CN101441775A - Texture synthesis method based on matching compatibility - Google Patents
Texture synthesis method based on matching compatibility Download PDFInfo
- Publication number
- CN101441775A CN101441775A CNA2008102394248A CN200810239424A CN101441775A CN 101441775 A CN101441775 A CN 101441775A CN A2008102394248 A CNA2008102394248 A CN A2008102394248A CN 200810239424 A CN200810239424 A CN 200810239424A CN 101441775 A CN101441775 A CN 101441775A
- Authority
- CN
- China
- Prior art keywords
- texture
- texture block
- vein
- block
- synthetic
- 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.)
- Granted
Links
Images
Landscapes
- Image Generation (AREA)
Abstract
The invention discloses a method for synthesizing a vein based on matching compatibility, which belongs to the fields of computer algorithm, computer image technology and vein generation technology, and comprises the following steps: 1) dividing a sample vein into vein blocks capable of reflecting the overall vein characteristics; 2) generating a corresponding matched compatible vein block set for each vein block; 3) selecting one vein block randomly, placing the vein block on an angle of a target vein to be synthesized, and selecting a matched vein block from the corresponding matched compatible vein block set to start synthesizing the vein; and 4) inhibiting and generating a part to be synthesized step by step according to the vein blocks of the synthesized parts until the target vein synthesis is finished. The method does not need to perform a complex similarity calculation on line, and can greatly reduce the amount of the veins required to be inspected when the vein blocks are selected each time, thus the method can better increase the speed for synthesizing the vein and can generate the vein with the size of 1024*1024 at an interactive speed.
Description
Technical field
The present invention relates to a kind of texture synthesis method, belong to computerized algorithm, computer graphics techniques, texture generation technique field, is a kind of composition algorithm that improves the texture combined coefficient based on matching compatibility specifically.
Background technology
The texture synthetic technology can generate visually very similar bulk sample according to the sample of a fritter, can effectively reuse the result of the calculating or the measurement of illumination and color like this, so that generate high-quality drawing result with less expense.This technology has very important in a lot of fields such as sense of reality drafting, virtual reality and uses widely.The texture synthetic technology mainly is to calculate according to the Markov chain probability model at present, and promptly the color of arbitrary position is determined by other COLOR COMPOSITION THROUGH DISTRIBUTION near the certain limit it.According to the range size of calculating, the texture synthetic technology is divided into local synthetic technology and overall synthetic technology.Local synthetic technology is exactly the color of inferring the adjacent domain position according to the COLOR COMPOSITION THROUGH DISTRIBUTION in the less scope, and these class methods in general velocity ratio are very fast, but are difficult to effective maintenance for the feature of overall importance of texture.And overall synthetic technology, then generate earlier a texture of target texture size at random, then based on the similarity measurement of textural characteristics of overall importance, target texture is carried out the progressively optimization of globality, to get target texture to the end, this method can effectively reflect the feature of overall importance of texture, but computing velocity is slow.At present, synthetic top-quality method is that overall synthetic method texture is optimized (textureoptimization) [2] (Kwatra V, Essa I, Bobick A, et al.Texture optimization for example-basedsynthesis.ACM Trans.Graph, 2005,24 (3): 795-802), and the fastest method of aggregate velocity is parallel controlled texture synthesis method (parallel controllable texture synthesis) [3] (Lefebvre S, HoppeH.Parallel controllable texture synthesis.ACM Trans.Graph., 2005,24 (3): 777-786).From the synthetic application demand of texture, it is necessary generating high-quality texture fast, and therefore, the texture synthetic technology is international hot research content in recent years always.
Summary of the invention
For addressing the above problem, the present invention proposes a kind of texture synthesis method based on matching compatibility.
Technical scheme of the present invention is as follows:
A kind of texture synthesis method based on matching compatibility may further comprise the steps:
1) sample texture is divided into the texture block that can reflect its textural characteristics of overall importance;
2) generate its corresponding matched compatible texture block set for each texture block;
3) select a texture block at random, be placed on the angle of target texture to be synthesized, the texture block that selection matches from the set of corresponding matched compatible texture block, the beginning texture is synthetic;
4) retrain step by step according to composite part texture block and generate part to be synthesized, synthetic until finishing target texture.
