CN102867331A - Graphics processing unit (GPU)-orientated large-scale terrain fast drawing method - Google Patents
Graphics processing unit (GPU)-orientated large-scale terrain fast drawing method Download PDFInfo
- Publication number
- CN102867331A CN102867331A CN2012103167178A CN201210316717A CN102867331A CN 102867331 A CN102867331 A CN 102867331A CN 2012103167178 A CN2012103167178 A CN 2012103167178A CN 201210316717 A CN201210316717 A CN 201210316717A CN 102867331 A CN102867331 A CN 102867331A
- Authority
- CN
- China
- Prior art keywords
- terrain
- landform
- block
- node
- lod
- 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
Images
Landscapes
- Image Generation (AREA)
Abstract
The invention provides a graphics processing unit (GPU)-orientated large-scale terrain fast drawing method. The method comprises the following steps of: generating an out-of-core file of original terrain, wherein the out-of-core file stores a plurality of terrain blocks containing original terrain data, an indexing serial number of each terrain block and compressed texture blocks corresponding to the terrain blocks, and the terrain data comprises texture data and altitude data; determining the level of detail (LOD) hierarchical distribution of the terrain according to a predetermined view point, establishing a terrain quadtree structure, traversing the terrain quadtree structure, selecting a corresponding quadtree node for different LODs, and adding the selected quadtree node to a rendering scheduling queue, wherein each node of the terrain quadtree structure corresponds to the corresponding compressed terrain block in the out-of-core file and is used for storing the LOD and the altitude data of the corresponding compressed terrain block; and acquiring the node to be scheduled from the rendering scheduling queue, and rendering the node to be scheduled to finish the drawing of the original terrain.
Description
Technical field
The present invention relates to field of Computer Graphics, relate in particular to a kind of large-scale terrain fast drawing method towards GPU.
Background technology
Landform is the modal scenery of occurring in nature, and it and people's life is closely related.The Drawing of Three-Dimensional Topography technology always is the study hotspot of computer graphics, and in recent years, its development has produced significant impact with progressive various aspects to national economy, is promoting constantly advancing of information age.
Along with photogrammetric and fast development satellite remote sensing technology, the available terrain data precision of people is more and more higher, the thing followed is the terrain data amount that sharply increases, and for computer hardware, although in the same period also experienced the fast-developing stage, but compare with the terrain data amount that increases without limitation, the development of hardware always seems and relatively lags behind, under certain frame frequency, the computer hardware per second can only be drawn the triangle surface of limited quantity, this has just caused limited computer hardware level with the contradiction between the massive terrains data, and therefore large-scale terrain is carried out high-precision drafting often is difficult to obtain gratifying drafting efficient.
A kind of method of effective lifting terrain rendering efficient is to reduce the precision of relief block, namely adopt the landform of the lower model representation the same area of a kind of precision and then realize drafting, level of detail (LOD) technology just with that in mind, by simplifying the complexity of whole scene, minimizing participates in the data volume of drafting, alleviates the contradiction of drawing between precision and the drafting efficient.The another kind of method that improves drafting efficient is external memory (out-of-core) technology, because terrain data is on a grand scale, can't disposable it be written into internal memory, consider that external deposit data reorganization is stored in hard disk, as required, its part is written into when drawing, this just requires to find a kind of data dispatch strategy certainly, at present the main scheduling strategy that adopts is to locate fast required data block by setting up index, then it is written into internal memory.Chinese scholars is in the research of large-scale terrain rendering, obtained a large amount of achievements in research, also proposed simultaneously many outstanding solutions, but because people are very high to drawing precision and drawing the two requirement of efficient, existing solution can not the two be taken into account ideally all.
Summary of the invention
The technical matters that the present invention mainly solves provides a kind of large-scale terrain fast drawing method towards GPU.In order to solve many deficiencies of prior art.
For addressing the above problem, a kind of technical scheme that the present invention adopts is: a kind of large-scale terrain fast drawing method towards GPU is provided, may further comprise the steps:
The external memory file of S1, generation initial landform; Wherein, store a plurality of terrain block that comprise the initial landform data, the index number of each described terrain block and the compressed textures piece corresponding with described terrain block in the described external memory file, described terrain data comprises data texturing and altitude figures;
S2, the default viewpoint parameter of basis determine that the LOD level of landform distributes, set up the landform quad-tree structure, traveling through described landform quad-tree structure is the different corresponding quadtree's nodes of LOD control of LOD, the quadtree's node of choosing is added to play up scheduling queue; Wherein, each node of described landform quad-tree structure is used for storing LOD level and the altitude figures of corresponding compression terrain block corresponding to the corresponding compression terrain block in the described external memory file;
S3, obtain the node that to be scheduled from the described scheduling queue of playing up, and will be played up by the node of height described, finish the drafting to initial landform.
