CN108830922A - A kind of profile tree constructing method based on multithreading - Google Patents
A kind of profile tree constructing method based on multithreading Download PDFInfo
- Publication number
- CN108830922A CN108830922A CN201810432235.6A CN201810432235A CN108830922A CN 108830922 A CN108830922 A CN 108830922A CN 201810432235 A CN201810432235 A CN 201810432235A CN 108830922 A CN108830922 A CN 108830922A
- Authority
- CN
- China
- Prior art keywords
- tree
- data set
- sub
- profile
- profile tree
- 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
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 5
- 238000004422 calculation algorithm Methods 0.000 claims description 18
- 239000000203 mixture Substances 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000000354 decomposition reaction Methods 0.000 claims description 2
- 230000002688 persistence Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 7
- 230000006870 function Effects 0.000 description 19
- 238000009877 rendering Methods 0.000 description 19
- 230000005540 biological transmission Effects 0.000 description 14
- 238000013461 design Methods 0.000 description 7
- 238000012546 transfer Methods 0.000 description 5
- 238000012800 visualization Methods 0.000 description 5
- 230000002452 interceptive effect Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000000007 visual effect Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000003814 drug Substances 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005266 casting Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 238000013079 data visualisation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000010422 painting Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000011112 process operation Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 210000003625 skull Anatomy 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000010937 topological data analysis Methods 0.000 description 1
- 238000007794 visualization technique Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/08—Volume rendering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline or look ahead
- G06F9/3867—Concurrent instruction execution, e.g. pipeline or look ahead using instruction pipelines
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computer Graphics (AREA)
- General Engineering & Computer Science (AREA)
- Image Analysis (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention discloses a kind of profile tree constructing method based on multithreading, firstly, obtain the slice map image set that image documentation equipment generates, according to from top to down or to it is lower and on sequentially form 3-D data set；Then 3-D data set is executed into data etc. point operation, it is divided into multiple Sub Data Sets according to range intervals；Then each Sub Data Set is operated respectively, utilize multithreading, the splay tree and threaded tree that 2 threads calculate profile tree are distributed for each Sub Data Set, and then merge as sub- profile tree, the sub- profile tree that each Sub Data Set is formed again merges, a complete profile tree is formed, finally the topological structure of the profile tree of formation is simplified, obtains final required profile tree.The present invention, which can not only obtain, clearly indicates effect, while can flexibly interact.
Description
Technical field
The present invention relates to Medical Image Processing and the technical fields of application, refer in particular to a kind of profile tree based on multithreading
Construction method.
Background technique
Volume drawing based on optical properties is the visual mainstream technology of medical data.Volume drawing simulates light in volume data
Transmission process, and then calculate the contribution done of each point in transmission path.It is distributed by the sampled point that transmission function is volume data
Color and transparency property, and the light projecting algorithm based on GPU is combined, vivid 3-D image can be reconstructed, such as
Document《Efficient Empty Space Skipping for Large-Scale Volume Rendering》(IEEE
transactions on visualization and computer graphics,2018,24(1):974-983).Therefore,
The quality of transmission function design determines the quality of volume drawing rendering.
