Search

This paper investigates methods of decision-making from uncertain and disparate data. The need for such methods arises in those sensing application areas in which multiple and diverse sensing modalities are available, but the information provided can be imprecise or only indirectly related to the effects to be discerned. Biological sensing for biodefense is an important instance of such applications. Information fusion in that context is the focus of a research program now underway at MIT Lincoln Laboratory. The paper outlines a multi-level, multi-classifier recognition architecture developed within this program, and discusses its components. Information source uncertainty is quantified and exploited for improving the quality of data that constitute the input to the classification processes. Several methods of sensor uncertainty exploitation at the feature-level are proposed and their efficacy is investigated. Other aspects of the program are discussed as well. While the primary focus of the paper is on biodefense, the applicability of concepts and techniques presented here extends to other multisensor fusion application domains.

Well-chosen background models are critical to accurately predict the performance of hyperspectral detection and classification algorithms and to evaluate the effects on system performance of variation in environmental or sensor parameters. Such models also have implications for the derivation of optimal algorithms. First-principal physical models and statistical models have been developed for these purposes. However, in many circumstances these models may not accurately represent hyperspectral data that are complicated by intra-class variability and subpixel mixing of materials as well as atmospheric, illumination, temperature (in the emissive regime) and sensor effects. In this paper we propose a statistical representation of hyperspectral data defined by class parameters and an abundance probability distribution. Various representations of the probability distribution function of the abundance values are developed and compared with data to determine if the estimated abundance distributions and intra-class variation explain the observed heavy tails in the data. The consequences of the Gaussian endmembers of the normal compositional model violating the non-negativity constraint are also investigated.

The normal compositional model (NCM) simultaneously models subpixel mixing and intra-class variation in multidimensional imagery. It may be used as the foundation for the derivation of supervised and unsupervised classification and detection algorithms. Results from applying the algorithm to AVIRIS SWIR data collected over Cuprite, Nevada are described. The NCM class means are compared with library spectra using the Tetracorder algorithm. Of the eighteen classes used to model the data, eleven are associated with minerals that are known to be in the scene and are distinguishable in the SWIR, five are identified with Fe-bearing minerals that are not further classifiable using SWIR data, and the remaining two are spatially diffuse mixtures. The NCM classes distinguish (1) high and moderate temperature alunites, (2) dickite and kaolinite, and (3) high and moderate aluminum concentration muscovite. Estimated abundance maps locate many of the known mineral features. Furthermore, the NCM class means are compared with corresponding endmembers estimated using a linear mixture model (LMM). Of the eleven identifiable (NCM class mean, LMM endmember) pairs, ten are consistently identified, while the NCM estimation procedure reveals a diagnostic feature of the eleventh that is more obscure in the corresponding endmember and results in conflicting identifications.

The normal compositional model (NCM) is a descriptive model that explicitly accounts for sub-pixel mixing and random variation of the spectrum of a material. In this paper the normal compositional model, defined in an earlier work, is extended to include an additive term that may represent path radiance and additive sensor noise. If the covariance matrix of the additive term is non-singular, as may be assumed since it includes the covariance matrix of the additive noise, the covariance matrix of the other classes need not be non-singular. Thus the current model synthesizes the linear unmixing and Gaussian clustering algorithms. Anomaly and matched target detection algorithms based on these three models are compared using ocean hyperspectral imagery, and for these data the NCM approach reduces the false alarm probability by more than an order of magnitude. The linear mixture and normal compositional models separate surface reflections and upwelling light more effectively than the Gaussian clustering algorithm. Furthermore, greater inter-band correlation is estimated using the subpixel covariance estimation methodology than using the pure pixel modeling approach.

A three year seeing measurement campaign was conducted at the 4.2-m William Herschel Telescope (WHT) using a Shack- Hartmann (S-H) wavefront sensor. An analysis of the spatio- temporal autocorrelation of the S-H centroid motion across the WHT pupil demonstrates the partial validity of Taylor's hypothesis, but at the same time highlights its limitations in the context of ground-based interferometry. So-called wavefront `boiling' is shown to play a major, perhaps dominant role in the evolution of phase errors on the timescales of practical interest. This conclusion is reached by comparing the coherence times measured from the fluctuation of reconstructed wavefronts with those inferred from Taylor's hypothesis using the wind speed determined from the S-H spatio-temporal autocorrelation. Nevertheless it appears that knowledge of the wind speed can be a reliable indicator of coherence time, provided a suitable calibration is established. The practical problem of measuring the turbulence coherent time using portable equipment is explored. The promising method of centroid velocity variance is investigated and the wind speeds measured this way are seen to agree with those determined from the S-H spatio-temporal autocorrelation.

Fiber optics is useful for connecting the element telescopes of a long-baseline interferometer. Experiments with a fiber- connected interferometer are carried out as a part of MIRA- I.2 interferometer under construction. MIRA-I.2 is an interferometer consisting of two 300 mm siderostats which are placed on a 30 m south-north baseline. For this baseline, it is necessary to use fibers of about 100 m long. Fiber optics will be essential for the interferometric connection of large telescopes, such as Subaru, Gemini, and Keck on the summit of Mauna Kea. In this case, fiber length of about 1 km is needed. We achieved white light fringes using 10 m and 100 m fibers. With the 10 m fiber, visibility of 0.7 was achieved for both He-Ne laser and white light sources. Visibility of 0.7 for white light was obtained by compensating the dispersion effect of fiber with glass plates. Experiments will be extended to longer fibers such as 1000 m.

There are difficult problems involved in building a near-infrared interferometer which uses more than two elements simultaneously. These problems have been overcome at the Cambridge Optical Aperture Synthesis Telescope(COAST). This has allowed us to make the first closure phase measurements on an astronomical source in the infrared. We describe a scheme for fast, low noise readout of an infrared focal-plane array detector, capable of adequately sampling pupil plane fringes on three simultaneous baselines, as well as a procedure for aligning a many-component beam-combiner in the infrared. Finally, the performance of the working COAST infrared system is discussed.

A low-resolution CCD spectrometer has been installed at COAST to provide multi-wavelength fringe measurements across the band 650-950 nm. The measurements are based on the analysis of time-series of channeled spectra. Laboratory tests and stellar observations are presented. The advantages and limitations of the system are discussed.

Visibility measurements obtained with optical astronomical interferometers are corrupted by random wavefront distortions of atmospheric origin. In this paper we discus how spatial filtering using pinholes can lead to increased measured visibility, improved signal-to-noise ratio and reduced dependence on seeing fluctuations. The potential for calibrating visibility measurements without resorting to a separate calibrator target is also discussed. Results of preliminary pinhole experiments carried out with the Cambridge Optical Aperture Synthesis Telescope are presented.

This paper reviews the current performance of the Cambridge Optical Aperture Synthesis Telescope as an imaging array. Tests of the hardware and methods of measuring fringe visibility and closure phase are described in the context of prospects for a Large Optical Array.