The Pairwise Region Comparison algorithm offers a highly efficient means of automatically segmenting salt bodies in seismic images, but its accuracy can suffer in areas with discontinuous boundaries or poor illumination. In such cases, altering the input amplitude data according an interpreter's manual picks can successfully guide the segmentation to a more accurate result, even in areas far from the manual picking locations. Combining interpreter expertise with automation in this manner could help alleviate a major bottleneck in the model-building and imaging workflow.
![[pdf]](icons/pdf.png){kind=icon}