Modeling Spectral–Temporal Data From Point Source Events

Abstract

In recent years, a great deal of effort has been invested in developing sensors to detect, locate, and identify "energetic" electromagnetic events. When observed through one type of imaging spectrometer, these events produce a data record that contains complete spectral and temporal information over the event's evolution. This article describes the development of a statistical model for the data produced by a particular spectral-temporal sensor. While the application is unique in some ways, this approach to model building may be useful in other related contexts. Several plots, estimated parameters, and some additional details for an equation are provided in the Appendix which is available as supplementary material online.

Modeling and discrimination for spectral-temporal data

“…Additionally, we will discuss practical issues that need serious consideration in moving forward with the work presented in Reising et al (2008a) and Reising et al (2008b). The sensor used for data collection by Reising et al(2008aReising et al( , 2008b The CTHIS sensor can be viewed as containing 2 system components, the chromotomgraphic sensor system and the digital post-processing and image construction system (Dearinger 2004). While the sensor has the capability of constructing hyperspectral images, our interest in the system and digital post-processing is in terms of how sub-pixel events are handled.…”

“…While refinement is needed, this methodology has potential to be used as a starting point in real-time discrimination between types of point source events. Reising et al (2008a) and 2008bwere generated by physically stationary point sources of energy. Further, we consider the implications of the fact that while the CTHIS is equipped with a rotating direct vision prism and rotating preprocessing algorithms, these were not utilized in the data collection and modeling in Reising et al (2008a).…”

“…Reising et al (2008a) and 2008bwere generated by physically stationary point sources of energy. Further, we consider the implications of the fact that while the CTHIS is equipped with a rotating direct vision prism and rotating preprocessing algorithms, these were not utilized in the data collection and modeling in Reising et al (2008a). Therefore, this article also discusses how data were manually extracted from the focal plane array.…”

“…2 Introduction Reising et al(2008a) state that in normal use, the Chromotomographic Hyperspectral Imaging Spectrometer (CTHIS) sensor used in data collection for their study makes use of a rotating prism and rotating preprocessing algorithm that can produce data matrices that are commonly visualized as "waterfall plots." However, for the purpose of inital development of modeling methods and a discrimination algorithm, the authors used the sensor in fixed-prism mode during data collection and collected raw (i.e.…”

“…The data collection of Reising et al (2008a) was done at a framing rate of 80 frames per second. We assume, but have not verified, that the change in the characteristic event evolution is smooth from one frame of data to the next at this framing rate.…”