Two new tools to extract faults from the seismic data cube

Pierre Jacquemin. ( 2006 )
in: 26th gOcad Meeting, ASGA

Abstract

This paper presents two new tools to be used in the fault extraction process from a seismic data cube. Last year we presented a method to extract points on faults using a double Hough transform. This method needs to use a fault property and until now we only used the negative of a coherency property which is not well adapted because it contains many other things than the fault attribute. Here, we consider the extremities of the strips from positive or negative values of the seismic signal in cross sections. The result is a very efficient method which gives a very good result in a very short time. After extracting the points corresponding to each fault by searching alignments using the double Hough transform we obtain clouds of points that may contain points coming from noisy data or parts of other faults aligned on the first one. So, we developed a very simple method to cut the cloud in several logical parts. The method consists in projecting the points on the axis the most parallel to the fault. After that we simply need to search for the biggest open spaces on this axis and to use their location as separator for the various parts

Download / Links

    BibTeX Reference

    @inproceedings{JacqueminRM2006,
     abstract = { This paper presents two new tools to be used in the fault extraction process from a seismic data cube. Last year we presented a method to extract points on faults using a double Hough transform. This method needs to use a fault property and until now we only used the negative of a coherency property which is not well adapted because it contains many other things than the fault attribute. Here, we consider the extremities of the strips from positive or negative values of the seismic signal in cross sections. The result is a very efficient method which gives a very good result in a very short time. After extracting the points corresponding to each fault by searching alignments using the double Hough transform we obtain clouds of points that may contain points coming from noisy data or parts of other faults aligned on the first one. So, we developed a very simple method to cut the cloud in several logical parts. The method consists in projecting the points on the axis the most parallel to the fault. After that we simply need to search for the biggest open spaces on this axis and to use their location as separator for the various parts },
     author = { Jacquemin, Pierre },
     booktitle = { 26th gOcad Meeting },
     month = { "june" },
     publisher = { ASGA },
     title = { Two new tools to extract faults from the seismic data cube },
     year = { 2006 }
    }