Ray-Tracing-Based Geospatial Optimization for Heterogeneous Architectures Enhancing Situational Awareness

This paper presents the implementation of ray tracing-based algorithms for multi-objective geospatial optimization targeting various many-core processing technologies such as graphics processing units, x86 multi-cores, and ARM processors. High performance is achieved through highly parallel core algorithms, executed on multiple compute devices across a heterogeneous architecture using low-level OpenCL kernels. Algorithms for calculating line-of-sight ballistic threat, visual observability, ground plane extraction, and Markov chain Monte Carlo optimization provide an augmented geospatial intelligence and situational awareness in three-dimensional urban environments.