Wherein, described S1 may further comprise the steps:
Obtain altitude figures and the data texturing of initial landform, and altitude figures and the data texturing of described initial landform carried out the hierarchical block processing, obtain the identical terrain block of a plurality of sizes and the texture block corresponding with described terrain block;
Described texture block is compressed, obtained the compressed textures piece, and make up corresponding index number for described terrain block;
Be stored in the file with all described terrain block and with the corresponding index number of each terrain block, compressed textures piece, obtain the external memory file.
Wherein, to described the node that is scheduled is played up among the described S3,, the step of finishing the drafting of initial landform may further comprise the steps:
According to the described terrain data that the node that is scheduled is obtained corresponding terrain block;
The altitude figures of described terrain block is written into video memory as vertex texture, the Plane Gridding Model of setting up landform by fixing vertex cache and indexed cache;
Utilize the Vertex Texture Fetch function of vertex shader to obtain the height value of terrain mesh, the transitional zone of eliminating crack, jump phenomena is set, and carry out the summit conversion and sheet unit processes, the final realization drawn.
Further, the layered approach that adopts of described S1 is: adopt interlacing to resample, adopt bilinear interpolation that the data texturing of described initial landform is resampled every the row method to the altitude figures of described initial landform.
The compression algorithm that adopts when in the above-described embodiment, described S1 compresses the data texturing of described initial terrain block is for adopting the DXT algorithm.
In the technical program, in the LOD of described S2 level distributes: when sighting distance doubled, the LOD rank reduced one-level.
Further, described S2 also carries out following step when the described landform quad-tree structure of traversal is the different corresponding quadtree's node of LOD control of LOD: described landform quad-tree structure is carried out the view frustums cutting.
In above-mentioned concrete embodiment, describedly described landform quad-tree structure is carried out the view frustums cutting be specially: bounding box and view frustums to current quadtree's node intersect test, finish quick view frustums cutting.
The invention has the beneficial effects as follows: be different from prior art and can not effectively promote terrain rendering efficient, the invention provides a kind of large-scale terrain fast drawing method towards GPU, by in conjunction with out-of-core technology, LOD technology, visibility cutting technology and GPU speed technology, can effectively improve the efficient of data dispatch, bring into play fully programmability and the acceleration capacity of GPU, alleviate the CPU computational burden, increase per second and can draw leg-of-mutton quantity, obtain simultaneously very high frame per second, improve the efficient of large-scale terrain rendering.
Description of drawings
Fig. 1 is towards the process flow diagram of the large-scale terrain fast drawing method of GPU in the embodiment;
Fig. 2 is towards the process flow diagram of the large-scale terrain fast drawing method of GPU in another embodiment;
Fig. 3 is an embodiment File storage organization schematic diagram;
Fig. 4 is an embodiment intermediate cam shape striping mode;
Fig. 5 is how much transition schematic diagrams in summit in the embodiment;
Fig. 6 is that an embodiment mesorelief piece transitional zone arranges schematic diagram.
Embodiment
By describing technology contents of the present invention, structural attitude in detail, being realized purpose and effect, below in conjunction with embodiment and cooperate accompanying drawing to give in detail explanation.
See also Fig. 1, present embodiment provides a kind of large-scale terrain fast drawing method towards GPU, may further comprise the steps:
The external memory file of S1, generation initial landform; Wherein, store a plurality of terrain block that comprise the initial landform data, the index number of each described terrain block and the compressed textures piece corresponding with described terrain block in the described external memory file, described terrain data comprises data texturing and altitude figures;
S2, the default viewpoint parameter of basis determine that the LOD level of landform distributes, set up the landform quad-tree structure, traveling through described landform quad-tree structure is the different corresponding quadtree's nodes of LOD control of LOD, the quadtree's node of choosing is added to play up scheduling queue; Wherein, each node of described landform quad-tree structure is used for storing LOD level and the altitude figures of corresponding compression terrain block corresponding to the corresponding compression terrain block in the described external memory file;
S3, obtain the node that to be scheduled from the described scheduling queue of playing up, and will be played up by the node of height described, finish the drafting to initial landform.
In embodiment as shown in Figure 2, described S1 comprises step S110, S120 and S130, and literary composition specific as follows is described.
S110, the altitude figures of obtaining initial landform and data texturing, and the altitude figures of described initial landform and data texturing are carried out hierarchical block process, the identical terrain block of a plurality of sizes and the texture block corresponding with described terrain block obtained.
