The Application Of Multi-structure Adaptive Morphological Filter In Processing Of Seismic Images

Seismic image is composed of many stripe lines, these lines are absolutelynecessary evidence for judging the seismic structure fault, so we need an enhancementtechnique to make the stripe lines more clear and to set up good foundation forexplaining seismic information, it is very meaningful for oil field exploration.Because of geophysical images and digital image have some similarity in generalsense, so image processing techniques can be used in geophysical images. This papercarried out a research in using the technology of mathematical morphology filtering toimprove SNR and clarity of the seismic data.This paper is divided into three parts. The first part introduces the theoreticalbasis of seismic signal noise and mathematical morphology; the second part designsmorphological filters for seismic signal de-noising by a large number of comparativeexperiments and the third part process the actual seismic profiles and analysis results.The part of theoretical introduction: Firstly analysis the sources of noise and theclassification of seismic signals, cited the methods of seismic de-noising and theiradvantages and disadvantages, then Focus on the theoretical knowledge ofmathematical morphology, figure ground saying the basic operations of binary andgrayscale morphology, and listing some practical algorithm.The part of filter designing: detailing the different results of filtering is due to thedifferent of the filter combinations, shape and size of the structural elements by alarge number of comparative experiments. Then improve the algorithm combined withtheoretical knowledge, on this basis, we designed the adaptive morphological filterbased on the multi-structure of OCCO.The part of using in actual data: using various forms of filter processing theseismic profile based on a certain place. It can be seen, this method is superior to other.it can de-noising at the greatest degree and retain the effective signal at the sametime.Therefore, this paper indicate the method of adaptive morphological filter based on the multi-structure is useful in interpretation of seismic data.The result of research indicates it is feasible to process seismic signals based onmathematical morphology. And mathematical morphology has the merits of simplealgorithm、fast operation and parallel running has prospect of popularizing andapplying.This specification is set for the theses of Chengdu University of Technology.