CN110458910A - A kind of intention method for generating pattern to be developed based on tree - Google Patents
A kind of intention method for generating pattern to be developed based on tree Download PDFInfo
- Publication number
- CN110458910A CN110458910A CN201910719347.4A CN201910719347A CN110458910A CN 110458910 A CN110458910 A CN 110458910A CN 201910719347 A CN201910719347 A CN 201910719347A CN 110458910 A CN110458910 A CN 110458910A
- Authority
- CN
- China
- Prior art keywords
- tree
- population
- pattern
- user
- sampled
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 4
- 238000003786 synthesis reaction Methods 0.000 claims abstract description 4
- 238000005070 sampling Methods 0.000 claims description 8
- 230000003993 interaction Effects 0.000 description 3
- 238000005034 decoration Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000000059 patterning Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention discloses a kind of intention method for generating pattern to develop based on tree, belongs to pattern generation field.Include: (1) using the corresponding dependency tree of pattern that user inputs as initially population G is enlivened, initializes empty background population B;(2) it uses Different Strategies to sample to initially enlivening population G and background population B two tree species group, obtains a collection of pattern dependency tree;(3) random grafting operation is carried out to pattern dependency tree obtained in the previous step, rapid synthesis is largely newly set;(4) the corresponding pattern of newly tree synthesized is presented to user, user is allowed to select more satisfied result;(5) result that user selects is put into and is enlivened in population G, and non-selected result is then put into background population B;It is equivalent to and has updated sampled probability, can control the direction that tree develops;(6) (2)-(5) step is repeated until user stops the evolution process of tree.It is selected according to user, automatically updates evolution direction, generated and have more diversity and creative pattern.
Description
Technical field
The present invention relates to pattern generation field more particularly to a kind of intention method for generating pattern to be developed based on tree.
Background technique
Pattern is seen everywhere with us, can all there is exquisite figure on the large-tonnage products such as dress ornament, craftwork, home decoration
Case.With the aesthetic raising of people, aesthetics and creative to pattern have higher demand.But design is one
Time-consuming work, designer need to take much time to adjust patterning and intention.Existing automated pattern generation method
It can be divided into that interaction is oriented to and goal-oriented.The method of interaction guiding, which is laid particular emphasis on, allows user faster and more easily to pattern element
It is arranged and is modified, but inspiration intention can not be provided for designer, it also can not Mass production pattern.Goal-oriented method is then
It is to be arranged by optimizing to element, to generate pattern similar with target is inputted, but these methods are generally difficult to generate layout again
Miscellaneous pattern.
Summary of the invention
In order to overcome the problems, such as tradition design design method, time-consuming, intention is limited, can not carry out iteratively faster design, this
Invention provides a kind of easy interaction, automation, diversified intention pattern generation means.
A kind of intention method for generating pattern to be developed based on tree characterized by comprising
(1) using the corresponding dependency tree of pattern of user's input as initially population G is enlivened, empty background population B is initialized;
(2) it uses Different Strategies to sample to initially enlivening population G and background population B two tree species group, obtains a batch
Pattern dependency tree;
(3) random grafting operation is carried out to pattern dependency tree obtained in the previous step, rapid synthesis is largely newly set;
(4) the corresponding pattern of newly tree synthesized is presented to user, user is allowed to select more satisfied result;
(5) result that user selects is put into and is enlivened in population G, and non-selected result is then put into background population B;
This, which is equivalent to, has updated sampled probability, can control the direction that tree develops;
(6) (2)-(5) step is repeated until user stops the evolution process of tree.
Further, the input pattern and the pattern of generation all have tree-shaped bottom data structure.
Further, it population G and background population B two tree species group is enlivened is sampled using Different Strategies for described pair, wrapped
It includes:
It enlivening population G, carrying out multiple repairing weld in two populations of background population B, number of samples 2n, n are that subsequent will give birth to
At new dependency tree quantity;Every time when sampling, from enliven the probability P (G) that is sampled in population G=| | G | |/(| | G | |+| | B | |),
And the probability P (B) sampled from background population B=1-P (G).;Wherein | | G | |, | | B | | respectively refer to the quantity set in population.
Further, from enliven sampled in population G when, the probability that all LTD trees enlivened in population G are sampled is consistent
's;When being sampled from background population B, it is intended to the tree that the tree in population G differs greatly is selected and enliven, so that it is more to increase result
Sample;For t ∈ B, the probability that t is sampledWherein E is cross entropy,T with enliven it is every in population G
Similarity between a tree, N are to be uniformly distributed.
Further, the similarity between the tree is measured by feature vector or path core.
Further, the random grafting operation of step (3) the pattern dependency tree, comprising the following steps:
(3.1) dependency tree for sampling previous step, random division are stock supplier and scion supplier;
(3.2) subtree of 1 height > 1 is randomly choosed from the tree as stock supplier as stock;
(3.3) nonleaf node is randomly choosed on stock as grafting node, its left subtree of random erasure or right son
Tree;
(3.4) 1 subtree is randomly choosed from the tree as scion supplier as scion;
(3.5) subtree as scion is added on stock, if grafting node be deleted be left subtree, scion
Left subtree as grafting node;If scion is as the right side for grafting node if what grafting node was deleted is right subtree
Subtree.Then stock and scion constitute a new complete dependency tree.
