Tu N105 16 A Novel Salt Body Detection Workflow

Salt body detection from seismic data is a very important task in the velocity model building process as well as salt interpretation. Unfortunately, this task can consume significant interpretation time, weeks or months, and it is a tedious labor for the interpreter due to they normally use a manually process. In this paper we present a novel workflow for detecting salt bodies from seismic data with the goal to automate it or part of it and reduce interpretation time. Our workflow consists in to obtain a good initial segmentation from our L1 sparse representation technique to define a region of interest into the seismic data. After that, we have used an edge detection attributes, mathematical morphological operations and different threshold techniques. We have used a real seismic data to test our workflow. The result shows a good detection in our case.