Title:  Computational Seismic Interpretation using Geometric Representation and Tensor-based Texture Analysis

Committee: 

Dr. AlRegib, Advisor     

Dr. McClellan, Chair

Dr. Fekri

Abstract:

The objective of the proposed research is to develop computational seismic interpretation methods using geometric representation and tensor-based texture analysis. In hydrocarbon exploration, seismic interpretation is a process that infers geological information from processed seismic data. With the dramatically growing sizes of collected seismic data, manual interpretation has become more time consuming and more labor intensive. To improve efficiency and effectiveness, in this proposal, we focus on interpretation methods using geometric representation and tensor-based texture analysis. Faults as signification displacements in the subsurface have line or curved shapes in seismic sections. The geometric features of faults can be effectively extracted by the Hough transform. We propose a double-threshold method to remove noisy features and delineate faults with geological constraints to increase accuracy. In addition, textures are also an important feature of geological structures. To extract texture features, we propose a tensor-based subspace learning method, tensor orthogonal locality discriminant projection with maximum margin criterion (TOLDP-MMC). The proposed TOLDP-MMC extract the features of tensor samples by preserving both the similarity and variability of neighboring samples, as well as the global structure of all samples. Texture features around the boundaries of salt domes can be extracted by the proposed TOLDP-MMC and will be utilized to implement salt-dome tracking in our future work.