CN106249286A - A kind of Gaussian beam offset method for seismic data with low signal-to-noise ratio and system - Google Patents

Abstract

The present invention relates to earthquake data offset imaging field, particularly relate to a kind of Gaussian beam offset method for Low SNR signal and system, the described Gaussian beam offset method for Low SNR signal includes: geological data is decomposed into part plan wave datum；Utilize similarity coefficient to carry out preferably to described part plan wave datum, pick out similarity coefficient more than the plane wave component setting threshold value；The direction of the plane wave component to pick out carries out Gaussian beam continuation as Gaussian beam original incident direction；And after carrying out Gaussian beam continuation, select the Gaussian beam to current imaging point contribution maximum to carrying out imaging.The present invention forms the characteristic Gaussian beam migration technology being suitable for complicated earth surface complicated structure seismic data with low signal-to-noise ratio, can improve structure imaging effect while being effectively improved imaging section signal to noise ratio.

Description

A kind of Gaussian beam offset method for seismic data with low signal-to-noise ratio and system

Technical field

The present invention relates to earthquake data offset imaging field, particularly relate to a kind of for Low SNR signal Gaussian beam offset method and system.

Background technology

Prestack migration image technology has played important function in the seismic prospecting of complex area, is seismic data Processing and a ring important during oil-gas exploration, the quality of imaging results directly affects follow-up explanation Location with oil well.But the south of China and western mountain front earth's surface are complicated and changeable, and data acquisition excites Poor with condition of acceptance, the signal to noise ratio causing seismic data is extremely low, additionally, due to the complexity of subsurface structure The highest, increase the difficulty of seism processing imaging the most further.Therefore, low signal-to-noise ratio problem is to work as One main bugbear of front pre-stack depth migration imaging, conventional processing method carries out denoising to initial data Rear reimaging, both are performed separately, but the low signal-to-noise ratio imaging effect of this conventional treatment method paying no attention to Think.

Summary of the invention

It is an object of the invention to provide a kind of Gaussian beam side of preferably offsetting for seismic data with low signal-to-noise ratio Method and system, for solving the pre-stack depth migration problem of low signal-to-noise ratio seismic data.

To achieve these goals, technical scheme provides a kind of for low signal-to-noise ratio earthquake number According to Gaussian beam preferably offset method, including: geological data is decomposed into part plan wave datum；Utilize Described part plan wave datum is carried out preferably by similarity coefficient, picks out similarity coefficient more than setting threshold value Plane wave component；The direction of the plane wave component to pick out carries out height as Gaussian beam original incident direction This Shu Yantuo；And after carrying out Gaussian beam continuation, select the Gaussian beam maximum to the contribution of current imaging point To carrying out imaging.

Preferably, described geological data is decomposed into part plan wave datum, specifically includes: with earthquake number According to big gun road collection be input, utilize linear local τ-p transformation for mula, will input by local dip superposition Big gun road collection be decomposed into part plan wave datum；Wherein, described linear local τ-p transformation for mula is

$F(\mathrm{\τ},p_x,p_y)=\underset{X_1}{\overset{X_2}{\∫}}f(t=\mathrm{\τ}+p_{x}x+p_{y}y,x,y)\mathrm{dX}$

Wherein, F (τ, p_x,p_y) representing the part plan wave datum after converting, (t, x y) represent seismic wave to f Big gun road collection, limit of integration X₁With X₂Determine the input range of local τ-p conversion, (p_x,p_y) represent conversion After the direction of plane wave component.

Preferably, described similarity coefficient is utilized to carry out described part plan wave datum preferably, specifically including: Set the threshold value of similarity coefficient；Calculate in described part plan wave datum between each plane wave component is similar Coefficient；And the plane wave component picking out the threshold value that similarity coefficient is more than the similarity coefficient set carries out height This Shu Yantuo.