Described step 1) adopts the packets of information capacitive metric parameter of texture block and periodicity metric parameter to divide texture block.
Described step 2) whether mates in Error Calculation two texture block of overlapping region according to texture block, find left and right, upper and lower 4 matched compatible texture block set of each texture block with the exhaustive search method.
Further, described error calculation method is:
I represents a pixel coordinates of overlapping region, P
iExpression is positioned at the color from the pixel of texture block P of i position, Q
iExpression is positioned at the color from the pixel of texture block Q of i position.
Further, set is optimized processing to the matched compatible texture block, and concrete steps are as follows:
To each texture block, left and right, upper and lower 4 matched compatible texture block set according to it generated generates that it is upper left, lower-left, upper right, 4 matched compatible texture block set in bottom right; And, 4 original matched compatible texture block set are optimized calculating according to these 4 newly-generated matched compatible texture block set, promptly leave out the texture block that those may cause matching conflict.Be without loss of generality, the step that is optimized calculating according to upper left set is as follows:
To the arbitrary texture block in the set of a left side,, then this texture block is left out from the set of a left side if its upper set and upper left set are not occured simultaneously; In like manner same processing is carried out in right, upper and lower set, finished optimization process.
Described step 3) begins texture when synthetic, according to the upper right matched compatible texture block set of this piece, therefrom select a texture block be placed on its upper right side to the position; The texture block that begins respectively left, carries out downwards progressively according to these two texture block is selected at last, and the texture of finishing a level is synthetic.
Described step 4) is dissolved some matching conflict by recalling to calculate, and concrete steps are as follows:
When a texture block to be synthesized is sought in constraint, if its adjacent synthetic texture block relevant common factor when constraint is calculated is empty, then to change,, can generate texture block to be synthesized so that relevant common factor is not empty to these adjacent synthetic texture block;
To the texture part of the generation that texture block retrains that is replaced by these, also to carry out again synthetic calculating to upgrade.
Advantage of the present invention and good effect:
By calculating the matched compatible texture block set that each texture block can amalgamation, can from such set, select suitable texture block easily for use when making synthetic calculating, do not calculate and do not need to carry out online complicated similarity, and can greatly reduce the texture block quantity that to investigate when at every turn selecting texture block for use so yet.Therefore, the present invention can improve the synthetic speed of texture well, and its time complexity is 0 (n), and n is the number of pixels of target texture here; Foundation based on the set of matched compatible texture block can greatly reduce the data search space that coupling is calculated, so the present invention can be with the texture of mutual speed generation 1024*1024 size, and this is that prior art is not accomplished.
Description of drawings
Fig. 1 is a method flow diagram of the present invention
Fig. 2 is the processing sequence synoptic diagram of the synthetic texture of the present invention;
Figure (a) shows that the part of posting texture is synthetic part, and dash area is to be about to synthetic part, and blank parts is the part that will synthesize;
Figure (b) shows, a texture on the earlier synthetic diagonal line synthesizes texture respectively left, downwards along the direction of arrow then, and is synthetic with the texture of finishing a level;
Figure (c) shows the synthetic order of the texture of many levels;
Fig. 3 is contrast computing time of the present invention and parallel controlled texture synthesis method;
Fig. 4 is the contrast that the present invention and texture optimization method, parallel controlled texture synthesis method synthesize quality;
Fig. 5 is the large texture of the present invention with the 1024*1024 of interactive speed generation a), time spent 350ms;
Fig. 5 b) is the large texture of the present invention, time spent 261ms with the 1024*1024 of interactive speed generation.
Embodiment
Below in conjunction with the drawings and specific embodiments the present invention is described in further detail:
As shown in Figure 1, method of the present invention comprises following two key steps: the phase one is a pretreatment stage, and the piecemeal that sample texture is optimized calculates, and sets up the corresponding matched compatible texture block set of optimizing for each texture block then; Subordinate phase is to adopt when calculating efficiently texture block to select to calculate according to the result of phase one in that texture is synthetic, with the high-quality texture of quick generation.
