A method for trace gas detection in hyperspectral data is demonstrated using the wavelet packet transform. This new
method, the Wavelet Packet Subspace (WPS), applies the wavelet packet transform and selects a best basis for pattern
matching. The wavelet packet transform is an extension of the wavelet transform, which fully decomposes a signal into a
library of wavelet packet bases. Application of the wavelet packet transform to hyperspectral data for the detection of
trace gases takes advantage of the ability of the wavelet transform to locate spectral features in both scale and location.
By analyzing the wavelet packet tree of specific gas, nodes of the tree are selected which represent an orthogonal best
basis. The best basis represents the significant spectral features of that gas. This is then used to identify pixels in the
scene using existing matching algorithms such as spectral angle or matched filter. Using data from the Airborne
Hyperspectral Imager (AHI), this method is compared to traditional matched filter detection methods. Initial results
demonstrate a promising wavelet packet subspace technique for hyperspectral trace gas detection applications.
It is often believed that quantum entanglement plays an important role in the speed-up of quantum algorithms. In
addition, a few research groups have found that Majorization behavior may also play an important role in some quantum
algorithms. In some of our previous work we showed that for a simple spin 1/2 system, consisting of two or three qubits,
the value of a Groverian entanglement (a rather useful measure of entanglement) varies inversely with the temperature.
In practical terms this means that more iterations of the Grover's algorithm may be needed when a quantum computer is
working at finite temperature. That is, the performance of a quantum algorithm suffers due to temperature-dependent
changes on the density matrix of the system. Most recently, we have been interested in the behavior of Majorization for
the same types of quantum system, and we are trying to determine the relationship between Groverian entanglement and
Majorization at finite temperature. As Majorization entails the probability distribution arising out of the evolving
quantum state from the probabilities of the final outcomes, our study will reveal how Majorization affects the evolution
of Grover's algorithm at finite temperature.
Some of our previous research showed some interesting results regarding the effect of non-zero temperature on a
specified quantum computation. For example, our analysis revealed that more Grover iterations are required to amplify
the amplitude of the solution in a quantum search problem when the system is found at some finite temperature. We want
to further study the effects of temperature on quantum entanglement using a finite temperature field theoretical
description. Such a framework could prove to be useful for the understanding of computational dynamics inside a
quantum computer. Other issues that we will address in our discussion include analytical descriptions of the effects of
the temperature in the Von Newman entropy and others as a measure of entanglement.
The engineering of practical quantum computers requires dealing with the so-called "temperature mismatch problem".
More specifically, analysis of quantum logic using ensembles of quantum systems typically assumes very low
temperatures, kT<< E, where T is the temperature, k is the Boltzmann's constant, and E is the energy separation used to
represent the two different states of the qubits. On the other hand, in practice the electronics necessary to control these
quantum gates will almost certainly have to operate at much higher temperatures. One solution to this problem is to
construct electronic components that are able to work at very low temperatures, but the practical engineering of these
devices continues to face many difficult challenges. Another proposed solution is to study the behavior of quantum gates
devices continues to face many difficult challenges. Another proposed solution is to study the behavior of quantum gates
different from the T=0 case, where collective interactions and stochastic phenomena are not taken into consideration. In
this paper we discuss several aspects of quantum logic at finite temperature. In particular, we present analysis of the
behavior of quantum systems undergoing a specified computation performed by quantum gates at nonzero temperature.
Our main interest is the effect of temperature on the practical implementation of quantum computers to solve potentially
large and time-consuming computations.