Preferably, similar system between each plane wave component in described calculating described part plan wave datum Number, employing below equation:

$s({\mathrm{\τ}}_i,p_j)=\frac{\underset{w}{\mathrm{\Σ}}{\left[\underset{k=1}{\overset{N}{\mathrm{\Σ}}}f({\mathrm{\τ}}_i+p_j{X}_k,X_k)\right]}^2}{N\underset{w}{\mathrm{\Σ}}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{\left[f({\mathrm{\τ}}_i+p_j{X}_k,X_k)\right]}^2}$

In formula: X refers to that offset distance, k are the subscripts of X, represent number of channels；N is total number of channels；I is time τ Subscript, express time sampling point；J is the subscript of ray parameter P, represents discrete plane wave；W is Weight coefficient.

Preferably, the direction of the described plane wave component to pick out is entered as Gaussian beam original incident direction Row Gaussian beam continuation, specifically includes: the direction of the plane wave component to pick out initially enters as Gaussian beam Penetrate direction；Utilize Gaussian beam original incident direction, asked for by kinesiology ray tracing occurring in Gaussian beam The path of heart ray and when walking；And utilize Gaussian beam original incident direction, pass through kinetics ray-tracing Ask for the kinetic parameters of Gaussian beam central ray, form single Gaussian beam continuation.

Preferably, the described kinetic parameters being asked for Gaussian beam central ray by kinetics ray-tracing, Specifically include: use below equation to ask for the kinetic parameters of Gaussian beam central ray,

$\left\{\begin{array}{c}\frac{\∂Q\left(s\right)}{\∂s}=v\left(s\right)P\left(s\right)\\ \frac{\∂P\left(s\right)}{\∂s}=\frac{1}{v^2\left(s\right)}V\left(s\right)Q\left(s\right)\end{array}\right.$

Wherein, P (s) and Q (s) is 2 × 2 complex valued matrices along central ray change, characterizes kinetics ray Follow the trail of parameter, determine width and the wave-front curvature of Gaussian beam propagation respectively；V (s) is ray center coordinate system Under 2 × 2 matrixes led about velocity field second order, s represents along ray integral step-length.

Preferably, current imaging point is contributed maximum Gaussian beam to carrying out imaging by described selection, specifically wraps Include: to current imaging point r, it is assumed that all to its contributive Gaussian beam to there being N pair, utilize Gaussian beam To imaginary part walk time to all of Gaussian beam to carrying out ascending sort, only use the forward N1 of sequence to entering Row imaging.

Preferably, N1 value is

It is preferred that technical scheme additionally provides a kind of Gaussian beam for seismic data with low signal-to-noise ratio Offset system, including: decomposing module, for geological data being decomposed into part plan wave datum；Preferably Module, is used for utilizing similarity coefficient to carry out described part plan wave datum preferably, picking out similarity coefficient More than the plane wave component setting threshold value；Gaussian beam continuation module, for the plane wave component picked out Direction carry out Gaussian beam continuation as Gaussian beam original incident direction；And image-forming module, for entering After row Gaussian beam continuation, select the Gaussian beam to current imaging point contribution maximum to carrying out imaging.

Preferably, described preferred module, specifically include: threshold value setting module, be used for setting similarity coefficient Threshold value；Similarity coefficient computing module, is used for calculating each plane wave component in described part plan wave datum Between similarity coefficient；And plane wave component determines module, it is used for picking out similarity coefficient more than setting The plane wave component of threshold value of similarity coefficient carry out Gaussian beam continuation.

The invention has the beneficial effects as follows: the present invention is by developing based on similarity coefficient local optimum lineups side Preferably offset method to the Gaussian beam identified, formed and be suitable for complicated earth surface complicated structure low signal-to-noise ratio earthquake The characteristic Gaussian beam migration technology of data, can effectively utilize the particular propagation direction character of characteristic wave, can be Improving structure imaging effect while being effectively improved imaging section signal to noise ratio, treatment effect is better than conventional imaging Technical finesse result.Therefore, the present invention strong adaptability to low SNR data imaging, it is particularly conducive to Solve complex area particularly mountain front seismic prospecting need badly solution, be also the low noise of where the shoe pinches Ratio imaging, improves its final imaging effect.