The invention has the advantages that:
1, using the element in input pattern, develop and generate the pattern of a large amount of different structures, achieve the purpose that quickly to design.
2, evolution direction can be automatically updated according to the user's choice, generates and has more diversity and creative pattern.
Detailed description of the invention
The basic flow chart for the intention method for generating pattern that Fig. 1 develops the present invention is based on tree;
Fig. 2 carries out random grafting operation flow chart to pattern dependency tree.
Specific embodiment
Present invention will be further explained below with reference to the attached drawings and examples.
As shown in Figure 1, a kind of intention method for generating pattern to be developed based on tree of the present invention, includes following implementation steps:
(1) using the corresponding dependency tree of pattern of user's input as initially population G is enlivened, empty background population B is initialized;
(2) it is sampled using Different Strategies to enlivening population G and background population B two tree species group, obtains a collection of pattern
Dependency tree;
It is sampled, is referred specifically to: at active kind using Different Strategies to two tree species groups of population G and background population B are enlivened
Multiple repairing weld is carried out in two group G, background population B populations, number of samples 2n, n are the subsequent new dependency tree quantity by generation;
Every time when sampling, from enliven the probability P (G) that is sampled in population G=| | G | |/(| | G | |+| | B | |), and the probability sampled from B
P (B)=1-P (G).Wherein | | G | |, | | B | | respectively refer to the quantity set in population.It is all to enliven in population G when being sampled from G
The probability that is sampled of LTD tree be consistent.When sampling from background population B, it can tend to select and enliven the tree in population G
The tree to differ greatly, to increase result diversity.For t ∈ B, the probability that t is sampledWherein E is to hand over
Entropy is pitched,It is t and enlivens the similarity between each tree in population G, N is to be uniformly distributed.Similarity between tree can be used
A variety of methods are measured, such as feature vector, path core etc..
(3) random grafting operation is carried out to pattern dependency tree obtained in the previous step, rapid synthesis is largely newly set;
Described carries out random grafting operation to pattern dependency tree, specific as shown in Figure 2:
(3.1) dependency tree for sampling previous step, random division are stock supplier and scion supplier;
(3.2) subtree of 1 height > 1 is randomly choosed from the tree as stock supplier as stock;
(3.3) nonleaf node is randomly choosed on stock as grafting node, its left subtree of random erasure or right son
Tree;
(3.4) 1 subtree is randomly choosed from the tree as scion supplier as scion;
(3.5) subtree as scion is added on stock, if grafting node be deleted be left subtree, scion
Left subtree as grafting node;If what grafting node was deleted is right subtree, right subtree of the scion as grafting node.
Then stock and scion constitute a new complete dependency tree.
(4) the corresponding pattern of newly tree synthesized is presented to user, user is allowed to select more satisfied result;
(5) result that user selects is put into and is enlivened in population G, and non-selected result is then put into background population B.
This, which is equivalent to, has updated sampled probability, can control the direction that tree develops;
(6) (2)-(5) step is repeated until user stops the evolution process of tree.
Claims (6)
1. a kind of intention method for generating pattern to be developed based on tree characterized by comprising
(1) using the corresponding dependency tree of pattern of user's input as initially population G is enlivened, empty background population B is initialized;
(2) it uses Different Strategies to sample to initially enlivening population G and background population B two tree species group, obtains a collection of pattern
Dependency tree;
(3) random grafting operation is carried out to pattern dependency tree obtained in the previous step, rapid synthesis is largely newly set;
(4) the corresponding pattern of newly tree synthesized is presented to user, user is allowed to select more satisfied result;
(5) result that user selects is put into and is enlivened in population G, and non-selected result is then put into background population B;Quite
In having updated sampled probability, the direction that tree develops can control;
(6) (2)-(5) step is repeated until user stops the evolution process of tree.
2. a kind of intention method for generating pattern to be developed based on tree according to claim 1, which is characterized in that the input
Pattern and the pattern of generation all have tree-shaped bottom data structure.
3. a kind of intention method for generating pattern to be developed based on tree according to claim 1, which is characterized in that described to G and B
Two tree species groups are sampled using Different Strategies, comprising:
It enlivening population G, carrying out multiple repairing weld in two populations of background population B, number of samples 2n, n are subsequent by generation
New dependency tree quantity;Every time when sampling, from enliven the probability P (G) that is sampled in population G=| | G | |/(| | G | |+| | B | |), and from
The probability P (B) sampled in background population B=1-P (G).;Wherein | | G | |, | | B | | respectively refer to the quantity set in population.