In the present embodiment, for the layering processing procedure, this method adopts interlacing to resample every the altitude figures of row method to described initial landform, can guarantee like this to remain unchanged in the height value of the point of same geographic location, and can reuse height value; Adopt bilinear interpolation that the data texturing of described initial landform is resampled, like this local detail is preserved better.In the art, for the piecemeal processing procedure, the rule of piecemeal mainly contains two kinds: equidistant piecemeal and homalographic piecemeal, because equidistantly division methods is simple, retrieval rate is fast, and this method adopts equidistant piecemeal.In the present invention, the size of piecemeal does not have concrete computing formula, need to consider according to drawing needs and hardware condition.In the present embodiment, a minute block operations is from the lower left corner of raw data, presses from left to right, and order is from top to bottom carried out, for the marginal date piece at the rightmost side and top, because they can not consist of a square block, just with empty data polishing.In other embodiments, the order of minute block operations can be from the upper left corner of raw data, by from left to right, from top to bottom order carries out.In other embodiment, the order of minute block operations can specifically be set according to actual conditions.
S120, described texture block is compressed, obtained the compressed textures piece, and make up corresponding index number for described terrain block.
Compare with altitude figures, the data texturing amount is relatively large, can take more external space after hierarchical block is processed, and texture is meticulousr, and shared storage space is just larger, when data texturing transmits, also can take more bandwidth, so this method is compressed reasonably to texture.The method of texture compression has a variety of, and in the present embodiment, the compression algorithm that adopts during texture compression is DXT(DirectX Texture Compress).DXT is divided into five kinds of compressed format: DXT1, DXT2, DXT3, DXT4 and DXT5, and wherein the ratio of compression of DXT1 form is 6:1, is applicable to not be with the texture of Alpha component.In the present embodiment, select DXT1 as the Final Format of texture compression, be applied to Direct3D and generate the compressed textures piece.
Because each terrain block can be by the unique expression of LOD rank, line number and column number, so the coding of data is also only needed to consider these three factors.In the present embodiment, this method adopts UINT32 type integer to make up the index coding of terrain block, its form is (DataType(1), BlockLevel(5), BlockX(13), BlockY(13)), i.e. the 1st record data type (altitude figures or data texturing), the 2nd to the 6th record LOD rank, the 7th to the 19th record line number, last 13 record column numbers.
S130, be stored in the file with all described terrain block and with the corresponding index number of each terrain block, compressed textures piece, obtain the external memory file.
In the present embodiment, store all terrain block data with document form, as shown in Figure 3, file is divided into file Head Section, piece index area and blocks of data district.Wherein the file Head Section comprised geography information description to initial landform, landform hierarchy number, every layer in length and breadth to information such as terrain block numbers.The piece index area comprises the index information of all terrain block, has recorded each terrain block side-play amount (take byte as unit), data block size (take byte as unit) hereof, i.e. reference position and the length of this piece in the ground shape file.In the blocks of data district, the metadata of terrain block comprises that terrain block coding, terrain block represent the information such as geographic range.
In the above-described embodiment, the S2 step mainly comprises step S210, S220, S230, and literary composition specific as follows is described.
S210, the default viewpoint parameter of basis determine that the LOD level of landform distributes.
Its detailed process is as follows: with viewpoint and internodal standard apart from distributing as definite LOD, if the Terrain LOD hierachy number is n, the 0th layer is detailed level, the n-1 layer is rough layer, during the Continuous Multiresolution Modeling of tectonic landform, centered by viewpoint, select the node than detailed level to represent from the zone of viewpoint close to, and select the node than rough layer to represent from the zone of viewpoint away from.Particularly, sighting distance is at l
0Select the 0th node layer to represent with interior zone, sighting distance is at l
0And the zone between the l selects the 1st node layer to represent, by that analogy, sighting distance is at l
I-1And l
iBetween the zone select the i node layer to represent, and satisfy l
i=2l
I-1(l≤i≤n-1), wherein l
N-1Being the maximum visual distance, is a customized parameter, and its size directly has influence on the overall level distribution situation of whole landform.Further, when sighting distance doubled, the LOD rank reduced one-level.
S220, set up the landform quad-tree structure.
Its detailed process is as follows: quad-tree structure is by landform altitude figure Information generation, and the degree of depth of tree is by the determine precision of landform, and it is corresponding with the LOD hierachy number of landform.In the present embodiment, the node of quad-tree structure is only preserved the minimum and maximum height value of each node, is used for the AABB bounding box of computing node, other data of node can be when the traversal quad-tree structure Dynamic Acquisition.In a single day quad-tree structure is set up, just memory-resident during the drafting, it just can not change if the initial landform data change.
S230, the described landform quad-tree structure of traversal are the different corresponding quadtree's nodes of LOD control of LOD, the quadtree's node of choosing is added to play up scheduling queue.
Said process is specific as follows: from the root node of quad-tree structure, and the node of minimum level of detail namely, the traversal quaternary tree is that different level of detail is selected corresponding quadtree's node.For the node of i layer, judge its bounding box with centered by viewpoint, l
iFor the spheroid of radius and between position relationship, if enclosure body fully outside spheroid, then withdraws from the judgement to this node, replace it with its father node, father node is joined plays up in the scheduling queue, and continue other child nodes of its father node are judged; If spheroid and enclosure body intersect or comprise, further judge again centered by viewpoint l
I-1Be the spheroid of radius and the position relationship between the enclosure body, if enclosure body fully outside spheroid, then joins this node and plays up in the scheduling queue, otherwise its child node is judged, recurrence repeats this process.