The performing step of above-mentioned phase one is as follows:
1) according to document [1] (Wang Yiping, Wang Wencheng, Wu Enhua. the computation optimization that the piece texture is synthetic. computer-aided design (CAD) and graphics journal, 2006,18 (10): method 1502-1507) is calculated the optimization of texture block and is divided size and dimension, promptly survey the reflection degree of the texture block division of different size and rectangular shape, decide the division of texture block with this to sample texture periodicity global characteristics.Particularly, calculate two kinds of metric parameter, the periodicity tolerance of the packets of information capacitive of texture block tolerance and texture block, and with all reasonable texture block size of these two kinds of metric parameter (this is comparatively speaking, is different to the requirement of different textures as the selection of division; General get preceding separately 20% preferably earlier, it is overlapping to see then whether the two selected block size has, if having, then further improves requirement, best such as preceding 10%; If no, then lower the requirement; Stop seldom the time until the number that finds so overlapping block size).The calculation procedure of these two kinds of parameters is as follows:
The packets of information capacitive tolerance of texture block
A kind of packets of information capacitive of texture block of size is meant the containing reflection degree of the texture block of such size to sample texture information.Its calculation procedure is as follows:
(1) grey level histogram of calculating sample texture;
(2) take out each texture block of size so successively, calculate its grey level histogram;
(3), and calculate Euclidean distance between them with the grey level histogram normalization of texture block and sample texture.Distance is near more, and then texture block has better reflection to the information of sample texture feature of overall importance.
(4) if all the feature of overall importance with sample texture is close texture block for the major part under this size (such as more than 90%), the texture block under this size just has good packets of information capacitive so.
The periodicity tolerance of texture block
Sample texture is divided into bigger grid equably, then to each may size piece, one of the picked at random reference block of size like this in each grid; Subsequently,, last to these reference blocks according to the processing of classifying of the similarity of texture and structural characteristic, the number of the reference block of each class averaged (purpose that is divided into grid is to reduce the texture block number that needs sampling when the periodicity metric calculation; Otherwise, to sample exhaustively, calculated amount is too big).If the reference block under a kind of size can both find similar piece, and the number of these pieces has higher mean value, and then the piece of this size is just relatively good to the periodically variable reflection of texture information.
2) according to 1) result of gained carries out the division of texture block to sample texture, and finds left and right, upper and lower 4 matched compatible texture block set of each texture block with the method for exhaustive search.At this, whether two texture block can mate amalgamation, are to decide according to their Error Calculation in the overlapping region at edge.If error is little, then can mate; Otherwise it is not all right.The error calculation method here is according to document [4] (Liang L, Liu C, Xu Y-Q, et al.Real-time texturesynthesis by patch-based sampling.ACM Trans.Graph., 2001,20 (3): the similarity measurement calculating 127-150) is carried out, that is:
Here, i represents a pixel coordinates of overlapping region, P
iExpression is positioned at the color from the pixel of texture block P of i position, Q
iExpression is positioned at the color from the pixel of texture block Q of i position.
3) according to 2) the result generate upper left, the lower-left of each texture block, the matched compatible texture block set of upper right, bottom right, and thus left and right, upper and lower 4 matched compatible texture block set of each texture block is optimized processing.Being without loss of generality, is that example illustrates that its generative process is as follows to generate upper left matched compatible texture block set:
1) to each texture block in the set of a left side, such as each texture block in A and the upper set, whether such as B, investigating the upper set of A and the left side set of B has common factor.If common factor is arranged, then set belongs to upper left matched compatible texture block set.So investigate various combinations, just can generate last upper left matched compatible texture block set.
2) optimization of matched compatible texture block set is exactly to remove the texture block that those can cause matching conflict.Such as, texture block A in the left side set, for the left side set of all texture block in the upper set, if the upper set of A and they equal common factors are not then left out A from gather on a left side.
The performing step of above-mentioned subordinate phase is as follows:
1) processing sequence of synthetic texture as shown in Figure 2, select a texture block to be placed on an angle of target texture to be synthesized at random, such as the lower left corner, shown in (a) among Fig. 2, then according to the upper right matched compatible texture block set of this piece, therefrom select a texture block be placed on its upper right side to the position; The texture block that begins respectively left, carries out downwards progressively according to these two texture block is selected at last, and the texture of finishing a level is synthetic, shown in (b) among Fig. 2.
