Seismic DNA - A Novel Seismic Feature Extraction Method Using Non-local and Multi-attribute Sets

We present a new approach to extract features from seismic data using a non-local and multi-attribute search method. By transforming seismic reflection data and attributes derived from seismic data into characters we can search for complex features in the seismic data volumes using a text-based search technology. The method presented here allows for feature patterns with a varying vertical extent, varying attribute values and a varying number of subfeatures in the feature signature. We describe the method and give two examples of how it can be used. The first one shows how the number of subfeatures can vary in a feature. The second one shows how we can combine different attributes in one search.