Further, in some embodiment of above-mentioned embodiment, in the traversal quad-tree structure, also can carry out the step that described landform quad-tree structure is carried out the view frustums cutting.Concrete implementation process is as follows: can determine the observability of node by the position relationship of determining bounding box and view frustums, namely bounding box and the view frustums of current quadtree's node be intersected test, and then finish the view frustums cutting.In one embodiment, realize the view frustums cutting by eight summits judging bounding box with the position relationship between six cutting planes of view frustums.
In the above-described embodiment, to described the node that is scheduled is played up among the described S3, the step of finishing the drafting of initial landform comprises step S310, S320, S330, and literary composition specific as follows is described.
S310, according to the described terrain data that the node that is scheduled is obtained corresponding terrain block.
This is to the above-mentioned scheduling process of choosing node, specific as follows: carry out data dispatch take terrain block as unit, open up in advance the data message that a buffer zone is used for receiving terrain block, and arrange one and play up formation and a buffer zone formation, employing LRU(Least Recently Used) mode is managed the terrain data in the buffer zone, specifically describes as follows:
(1) partial information of file Head Section, terrain block index area and terrain block metadata are read in internal memory, set up simultaneously mapping std::map＜uint32, Block*〉BlockMap, with terrain block coding as key assignments, with the terrain block object as the mapping (enum) data type;
(2) will determine that the terrain block that need to play up joins plays up in the formation;
(3) traversal is played up formation, extract each the terrain block coding in the formation, whether each terrain block is searched its coding is present in the buffer zone formation, if there is its coding in the buffer zone formation, in the BlockMap mapping, obtain corresponding terrain block object by encoded radio, and the number of times quantity of state of quoting of terrain block added 1, the coding with this terrain block is set to-1 simultaneously; If do not find its coding in the buffer zone formation, the step below then carrying out is to be written into the terrain block data:
(3a) according to the coding of terrain block, calculate LOD rank corresponding to this terrain block, line number, column number;
(3b) particular location in the calculating terrain block index indexed file is by index acquisition terrain block side-play amount and the landform block size hereof of terrain block;
(3c) utilize the information architecture terrain block object B lock such as coding, altitude figures and data texturing of terrain block, and the Block object is joined among the Mapping B lockMap.
Playing up after formation traversal finishes, the step below carrying out is to upgrade buffer zone:
(A) traversal buffer zone formation is in the buffer zone formation, if the terrain block encoded radio is greater than 0, then with its deletion, discharge terrain block object memory headroom, and the deletion key assignments is the element of this coding in mapping, deletes at last this code element in the buffer zone formation.Less than 0 terrain block, quote number of times according to it for encoded radio in the buffer zone formation, whether determine its deletion is paid the utmost attention to its release for the terrain block of quoting least number of times;
(B) will play up in the formation exist and in the buffer zone formation non-existent terrain block coding be inserted in the buffer zone formation, these codings will be as the part of the buffer zone formation of next frame.
S320, the altitude figures of described terrain block is written into video memory as vertex texture, the Plane Gridding Model of setting up landform by fixing vertex cache and indexed cache.
The texture buffer that the landform altitude data are written into video memory calls in the method the IDirect3DDevice9::CreateTexture method and generates vertex texture as vertex texture.Further, in order to reduce the number of times of drawing function call, we are converted into triangle strip with the grid of model, regular piece take the leaf node size is played up by the form realization of triangle strip as unit, is illustrated in figure 4 as the terrain block triangle strip traversal mode of 5 * 5 sizes and corresponding sequential index tabulation.
Because all terrain block have the summit of identical topological structure and equal number, and the position on summit is to calculate in GPU, in order further to improve the data transmission efficiency of GPU, all terrain block can be used a fixing summit and indexed cache, the corresponding one group of convergent-divergent of each terrain block and migration parameter are used for determining actual range that terrain block is represented and at the particular location in whole landform zone.Simultaneously, same summit connected mode also can be adopted in the inside of terrain block, so the terrain block grid all can generate the network of triangle with regulation summit annexation by fixing index sequence arbitrarily, and then realizes playing up of terrain block.
S330, utilize the Vertex Texture Fetch function of vertex shader to obtain the height value of terrain mesh, the transitional zone of eliminating crack, jump phenomena is set, and carry out the summit conversion and sheet unit processes, the final realization drawn.As shown in Figure 5, V
1And V
0Be respectively in two adjacent levels corresponding to the summit of same level coordinate, in two level handoff procedures, utilize transition weights α to make the summit at V
1And V
0Between the interpolation transition, the vertex position V of any time can be expressed as in the transient process: V=V
1-α (V
1-V
0).