Other features and advantages of the present invention will be described in detail in detailed description of the invention part subsequently.

Accompanying drawing explanation

Accompanying drawing is used to provide a further understanding of the present invention, and constitutes the part of description, with Detailed description below is used for explaining the present invention together, but is not intended that limitation of the present invention.? In accompanying drawing:

Fig. 1 is the Gaussian beam side of preferably offsetting in embodiments of the present invention for seismic data with low signal-to-noise ratio The schematic flow sheet of method；

Fig. 2 is the schematic diagram of kinesiology ray tracing coordinate system definition in embodiments of the present invention；

Fig. 3 is the process schematic of Gaussian beam continuation in embodiments of the present invention；

Fig. 4 is the preferred process schematic of Gaussian beam imaging in embodiments of the present invention；

Fig. 5 is that in embodiments of the present invention, Gaussian beam for seismic data with low signal-to-noise ratio preferably offsets and is The structural representation of system；

Fig. 6 is to use the theoretical model schematic diagram of application examples in embodiments of the present invention；

Fig. 7 is the big gun road collection schematic diagram that the theoretical model for Fig. 6 extracts；

Fig. 8 is the schematic diagram that the theoretical model to Fig. 6 adds noise big gun road collection；

The normal Gaussian bundle that Fig. 9 is corresponding for Fig. 8 preferably offsets generalized section；

Figure 10 is the Gauss that the Gaussian beam of the employing present invention corresponding for Fig. 8 in application examples preferably offsets method Bundle migrated section schematic diagram；

Figure 11 is any big gun of real data road collection schematic diagram in application examples；

Figure 12 is the schematic diagram of normal Gaussian bundle migrated section corresponding for Figure 11；

Figure 13 is the Gauss that the Gaussian beam of the employing present invention corresponding for Figure 11 in application examples preferably offsets method Bundle migrated section schematic diagram, wherein preferably coefficient is 0.1.

Figure 14 is the Gauss that the Gaussian beam of the employing present invention corresponding for Figure 11 in application examples preferably offsets method Bundle migrated section schematic diagram, wherein preferably coefficient is 0.2.

Detailed description of the invention

Below in conjunction with accompanying drawing, the detailed description of the invention of the present invention is described in detail.It should be appreciated that Detailed description of the invention described herein is merely to illustrate and explains the present invention, is not limited to this Bright.

For the pre-stack depth migration problem of implementation of seismic data with low signal-to-noise ratio (data), present embodiment Provide a kind of Gaussian beam for seismic data with low signal-to-noise ratio and preferably offset method, as it is shown in figure 1, bag Include: geological data is decomposed into part plan wave datum；Utilize similarity coefficient to described part plan wave number According to carrying out preferably, pick out similarity coefficient and carry out Gaussian beam continuation more than the plane wave component setting threshold value； The direction of the plane wave component to pick out carries out Gaussian beam continuation as Gaussian beam original incident direction；With And after carrying out Gaussian beam continuation, select the Gaussian beam to current imaging point contribution maximum to carrying out imaging.

In present embodiment, similarity coefficient is used to weigh part plan wave datum, its definition be along The axial degree of coherence of data homophase, degree of coherence is the biggest, and similarity coefficient is the biggest, and vice versa. Therefore, just can preferably be gone out data lineups by similarity coefficient and be concerned with best direction, i.e. optimum is flat Ripple direction, face.

In present embodiment, pick out local optimum plane wave direction first with similarity coefficient, abandon and make an uproar Sound composition, effective reflection that prominent similarity is high, only effective reflected signal is carried out Gaussian beam continuation；So After during continuation preferably to the maximum Gaussian beam imaging of imaging point contribution to carrying out imaging, thus realize Improve imaging noise and when improve the purpose of high-dip structure imaging effect.Present embodiment carries out Gaussian beam Preferably offsetting result can directly output offset section and imaging road collection.

