Generation of a fundamental data set for hard/soft information fusion

Abstract

We report here on our effort to investigate the types of hard/soft information that can be realistically collected in an urban operational environment and to generate a data set that can be used for the development of hard/soft data fusion algorithms. Specifically, we discuss: 1) sources of ldquohard informationldquo (i.e. information from physics-based sources) and ldquosoft informationrdquo (i.e. information from human-based sources), 2) conceptual examples of hard/soft fusion, 3) data collection requirements and strategy, and 4) conceptual requirements for future algorithms with reference points in ongoing work. Specific techniques or frameworks for automated hard/soft fusion are beyond the scope of the effort being reported here. The presentation focuses on describing the context and requirements for hard/soft fusion at Joint Directors of Laboratories (JDL) Levels 1 and 2, and providing information about initial steps to generate a first-of-its-kind hard/soft data set that will serve as a foundation and a verification/validation resource for future research.