In terrain block, be in the summit in the zones of different, the effect of its transition weights α also is not identical, be in the summit of terrain block boundary, its α is used for guaranteeing the seamless link between terrain block, and is in the summit of terrain block inside, and its α is used for guaranteeing seamlessly transitting between the LOD level.
Boundary in terrain block arranges a transitional zone H, and as shown in Figure 6, the summit at the planimetric coordinates v place at a time transition weights α (t, v) of t can be determined by following formula:
H is the width of transitional zone in the formula, and dist is that planimetric coordinates v is to the distance of transitional zone inner boundary, α
tBe time transition weights, α
sSpace transition weights for the outer boundary place.
For time transition weights α
t, calculate take whole terrain block as unit, for simplified operation, ignore time parameter to α
tImpact, the distance parameter of only choosing between terrain block and viewpoint is asked for α
t, then
α
t=min(max((D-D
min)/(D
max-D
min),0),1)
D is the minimum distance between terrain block bounding box and the viewpoint in the formula, D
MinAnd D
MaxParticipate in two critical distances of network forming for the terrain block of this level of detail.
For space transition weights α
s, at first investigate the stratum level of four adjacent terrain block, if stratum level is identical, the space transition weights α of corresponding outer boundary is set then
sBe 0, namely the vertex position of the outside boundary of current block is constant, if the rank of adjacent terrain block, then arranges the space transition weights α of corresponding outer boundary less than the rank of current block
sBe 1, guarantee the vertex position of this boundary with adjacent identical than rough layer, thereby effectively avoided crack phenomenon.
In sum, be different from prior art and can not effectively promote terrain rendering efficient, the invention provides a kind of large-scale terrain fast drawing method towards GPU, by in conjunction with out-of-core technology, LOD technology, visibility cutting technology and GPU speed technology, can effectively improve the efficient of data dispatch, bring into play fully programmability and the acceleration capacity of GPU, alleviate the CPU computational burden, increase per second and can draw leg-of-mutton quantity, obtain simultaneously very high frame per second, improve the efficient of large-scale terrain rendering.
The above only is embodiments of the invention; be not so limit claim of the present invention; every equivalent structure or equivalent flow process conversion that utilizes instructions of the present invention and accompanying drawing content to do; or directly or indirectly be used in other relevant technical fields, all in like manner be included in the scope of patent protection of the present invention.
Claims (8)
1. large-scale terrain fast drawing method towards GPU is characterized in that may further comprise the steps:
The external memory file of S1, generation initial landform; Wherein, store a plurality of terrain block that comprise the initial landform data, the index number of each described terrain block and the compressed textures piece corresponding with described terrain block in the described external memory file, described terrain data comprises data texturing and altitude figures;
S2, the default viewpoint parameter of basis determine that the LOD level of landform distributes, set up the landform quad-tree structure, traveling through described landform quad-tree structure is the different corresponding quadtree's nodes of LOD control of LOD, the quadtree's node of choosing is added to play up scheduling queue; Wherein, each node of described landform quad-tree structure is used for storing LOD level and the altitude figures of corresponding compression terrain block corresponding to the corresponding compression terrain block in the described external memory file;
S3, obtain the node that to be scheduled from the described scheduling queue of playing up, and will be played up by the node of height described, finish the drafting to initial landform.
2. the large-scale terrain fast drawing method towards GPU according to claim 1 is characterized in that described S1 may further comprise the steps:
Obtain altitude figures and the data texturing of initial landform, and altitude figures and the data texturing of described initial landform carried out the hierarchical block processing, obtain the identical terrain block of a plurality of sizes and the texture block corresponding with described terrain block;
Described texture block is compressed, obtained the compressed textures piece, and make up corresponding index number for described terrain block;
Be stored in the file with all described terrain block and with the corresponding index number of each terrain block, compressed textures piece, obtain the external memory file.
3. the large-scale terrain fast drawing method towards GPU according to claim 1 and 2 is characterized in that to described the node that is scheduled being played up among the described S3, and the step of finishing the drafting of initial landform may further comprise the steps:
According to the described terrain data that the node that is scheduled is obtained corresponding terrain block;
The altitude figures of described terrain block is written into video memory as vertex texture, the Plane Gridding Model of setting up landform by fixing vertex cache and indexed cache;
Utilize the Vertex Texture Fetch function of vertex shader to obtain the height value of terrain mesh, the transitional zone of eliminating crack, jump phenomena is set, and carry out the summit conversion and sheet unit processes, the final realization drawn.
4. the large-scale terrain fast drawing method towards GPU according to claim 2 is characterized in that the layered approach that described S1 adopts is: adopt interlacing to resample, adopt bilinear interpolation that the data texturing of described initial landform is resampled every the row method to the altitude figures of described initial landform.
