Research On Seismic Preservation Algorithm Of River Facies Reservoir

In the prediction of fluvial facies reservoir,It is important to determine the discontinuous character of the sedimentary characteristics of sand bodies in fluvial facies.It is necessary for extracting the feature information to use a kind of seismic attribute technique which is more sensitive to discontinuities.The noise is easy to affect the accuracy of continuous attributes,what’s worse,the discontinuity attribute is easy to produce false detection and miss detection and ultimately attributes can not correctly identify the boundary of sand body,so we need to use the filter technique to suppress noise in seismic data.However,it is difficult to remove the noise by using the conventional seismic filtering method.According to the similar properties between edge feature in image and discontinuity of seismic,we could use edge preserving smoothing in image processing to suppress noise,which could improve the accuracy of seismic attribute discontinuity detection.In this paper,two kinds of edge preserving filtering methods are discussed.The first method,anisotropic diffusion filter,is based on partial differential equations,which originated from the thermal diffusion phenomena.On the basis of the thermal diffusion equation,the diffusion coefficient is constructed by using the physical property of discontinuity.As the initial value of the data,we can get the result of diffusion filtering by solving the diffusion equation.filtering by solving the diffusion equation.The second kind of filter is based on fuzzy theory.Edge-preserving fuzzy filter is advantage of fuzzy theory specialized in analyzing uncertainty.Frist the edge-preserving fuzzy filterdicusses the different between the points in the time window,then,according to the degree of different,the different weights are assigned,finally,result is obtained.In this paper,On the basis of edge-preserving fuzzy filtering,a fuzzy control filter based on scanning edge structure is proposed.The method takes advantage of the fuzzy and structural properties of edge features,which could distinguishe the edge region or non edge region,and obtains the filtering results in form of weight.Through the model test,the paper discusses the influence of the key parameters on the filtering results of the edge preserving fuzzy filter and the fuzzy control filter based on the edge structure scanning.We use a channel sandbody model to analyze the ability of edge-preserving and clearing noise of the traditonal filter and the three filtersmentioned in the paper.The experimental result shows that the fuzzy control filter basedon scanning edge structure is the best filter.Finally,the data in BZ34-1 is used to further verify the conclusions obtained in the model test.