Bayesian Urban Degradation Assessment

To address the NGA need for algorithms that fuse observables from over-flight operations and from ground sources to automatically estimatethe degradation of urban environments due to battle damage or natural disasters, Intellisense Systems, Inc. (ISS) proposes to develop a newBayesian Urban Degradation Assessment (BUDA) software system. It is based on the integration of multiple damage assessment algorithmsutilizing different data sources and implementation of a Bayesian inference engine for optimal data fusion of heterogeneous and multiscaledata for damage estimation of urban environments. Specifically, the novel system architecture integrating multiple concurrent algorithmsallows optimal 3D assessment of real-world observables from a wide variety of input data sources, and a new implementation of a Bayesianinference engine implements optimal data fusion, time-critical knowledge management/updates, and confidence estimation. In Phase I, ISSwill demonstrate the feasibility of BUDA by developing a proof-of-concept BUDA system and showing its capability to achieve optimal datafusion from different sources of overhead and ground imagery. In Phase II, ISS will develop a fully functional BUDA system capable ofautomatic assessment of urban environment degradation. We will demonstrate the increased accuracy of the fused data estimates ascompared to those derived from a single source.