5. it is characterized in that according to claim 2 or 4 described large-scale terrain fast drawing methods towards GPU: the compression algorithm that described S1 adopts when compressing the data texturing of described initial terrain block is for adopting the DXT algorithm.
6. the large-scale terrain fast drawing method towards GPU according to claim 1 is characterized in that in the LOD of described S2 level distributes: when sighting distance doubled, the LOD rank reduced one-level.
7. the large-scale terrain fast drawing method towards GPU according to claim 1 is characterized in that described S2 also carries out following step when the described landform quad-tree structure of traversal is the different corresponding quadtree's node of LOD control of LOD: described landform quad-tree structure is carried out the view frustums cutting.
8. the large-scale terrain fast drawing method towards GPU according to claim 7, it is characterized in that describedly described landform quad-tree structure is carried out the view frustums cutting being specially: bounding box and view frustums to current quadtree's node intersect test, finish quick view frustums cutting.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012103167178A CN102867331A (en) | 2012-08-31 | 2012-08-31 | Graphics processing unit (GPU)-orientated large-scale terrain fast drawing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012103167178A CN102867331A (en) | 2012-08-31 | 2012-08-31 | Graphics processing unit (GPU)-orientated large-scale terrain fast drawing method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102867331A true CN102867331A (en) | 2013-01-09 |
Family
ID=47446187
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012103167178A Pending CN102867331A (en) | 2012-08-31 | 2012-08-31 | Graphics processing unit (GPU)-orientated large-scale terrain fast drawing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102867331A (en) |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104519353A (en) * | 2013-09-29 | 2015-04-15 | 联想(北京)有限公司 | Image processing method and electronic apparatus |
| CN104679509A (en) * | 2015-02-06 | 2015-06-03 | 腾讯科技（深圳）有限公司 | Graph rendering method and device |
| CN106485776A (en) * | 2016-11-23 | 2017-03-08 | 北京三体高创科技有限公司 | A kind of method and system of 3D game real-time rendering large scale scene |
| CN106997612A (en) * | 2016-01-13 | 2017-08-01 | 索尼互动娱乐股份有限公司 | The apparatus and method of image rendering |
| CN104318509B (en) * | 2014-10-30 | 2017-09-26 | 中国电子科技集团公司第五十四研究所 | A kind of high-volume targeted cache drawing method based on GPU |
| CN107368612A (en) * | 2017-08-24 | 2017-11-21 | 中煤航测遥感集团有限公司 | Quaternary tree index structuring method and device |
| CN107833269A (en) * | 2017-09-19 | 2018-03-23 | 浙江科澜信息技术有限公司 | The transparent method of landform is realized in a kind of three-dimensional scenic |
| CN108027983A (en) * | 2015-09-23 | 2018-05-11 | 皇家飞利浦有限公司 | For the generation of the triangular mesh of 3-D view |
| CN108830929A (en) * | 2018-05-21 | 2018-11-16 | 东南大学 | Multi-resolution Terrain pyramid model generation method and system based on database |
| CN108932742A (en) * | 2018-07-10 | 2018-12-04 | 北京航空航天大学 | A kind of extensive infrared terrain scene real-time rendering method based on remote sensing image classification |
| CN108986212A (en) * | 2018-06-21 | 2018-12-11 | 东南大学 | Three-dimensional virtual terrain L OD model generation method based on crack elimination |
| CN109242967A (en) * | 2018-08-07 | 2019-01-18 | 云南电网有限责任公司曲靖供电局 | A kind of dimensional topography rendering method and device |
| CN109299548A (en) * | 2018-09-28 | 2019-02-01 | 宁波市安贞信息科技有限公司 | A kind of the Building Information Model data publication and method for visualizing of network-oriented |
| CN109499067A (en) * | 2018-11-16 | 2019-03-22 | 网易（杭州）网络有限公司 | A kind of method for drafting and device, electronic equipment, storage medium of terrain texture |
| CN109559376A (en) * | 2018-11-21 | 2019-04-02 | 北京理工大学 | A kind of dimensional topography generation method and device |
| CN109568953A (en) * | 2018-11-26 | 2019-04-05 | 北京像素软件科技股份有限公司 | A kind of terrain texture processing method, device, computer equipment and readable storage medium storing program for executing |
| CN110738721A (en) * | 2019-10-12 | 2020-01-31 | 四川航天神坤科技有限公司 | Three-dimensional scene rendering acceleration method and system based on video geometric analysis |