Below in conjunction with above-mentioned basic step and following each formula, illustrate this Gaussian beam side of preferably offsetting The specific implementation process of method.

One, local optimum plane wave direction discernment based on similarity coefficient.

Local optimum plane wave direction discernment is implemented in carrying out local dip additive process, i.e. make use of line Property local τ-p transformation for mula, then geological data is decomposed into part plan wave datum be mainly with earthquake number According to big gun road collection be input, utilize linear local τ-p transformation for mula, will input by local dip superposition Big gun road collection be decomposed into part plan wave datum；

Wherein, described linear local τ-p transformation for mula is

$F(\mathrm{\τ},p_x,p_y)=\underset{X_1}{\overset{X_2}{\∫}}f(t=\mathrm{\τ}+p_{x}x+p_{y}y,x,y)\mathrm{dX}---\left(1\right)$

Wherein, F (τ, p_x,p_y) representing the part plan wave datum after converting, (t, x y) represent seismic wave to f Big gun road collection, limit of integration X₁With X₂Determine the input range of local τ-p conversion, (p_x,p_y) represent conversion After the direction of plane wave component.It addition, (p_x,p_y) also determine the height that this plane wave component is corresponding simultaneously This restraints original incident angle, and this Gaussian beam original incident angle plays an important role in subsequent schedule.

Input data (big gun road collection) is decomposed into Local plane wave by local dip superposition by present embodiment Data, have also suppressed the random noise of initial data, have improve data SNR while decomposition.

After input data (big gun road collection) is decomposed into part plan wave datum by local dip superposition, Consider part plan wave datum is carried out preferably.Based on the relevant journey judging initial data local lineups The principle of degree, it is known that between adjacent seismic channel, signal has dependency, and noise does not has or is concerned with Property less, therefore it is contemplated that distinguish obtain part plan wave datum be " noise " or " signal ", Only " signal " therein is carried out Gaussian beam continuation and imaging, to improve Gaussian beam imaging signal to noise ratio, reach To preferred purpose.

According to this thinking, present embodiment proposes and utilizes similarity coefficient to enter described part plan wave datum Row preferably, specifically includes: set the threshold value of similarity coefficient；Calculate in described part plan wave datum each flat Similarity coefficient between the wave component of face；And pick out the similarity coefficient threshold value more than the similarity coefficient set Plane wave component carry out Gaussian beam continuation.

Wherein, formula (2) is used to calculate in described part plan wave datum between each plane wave component Similarity coefficient:

$s({\mathrm{\τ}}_i,p_j)=\frac{\underset{w}{\mathrm{\Σ}}{\left[\underset{k=1}{\overset{N}{\mathrm{\Σ}}}f({\mathrm{\τ}}_i+p_j{X}_k,X_k)\right]}^2}{N\underset{w}{\mathrm{\Σ}}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}{\left[f({\mathrm{\τ}}_i+p_j{X}_k,X_k)\right]}^2}---\left(2\right)$

In formula: X refers to that offset distance, k are the subscripts of X, represent number of channels；N is total number of channels；I is time τ Subscript, express time sampling point；J is the subscript of ray parameter P, represents discrete plane wave；W is Selectable weight coefficient, is 1 when not dealing with.

Similarity coefficient is in theory between 0-1.0.For gross data, it is considered that there is no noise, Thus the threshold value of similarity coefficient takes 0, i.e. represent and data are not done any screening；For real data, root Signal to noise ratio according to different pieces of information is different, carries out testing to select optimum parameter, usually, initial testing Similarity coefficient cut-off be not higher than 0.1, this is in order to avoid value is too high and damage useful signal, afterwards Can be offset by score and progressively heighten this coefficient, finally come according to imaging noise when section resolution Determine concrete value.The threshold value of similarity coefficient is the biggest, and imaging signal to noise ratio is the highest, but the resolution of section is more Low, vice versa, therefore don't fail to select suitable threshold value according to the actual requirements.In present embodiment, Only similarity coefficient is considered as " signal " more than the plane wave component of this threshold values, and then participates in Gauss Shu Yantuo and imaging, can significantly improve the imaging signal to noise ratio of low signal-noise ratio data by this operation.

