The identification of faults in 3-D data volumes can be enhanced through the use of simple least-squares based edge-detection methods. The use of dip-magnitude and coherency based on local dip estimates is shown to be an effective approach to detection and identification of both high and low relief fault patterns. Three-dimensional dip estimates can be an effective tool for the interpretation of 3-D volumes.
Thanks go to Jon Claerbout for the encouragment to do the work and for providing the opportunity to participate in SEP during the summer of 1997. Appreciation is also due to Sergey Fomel for his numerous suggestions and help with LATEX. Christine Ecker made sure that each revision of the article was compatible with the gmake environment of SEP-95.