Effective Volume Rendering Techniques need designed transmission function, have had many scholars to be made that at present rather good
Achievement, such as document《State of the Art in Transfer Functions for Direct Volume
Rendering》(Computer Graphics Forum,2016,35(3):669-691).Use value-histogram of gradients guidance
Transmission function can be such that volume data quickly clearly shows.Using 2D function come classification data and feature, rendering figure can be increased
The quality of picture.Measure data local shape attribute first is that curvature, document《Non-uniform sampling of
geometry for the numeric simulation of head-related transfer functions》
(Proceedings of the 21st International Congress of Sound and Vibration,
(Beijing, CN) .2014) design of transmission function is instructed using curvature as parameter.The image segmentation of two-dimensional histogram and spy
Sign measurement can be improved volume drawing effect, such as document《Texture-based transfer functions for direct
volume rendering》(IEEE Transactions on Visualization&Computer Graphics,2008,
14(6):1364)。
Existing transmission function design can effectively instruct the rendering of volume drawing.However, complicated for space structure
Volume data, it is difficult to clearly give expression to the frontier properties of substance.Medical application needs finer structure to carry out accurately
Diagnosis, this just needs to have institutional framework better differentiation.For the volume data of labyrinth, topological structure can analyze its sky
Between topology, and then instruct the design of transmission function.There are many effective topology methods to be used in volume drawing rendering pipe
In line.For the topological analysis of volume data, Reeb figure, Morse-Smale complex, profile tree scheduling algorithm have been applied to many
In research field, such as document《Direct Feature Visualization Using Morse-Smale Complexes》
(IEEE Transactions on Visualization&Computer Graphics,2012,18(9):1549-1562)、
《Measuring distance between Reeb graphs》(Proceedings of the thirtieth annual
symposium on Computational geometry.ACM,2014:464).Profile tree algorithm has more with respect to other algorithms
There is practicability.However, increasing as the continuous growth of volume data resolution ratio and profile tree algorithm need multiple ergodic data collection
The technical problem of data topology analysis, including computing cost.Especially for can be interacted in data analysis process
Operation, the experience for needing efficient algorithm that can just be got well.Parallel algorithm is introduced into the calculating of profile tree, it can be effective
Raising arithmetic speed.Profile tree developing algorithm based on distributed environment, such as document《Distributed Contour
Trees》(Topological Methods in Data Analysis and Visualization III.Springer,
Cham,2014:89-102), still, multiple serial computing is needed in overall process, reduces operational performance.
Above-mentioned algorithm is all that the calculating of part stage has used parallel, and the building speed of profile tree is slow, and due to profile
Tree rendering needs complicated pretreatment, and interactive operation is not satisfactory.In order to overcome the defect, the present invention carries out in terms of two
It improves：1) the multicore ability for utilizing CPU, proposes a kind of profile tree developing algorithm based entirely on multithreading；2) based on quickly wheel
Wide tree algorithm, using new render mode, realize it is a kind of can real-time, interactive volume drawing.It, can not only be high by improving above
The drafting effect of the display volume data of effect, while real time interactive operation can be more easily carried out, find interested feature.
Summary of the invention
The purpose of the present invention is to overcome the shortcomings of the existing technology and deficiency, proposes a kind of profile tree based on multithreading
Construction method can complete the parallel computation of profile tree, especially based on the partition strategy of range driving using shared drive
For complexity and comprising the data of noise, parallel implementation can reach in speed in the load and calculating of redundant data
Equilibrium, meanwhile, the profile tree of method building can explicitly divide volume data with good level, thus energy
The design for enough instructing body painting modulation trnasfer function can effectively be added on the basis of the segmentation of profile tree using new render mode
The rendering of fast profile tree branch, and to the flexible operation of profile tree, interactivity is strong, can solve medicine volume rendered images boundary
Fuzzy problem.
To achieve the above object, technical solution provided by the present invention is：A kind of profile tree building side based on multithreading
Method, firstly, obtain image documentation equipment generate slice map image set, according to from top to down or to it is lower and on sequentially form three dimensions
According to collection；Then 3-D data set is executed into data etc. point operation, it is divided into multiple Sub Data Sets according to range intervals；Then divide
It is other that each Sub Data Set is operated, using multithreading, the division that 2 threads calculate profile tree is distributed for each Sub Data Set
Tree and threaded tree, and then merge and merged as sub- profile tree, then the sub- profile tree that each Sub Data Set is formed, form one
A complete profile tree, finally simplifies the topological structure of the profile tree of formation, obtains final required profile tree；It is wrapped
Include following steps：
1) according to the value range of the size of data set and data, in conjunction with the multithreading number n of computer supported, by data set
It is divided into n/2 one's share of expenses for a joint undertaking data set；
2) splay tree and threaded tree that 2 threads are used to carry out calculating Sub Data Set are distributed for each Sub Data Set, and will
Splay tree and threaded tree merge into sub- profile tree；
3) the sub- profile tree on all Sub Data Sets is merged into a complete profile tree；
4) topological structure of the profile tree after merging to step 3) simplifies, and obtains final required profile tree.
