The Study Of Seismic Geometry Attribute Based On Eccentric Windows Scanning Method

Seismic geometry attribute is an expression of subsurface geological structure that we see from the reflection seismic data.It is a data driven technology and an effective assistance in 3D seismic interpretation.It mainly includes dip attribute,azimuth attribute,coherence attribute,curvature attribute,texture attribute and so on.Dip attribute describes the variation of the formation dip angle.Azimuth attribute informs us the azimuth of strata.These two attributes tell us the orientation and complexity of geological structure.Then coherence attribute is used to faults and channels imaging and curvature attribute is used to identify folds,curve and fracture reservoirs.Moreover,Texture attribute characterizes geometry texture and so on.The reasonable application and combination of those geometry attribute are critical in improving the interpretation precision and enriching its contents.This paper carries out a system and in-depth research in dip,azimuth,curvature and coherence attribute,and combining curvelet transform to geometry attribute.Firstly this paper proposed a dip and azimuth scan method based on eccentric windows,which can help us see the detailed information hided in the seismic profiles and the geological framework changes in time slices or layer sections.Then this paper uses this scanning method to curvature and coherence attributes calculation.Both improvements enhance the accuracy of the two attribute and increase the signal to noise ratio.In coherence attribute calculation,it can also protect the imagination of faults.In a addition,this paper use the curvelet transform with scale and angle characteristic features to the calculation of seismic geometry attribute.We put forward a new method to calculate dip and azimuth attribute based on curvelet function which avoids the limitation that dip scan increment cannot be larger than sample rate.We can freely set the dip scan increment and obviously increases the accuracy of the results,especially in some flat area.This method proves great advantages in seismic 2d profile and 3d volume data test.But the current algorithm remains to be modified in order to enhance the computational efficiency.