4. a kind of intention method for generating pattern to be developed based on tree according to claim 3, it is characterised in that from active kind
When sampling in group G, the probability that all LTD trees enlivened in population G are sampled is consistent;When being sampled from background population B, incline
To the tree to differ greatly in the tree for selecting and enlivening in population G, to increase result diversity;T ∈ B, t are sampled general
RateWherein E is cross entropy,It is the similarity between each tree in t and G, N is to be uniformly distributed.
5. a kind of intention method for generating pattern to be developed based on tree according to claim 4, it is characterised in that the tree
Between similarity measured by feature vector or path core.
6. a kind of intention method for generating pattern to be developed based on tree according to claim 1, which is characterized in that the step
(3) the random grafting operation of pattern dependency tree, comprising the following steps:
(3.1) dependency tree for sampling previous step, random division are stock supplier and scion supplier;
(3.2) subtree of 1 height > 1 is randomly choosed from the tree as stock supplier as stock;
(3.3) nonleaf node is randomly choosed on stock as grafting node, its left subtree of random erasure or right subtree;
(3.4) 1 subtree is randomly choosed from the tree as scion supplier as scion;
(3.5) subtree as scion is added on stock, if grafting node be deleted be left subtree, scion conduct
Graft the left subtree of node;If scion is as the right subtree for grafting node if what grafting node was deleted is right subtree;
Then stock and scion constitute a new complete dependency tree.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910719347.4A CN110458910A (en) | 2019-08-05 | 2019-08-05 | A kind of intention method for generating pattern to be developed based on tree |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910719347.4A CN110458910A (en) | 2019-08-05 | 2019-08-05 | A kind of intention method for generating pattern to be developed based on tree |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110458910A true CN110458910A (en) | 2019-11-15 |
Family
ID=68484970
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910719347.4A Pending CN110458910A (en) | 2019-08-05 | 2019-08-05 | A kind of intention method for generating pattern to be developed based on tree |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110458910A (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101826126A (en) * | 2010-04-02 | 2010-09-08 | 浙江大学 | Garment design method based on three-dimensional sketch in any sight |
CN109086526A (en) * | 2018-08-03 | 2018-12-25 | 贵州大学 | National Ridge tracing method based on the tiered form syntax |
-
2019
- 2019-08-05 CN CN201910719347.4A patent/CN110458910A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101826126A (en) * | 2010-04-02 | 2010-09-08 | 浙江大学 | Garment design method based on three-dimensional sketch in any sight |
CN109086526A (en) * | 2018-08-03 | 2018-12-25 | 贵州大学 | National Ridge tracing method based on the tiered form syntax |
Non-Patent Citations (1)
Title |
---|
XINWEI ZHANG等: "Pattern understanding and synthesis based on layout tree descriptor", 《SPRINGER-VERLAG GMBH GERMANY, PART OF SPRINGER NATURE 2019》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Iwata et al. | AntMap: constructing genetic linkage maps using an ant colony optimization algorithm | |
CN104035816B (en) | Cloud computing task scheduling method based on improved NSGA-II | |
US20070005522A1 (en) | Resource assignment optimization using direct encoding and genetic algorithms | |
CN110909787A (en) | Method and system for multi-objective batch scheduling optimization based on clustering evolutionary algorithm | |
CN109991950A (en) | The balance ameliorative way of cooperation robotic asssembly production line based on genetic algorithm | |
Kielarova et al. | New interactive-generative design system: hybrid of shape grammar and evolutionary design-an application of jewelry design | |
CN105590238A (en) | Method, system and terminal device of customizing shoes | |
CN103164631B (en) | A kind of intelligent coordinate expression gene analyser | |
CN110727969B (en) | Method, device and equipment for automatically adjusting workflow and storage medium | |
McCarthy et al. | Early consequences of allopolyploidy alter floral evolution in Nicotiana (Solanaceae) | |
CN110458910A (en) | A kind of intention method for generating pattern to be developed based on tree | |
Tang et al. | A hybrid discrete imperialist competition algorithm for gene selection for microarray data | |
Rosvall et al. | Introduction to breeding strategies and evaluation of alternatives | |
Jin et al. | Meta optimization of stand management with population-based methods | |
Ong et al. | Genetic algorithm approach in PCB assembly | |
CN113901728A (en) | Computer second-class assembly line balance optimization method based on migration genetic algorithm | |
CN108230121A (en) | A kind of product design method based on Recognition with Recurrent Neural Network | |
CN103143123A (en) | System and method for beam direction multi-target optimization based on ant colony algorithm | |
CN101331505A (en) | Method and apparatus for an algorithm development environment for solving a class of real-life combinatorial optimization problems | |
Santantonio et al. | A hybrid optimal contribution approach to drive short-term gains while maintaining long-term sustainability in a modern plant breeding program | |
Semnani et al. | Jacquard pattern optimizing in weft knitted fabrics via interactive genetic algorithm | |
CN109241134A (en) | A kind of grouping of commodities mode multiple target method for digging based on agent model | |
Ahmadi et al. | Rethinking plant breeding | |
CN104598591B (en) | A kind of model element matching process for type attribute graph model | |
Astola et al. | Inferring the gene network underlying the branching of tomato inflorescence |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191115 |