Janet L. Jensen

Automatic target recognition for hyperspectral imagery using high-order statistics

Due to recent advances in hyperspectral imaging sensors many subtle unknown signal sources that cannot be resolved by multispectral sensors can be now uncovered for target detection, discrimination, and identification. Because the information about such sources is generally not available, automatic target recognition (ATR) presents a great challenge to hyperspectral image analysts. Many approaches developed for ATR are based on second-order statistics in the past years. This paper investigates ATR techniques using high order statistics. For ATR in hyperspectral imagery, most interesting targets usually occur with low probabilities and small population and they generally cannot be described by second-order statistics. Under such circumstances, using high-order statistics to perform target detection have been shown by experiments in this paper to be more effective than using second order statistics. In order to further address a challenging issue in determining the number of signal sources needed to be detected, a recently developed concept of virtual dimensionality (VD) is used to estimate this number. The experiments demonstrate that using high-order statistics-based techniques in conjunction with the VD to perform ATR are indeed very effective

Spectral abundance fraction estimation of materials using Kalman filters

Kalman filter has been widely used in statistical signal processing for parameter estimation. Although a Kalman filter approach has been recently developed for spectral unmixing, referred to as Kalman filter-based linear unmixing (KFLU), its applicability to spectral characterization within a single pixel vector has not been explored. This paper presents a new application of Kalman filtering in spectral estimation and quantification. It develops a Kalman filter-based spectral signature esimator (KFSSE) which is different from the KFLU in the sense that the former performs a Kalman filter wavelength by wavelength across a spectral signature as opposed to the latter which implements a Kalman filter pixel vector by pixel vector in an image cube. The idea of the KFSSE is to implement the state equation to characterize the true spectral signature, while the measurement equation is being used to describe the spectral signature to be processed. Additionally, since a Kalman filter can accurately estimate spectral abundance fraction of a signature, our proposed KFSSE can further used for spectral quantification for subpixel targets and mixed pixel vectors, called Kalman filter-based spectral quantifier (KFSQ). Such spectral quantification is particularly important for chemical/biological defense which requires quantification of detected agents for damage control assessment. Several different types of hyperspectral data are used for experiments to demonstrate the ability of the KFSSE in estimation of spectral signature and the utility of the KFSQ in spectral quantification.

Anomalous thermal expansion of mercurous halides

Mercurous halides; mercurous iodide (Hg2I2), mercurous bromide (Hg2Br2) and mercurous chloride (Hg2Cl2) are high figure of merit materials for fabricating Acousto-optic (AO) devices that operate in the visible and infrared regions. Single crystals of mercurous halides were grown by physical vapor transport (PVT) method. Thermal expansion as well as the effect of annealing on the material are discussed. Mercurous halides show positive thermal expansion along a axis whereas no significant thermal expansion along c axis. The coefficients of thermal expansion of Hg2Cl2, Hg2Br2 and Hg2I2 are 6.72 × 10-5 °C-1, 6.44× 10-5 °C-1 and 6.08 × 10-5 °C-1, respectively. The optical band gap of Hg2Cl2 was calculated using the transmission spectra as 2.9 eV.

Long wavelength infrared (LWIR) AOTF and AOM modulators using Hg2Br2 crystals

An acousto-optic devices were designed and fabricated using high quality single crystals of mercurous halide (Hg2X2) that were grown by physical vapor transport method (PVT). The orientation and the crystalline quality of the grown crystals were determined using high resolution x-ray diffraction (HRXRD) technique. The full width at half maximum (FWHM) of the grown mercurous bromide (Hg2Br2) crystals was measured to be 0.13 degrees for (004) reflection, which is the best that has been achieved so far for PVT grown mercurous halide single crystals. The extended defects of the crystals were also analyzed using high resolution x-ray diffraction topography. Preliminary studies were carried out to evaluate the performance of the crystals on acousto-optic modulator (AOM) and acousto-optic tunable filter (AOTF) applications. The results indicate the grown mercurous halide crystals are excellent materials for acousto-optic modulator as well as acousto-optic tunable filter device fabrications. The diffraction efficiencies of the fabricated AOM device with 1152 and 1523nm wavelength lasers polarizing parallel to the acoustic wave were found to be 35% and 28%, respectively. The diffraction efficiencies of the fabricated AOTF device with 10600 nm wavelength laser found to be 26%.

Nanoengineered Bioplatforms Based on DNA Origami [Point of View]

DNA is generally associated with the storage of genetic information. However, in many ways, it is also an ideal building material. The shape of a DNA structure is determined by the sequences of the DNA strands within the structure. DNA origami [1] has recently evolved as a method for producing programmable structures at the nanoscale. In a DNA origami, a long single-stranded DNA molecule is folded and held in place with shorter DNA strands. This process can be visualized by taking a very long tube or hose and folding it into a desired shape. Smaller strands of the same material can then be used to tie the large tube into a space-filling structure. In the case of DNA origami the shorter strands are called staples. The staples crosslink and stabilize the entire structure, enabling the formation of complex and programmable 2-D and 3-D shapes. Structures with considerable complexity can be designed and produced.

An embedded system developed for hand held assay used in water monitoring

The US Army Joint Service Agent Water Monitor (JSAWM) program is currently interested in an approach that can implement a hardware- designed device in ticket-based hand-held assay (currently being developed) used for chemical/biological agent detection. This paper presents a preliminary investigation of the proof of concept. Three components are envisioned to accomplish the task. One is the ticket development which has been undertaken by the ANP, Inc. Another component is the software development which has been carried out by the Remote Sensing Signal and Image Processing Laboratory (RSSIPL) at the University of Maryland, Baltimore County (UMBC). A third component is an embedded system development which can be used to drive the UMBC-developed software to analyze the ANP-developed HHA tickets on a small pocket-size device like a PDA. The main focus of this paper is to investigate the third component that is viable and is yet to be explored. In order to facilitate to prove the concept, a flatbed scanner is used to replace a ticket reader to serve as an input device. The Stargate processor board is used as the embedded System with Embedded Linux installed. It is connected to an input device such as scanner as well as output devices such as LCD display or laptop etc. It executes the C-Coded processing program developed for this embedded system and outputs its findings on a display device. The embedded system to be developed and investigated in this paper is the core of a future hardware device. Several issues arising in such an embedded system will be addressed. Finally, the proof-of-concept pilot embedded system will be demonstrated.

Pyrolysis gas-chromatography ion-mobility spectrometry data analysis for chemical detection in water monitoring

Pyrolysis-Gas Chromatography-Ion Mobility Spectrometry (Py-GC-IMS) has been used to detect chemical resident in water by obtaining the characteristic information on the structure and molecular weight. It is a great challenge to analyze this 2-dimensional signal because the noise is not stationary. In this paper, one approach has been designed to remove noise baseline and suppress noise power. The provided data is used for analyze the improvement under controlled experiment. Receiver Operating Characteristic (ROC) curve is also used to analyze the performance and it shows the proposed method improves the detection results.