Through the centuries gem materials have been highly prized and sought after. The varieties of gem materials run into the hundreds if not thousands, characterized by a gamut of material classes running from organic to inorganic and from crystalline to amorphous. All consisting of numerous chemical compositions and characterized by various physical and optical properties. In addition, most gem materials have been subject to numerous modifications to enhance and imitate the most pleasing of esthetic qualities, e.g., dyeing, impregnation, heating, reconstruction, high pressure and temperature, irradiation, and diffusion. Of concern is the ability not only to identify the gem material in question, but if applicable, the treatment. Up until recent, the main instruments utilized to detect these have been simple but quite effective such as a binocular microscope, refractometer, hand spectroscope, dichroscope, and measuring of specific gravity. New gem materials and techniques involved in treatments have become increasingly sophisticated such as ultraviolet-visible-infrared and Raman spectroscopy. In certain cases, some of the most recent techniques have become time consuming and expensive. Here is the opportunity to overview and utilize a powerful technology found in the field of remote sensing, i.e., Hyperspectral Imaging. This technology has been in effect for many years but only recently has it been used to focus on areas similar to the ones in this paper. In particular, hyperspectral imaging technology and its potential application to gem identification and authentication are covered in this paper.
This paper surveys the potential use of hyperspectral imaging technology for standoff detection of chemical and biological agents in terrorism defense applications. In particular it focuses on the uses of hyperspectral imaging technology to detect and monitor chemical and biological attacks. In so doing it examines current technologies, their advantages and disadvantages, and investigates the possible role of hyperspectral imaging for homeland security applications. The study also addresses and provides applicable solutions for several of the potential challenges that currently create barriers to the full use of hyperspectral technology in the standoff detection of likely available chemical and biological agents.
As interest in quantum computing evolves, consideration must be given to the development of new methods to improve the current design of quantum computers. Such ideas are not only helping the advance towards practical quantum computation applications, but are also providing clearer understanding of quantum computation itself. Eventually, several new exploratory efforts to increase the efficiency beyond the inherent advantage of quantum computational systems to classical systems will materialize. As a part of these exploration efforts, this paper presents a modified version of the qubit, which we refer to as a "Qubit", that allows a smaller number of Qubits than qubits to reach the same result in applications such as Shor’s algorithm for the factorization of large numbers. The current model of the qubit consists of a quantum bit with two states, a zero and a one in a quantum superposition state. The Qubit, which consists of more than two states, is introduced and explained. A mathematical analysis of the Qubit within Hilbert space is given. We present examples of applications of the Qubit to several quantum computing algorithms, including discussion of the advantages and disadvantages that are involved. Finally a physical model to construct such a Qubit is considered.
Recent research on the topic of quantum computation provides us with some quantum algorithms with higher efficiency and speedup compared to their classical counterparts. In this paper, it is our intent to provide the results of our investigation of several applications of such quantum algorithms - especially the Grover's Search algorithm - in the analysis of Hyperspectral Data. We found many parallels with Grover's method in existing data processing work that make use of classical spectral matching algorithms. Our efforts also included the study of several methods dealing with hyperspectral image analysis work where classical computation methods involving large data sets could be replaced with quantum computation methods. The crux of the problem in computation involving a hyperspectral image data cube is to convert the large amount of data in high dimensional space to real information. Currently, using the classical model, different time consuming methods and steps are necessary to analyze these data including: Animation, Minimum Noise Fraction Transform, Pixel Purity Index algorithm, N-dimensional scatter plot, Identification of Endmember spectra - are such steps. If a quantum model of computation involving hyperspectral image data can be developed and formalized - it is highly likely that information retrieval from hyperspectral image data cubes would be a much easier process and the final information content would be much more meaningful and timely. In this case, dimensionality would not be a curse, but a blessing.
Oil pollution is a very important aspect in the environmental field. Oil pollution is an important subject due to its capacity to adversely affect animals, aquatic life, vegetation and drinking water. The movement of open water oil spills can be affected by mind, waves and tides. Land based oil spills are often affected by rain and temperature. It is important to have an accurate management of the cleanup. Remote sensing and in particular hyper-spectral capabilities, are being use to identify oil spills and prevent worse problems. In addition to this capability, this technology can be used for federal and state compliance of petroleum related companies. There are several hyper-spectral sensors used in the identification of oil spills. One commonly use sensor is the Airborne Imaging Spectroradiometer for Applications (AISA). The main concern associated with the use of these sensors is the potential for false identification of oil spills. The use of AISA to identify an oil spill over the Patuxent River is an example of how this tool can assist with investigating an oil pipeline accident, and its potential to affect the surrounding environment. A scenario like this also serves as a good test of the accuracy with which spills may be identified using new airborne sensors.