Two, Gaussian beam continuation

Utilize the similarity coefficient can be by this plane wave behind some plane wave direction preferred when slant stack Direction carries out Gaussian beam continuation as Gaussian beam original incident direction, and continuation process includes two steps: utilize Gaussian beam original incident direction, asks for path Gaussian beam central ray occur by kinesiology ray tracing And when walking；And utilize Gaussian beam original incident direction, asked in Gaussian beam by kinetics ray-tracing The kinetic parameters of heart ray, forms single Gaussian beam continuation.

(1) kinesiology ray tracing

First-order ordinary differential equation system as defined in formula (3a) and formula 3 (b) carries out kinesiology ray and chases after Track, generally uses Runge Kutta method to solve for this formula (3a) and formula 3 (b).

$\frac{\mathrm{dz}}{\mathrm{d\τ}}=v^2{p}_z=v\cos \mathrm{\θ}$

Wherein θ andRepresenting inclination angle and the azimuth of ray respectively, definition is as shown in Figure 2.X, y, z generation The coordinate of table current ray point, v is speed, and p then represents ray slowness.By kinesiology ray tracing When having asked for the path of Gaussian beam central ray and walked.

(2) kinetics ray-tracing

First-order ordinary differential equation system as defined in formula (4) is carried out, and the most also uses Runge Kutta side Method solves.

$\left\{\begin{array}{c}\frac{\∂Q\left(s\right)}{\∂s}=v\left(s\right)P\left(s\right)\\ \frac{\∂P\left(s\right)}{\∂s}=\frac{1}{v^2\left(s\right)}V\left(s\right)Q\left(s\right)\end{array}\right.---\left(4\right)$

Wherein, P (s) and Q (s) is 2 × 2 complex valued matrices along central ray change, characterizes kinetics ray Follow the trail of parameter, determine width and the wave-front curvature of Gaussian beam propagation respectively；V (s) is ray center coordinate system Under 2 × 2 matrixes led about velocity field second order, s represents along ray integral step-length.

Kinetics ray-tracing combines kinesiology ray tracing and is the formation of single Gaussian beam continuation, such as Fig. 3 Shown in, give the process signal of Gaussian beam continuation.

Three, Gaussian beam preferably offsets

After utilizing similarity coefficient some plane wave direction preferred participating in Gaussian beam continuation when slant stack, Underground imaging point can be carried out Gaussian beam imaging.In imaging process, for current imaging point, and The Gaussian beam got off from shot point and geophone station of not all is to all carrying out imaging, but selects current imaging The Gaussian beam of some contribution maximum is to carrying out imaging, to avoid imaging noise, improves imaging signal to noise ratio further, Improve high-dip structure image quality.

Accordingly, described selection to the maximum Gaussian beam of current imaging point contribution to carrying out imaging, such as Fig. 4 institute Show, specifically include: to current imaging point r, it is assumed that all to its contributive Gaussian beam to there being N pair, When the imaginary part utilizing Gaussian beam pair is walked, (imaginary part of shot point Gaussian beam adds the imaginary part of geophone station Gaussian beam when walking When walking) to all of Gaussian beam to carrying out ascending sort (when walking and from small to large sequence) by imaginary part, only Use the forward N1 that sorts to carrying out imaging.

N1 is by user's self-defining, N1 < N, and generally, N1 takesSubstantially can meet imaging to want Ask.

