CN103810732A - Dynamic texture generating method - Google Patents
Dynamic texture generating method Download PDFInfo
- Publication number
- CN103810732A CN103810732A CN201410067051.6A CN201410067051A CN103810732A CN 103810732 A CN103810732 A CN 103810732A CN 201410067051 A CN201410067051 A CN 201410067051A CN 103810732 A CN103810732 A CN 103810732A
- Authority
- CN
- China
- Prior art keywords
- isotropy
- texture
- texture image
- alpha
- random
- 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 provides a dynamic texture generating method which includes the following steps that S1. the size x*y and image quantity n of generated texture images are defined; S2. a plurality of stationary increments X complying with alpha stable distribution S alpha (sigma, beta, mu) are generated randomly; S3. x*y variables are extracted from the randomly generated stationary increments and written into a preset two-dimensional matrix, the matrix is processed through a random middle point substitution method, and isotropy texture images with the size of x*y are obtained; S4. steps of S2-S3 are repeated till n frames of isotropy texture images are generated; S5. the n frames of isotropy texture images obtained in the S4 are written into a video file, and isotropy dynamic textures are generated. The isotropy texture images can be input into a structure filter, the isotropy texture images are obtained, and the isotropy dynamic textures are generated. By means of the method, rich dynamic textures can be generated.
Description
Technical field
The invention belongs to virtual reality field, be specifically related to a kind of dynamic texture generation method.
Background technology
Texture is the natural surface of a class complexity, and the key of its expression is the mathematical modeling to texture.At present, manually generate texture and be mainly the texture producing based on FBM model, FBM model hypothesis texture has desirable self-similarity, its increment Gaussian distributed, and there is isotropic feature.
But, in to actual natural texture analytic process, find that very eurypalynous texture does not have the statistical property of FBM model, as rock surface, be with alveolate seawater background etc., only there is minority to meet FBM model as hydrothermal solution texture.In the FBM model that has its source in of problem, suppose: the increment Gaussian distributed of texture, this is confined in certain scope its description to texture, can not be described texture widely.
Summary of the invention
For the deficiencies in the prior art, the invention provides a kind of dynamic texture generation method, can generate abundant dynamic texture.
For achieving the above object, the present invention is achieved by the following technical programs:
A kind of dynamic texture generation method, the method comprises the steps:
S1. definition generates big or small x*y and the picture number n of texture image;
S2. generate at random several obediences alpha and stablize distribution S
αthe stationary increment X of (σ, β, μ), wherein α is that characterization factor, σ are that scale parameter, β are position migration parameter centered by the deflection factor, μ;
S3. from the stationary increment X of random generation, choose x*y variable and write in the 2 dimension matrixes that set in advance, utilize random midpoint displacement method to process this matrix, obtain the isotropy texture image of size for x*y;
S4. repeating step S2-S3, until generate n frame isotropy texture image;
S5. n two field picture step S4 being obtained writes video file, generates isotropy dynamic texture.
Wherein, the described random midpoint displacement method of utilizing is processed this matrix, obtains the big or small isotropy texture image for x*y to be:
Get the mean value of four end points pixel values in matrix and add the random file amount setting in advance as central point pixel value, all points in matrix are processed in a manner described, obtain the isotropy texture image that size is x*y, the random file amount wherein setting in advance is obeyed the stable distribution of alpha.
Wherein, after step S4, comprise n two field picture is done to normalized, make its pixel value for [0,255].
Wherein, before step S5, comprise n two field picture is done to correlativity processing.
Further, the method also comprises:
S6. every frame isotropy texture image input structure wave filter step S4 being generated, obtains n frame anisotropic texture image;
S7. n two field picture step S6 being obtained writes video file, generates anisotropy dynamic texture.