2) then as 1) account form, successively generate texture, until the generation of finishing target texture, shown in (c) among Fig. 2 (texture of the corresponding level of texture block on each diagonal line).
3) in above each step, carry out in the selection calculating of texture block, matching conflict may take place, promptly (' constraint ' is meant texture block to be synthesized and texture block with contiguous concatenation according to the constraint of adjacent texture block, the gap of their overlapping regions, border between them is very little) when generating a texture block to be synthesized, their common factor is empty, can not find suitable texture block.At this moment, will recall calculating, dissolve some matching conflict, reduce the matching conflict in the building-up process as far as possible, synthetic calculating can be carried out smootherly by recalling to calculate.Concrete, recall calculating and relate to two kinds of following calculating:
1) when a texture block to be synthesized is sought in constraint, if its adjacent synthetic texture block relevant common factor when constraint is calculated is empty, then to change,, can generate texture block to be synthesized so that relevant common factor is not empty to these adjacent synthetic texture block.
2) because 1) calculating, make some synthetic texture block be replaced.So,, also to carry out again synthetic calculating to upgrade to the texture part of the generation that texture block retrains that is replaced by these.
Finish the generation of target texture through above-mentioned steps
The theoretical analysis and experiment show that time complexity of the present invention is 0 (n), and n is the number of picture elements of target texture here.When generating the texture of sizes such as 128*128,256*256,512*512, the present invention is suitable with at present the fastest parallel controlled texture synthesis method the computing time on the lower processor of operation efficiency, and can alternatively generate the texture of 1024*1024 size, this is that existing method is unapproachable.And aspect synthetic quality, the texture quality that the present invention generates can match in excellence or beauty at present synthetic top-quality texture optimization method.Time complexity analysis of the present invention and experimental result are as follows.
Time complexity is analyzed:
Be without loss of generality, that establishes each piece has k texture block in abutting connection with ensemble average, and target texture can be divided into m
2Individual piece.Generated time of the present invention is made up of two parts, and a part is to handle current region when synthetic, and search draws current to be synthesized common factor from neighborhood; Another part is to recall because of other zone to handle this zone again.Because the time complexity of the set cap here is k
2So the time complexity of first is 0 (k
2* m
2); And second portion causes that at most the number of times of being handled again in a zone is k because recall, so the time complexity of this part is 0 (k*k
2* m
2).Integrate, the time complexity of new method is 0 (k
3* m
2).A large amount of experiments show that k is less than 20, and are little to the influence of complexity because k is smaller, so the piece number (m of the time complexity of this paper method and target texture
2) be roughly linear.Again because the linear relationship between the pixel number of the number of piece and target texture, so the complexity of this paper method is linear with the pixel number of target texture.
Experiment 1: the present invention contrasted with the computing time of parallel controlled texture synthesis method, as shown in Figure 3, the ms express time, pixel represents pixel.Data of the present invention are to obtain on its IntelCore Duo CPU that to run on a dominant frequency be 1.86GHz, and the computing time of parallel controlled texture synthesis method be from its document, extract referring to document [3], its experiment runs on (treatment effeciency of this GPU is better than the present invention and tests used CPU) on a NVIDIA GeForce 6800 Ultra GPU.
Experiment 2: the contrast of the present invention and texture optimization method, the synthetic quality of parallel controlled texture synthesis method, as shown in Figure 4.
Experiment 3: the large texture of the 1024*1024 that the present invention generates with interactive speed, as Fig. 5 a), 5b) shown in, this is that existing method is unapproachable.
Experiment shows, generate 128*128,256*256, during the texture of 512*512 size, the present invention have similar speed, and the present invention is because the foundation of matched compatible texture block set has greatly reduced the size of data space to be processed when coupling is calculated under the processor work efficiency of the parallel controlled texture synthesis method [3] of the efficiency ratio of processor work is wanted slow situation, thereby can be with the texture of mutual speed generation 1024*1024 size, this is that existing method is unapproachable.