| CN110852952A (en) * | 2019-11-08 | 2020-02-28 | 四川大学 | GPU-based large-scale terrain real-time rendering method |
| CN110930509A (en) * | 2019-11-27 | 2020-03-27 | 王程 | Linear quad-tree multilayer model driven DEM (digital elevation model) real-time visualization method |
| CN111210516A (en) * | 2019-12-30 | 2020-05-29 | 成都赫尔墨斯科技股份有限公司 | Software platform for comprehensive display control of avionics equipment |
| CN111494944A (en) * | 2020-04-22 | 2020-08-07 | 浙江无端科技股份有限公司 | Terrain texture loading method and related device |
| CN111563948A (en) * | 2020-03-30 | 2020-08-21 | 南京舆图科技发展有限公司 | Virtual terrain rendering method for dynamically processing and caching resources based on GPU |
| WO2021243862A1 (en) * | 2020-06-02 | 2021-12-09 | 完美世界(北京)软件科技发展有限公司 | Data processing method and apparatus |
| CN115953541A (en) * | 2023-03-10 | 2023-04-11 | 中国地质大学（武汉） | Quad-tree LOD terrain generation method, device, equipment and storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1547162A (en) * | 2003-12-05 | 2004-11-17 | 浙江大学 | Texture mapping method based on pyramid quad tree |
-
2012
- 2012-08-31 CN CN2012103167178A patent/CN102867331A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1547162A (en) * | 2003-12-05 | 2004-11-17 | 浙江大学 | Texture mapping method based on pyramid quad tree |
Non-Patent Citations (5)
| Title |
|---|
| 康宁: "基于GPU的全球地形实时绘制技术", 《中国优秀硕士学位论文全文数据库》 * |
| 杨佩理: "地理数据可视化的数据库模型结构与算法", 《沙洲职业工学院学报》 * |
| 温菊屏: "三维地形数据存储和调度方法研究", 《中国优秀硕士学位论文全文数据库》 * |
| 王冬 等: "一种基于GPU的瓦片块四叉树地形绘制算法", 《系统仿真学报》 * |
| 韩永红 等: "利用GPU进行实时大规模3D地形渲染的方法研究", 《计算机工程与应用》 * |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104519353B (en) * | 2013-09-29 | 2019-02-05 | 联想(北京)有限公司 | Image processing method and electronic equipment |
| CN104519353A (en) * | 2013-09-29 | 2015-04-15 | 联想(北京)有限公司 | Image processing method and electronic apparatus |
| CN104318509B (en) * | 2014-10-30 | 2017-09-26 | 中国电子科技集团公司第五十四研究所 | A kind of high-volume targeted cache drawing method based on GPU |
| CN104679509A (en) * | 2015-02-06 | 2015-06-03 | 腾讯科技（深圳）有限公司 | Graph rendering method and device |
| CN108027983A (en) * | 2015-09-23 | 2018-05-11 | 皇家飞利浦有限公司 | For the generation of the triangular mesh of 3-D view |
| CN106997612A (en) * | 2016-01-13 | 2017-08-01 | 索尼互动娱乐股份有限公司 | The apparatus and method of image rendering |
| CN106997612B (en) * | 2016-01-13 | 2022-08-30 | 索尼互动娱乐股份有限公司 | Image rendering apparatus and method |
| CN106485776A (en) * | 2016-11-23 | 2017-03-08 | 北京三体高创科技有限公司 | A kind of method and system of 3D game real-time rendering large scale scene |
| CN107368612A (en) * | 2017-08-24 | 2017-11-21 | 中煤航测遥感集团有限公司 | Quaternary tree index structuring method and device |
| CN107833269A (en) * | 2017-09-19 | 2018-03-23 | 浙江科澜信息技术有限公司 | The transparent method of landform is realized in a kind of three-dimensional scenic |
| CN108830929A (en) * | 2018-05-21 | 2018-11-16 | 东南大学 | Multi-resolution Terrain pyramid model generation method and system based on database |
| CN108986212B (en) * | 2018-06-21 | 2022-05-13 | 东南大学 | Three-dimensional virtual terrain LOD model generation method based on crack elimination |
| CN108986212A (en) * | 2018-06-21 | 2018-12-11 | 东南大学 | Three-dimensional virtual terrain L OD model generation method based on crack elimination |
| CN108932742B (en) * | 2018-07-10 | 2022-09-09 | 北京航空航天大学 | Large-scale infrared terrain scene real-time rendering method based on remote sensing image classification |
| CN108932742A (en) * | 2018-07-10 | 2018-12-04 | 北京航空航天大学 | A kind of extensive infrared terrain scene real-time rendering method based on remote sensing image classification |
| CN109242967A (en) * | 2018-08-07 | 2019-01-18 | 云南电网有限责任公司曲靖供电局 | A kind of dimensional topography rendering method and device |
| CN109242967B (en) * | 2018-08-07 | 2023-04-21 | 云南电网有限责任公司曲靖供电局 | Three-dimensional terrain rendering method and device |
| CN109299548A (en) * | 2018-09-28 | 2019-02-01 | 宁波市安贞信息科技有限公司 | A kind of the Building Information Model data publication and method for visualizing of network-oriented |
| CN109299548B (en) * | 2018-09-28 | 2023-09-01 | 中航安贞(浙江)信息科技有限公司 | Network-oriented building information model data publishing and visualizing method |
| CN109499067A (en) * | 2018-11-16 | 2019-03-22 | 网易（杭州）网络有限公司 | A kind of method for drafting and device, electronic equipment, storage medium of terrain texture |
| CN109499067B (en) * | 2018-11-16 | 2022-05-31 | 网易（杭州）网络有限公司 | Terrain texture drawing method and device, electronic equipment and storage medium |