This paper analyzes the feasibility and performance of HSI systems for medical diagnosis as well as for food safety. Illness prevention and early disease detection are key elements for maintaining good health. Health care practitioners worldwide rely on innovative electronic devices to accurately identify disease. Hyperspectral imaging (HSI) is an emerging technique that may provide a less invasive procedure than conventional diagnostic imaging. By analyzing reflected and fluorescent light applied to the human body, a HSI system serves as a diagnostic tool as well as a method for evaluating the effectiveness of applied therapies. The safe supply and production of food is also of paramount importance to public health illness prevention. Although this paper will focus on imaging and spectroscopy in food inspection procedures -- the detection of contaminated food sources -- to ensure food quality, HSI also shows promise in detecting pesticide levels in food production (agriculture.)
A wide variety of hyper-spectral (HS) sensors and collection platforms are in existence. This paper investigates hyper-spectral imaging systems (HIS) worldwide in order to compose a comprehensive listing of these systems. A meta-data structure was developed to identify basic parameter information for all sensors that were reviewed. Systems were grouped into two primary categories of space borne and air borne. Sensors were further grouped into three types of imaging spectrometers; whiskbroom line array band interleaved by pixel, push broom area array band interleaved by line and framing camera band sequential methodology. Several sensor systems are presented using the meta-data structure and parameters developed for the analysis. A summary table identifying all sensor systems that were evaluated is presented. Applications include geo-environmental studies, aerosol release, materials identification, agricultural studies, atmospheric studies, and many others.
The Squeezed Signature Analysis (SSA) hyperspectral classification method is presented as a fast method to compare the target spectral signature to all the signatures in the spectral library. This discussion includes the possible use of this technique on-board a hypothetical remote sensing system that could take advantage of parallel computations.
Public Works facilities require up-to-date information on the health status of the road network they maintain. However, roadway maintenance and rehabilitation involves the greatest portion of a municipality's annual operating budget. Government officials use various technologies such as a pavement management system to assist in making better decisions about their roadways systems, pavement condition, history, and projects. Traditionally, manual surveying has served as the method of obtaining this information. To better assist in decision-making, a regionally specific spectral library for urban areas is being developed and used in conjunction with hyperspecrtal imaging, to map urban materials and pavement conditions. A Geographical Information and Positioning System (GIS/GPS) will also be implemented to overlay relative locations. This paper will examine the benefits of using hyperspectral imaging over traditional methods of roadway maintenance and rehabilitation for pavement management applications. In doing so, we will identify spatial and spectral requirements for successful large-scale road feature extraction.
Precision farming relies on the cost effectiveness of collecting and interpreting data, which describes the variations of agricultural conditions such as crop stresses, nutrient deficiencies, water stresses, or pest infestation. Hyperspectral remote sensing from satellites and airborne sensors can be a way to obtain data needed to develop site-specific farming management strategies. The primary objective of the hyperspectral applications in precision farming is to provide farmers with a technology, which can detect specific crop conditions that can be used to program variable-rate applications. Applications of water, pesticides, and fertilizer can be tailored to the needs of the agricultural crops, based on the conditions reflected on the imagery. This paper presents an experimental study performed in Beltsville, Maryland for assessing the plant density and nutrient uptake of corn using a simple photographic method from a model airplane versus obtaining hyperspectral imagery from an airborne sensor. The hyperspectral sensor utilized in this study was the AISA sensor. These remote sensors can measure the temperature of plants; or to be more specific, they can measure how much energy plants emit at the visible and near-infrared wavelengths of the spectrum, such as water and vegetation.
A wide variety of hyper-spectral (HS) sensors and collection platforms are in existence. This paper investigates hyper-spectral imaging systems (HIS) worldwide in order to address issues associated with the better both airborne and space based systems are included in the review. Examples of the sensors include ENVISAT, SCIAMACHY, TERRA, AQUA, MOPITT, MIPAS, AVIRIS, LIDAR, Landsat 7 and others. Applications include geo-environmental studies, aerosol release, materials identification, agricultural studies, atmospheric studies, and many others. Two case studies are presented that address the evaluation of African smog and its effect on the African ecosystem and the evaluation of aerosol pollution in the northeastern region of the United states with particular attention to particulate matter.