Aspect synthetic quality, the piecemeal that the present invention utilizes the work of document [1] that sample texture is optimized, make the texture block feature of overall importance of reflected sample texture efficiently, thus the present invention can generate based on the synthetic calculating of part can efficient reflected sample texture feature of overall importance texture.Simultaneously, the present invention adopts and recalls calculating to dissolve issuable matching conflict in the building-up process, so that synthetic calculating can be carried out smooth-goingly.Experiment shows, synthetic texture result of the present invention can match in excellence or beauty in overall synthetic method, as texture optimization method [2].
Claims (7)
1, a kind of texture synthesis method based on matching compatibility may further comprise the steps:
1) sample texture is divided into the texture block that can reflect its textural characteristics of overall importance;
2) generate its corresponding matched compatible texture block set for each texture block;
3) select a texture block at random, be placed on the angle of target texture to be synthesized, the texture block that selection matches from the set of corresponding matched compatible texture block, the beginning texture is synthetic;
4) retrain step by step according to composite part texture block and generate part to be synthesized, synthetic until finishing target texture.
2, the method for claim 1 is characterized in that, described step 1) adopts the packets of information capacitive metric parameter of texture block and periodicity metric parameter to divide texture block.
3, the method for claim 1, it is characterized in that, described step 2) whether mates in Error Calculation two texture block of overlapping region according to texture block, find left and right, upper and lower 4 matched compatible texture block set of each texture block with the exhaustive search method.
4, method as claimed in claim 3 is characterized in that, described error calculation method is: the error=∑
i(P
i-Q
i)
2, i represents the pixel coordinates of overlapping region, P
iExpression is positioned at the color of pixel of the texture block P of i position, Q
iExpression is positioned at the color of pixel of the texture block Q of i position.
5, method as claimed in claim 3 is characterized in that, generates upper left, lower-left, upper right, 4 matched compatible texture block set in bottom right, and left and right, upper and lower 4 the matched compatible texture block set of rows optimization process to generating is specially:
To the upper set of the arbitrary texture block in the set of a left side, do not occur simultaneously with upper left set, then this texture block is left out from the set of a left side, in like manner same processing is carried out in right, upper and lower set.
6, the method for claim 1 is characterized in that, described step 3) begins texture when synthetic, progressively selects texture block respectively left, downwards according to two known texture block, and the texture of finishing a level is synthetic.
7, the method for claim 1, it is characterized in that, when the constraint of described step 4) generates part to be synthesized, its adjacent synthetic texture block relevant common factor when constraint is calculated is not empty, otherwise adjacent synthetic texture block is changed, and to the texture part of the generation that texture block retrains changed, synthetic again calculating is upgraded.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102394248A CN101441775B (en) | 2008-12-09 | 2008-12-09 | Texture synthesis method based on matching compatibility |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102394248A CN101441775B (en) | 2008-12-09 | 2008-12-09 | Texture synthesis method based on matching compatibility |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101441775A true CN101441775A (en) | 2009-05-27 |
| CN101441775B CN101441775B (en) | 2010-09-08 |
Family
ID=40726194
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008102394248A Expired - Fee Related CN101441775B (en) | 2008-12-09 | 2008-12-09 | Texture synthesis method based on matching compatibility |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101441775B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101807304A (en) * | 2010-03-23 | 2010-08-18 | 中国科学院软件研究所 | Texture synthesis method based on multiplexing |
| CN102298791A (en) * | 2011-09-26 | 2011-12-28 | 清华大学 | Gradient volumetric texture synthesis method |