In step 1), data set is subjected to equal part, enabling Thread Count is n, data set is divided into n/2 Sub Data Set, often
A Sub Data Set contains the vertex of equal number：
Wherein, R indicates that real number, μ indicate that the element of data set, i, j indicate the order of Sub Data Set, indicate ξiIndicate i-th
A Sub Data Set, | δ0|iIndicate ξiIn vertex number；
Divide this operation to complete data etc., needs following steps：
1.1) quick sorting algorithm is used, with vertex all on sequence arrangement data set from small to large, uses array
Storage element；
1.2) sorted array is subjected to equal part, is divided into n/2 array θi, θiCorresponding Sub Data Set is ξi；
1.3) θ is extendediIt is expanded into θi', θi' it is by θiIn additionWithConnected vertex composition, enables fi +、fi -For ξiEnd
Point, i.e. ξi=(fi -,fi +), definitionIt is and fi +The side of connection,It is and fi -The side of connection.
In step 2), the local calculation for carrying out profile tree to each Sub Data Set is operated, to calculate the profile tree that this is walked,
To each data set ξiIt performs the following operations：
2.1) Sub Data Set ξ is scanned with increasing orderiExpanded set θi', wheel is constructed using classical Union-find Sets algorithm
The threaded tree of exterior feature tree；
2.2) Sub Data Set ξ is scanned with descending orderiExpanded set θi', profile is constructed using classical Union-find Sets algorithm
The splay tree of tree；
2.3) merge threaded tree and splay tree that first two steps obtain, the node that degree is 2 is merged, profile tree is formed.
In step 3), the sub- profile tree on each Sub Data Set is as input, by the sub- profile tree of all Sub Data Sets
Merging becomes a complete profile tree, remembers that the sub- profile tree on Sub Data Set is γ (f)i, data vertex set thereon isArc between the sub- profile tree in front and back is denoted as The collection of all arc compositions is combined into arc={ arc0,
arc1···arcn/2-1, since there are one-to-many relationships between Sub Data Set, arc is traversed using binary chop, will be met
The arc of relationship connects, and when having traversed all elements in arc, can form a complete profile tree.
In step 4), the topological structure of the profile tree after merging to step 3) simplifies, and comprises the steps of：
4.1) the geometrical characteristic attribute of profile tree is chosen：Persistence, volume, hypervolume；
4.2) calculating of multiresolution is carried out using branch's decomposition method to the geometrical characteristic of selection；
4.3) branch that resolution ratio is less than setting value is merged, the profile tree being simplified.
Compared with prior art, the present invention having the following advantages that and beneficial effect：
1, the present invention solves the object boundary fuzzy problem of conventional transmission function division, the transmitting instructed using profile tree
Function can clearly indicate the marginal information of object, help to highlight interested position.
2, the drawbacks of the present invention overcomes traditional multithreading profile tree methods, it is fully parallelized, improve the building of method
Speed.
3, this invention removes the interactions between thread, have given full play to mostly the parallel ability of thread, and the speed of operation reaches
To live effect.
4, the present invention combines profile tree method and the ray casting volume rendering algorithm based on GPU for the first time, realizes efficient
Medicine volume drawing.
5, the present invention uses new delay render mode, and using the segmented characterizing of profile tree, interactive operation is more abundant, wash with watercolours
It is more preferable to contaminate effect.
Detailed description of the invention
Fig. 1 is the logical flow chart of the method for the present invention.
Fig. 2 is instance graph of the present invention using a simple profile tree building.
Fig. 3 is ICRM and DCRM mode comparison diagram.
Fig. 4 is to be compared using the rendering result of the two-dimentional transfer function method of traditional density-gradient value and the method for the present invention
Figure.
Fig. 5 is visual runnable interface of the invention.
Specific embodiment
Below with reference to specific implementation case, the present invention is further illustrated.