Corresponding said method, present embodiment additionally provides a kind of Gauss for seismic data with low signal-to-noise ratio Bundle preferably offsets system, as it is shown in figure 5, include: decomposing module, for geological data is decomposed into office Facial planes wave datum；Preferred module, is used for utilizing similarity coefficient to carry out excellent to described part plan wave datum Choosing, picks out similarity coefficient and carries out Gaussian beam continuation more than the plane wave component setting threshold value；Gaussian beam prolongs Open up module, carry out height for the direction of the plane wave component to pick out as Gaussian beam original incident direction This Shu Yantuo；And image-forming module, for after carrying out Gaussian beam continuation, select current imaging point tribute Offer the Gaussian beam of maximum to carrying out imaging.

Wherein, described preferred module includes: threshold value setting module, for setting the threshold value of similarity coefficient； And similarity coefficient computing module, it is used for calculating in described part plan wave datum between each plane wave component Similarity coefficient；And plane wave component determines module, for picking out similarity coefficient more than the phase set Gaussian beam continuation is carried out like the plane wave component of the threshold value of coefficient.

This Gaussian beam preferably offsets the corresponding Gaussian beam of system and preferably offsets being embodied as of method Journey is the same, is described again here.

Below by a concrete application examples, illustrate that the Gaussian beam of present embodiment preferably offsets method And the specific implementation process of system and the technique effect that can obtain.

As shown in Figure 6, a theoretical model comprising salt dome, this model horizontal 600 should be devised by use-case Individual grid (CDP number), mesh spacing 10m, longitudinally 150 grids, mesh spacing 10m.Should Model is just being drilled for follow-up and verifying offset effect.

As it is shown in fig. 7, any 7 Ge Bao road collection of forward modeling based on Fig. 6 extraction, it is laterally number of channels, Longitudinally count for time sampling, sampling interval 1ms.

As shown in Figure 8, the gross data of Fig. 7 being added random noise, signal to noise ratio is 0.5, vertical coordinate table Showing time sampling point, abscissa represents number of channels.Low signal-to-noise ratio causes useful signal to be difficult to.

As it is shown in figure 9, Fig. 8 is added noise data carry out normal Gaussian bundle skew, it can be seen that salt dome The border of model is substantially achieved imaging, but steep structure and salt bottom imaging are unclear.Wherein, vertical coordinate table Showing degree of depth sampling point, abscissa represents number of channels.

As shown in Figure 10, Fig. 8 adds noise data carry out Gaussian beam and preferably offset, it can be seen that salt The border imaging of mound model understands, steep structure and salt bottom obtain blur-free imaging.Wherein, vertical coordinate table Showing degree of depth sampling point, abscissa represents number of channels.

As shown in figure 11, any 5 Ge Bao road collection of certain real data extraction, are laterally number of channels, horizontal seat Parameter illustrates from 8000 roads to 12000 roads, is longitudinally the time (s), and vertical coordinate illustrates from 0～8 Second, sampling interval 2ms.As can be seen from the figure data SNR is relatively low, and wave field is complicated.

As shown in figure 12, the low signal-noise ratio data utilizing Figure 11 to show carries out normal Gaussian bundle skew, from Migrated section can be seen that this imaging section signal to noise ratio is the highest, and imaging axis obscures, the target zone under steep structure Imaging is unintelligible.Wherein, being laterally number of channels, abscissa illustrates 9969～12499 roads, is longitudinally deep Degree (m), vertical coordinate illustrates-2000～4000m, sampling interval 10m.

As shown in figure 13, the low signal-noise ratio data utilizing Figure 11 to show carries out Gaussian beam and preferably offsets, excellent Select coefficient 0.1, can be seen that this imaging signal to noise ratio has obtained certain lifting from migrated section, suddenly construct Under target zone imaging become apparent from.Wherein, being laterally number of channels, abscissa illustrates 9969～12499 Road, is longitudinally the degree of depth (m), and vertical coordinate illustrates-2000～4000m, sampling interval 10m.

