
The assembling of logical relation is a kind of foundation to internal logic between the parallel data, adopts the good logical term of predefine, and each data all is mapped to a logic of class type, also need state its required continuous logical term in front and back simultaneously;
According to the call format of different display devices, change into different data layouts to data.
6. claim 5 is described carries out the method for visualization processing to data, it is characterized in that said visualized data is shown comprises:
Visual data of needs and specific are played up the preprocessing process that technology combines;
Adopt the OpenGL technology to realize the transplantability of data;
Hardware resource is abstracted into an object, and shielding top layer details makes the upper strata developer needn't know the various ins and outs of bottom, and provides a kind of unified DLL;
Data are transferred to hardware to be shown.
7. one kind is carried out the system of visualization processing to data, it is characterized in that, comprising:
The data gridding module is used to adopt trend surface and residual error method of superposition to realize the gridding of data;
Visualization model is used for the corresponding visualization technique of the The data of gridding;
Simplify module, be used to adopt logic simplifying, texture simplification and encoding compression that data are gone heavily to stay essence;
Knockdown block is used for the parsing to the data form, the assembling of logical relation and the generation of visualized data;
The basic display module is used for visualized data is shown.
8. system as claimed in claim 7 is characterized in that, the visualization technique that said visualization model adopts comprises: parallel visual, visual based on signature analysis of parallel visual, the vector of scalar data field and tensor data fields.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012100493758A CN102622198A (en) | 2012-02-29 | 2012-02-29 | Method and system for visualizing data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012100493758A CN102622198A (en) | 2012-02-29 | 2012-02-29 | Method and system for visualizing data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102622198A true CN102622198A (en) | 2012-08-01 |
Family
ID=46562131
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012100493758A Pending CN102622198A (en) | 2012-02-29 | 2012-02-29 | Method and system for visualizing data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102622198A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111880918A (en) * | 2020-07-28 | 2020-11-03 | 南京市城市与交通规划设计研究院股份有限公司 | Road network front end rendering method and device and electronic equipment |
| CN112214565A (en) * | 2020-10-15 | 2021-01-12 | 厦门市美亚柏科信息股份有限公司 | Map visual display method, terminal equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101505267A (en) * | 2009-02-24 | 2009-08-12 | 南京联创科技股份有限公司 | Application method of secondary buffer in large concurrent real-time credit control |
| CN102065293A (en) * | 2010-11-23 | 2011-05-18 | 无锡港湾网络科技有限公司 | Image compression method based on space domain predictive coding |
| CN102314711A (en) * | 2010-07-01 | 2012-01-11 | 中国地质科学院矿产资源研究所 | Three-dimensional visualization method and device for mineral resource evaluation information |
-
2012
- 2012-02-29 CN CN2012100493758A patent/CN102622198A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101505267A (en) * | 2009-02-24 | 2009-08-12 | 南京联创科技股份有限公司 | Application method of secondary buffer in large concurrent real-time credit control |
| CN102314711A (en) * | 2010-07-01 | 2012-01-11 | 中国地质科学院矿产资源研究所 | Three-dimensional visualization method and device for mineral resource evaluation information |
| CN102065293A (en) * | 2010-11-23 | 2011-05-18 | 无锡港湾网络科技有限公司 | Image compression method based on space domain predictive coding |
Non-Patent Citations (1)
| Title |
|---|
| 王德清: "基于OpenGL的有限元分析数据可视化系统开发", 《中国优秀硕士学位论文全文数据库（电子期刊）-信息科技辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111880918A (en) * | 2020-07-28 | 2020-11-03 | 南京市城市与交通规划设计研究院股份有限公司 | Road network front end rendering method and device and electronic equipment |
| CN111880918B (en) * | 2020-07-28 | 2021-05-18 | 南京市城市与交通规划设计研究院股份有限公司 | Road network front end rendering method and device and electronic equipment |
| CN112214565A (en) * | 2020-10-15 | 2021-01-12 | 厦门市美亚柏科信息股份有限公司 | Map visual display method, terminal equipment and storage medium |
| CN112214565B (en) * | 2020-10-15 | 2022-06-21 | 厦门市美亚柏科信息股份有限公司 | Map visual display method, terminal equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5154551B2 (en) | Fast reconstruction of graphics pipeline state | |
| US10262392B2 (en) | Distributed and parallelized visualization framework | |
| EP2308224B1 (en) | Gpu scene composition and animation | |
| CN110458905A (en) | Device and method for the adaptive tessellation of level | |
| US8819016B2 (en) | Apparatus, method, and program for structuring visualization object data; and apparatus, method, and program for visualizing visualization object data | |
| Boubekeur et al. | A flexible kernel for adaptive mesh refinement on GPU | |
| CN104036537A (en) | Multiresolution Consistent Rasterization | |
| CN105556565A (en) | Fragment shaders perform vertex shader computations | |
| KR20060044935A (en) | Systems and methods for providing an enhanced graphics pipeline | |
| Sakamoto et al. | KVS: A simple and effective framework for scientific visualization | |
| McDonnel et al. | Towards utilizing gpus in information visualization: A model and implementation of image-space operations | |
| CN111932663B (en) | Parallel drawing method based on multi-level asymmetric communication management | |
| CN103871019A (en) | Optimizing triangle topology for path rendering | |
| US7394464B2 (en) | Preshaders: optimization of GPU programs | |
| CN105303506A (en) | Data parallel processing method and system based on HTML5 | |
| Cottam et al. | Abstract rendering: out-of-core rendering for information visualization | |
| Cottam et al. | Overplotting: Unified solutions under abstract rendering | |
| Okuyan et al. | Direct volume rendering of unstructured tetrahedral meshes using CUDA and OpenMP | |
| CN102622198A (en) | Method and system for visualizing data | |
| Song et al. | A real-time interactive data mining and visualization system using parallel computing | |
| Ioannidis et al. | Multithreaded rendering for cross-platform 3D visualization based on Vulkan Api | |
| CN111681307B (en) | Implementation method of dynamic three-dimensional coordinate axis applied to three-dimensional software | |
| Huang et al. | Parallel‐optimizing SPH fluid simulation for realistic VR environments | |
| Koliha et al. | Towards online visualization and interactive monitoring of real-time CFD simulations on commodity hardware | |
| KR101281156B1 (en) | Ray tracing core and processing mehtod for ray tracing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120801 |