| CN102385757A (en) * | 2011-10-25 | 2012-03-21 | 北京航空航天大学 | Semantic restriction texture synthesis method based on geometric space |
| CN106157291A (en) * | 2015-04-22 | 2016-11-23 | 阿里巴巴集团控股有限公司 | Identify the method and apparatus repeating texture |
| CN108364276A (en) * | 2018-03-13 | 2018-08-03 | 重庆大学 | Texture image synthetic method based on tag database |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6762769B2 (en) * | 2002-01-23 | 2004-07-13 | Microsoft Corporation | System and method for real-time texture synthesis using patch-based sampling |
| US7012624B2 (en) * | 2003-12-29 | 2006-03-14 | Arcsoft, Inc. | Texture synthesis for repairing damaged images |
| JP4415699B2 (en) * | 2004-02-20 | 2010-02-17 | 凸版印刷株式会社 | Seamless texture composition method, image composition program, and image composition system |
| CN1256707C (en) * | 2004-05-09 | 2006-05-17 | 北京航空航天大学 | Grain synthesizing method based on multiple master drawings |
| US7602398B2 (en) * | 2005-01-28 | 2009-10-13 | Microsoft Corporation | Decorating surfaces with textures |
-
2008
- 2008-12-09 CN CN2008102394248A patent/CN101441775B/en not_active Expired - Fee Related
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101807304A (en) * | 2010-03-23 | 2010-08-18 | 中国科学院软件研究所 | Texture synthesis method based on multiplexing |
| CN102298791A (en) * | 2011-09-26 | 2011-12-28 | 清华大学 | Gradient volumetric texture synthesis method |
| CN102298791B (en) * | 2011-09-26 | 2013-05-08 | 清华大学 | Gradient volumetric texture synthesis method |
| CN102385757A (en) * | 2011-10-25 | 2012-03-21 | 北京航空航天大学 | Semantic restriction texture synthesis method based on geometric space |
| CN102385757B (en) * | 2011-10-25 | 2013-06-05 | 北京航空航天大学 | Semantic restriction texture synthesis method based on geometric space |
| CN106157291A (en) * | 2015-04-22 | 2016-11-23 | 阿里巴巴集团控股有限公司 | Identify the method and apparatus repeating texture |
| CN106157291B (en) * | 2015-04-22 | 2019-07-12 | 阿里巴巴集团控股有限公司 | The method and apparatus that identification repeats texture |
| CN108364276A (en) * | 2018-03-13 | 2018-08-03 | 重庆大学 | Texture image synthetic method based on tag database |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101441775B (en) | 2010-09-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101441775B (en) | Texture synthesis method based on matching compatibility | |
| Lu et al. | Context-aware textures | |
| CN111127610A (en) | Point cloud data three-dimensional visualization rendering method and calculation method | |
| US8130223B1 (en) | System and method for structuring an A-buffer to support multi-sample anti-aliasing | |
| Vasilakis et al. | S-buffer: Sparsity-aware Multi-fragment Rendering. | |
| CN102822870B (en) | The process of 3D computer graphics data on multiple shadowed engines | |
| Lasram et al. | Parallel patch-based texture synthesis | |
| Krishnamurthy et al. | GPU-accelerated Hausdorff distance computation between dynamic deformable NURBS surfaces | |
| Sharma et al. | Point cloud upsampling and normal estimation using deep learning for robust surface reconstruction | |
| CN110660018B (en) | Image-oriented non-uniform style migration method | |
| CN101807304B (en) | Texture synthesis method based on multiplexing | |
| CN102867290B (en) | Texture optimization-based non-homogeneous image synthesis method | |
| Wu | Saliency-aware generative art | |
| Shesh et al. | Efficient and dynamic simplification of line drawings | |
| CN106815879A (en) | A kind of quick texture synthesis method based on LBP features | |
| CN114677468B (en) | Model correction method, device, equipment and storage medium based on reverse modeling | |
| Zakaria et al. | Interactive stylized silhouette for point-sampled geometry | |
| CN102426708B (en) | Texture design and synthesis method based on element reorganization | |
| Parow-Souchon et al. | Early Ahmarian lithic techno-economy and mobility at Al-Ansab 1, Wadi Sabra, southern Jordan | |
| CN101620737B (en) | Texture synthesis method based on interval distribution of texture blocks | |
| Kang et al. | A parallel framework for fast photomosaics | |
| Zhang et al. | Texture synthesis based on direction empirical mode decomposition | |
| Carter et al. | 3-axis cnc path planning using depth buffer and fragment shader | |
| Wang et al. | Texture synthesis via the matching compatibility between patches | |
| Lu et al. | Fast variable center-biased windowing for high-speed stereo on programmable graphics hardware |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100908 Termination date: 20141209 |
|
| EXPY | Termination of patent right or utility model |