| CN109559376A (en) * | 2018-11-21 | 2019-04-02 | 北京理工大学 | A kind of dimensional topography generation method and device |
| CN109568953A (en) * | 2018-11-26 | 2019-04-05 | 北京像素软件科技股份有限公司 | A kind of terrain texture processing method, device, computer equipment and readable storage medium storing program for executing |
| CN110738721A (en) * | 2019-10-12 | 2020-01-31 | 四川航天神坤科技有限公司 | Three-dimensional scene rendering acceleration method and system based on video geometric analysis |
| CN110738721B (en) * | 2019-10-12 | 2023-09-01 | 四川航天神坤科技有限公司 | Three-dimensional scene rendering acceleration method and system based on video geometric analysis |
| CN110852952A (en) * | 2019-11-08 | 2020-02-28 | 四川大学 | GPU-based large-scale terrain real-time rendering method |
| CN110930509A (en) * | 2019-11-27 | 2020-03-27 | 王程 | Linear quad-tree multilayer model driven DEM (digital elevation model) real-time visualization method |
| CN111210516A (en) * | 2019-12-30 | 2020-05-29 | 成都赫尔墨斯科技股份有限公司 | Software platform for comprehensive display control of avionics equipment |
| CN111210516B (en) * | 2019-12-30 | 2023-04-18 | 成都赫尔墨斯科技股份有限公司 | Software platform for integrated display control of avionics equipment |
| CN111563948A (en) * | 2020-03-30 | 2020-08-21 | 南京舆图科技发展有限公司 | Virtual terrain rendering method for dynamically processing and caching resources based on GPU |
| CN111494944A (en) * | 2020-04-22 | 2020-08-07 | 浙江无端科技股份有限公司 | Terrain texture loading method and related device |
| WO2021243862A1 (en) * | 2020-06-02 | 2021-12-09 | 完美世界(北京)软件科技发展有限公司 | Data processing method and apparatus |
| CN115953541A (en) * | 2023-03-10 | 2023-04-11 | 中国地质大学（武汉） | Quad-tree LOD terrain generation method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102867331A (en) | Graphics processing unit (GPU)-orientated large-scale terrain fast drawing method | |
| CN108133044B (en) | Spatial big data three-dimensional visualization method and platform based on attribute separation | |
| CN103093499B (en) | A kind of city three-dimensional model data method for organizing being applicable to Internet Transmission | |
| CN101441779B (en) | Rapid dynamic drafting method of magnanimity digital terrain | |
| CN102663801B (en) | Method for improving three-dimensional model rendering performance | |
| CN105760529A (en) | Spatial index and cache construction method for vector data of mobile terminal | |
| CN105261066B (en) | A kind of three-dimensional geographic information system real-time rendering multithreading distribution and control method | |
| CN105956165B (en) | A kind of big file storage organization method of the tile type of three-dimensional modeling data | |
| CN107423445A (en) | A kind of map data processing method, device and storage medium | |
| CN102117500B (en) | Three-dimensional modeling method and system | |
| CN102930594B (en) | Rendering method for computer three-dimensional terrain grid based on boundary maintenance | |
| CN102306180A (en) | Modeling method based on mass laser radar grid point cloud data | |
| CN113628314B (en) | Visualization method, device and equipment for photographic measurement model in illusion engine | |
| CN110992469A (en) | Visualization method and system for massive three-dimensional model data | |
| CN103412962A (en) | Storage method and reading method for mass tile data | |
| CN101963996A (en) | Global raster data storing method | |
| CN103077549A (en) | Real-time large-scale terrain visualization implementation method based on kd tree | |
| CN107590226A (en) | A kind of map vector rendering intent based on tile | |
| CN113628331B (en) | Data organization and scheduling method for photogrammetry model in illusion engine | |
| CN101593361A (en) | A kind of large-scale terrain rendering system based on double-layer nested grid | |
| CN103268342A (en) | DEM dynamic visualization accelerating system and method based on CUDA | |
| CN109636889A (en) | A kind of Large Scale Terrain model rendering method based on dynamic suture zone | |
| CN112380302B (en) | Thermodynamic diagram generation method and device based on track data, electronic equipment and storage medium | |
| CN112070909A (en) | Engineering three-dimensional model LOD output method based on 3D Tiles | |
| CN108830929A (en) | Multi-resolution Terrain pyramid model generation method and system based on database |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130109 |