As shown in Figure 1, the profile tree constructing method based on multithreading provided by this example, detailed process are：Firstly, obtaining
Take image documentation equipment generate slice map image set, according to from top to down or to it is lower and on sequentially form 3-D data set；Then
3-D data set is executed into data etc. point operation, it is divided into multiple Sub Data Sets according to range intervals；Then respectively to each
Sub Data Set operation distributes splay tree and the connection that 2 threads calculate profile tree using multithreading for each Sub Data Set
Tree, and then merge as sub- profile tree.The sub- profile tree that finally each Sub Data Set is formed merges, and forms complete wheel
Exterior feature tree, finally simplifies the topological structure of the profile tree of formation, obtains final required profile tree, so far, profile tree
Building process terminates.Fig. 2 shows the building diagram of profile tree by a simple example, specifically includes following steps：
1) according to the value range of the size of data set and data, data set is divided into Sub Data Set, is included the following steps：
1.1) data set is ranked up, highly represents the value range of data, the data set after sequence be [a, b, c, d, e,
f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v]。
1.2) by the data after sequence be divided into 2 Sub Data Sets [a, b, c, d, e, f, g, h, i, j, k] and [l, m, n, o,
p,q,r,s,t,u,v]。
1.3) Sub Data Set is extended, m, o, l, k, n is the edge vertices of adjacent Sub Data Set, adds it to 2 1.2)
In a Sub Data Set, formed new Sub Data Set [a, b, c, d, e, f, g, h, i, j, k, l, m, n, o] and [l, m, n, o, p, q, r,
s,t,u,v]。
2) two threads are distributed, calculate separately the sub- profile tree on Sub Data Set, steps are as follows for execution：
2.1) increasing scans Sub Data Set, constructs threaded tree using Union-find Sets；
2.2) backward scans Sub Data Set, constructs splay tree using Union-find Sets；
2.3) splay tree and threaded tree are merged into sub- profile tree.
It is the sub- profile tree that Sub Data Set [a, b, c, d, e, f, g, h, i, j, k, l, m, n, o] is formed on the right of (a) in Fig. 2.
It is the sub- profile tree that Sub Data Set [l, m, n, o, p, q, r, s, t, u, v] is formed on the right of (b) in Fig. 2.
3) the sub- profile tree for merging each Sub Data Set, forms complete profile tree, includes the following steps：
3.1) arc of each Sub Data Set and edge vertices formation is searched；[a,b,c,d,e,f,g,h,i,j,k,l,m,n,
O] arc be (j, k), the arc of [l, m, n, o, p, q, r, s, t, u, v] is (k, j).
3.2) the corresponding arc of adjacent Sub Data Set is connected, forms complete profile tree.
(c) is the integrity profile tree after 2 sub- data sets merge in Fig. 2.
Volume drawing needs to carry out real time communication between CPU and GPU, and the data volume of volume data is usually bigger, this passes CPU
Transmission of data causes very big load.In traditional Immediate Context render mode (abbreviation ICRM), between CPU and GPU
Interaction can only be completed by core cpu, other core cpus can not be utilized, and caused and greatly calculated money
Source waste.For this purpose, selecting newest Deferred Contexts render mode (abbreviation DCRM) here to complete communication scheduling.
DCRM can use the drawing command that multi-core CPU completes GPU.As shown in figure 3, ICRM can only using monokaryon CPU to volume data into
Row processing, gives GPU then to complete actual Rendering operations.It is segmented characterizing that DCRM can make full use of profile tree, will be segmented
Data afterwards are distributed in each core of CPU and go to execute, and finally uniformly give GPU rendering.It can not only accelerate to render in this way
Efficiency, moreover it is possible to it is independent that each segment data is operated, choose desired part and highlight, to the valence of clinical application
It is worth larger.
In order to test the practical application effect of the above-mentioned profile tree constructing method based on multithreading of the present embodiment, we are used
Visualization method applies the invention on true medical data, and the present invention is based on the quick profile tree structures of multithreading for verifying
The validity and feasibility of construction method.
