Seeded Growing Attribute for Salt Body Segmentation in Post-Stack Seismic Data

For a more accurate geological model, it is crucial to have accurate segmentation and delineation of salt structures. These are known to be difficult to segment given their chaotic nature. However, there are attributes that isolate the salt boundaries well. These attributes usually are a good indication of the flanks of the salt structure but struggle in segmenting the salt body as a whole. In our work, we propose an automated seeded growing method that will segment the salt bodies from the other structures in post-stack seismic data. Here, edge volumes are used as inputs to give a good indication of the flanks and borders of our salt structures. We based our method upon the conventional seeded growing approach, and extended this with a novel hybrid smoothing technique and boundary termination detection. the hybrid smoothing technique uses the textural attribute of chaos as the alternating criteria between the mean, gausian and median noise reduction attributes. Finally we detect discontinuous boundaries, as this can sometimes be the case, by evaluating the growing directions of our seed points and terminate based on unbalanced growing behavior.