Key points are not available for this paper at this time.
Historically, the automatic target recognition (ATR) and sensor fusion (SF) research and development community has suffered from a lack of consolidated, high quality, truthed signature data to support a variety of target and background conditions necessary to represent operational conditions. The sensor data management system (SDMS) has been developed to address this issue. The SDMS, a third-generation database, is a repository for all legacy and current image data sources at Wright Laboratory. This large scale (Terabyte-class) database was developed to provide comprehensive, high quality data to support the development of ATR and SF algorithms and architectures. Currently, the SDMS contains primarily synthetic aperture radar (SAR) data sources; however, the database design is flexible to accommodate other resources, such as high range resolution (HRR), electro-optical (EO), or infrared (IR) data sources. In addition, the SDMS includes auxiliary information such as ground truth, algorithm results, target photographs, data characterization metrics, and data manipulation tools. This paper will discuss the SDMS and future trends that will make sensor data available faster, cheaper and to a wider group of researchers. Our "data factory" approach will be discussed to implement the SDMS, which reduces "re-inventing the wheel".
A. Zembower (2002) studied this question.