The present invention realizes that experiment porch is Dell M4800 work station, I7 using C Plus Plus and Directx shape library
4810 processors, 16G memory, Quadro k2100 video card.This experiment porch the method for the present invention and system are to 3 D medical figure
As data are handled, rendering image is finally obtained.
1, the validity of the profile tree constructing method based on multithreading to illustrate the invention uses open source library
The serial implementation (SITour) and Parallel Implementation (PITour) and the method for the present invention of libtourtre is as a comparison.The data of experiment
Collection is Tooth (256x256x161), Foot (256x256x256), Head (256x256x225), Skull respectively
(256x256x256), Engine (256x256x256), the size of the internal representation data collection in bracket.
Table 1 shows five groups of data in the comparing result of SITour, PITour method and the method for the present invention, and n indicates thread
Number.As can be seen that the method for the present invention has preferable speed to be promoted in table.With the promotion of Thread Count, the advantage of the method for the present invention
It is more obvious.
Table 1 constructs rate results and compares (unit:Second)
Fig. 4 is using the two-dimentional transfer function method of traditional density-gradient value and using the rendering result of the method for the present invention
Comparison.It can be seen from the figure that more preferable using the picture quality that the method for the present invention renders, details is clear, the division of material boundary
It is relatively obvious.
2, it is the feasibility for verifying the method for the present invention, compares the rendering frame frequency of both of which.The visualization interface of its mode,
As shown in Figure 5.It is the design panel of transmission function on the left of interface, is mainly used to specify suitable transmission function for segment data.
Right side is illumination setting panel, specifies illumination parameter, enhances the sense of reality of rendering.Intermediate panel is used to show the knot of rendering
Fruit.The panel of lower-left is the display panel of the setting panel and profile tree of profile tree, by flexibly changing the branch of profile tree,
Any interested region can be chosen.
Table 2 is the operation frame frequency comparing result of the DCRM mode that the method for the present invention uses and traditional ICRM mode.From table
In result in as can be seen that the method for the present invention use mode have 40%~50% speed promoted, can reach preferable
Visual effect.
2 frame frequency comparison result of table
In conclusion the present invention provides new method for the clinical diagnosis and treatment of medical image processing.It is examined in auxiliary
In disconnected, the design of transmission function is instructed with profile tree, image clear in structure can be generated, is a kind of effective mode.
And the method for the present invention takes full advantage of the characteristic of profile tree, can generate the image of high quality, has actual promotional value, value
It must promote.
Embodiment described above is only the preferred embodiments of the invention, and but not intended to limit the scope of the present invention, therefore
All shapes according to the present invention change made by principle, should all be included within the scope of protection of the present invention.
Claims (5)
1. a kind of profile tree constructing method based on multithreading, it is characterised in that：Firstly, obtaining the slice map that image documentation equipment generates
Image set, according to from top to down or to it is lower and on sequentially form 3-D data set；Then 3-D data set is executed into data etc.
Divide operation, it is divided into multiple Sub Data Sets according to range intervals；Then each Sub Data Set is operated respectively, utilizes multithreading
Technology is distributed the splay tree and threaded tree that 2 threads calculate profile tree for each Sub Data Set, and then is merged as sub- profile
Tree, then the sub- profile tree that each Sub Data Set is formed merge, and a complete profile tree are formed, finally to the wheel of formation
The topological structure of exterior feature tree is simplified, and final required profile tree is obtained；It includes the following steps：
1) according to the value range of the size of data set and data, in conjunction with the multithreading number n of computer supported, by data set equal part
At n/2 one's share of expenses for a joint undertaking data set；
2) splay tree and threaded tree that 2 threads are used to carry out calculating Sub Data Set are distributed for each Sub Data Set, and will division
Tree and threaded tree merge into sub- profile tree；
3) the sub- profile tree on all Sub Data Sets is merged into a complete profile tree；
4) topological structure of the profile tree after merging to step 3) simplifies, and obtains final required profile tree.
