Remotely sensed satellite imagery of an earthquake-affected area can significantly assist in estimating the severity of infrastructure damage. Modern high-resolution satellite systems have been launched to provide users optical or Synthetic Aperture radar (SAR) data with sub-meter accuracy, which enable the possibility of sensing damage for individual infrastructure by means of pre- and post-event imagery. Herein, we focus our study on the region of Bam, Iran, which was devastated by a moment magnitude Mw = 6.6 earthquake on December 26, 2003, causing approximately 43,200 lives lost. To recognize houses within the Bam region before the earthquake, the boundary of houses are located using a combination of morphological gray-level open and intensity threshold operators. The unique aspect of this paper, as demonstrated with satellite imagery data from this event, is the use of an probabilistic framework for determining the optimal combination of morphological and intensity threshold parameters, which results in an estimated ground truth (EGT). By overlaying the EGT obtained from images before the earthquake with images of the same region after the earthquake, two statistical damage indices, including a new boundary-compactness based index proposed in this study, are compared. This comparison is presented using easily interpretable damage maps, where individual houses are rendered with colors representing the severity of damage.