Share Email Print
cover

Proceedings Paper

Features extraction and image analysis of seismic sections
Author(s): Maria A. Zoran; Liviu Florin Zoran
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

A seismic section is intermediate step between the material response of the Earth surface to seismic activity. As derived from materials properties (density, elastic modules and attenuation) the seismic response is numerical in nature. Data sets are transformed into plots or are exhibited as images. Geological interpreting of cross section is based on regions of lithological affinity. An interpreter considers the section as a transitional step from a wavefield to a map and views it as an image. Understanding the section is a complex process of features extraction and image analysis. Seismic sections contain 105 times more data that is needed to fully describe a completely interpreted section. This paper suggests three ways to reduce the data set size for seismic section analysis by a global factor of 100 or more. The interpretation becomes a problem of a pattern classification using clustering algorithms that highlight the important statigraphic attributes by operating on various transformed domains. The evaluation of seismic sections means processing the data and the new image interpreting. Could be identified: location and dip of faults, sense of motion along them, boundaries, reflection geometry and the degree of continuity. Such model can be applied for seismic sections in Vrancea seismic area, Romania.

Paper Details

Date Published: 13 February 2004
PDF: 7 pages
Proc. SPIE 5239, Remote Sensing for Environmental Monitoring, GIS Applications, and Geology III, (13 February 2004); doi: 10.1117/12.511240
Show Author Affiliations
Maria A. Zoran, National Institute of Research and Development for Optoelectronics (Romania)
Liviu Florin Zoran, Wyeth Whitehall (Romania)


Published in SPIE Proceedings Vol. 5239:
Remote Sensing for Environmental Monitoring, GIS Applications, and Geology III
Manfred Ehlers; Hermann J. Kaufmann; Ulrich Michel, Editor(s)

© SPIE. Terms of Use
Back to Top