2. a kind of profile tree constructing method based on multithreading according to claim 1, it is characterised in that：In step 1)
In, data set is subjected to equal part, enabling Thread Count is n, data set is divided into n/2 Sub Data Set, each Sub Data Set contains phase
Etc. numbers vertex：
Wherein, R indicates that real number, μ indicate that the element of data set, i, j indicate the order of Sub Data Set, ξiIndicate i-th of subdata
Collection, | δ0|iIndicate ξiIn vertex number；
Divide this operation to complete data etc., needs following steps：
1.1) quick sorting algorithm is used, with vertex all on sequence arrangement data set from small to large, uses storage of array
Element；
1.2) sorted array is subjected to equal part, is divided into n/2 array θi, θiCorresponding Sub Data Set is ξi；
1.3) θ is extendediIt is expanded into θ 'i, θ 'iIt is by θiIn additionWithConnected vertex composition, enables fi +、fi -For ξiEndpoint, i.e.,
ξi=(fi -,fi +), definitionIt is and fi +The side of connection,It is and fi -The side of connection.
3. a kind of profile tree constructing method based on multithreading according to claim 1, it is characterised in that：In step 2)
In, the local calculation for carrying out profile tree to each Sub Data Set operates, to calculate the profile tree that this is walked, to each data set ξiIt holds
The following operation of row：
2.1) Sub Data Set ξ is scanned with increasing orderiExpanded set θ 'i, profile tree is constructed using classical Union-find Sets algorithm
Threaded tree；
2.2) Sub Data Set ξ is scanned with descending orderiExpanded set θ 'i, utilize classical Union-find Sets algorithm building profile tree
Splay tree；
2.3) merge threaded tree and splay tree that first two steps obtain, the node that degree is 2 is merged, profile tree is formed.
4. a kind of profile tree constructing method based on multithreading according to claim 1, it is characterised in that：In step 3)
In, for the sub- profile tree on each Sub Data Set as input, the sub- profile tree of all Sub Data Sets, which is merged, becomes one completely
Profile tree, remember Sub Data Set on sub- profile tree be γ (f)i, data vertex set thereon is [fi -,fi +], with front and back sub- wheel
Arc between exterior feature tree is denoted asThe collection of all arc compositions is combined into arc={ arc0,arc1…arcn/2-1, due to
There are one-to-many relationship between Sub Data Set, arc is traversed using binary chop, the arc for the relationship that meets is connected, when time
The all elements in arc have been gone through, a complete profile tree can be formed.
5. a kind of profile tree constructing method based on multithreading according to claim 1, it is characterised in that：In step 4)
In, the topological structure of the profile tree after merging to step 3) simplifies, and comprises the steps of：
4.1) the geometrical characteristic attribute of profile tree is chosen：Persistence, volume, hypervolume；
4.2) calculating of multiresolution is carried out using branch's decomposition method to the geometrical characteristic of selection；
4.3) branch that resolution ratio is less than setting value is merged, the profile tree being simplified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810432235.6A CN108830922B (en) | 2018-05-08 | 2018-05-08 | Contour tree construction method based on multiple threads |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810432235.6A CN108830922B (en) | 2018-05-08 | 2018-05-08 | Contour tree construction method based on multiple threads |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108830922A true CN108830922A (en) | 2018-11-16 |
| CN108830922B CN108830922B (en) | 2022-02-11 |
Family
ID=64148365
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810432235.6A Active CN108830922B (en) | 2018-05-08 | 2018-05-08 | Contour tree construction method based on multiple threads |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108830922B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109977715A (en) * | 2019-03-05 | 2019-07-05 | 哈尔滨工业大学（深圳） | Two-dimensional code identification method and two dimensional code based on outline identification |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050137726A1 (en) * | 2003-12-22 | 2005-06-23 | Siemens Aktiengesellschaft | System and method for generating a control program |