Wherein, described step S6 comprises:
If the system function of Structure Filter is:
H
φ(ω
x,ω
y)=(1+α-2αcos
2(ω
θ-θ
0))
-1,
The n frame isotropy texture image that utilizes 2 dimension FFT conversion that step S4 is generated transforms to frequency domain successively, and is designated as S
i(ω
x, ω
y);
Every frame isotropy texture image is input to Structure Filter, obtains anisotropic texture image S
a(ω
x, ω
y), wherein S
a(ω
x, ω
y)=S
i(ω
x, ω
y) H
φ(ω
x, ω
y);
By S
a(ω
x, ω
y) by the IFFT conversion of 2 dimensions, obtain the anisotropic texture image of time domain.
Wherein, after step S6, comprise n two field picture is done to normalized, make its pixel value for [0,255].
Wherein, before step S7, comprise n two field picture is done to correlativity processing.
The present invention at least has following beneficial effect:
Dynamic texture generation method provided by the invention, can generate more horn of plenty and fine and smooth natural texture, such as rock surface or be with isotropy or the anisotropic natural textures such as alveolate seawater background, for traditional texture that can not generate based on FBM model, the invention provides a kind of new mode and carry out the generation of complex texture, met people's needs.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, to the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described below, apparently, accompanying drawing in the following describes is some embodiments of the present invention, for those of ordinary skills, do not paying under the prerequisite of creative work, can also obtain according to these accompanying drawings other accompanying drawing.
Fig. 1 is the process flow diagram of dynamic texture generation method in the embodiment of the present invention;
Fig. 2 is the isotropy texture schematic diagram that dynamic texture generation method generates described in the embodiment of the present invention;
Fig. 3 is the anisotropic texture schematic diagram that dynamic texture generation method generates described in the embodiment of the present invention.
Embodiment
For making object, technical scheme and the advantage of the embodiment of the present invention clearer, below in conjunction with the accompanying drawing in the embodiment of the present invention, technical scheme in the embodiment of the present invention is carried out to clear, complete description, obviously, described embodiment is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment in the present invention, those of ordinary skills, not making the every other embodiment obtaining under creative work prerequisite, belong to the scope of protection of the invention.
The embodiment of the present invention has proposed a kind of dynamic texture generation method, referring to Fig. 1, comprises the following steps:
Step 101: definition generates big or small x*y and the picture number n of texture image.
In this step, definition generates big or small x*y and the picture number n of texture image, and the texture image size that for example definition generates is 128*128, altogether 129 frames.
Step 102: generate at random several obediences alpha and stablize distribution S
αthe stationary increment X of (σ, β, μ).
Usually, the stable probability density function distributing of alpha, except minority special case, does not exist enclosed to express, and generally describes its distribution character with fundamental function.Stochastic variable X obeys the stable distribution of alpha, and its fundamental function form is:
Wherein, sign θ is-symbol function, 0 < α≤2, σ >=0 ,-1≤β≤1 and real number μ.Meet four parameters that this fundamental function is expressed, be referred to as the stable canonical parameter system distributing of alpha.Wherein, four parameters respectively: characterization factor α, scale parameter σ, deflection factor-beta, center migration parameter μ, the stable brief note that distributes of alpha is for S
α(σ, β, μ).
In this step, generate at random several obediences alpha and stablize distribution S
αthe stationary increment X of (σ, β, μ).Here take average (scale parameter) as 0, to be that 1 alpha is stable be distributed as example to dispersion degree (center migration parameter), and it also has two parameters, is respectively characterization factor α, and deflection factor-beta, is designated as S
α(1, β, 0), concrete generating mode is as follows:
Making γ is the equally distributed stochastic variable on (pi/2, pi/2), and W is an exponential random variable that average is 1, and γ and W separate, in the time of α ≠ 1, obey alpha stablize distribution S
αthe stationary increment X of (0, β, 1) can be described as:
X obeys S
α(1, β, 0) distributes, wherein
k (α) is sign function, in the time of α > 1, and K (α)=α; In the time of α < 1, K (α)=α-2.
In the time of α=1, obey alpha and stablize distribution S
αthe stationary increment X of (0, β, 1) can be described as:
In the present embodiment, getting the stable parameter distributing of alpha is α=1.7, and therefore β=0 generates obedience alpha and stablize distribution S
1.7the stationary increment of (1,0,0), due to when α ≠ 1,
Wherein
k (α) is sign function, and in the time of α > 1, K (α)=α, calculates several stationary increments according to the expression formula of above-mentioned X, and the stationary increment number calculating is here greater than x*y.
Step 103: choose x*y variable and write in the 2 dimension matrixes that set in advance from the stationary increment X of random generation, utilize random midpoint displacement method to process this matrix, obtain the isotropy texture image of size for x*y.
In this step, from the stationary increment X of random generation, choosing x*y variable writes in the 2 dimension matrixes that set in advance, then get the mean value of four end points pixel values in matrix and add the random file amount setting in advance as central point pixel value, all points in matrix are processed in a manner described, obtain the isotropy texture image that size is x*y, the random file amount wherein setting in advance is obeyed the stable distribution of alpha.Finally the data texturing generating is carried out to dimensional variation, all pixel values are changed to [0,255].
Step 104: judge whether to generate anisotropic texture, if desired, execution step 105, otherwise execution step 106.
Step 105: isotropy texture image is generated to anisotropic texture image by Structure Filter.
In this step, described Structure Filter is for characterizing anisotropic Structure Filter, portray the anisotropic character of texture completely, from isotropy texture to anisotropic texture, can be equivalent to isotropy texture by a Structure Filter that meets ad hoc structure function.Anisotropic texture F
a(x, y) can be considered isotropy texture F
i(x, y) obtains by wave filter h (u, v), that is:
F
A(x,y)=∫F
I(x-u,y-v)h(u,v)d(u,v);
Corresponding frequency spectrum designation is:
S
A(ω
x,ω
y)=S
I(ω
x,ω
y)H
φ(ω
x,ω
y),
Wherein, H
φ(ω
x, ω
y) be the frequency domain representation of wave filter, Structure Filter concrete form is: H
φ(ω
x, ω
y)=(1+ α-2 α cos
2(ω
θ-θ
0))
-1, wherein,
the intensity that parameter alpha ∈ [0,1] is anisotropic texture, parameter θ
0the direction that ∈ [0, π] is anisotropic texture.
Structure Filter does not change the autocorrelation of texture, just changes its Direction Distribution Characteristics, the deflection parameter of setting structure wave filter in the present embodiment
intensive parameter α=0.7.
First utilize 2 dimension FFT conversion that the isotropy texture image having obtained is transformed to frequency domain, and be designated as S
i(ω
x, ω
y).Anisotropic texture can be considered that each tropism's texture is by the response of Structure Filter, and every frame isotropy texture image is input to Structure Filter, obtains anisotropic texture image S
a(ω
x, ω
y), wherein S
a(ω
x, ω
y)=S
i(ω
x, ω
y) H
φ(ω
x, ω
y);
By S
a(ω
x, ω
y) by the IFFT conversion of 2 dimensions, obtain the anisotropic texture image of time domain;
Finally, the data texturing generating is carried out to dimensional variation, all pixel values are changed to [0,255].
Step 106: whether the number that judges the texture image generating equals n, if equal, execution step 107, otherwise execution step 102.
Step 107: n frame texture image is done to correlativity processing.
In this step, in order to make the texture generating have more continuity, need to increase image sequence correlativity each other, concrete operations are adjacent some two field pictures summations to be averaged rear for example, as a two field picture (5 adjacent two field pictures summations being averaged), in this manner all images are increased to correlativity successively.
Step 108: n frame texture image is write to video file, generate dynamic texture.
In this step, image after treatment correlativity is write to video file, generate dynamic texture.
The dynamic texture generation method that the embodiment of the present invention provides, can generate more horn of plenty and fine and smooth natural texture, such as rock surface or be with isotropy or the anisotropic natural textures such as alveolate seawater background, for traditional texture that can not generate based on FBM model, the invention provides a kind of new mode and carry out the generation of complex texture, met people's needs.
Above embodiment only, for technical scheme of the present invention is described, is not intended to limit; Although the present invention is had been described in detail with reference to previous embodiment, those of ordinary skill in the art is to be understood that: its technical scheme that still can record aforementioned each embodiment is modified, or part technical characterictic is wherein equal to replacement; And these modifications or replacement do not make the essence of appropriate technical solution depart from the spirit and scope of various embodiments of the present invention technical scheme.
Claims (8)
1. a dynamic texture generation method, is characterized in that, the method comprises the steps:
S1. definition generates big or small x*y and the picture number n of texture image;
S2. generate at random several obediences alpha and stablize distribution S
αthe stationary increment X of (σ, β, μ), wherein α is that characterization factor, σ are that scale parameter, β are position migration parameter centered by the deflection factor, μ;
S3. from the stationary increment X of random generation, choose x*y variable and write in the 2 dimension matrixes that set in advance, utilize random midpoint displacement method to process this matrix, obtain the isotropy texture image of size for x*y;
S4. repeating step S2-S3, until generate n frame isotropy texture image;
S5. n two field picture step S4 being obtained writes video file, generates isotropy dynamic texture.
2. method according to claim 1, is characterized in that, the described random midpoint displacement method of utilizing is processed this matrix, obtains the big or small isotropy texture image for x*y to be:
Get the mean value of four end points pixel values in matrix and add the random file amount setting in advance as central point pixel value, all points in matrix are processed in a manner described, obtain the isotropy texture image that size is x*y, the random file amount wherein setting in advance is obeyed the stable distribution of alpha.
3. method according to claim 2, is characterized in that, after step S4, comprises n two field picture is done to normalized, makes its pixel value for [0,255].
4. method according to claim 3, is characterized in that, before step S5, comprises n two field picture is done to correlativity processing.
5. method according to claim 4, is characterized in that, the method also comprises:
S6. every frame isotropy texture image input structure wave filter step S4 being generated, obtains n frame anisotropic texture image;
S7. n two field picture step S6 being obtained writes video file, generates anisotropy dynamic texture.
6. method according to claim 5, is characterized in that, described step S6 comprises:
The system function of described Structure Filter is:
H
φ(ω
x,ω
y)=(1+α-2αcos
2(ω
θ-θ
0))
-1,
The n frame isotropy texture image that utilizes 2 dimension FFT conversion that step S4 is generated transforms to frequency domain successively, and is designated as S
i(ω
x, ω
y);
Every frame isotropy texture image is input to Structure Filter, obtains anisotropic texture image S
a(ω
x, ω
y), wherein S
a(ω
x, ω
y)=S
i(ω
x, ω
y) H
φ(ω
x, ω
y);
By S
a(ω
x, ω
y) by the IFFT conversion of 2 dimensions, obtain the anisotropic texture image of time domain.
7. method according to claim 6, is characterized in that, after step S6, comprises n two field picture is done to normalized, makes its pixel value for [0,255].
8. method according to claim 7, is characterized in that, before step S7, comprises n two field picture is done to correlativity processing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410067051.6A CN103810732B (en) | 2014-02-26 | A kind of dynamic texture generation method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410067051.6A CN103810732B (en) | 2014-02-26 | A kind of dynamic texture generation method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103810732A true CN103810732A (en) | 2014-05-21 |
CN103810732B CN103810732B (en) | 2016-11-30 |
Family
ID=
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109448097A (en) * | 2018-10-30 | 2019-03-08 | 汕头大学 | A kind of ink and wash rendering method based on the conversion of adaptive style |
CN110427817A (en) * | 2019-06-25 | 2019-11-08 | 浙江大学 | A kind of hydrofoil cavitation feature extracting method based on vacuole framing Yu sound texture analysis |
CN112686983A (en) * | 2019-10-17 | 2021-04-20 | 畅想科技有限公司 | Texture filtering |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009009179A (en) * | 2007-06-26 | 2009-01-15 | Univ Of Tokyo | Image processor and image processing program |
CN102074050A (en) * | 2011-03-01 | 2011-05-25 | 哈尔滨工程大学 | Fractal multi-resolution simplified method used for large-scale terrain rendering |
CN102568032A (en) * | 2010-12-09 | 2012-07-11 | 中国科学院软件研究所 | Method for generation of reality surface deformation model |
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009009179A (en) * | 2007-06-26 | 2009-01-15 | Univ Of Tokyo | Image processor and image processing program |
CN102568032A (en) * | 2010-12-09 | 2012-07-11 | 中国科学院软件研究所 | Method for generation of reality surface deformation model |
CN102074050A (en) * | 2011-03-01 | 2011-05-25 | 哈尔滨工程大学 | Fractal multi-resolution simplified method used for large-scale terrain rendering |
Non-Patent Citations (3)
Title |
---|
LIFENG JIA ET AL.: "Fabric Pattern Modeling by Fractional Levy Stable Motion", 《2ND INTERNATIONAL SYMPOSIUM ON SYSTEMS AND CONTROL IN AEROSPACE AND ASTRONAUTICS》 * |
XUTAO LI ET AL.: "Fractional Levy Stable Motion for Modeling Speckle Image", 《ELECTRONIC IMAGING AND MULTIMEDIA TECHNOLOGY V》 * |
李旭涛 等: "各向异性多尺度自相似随机场与地形构建", 《中国图象图形学报》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109448097A (en) * | 2018-10-30 | 2019-03-08 | 汕头大学 | A kind of ink and wash rendering method based on the conversion of adaptive style |
CN109448097B (en) * | 2018-10-30 | 2022-12-06 | 汕头大学 | Ink and wash painting rendering method based on self-adaptive style conversion |
CN110427817A (en) * | 2019-06-25 | 2019-11-08 | 浙江大学 | A kind of hydrofoil cavitation feature extracting method based on vacuole framing Yu sound texture analysis |
CN110427817B (en) * | 2019-06-25 | 2021-09-07 | 浙江大学 | Hydrofoil cavitation feature extraction method based on cavitation image positioning and acoustic texture analysis |
CN112686983A (en) * | 2019-10-17 | 2021-04-20 | 畅想科技有限公司 | Texture filtering |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101950365B (en) | Multi-task super-resolution image reconstruction method based on KSVD dictionary learning | |
Li et al. | Image encryption based on dynamic filtering and bit cuboid operations | |
Liu et al. | Fractional Krawtchouk transform with an application to image watermarking | |
CN102156875B (en) | Image super-resolution reconstruction method based on multitask KSVD (K singular value decomposition) dictionary learning | |
CN110060204B (en) | Single image super-resolution method based on reversible network | |
CN103839292B (en) | Method for sampling contour surface and generating high-quality triangular mesh | |
CN105404817B (en) | It is based on(k, n)The user-friendly visual secret sharing scheme method of thresholding | |
Dai et al. | Chaotic medical image encryption algorithm based on bit-plane decomposition | |
CN102184519B (en) | Method for embedding and extracting watermark images | |
Hou et al. | Image companding and inverse halftoning using deep convolutional neural networks | |
CN104217388B (en) | A kind of image watermark insertion based on FSSVM, the method and apparatus extracted | |
CN103839260A (en) | Method for matching simulation grating based on hidden image lattice point area ratio | |
CN104299250A (en) | Front face image synthesis method and system based on prior model | |
CN101540055B (en) | Cartoon stylization method facing online real-time application | |
US20230214953A1 (en) | Systems and Methods for Message Embedding in Three-Dimensional Image Data | |
CN104299201B (en) | Image reconstruction method based on heredity sparse optimization | |
CN103632375B (en) | A kind of image scrambling evaluation method and device | |
CN113096015B (en) | Image super-resolution reconstruction method based on progressive perception and ultra-lightweight network | |
CN104660269A (en) | Generation method of sensing matrix for signal compressive sensing | |
CN109544435B (en) | Gyrater transformation chaotic random phase coding digital image encryption method | |
Wang et al. | Least-squares images for edge-preserving smoothing | |
CN103824264A (en) | Compressive sensing method based on random Fourier characteristics | |
CN103810732A (en) | Dynamic texture generating method | |
Li et al. | Image stylization with enhanced structure on GPU | |
CN103810732B (en) | A kind of dynamic texture generation method |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20161130 Termination date: 20190226 |