As shown in figure 14, the low signal-noise ratio data utilizing Figure 11 to show carries out Gaussian beam and preferably offsets, excellent Select coefficient 0.2, can be seen that this imaging signal to noise ratio has obtained bigger lifting from migrated section, suddenly construct Under target zone imaging definition promoted further, overall offset quality of profile has bigger improvement.Horizontal To for number of channels, abscissa illustrates 9969～12499 roads, is longitudinally the degree of depth (m), and vertical coordinate illustrates -2000～4000m, sampling interval 10m.

The present invention is preferred by developing Gaussian beam based on similarity coefficient local optimum lineups direction discernment Offset method, the characteristic Gaussian beam that formation is suitable for complicated earth surface complicated structure seismic data with low signal-to-noise ratio is inclined Shifting technology, can effectively utilize the particular propagation direction character of characteristic wave, can be effectively improved imaging section letter Making an uproar and than while improve structure imaging effect, treatment effect is better than conventional imaging technique result.Therefore, The present invention strong adaptability to low SNR data imaging, is particularly conducive to solve particularly mountain, complex area Front band seismic prospecting need badly solution, be also the low signal-to-noise ratio imaging of where the shoe pinches, improve it final Imaging effect.

The preferred embodiment of the present invention is described in detail above in association with accompanying drawing, but, the present invention does not limit Detail in above-mentioned embodiment, in the technology concept of the present invention, can be to the present invention Technical scheme carry out multiple simple variant, these simple variant belong to protection scope of the present invention.

It is further to note that each the concrete technology described in above-mentioned detailed description of the invention is special Levy, in the case of reconcilable, can be combined by any suitable means, in order to avoid need not The repetition wanted, various possible compound modes are illustrated by the present invention the most separately.

Additionally, combination in any can also be carried out between the various different embodiment of the present invention, as long as its Without prejudice to the thought of the present invention, it should be considered as content disclosed in this invention equally.

Claims (10)

1. one kind preferably offsets method for the Gaussian beam of seismic data with low signal-to-noise ratio, it is characterised in that bag Include:

Geological data is decomposed into part plan wave datum；

Utilize similarity coefficient to carry out preferably to described part plan wave datum, pick out similarity coefficient more than setting Determine the plane wave component of threshold value；

The direction of the plane wave component to pick out carries out Gaussian beam as Gaussian beam original incident direction and prolongs Open up；And

After carrying out Gaussian beam continuation, select the Gaussian beam to current imaging point contribution maximum to becoming Picture.

Gaussian beam the most according to claim 1 preferably offsets method, it is characterised in that described general Geological data is decomposed into part plan wave datum, specifically includes:

With the big gun road collection of geological data for input, utilize linear local τ-p transformation for mula, inclined by local Tiltedly the big gun road collection of input is decomposed into part plan wave datum by superposition；

Wherein, described linear local τ-p transformation for mula is

$F(\mathrm{\&tau;},p_x,p_y)=\underset{X_1}{\overset{X_2}{\&Integral;}}f(t=\mathrm{\&tau;}+p_{x}x+p_{y}y,x,y)\mathrm{dX}$

Wherein, F (τ, p_x,p_y) representing the part plan wave datum after converting, (t, x y) represent seismic wave to f Big gun road collection, limit of integration X₁With X₂Determine the input range of local τ-p conversion, (p_x,p_y) represent conversion After the direction of plane wave component.

Gaussian beam the most according to claim 1 preferably offsets method, it is characterised in that described profit Carry out preferably, specifically including to described part plan wave datum with similarity coefficient:

Set the threshold value of similarity coefficient；

Calculate the similarity coefficient between each plane wave component in described part plan wave datum；And

Pick out similarity coefficient and carry out Gaussian beam more than the plane wave component of the threshold value of the similarity coefficient set Continuation.