| US20070036434A1 (en) * | 2005-08-15 | 2007-02-15 | Peter Saveliev | Topology-Based Method of Partition, Analysis, and Simplification of Dynamical Images and its Applications |
| JPWO2005111944A1 (en) * | 2004-05-19 | 2008-03-27 | 国立大学法人京都大学 | Data display method, data display device, data display program, and computer-readable recording medium recording the same |
| US20170365094A1 (en) * | 2016-04-04 | 2017-12-21 | University Of Cincinnati | Localized Contour Tree Method for Deriving Geometric and Topological Properties of Complex Surface Depressions Based on High Resolution Topographical Data |
-
2018
- 2018-05-08 CN CN201810432235.6A patent/CN108830922B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050137726A1 (en) * | 2003-12-22 | 2005-06-23 | Siemens Aktiengesellschaft | System and method for generating a control program |
| JPWO2005111944A1 (en) * | 2004-05-19 | 2008-03-27 | 国立大学法人京都大学 | Data display method, data display device, data display program, and computer-readable recording medium recording the same |
| US20070036434A1 (en) * | 2005-08-15 | 2007-02-15 | Peter Saveliev | Topology-Based Method of Partition, Analysis, and Simplification of Dynamical Images and its Applications |
| US20170365094A1 (en) * | 2016-04-04 | 2017-12-21 | University Of Cincinnati | Localized Contour Tree Method for Deriving Geometric and Topological Properties of Complex Surface Depressions Based on High Resolution Topographical Data |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109977715A (en) * | 2019-03-05 | 2019-07-05 | 哈尔滨工业大学（深圳） | Two-dimensional code identification method and two dimensional code based on outline identification |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108830922B (en) | 2022-02-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gansner et al. | Multilevel agglomerative edge bundling for visualizing large graphs | |
| DE102021121109A1 (en) | RECOVERY OF THREE-DIMENSIONAL MODELS FROM TWO-DIMENSIONAL IMAGES | |
| DE112020000902T5 (en) | PRE-CALL DATA FOR GRAPHIC DATA PROCESSING | |
| Stojanovic et al. | Service-oriented semantic enrichment of indoor point clouds using octree-based multiview classification | |
| CN1430185A (en) | Ultralarge scale medical image surface reconstruction method based on single-layer surface tracking | |
| Guo et al. | 3D semantic segmentation based on spatial-aware convolution and shape completion for augmented reality applications | |
| RU2653634C2 (en) | Visualisation of flow patterns | |
| DE102019112595A1 (en) | GUIDED HALLUCATION FOR MISSING PICTURE CONTENT USING A NEURONAL NETWORK | |
| Rasoulzadeh et al. | Strokes2Surface: Recovering Curve Networks From 4D Architectural Design Sketches | |
| DE102021114013A1 (en) | TECHNIQUES FOR EFFICIENT SCANNING OF AN IMAGE | |
| CN108830922A (en) | A kind of profile tree constructing method based on multithreading | |
| Camahort et al. | Integrating volume data analysis and rendering on distributed memory architectures | |
| Kaufman et al. | A survey of architectures for volume rendering | |
| CN103745495A (en) | Medical volume data based volume rendering method | |
| Srivastava et al. | xcloth: Extracting template-free textured 3d clothes from a monocular image | |
| RoBler et al. | Dynamic shader generation for flexible multi-volume visualization | |
| Xiang et al. | Single-image mesh reconstruction and pose estimation via generative normal map | |
| DE102023105068A1 (en) | Motion vector optimization for multiple refractive and reflective interfaces | |
| Huang et al. | Accelerated parallel texture optimization | |
| Sharma et al. | Neural shape parsers for constructive solid geometry | |
| AT&T | ||
| Perez-Monte et al. | Modelling frame losses in a parallel alternate frame rendering system with a computational best-effort scheme | |
| Li et al. | Pipeline 3D Modeling Based on High‐Definition Rendering Intelligent Calculation | |
| Kumar et al. | Gpu-accelerated interactive visualization of 3D volumetric data using CUDA | |
| Levus et al. | Using software agents in a distributed computing system for procedural planetoid terrain generation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |