Automatic Enhancement and Detection of Active Sea Faults from Bathymetry

In this paper, we propose a framework for automatic enhancement and identification of the active sea faults, usually connected with seismic activity using bathymetric images. To our knowledge, the problem of automatic identification of sea faults based on an image analysis method is faced for the first time. First, we compute the Slope and Aspect images as well their derivatives and then we efficiently combine a rotation and scale invariant filter and a pixel labeling method, providing a robust detection of the sea faults. Experimental results on offshore real topographic images and their correlation with recent earthquake activity indicate the reliable performance of the proposed scheme.