Gaussian beam the most according to claim 3 preferably offsets method, it is characterised in that described meter Calculate the similarity coefficient between each plane wave component in described part plan wave datum, employing below equation:

$s({\mathrm{\&tau;}}_i,p_j)=\frac{\underset{w}{\mathrm{\&Sigma;}}{\left[\underset{k=1}{\overset{N}{\mathrm{\&Sigma;}}}f({\mathrm{\&tau;}}_i+p_j{X}_k,X_k)\right]}^2}{N\underset{w}{\mathrm{\&Sigma;}}\underset{k=1}{\overset{N}{\mathrm{\&Sigma;}}}{\left[f({\mathrm{\&tau;}}_i+p_j{X}_k,X_k)\right]}^2}$

In formula: X refers to that offset distance, k are the subscripts of X, represent number of channels；N is total number of channels；I is time τ Subscript, express time sampling point；J is the subscript of ray parameter P, represents discrete plane wave；W is Weight coefficient.

Gaussian beam the most according to claim 1 preferably offsets method, it is characterised in that described with The direction of the plane wave component picked out carries out Gaussian beam continuation as Gaussian beam original incident direction, specifically Including:

Using the direction of plane wave component picked out as Gaussian beam original incident direction；

Utilize Gaussian beam original incident direction, ask for occurring that Gaussian beam center is penetrated by kinesiology ray tracing The path of line and when walking；And

Utilize Gaussian beam original incident direction, ask for Gaussian beam central ray by kinetics ray-tracing Kinetic parameters, forms single Gaussian beam continuation.

Gaussian beam the most according to claim 5 preferably offsets method, it is characterised in that described logical Cross kinetics ray-tracing and ask for the kinetic parameters of Gaussian beam central ray, specifically include:

Below equation is used to ask for the kinetic parameters of Gaussian beam central ray,

$\left\{\begin{array}{c}\frac{\&PartialD;Q\left(s\right)}{\&PartialD;s}=v\left(s\right)P\left(s\right)\\ \frac{\&PartialD;P\left(s\right)}{\&PartialD;s}=\frac{1}{v^2\left(s\right)}V\left(s\right)Q\left(s\right)\end{array}\right.$

Wherein, P (s) and Q (s) is 2 × 2 complex valued matrices along central ray change, characterizes kinetics ray Follow the trail of parameter, determine width and the wave-front curvature of Gaussian beam propagation respectively；V (s) is ray center coordinate system Under 2 × 2 matrixes led about velocity field second order, s represents along ray integral step-length.

Gaussian beam the most according to claim 1 preferably offsets method, it is characterised in that described choosing Select and contribute maximum Gaussian beam to carrying out imaging current imaging point, specifically include:

To current imaging point r, it is assumed that all to its contributive Gaussian beam to there being N pair, utilize Gaussian beam To imaginary part walk time to all of Gaussian beam to carrying out ascending sort, only use the forward N1 of sequence to entering Row imaging.

Gaussian beam the most according to claim 7 preferably offsets method, it is characterised in that N1 value For

9. one kind preferably offsets system for the Gaussian beam of seismic data with low signal-to-noise ratio, it is characterised in that Including:

Decomposing module, for being decomposed into part plan wave datum by geological data；

Preferred module, is used for utilizing similarity coefficient to carry out described part plan wave datum preferably, picking out Similarity coefficient is more than the plane wave component setting threshold value；

Gaussian beam continuation module, initially enters as Gaussian beam for the direction of the plane wave component to pick out Penetrate direction and carry out Gaussian beam continuation；And

Image-forming module, for after carrying out Gaussian beam continuation, selects the height maximum to the contribution of current imaging point This bundle is to carrying out imaging.

Gaussian beam the most according to claim 9 preferably offsets system, it is characterised in that described excellent Modeling block, specifically includes:

Threshold value setting module, for setting the threshold value of similarity coefficient；

Similarity coefficient computing module, is used for calculating in described part plan wave datum between each plane wave component Similarity coefficient；And

Plane wave component determines module, for picking out the similarity coefficient threshold value more than the similarity coefficient set Plane wave component carry out Gaussian beam continuation.