Hyperspectral imagining has been recently been used to obtain several water quality parameters in water bodies either inland or in oceans. Optical and thermal have proven that spatial and temporal information needed to track and understand trend changes for these water quality parameters will result in developing better management practices for improving water quality of water bodies. This paper will review water quality parameters Chlorophyll (Chl), Dissolved Organic Carbon (DOC), and Total Suspended Solids (TSS) obtained for the Sakonnet River in Narragansett Bay, Rhode Island using the AVIRIS Sensor. The AVIRIS Sensor should improve the assessment and the definition of locations and pollutant concentrations of point and non-point sources. It will provide for necessary monitoring data to follow the clean up efforts and locate the necessary water and wastewater infrastructure to eliminate these point and non-point sources. This hyperspectral application would enhance the evaluation by both point and non-point sources, improve upon and partially replace expenses, labor intensive field sampling, and allow for economical sampling and mapping of large geographical areas.
Despite the considerable slowdown of wetlands loss in conterminous U.S., management of these valuable resources continues to be an area of interest for environmental professionals. The development of remote sensing technologies, particularly hyperspectral, offers an alternative for ecological and functional assessment of these sites. Extensive hyperspectral data image collected from the various sensor types can be analyzed by discriminatory techniques for reflectance analysis. Although data processing can become tedious, it enables scientists to target the various inherited characteristics of large wetland areas such as vegetative species and habitats. This information can be applied to determine the health and functionality of the nation's wetlands for means of wetland characterization, assessment, management and possible restorative efforts to bring a consistent and fundamental change on how these are managed today.
In recent years, computer graphics has emerged as a critical component of the scientific and engineering process, and it is recognized as an important computer science research area. Computer graphics are extensively used for a variety of aerospace and defense training systems and by Hollywood's special effects companies. All these applications require the computer graphics systems to produce high quality renderings of extremely large data sets in short periods of time. Much research has been done in "classical computing" toward the development of efficient methods and techniques to reduce the rendering time required for large datasets. Quantum Computing's unique algorithmic features offer the possibility of speeding up some of the known rendering algorithms currently used in computer graphics. In this paper we discuss possible implementations of quantum rendering algorithms. In particular, we concentrate on the implementation of Grover's quantum search algorithm for Z-buffering, ray-tracing, radiosity, and scene management techniques. We also compare the theoretical performance between the classical and quantum versions of the algorithms.
Many researchers consider coral reefs the 'rainforests of the oceans' because they cover such a small area and yet provide homes for literally thousands of unique marine species. A multispectral or hyperspectral remote sensing satellite, with its spectral coverage, offers iadvantages over traditional methodologies for coral reef surveying, monitoring, and mapping. This apper presents research into the suitabilty of spectral remote sensing for coral reed surveying, monitoring and mapping. This paper presents research into the suitability of spectral remote sensing for coral reef surveying, monitoring and mapping using the SeaWiFS multispectral ocean color data for illustration. We describe the information technology developed to support this research and provide an overview of the database driven web application, which was developed to allow live interaction with the data. A database of in situ observations from the ReefBase web site was used as validation data as part of this investigation. This discussion includes details on the XML representation of the satellite and in situ data and metadat. It also introduces a dynamic Java Visualization applet developed to allow the users to visually interact wiht the data. The paper concludes wiht a discussion of the suitability and additional advantages of using hyperspectral remote sensing technology for this application that exploits the full spectral characteristics of submerged coral reefs.
Virginia Access (VAccess) is a regional, remote sensing and Geographical Information Sciences project among several educational institutions. It is a prototype for regional projects in other states and other countries, and is funded by NASA's applications program. The user communities VAccess serves are the Commonwealth of Virginia and State of Maryland, local and regional users represented in a Technical Advisory Committee. Remote sensing data include global NASA and NOAA data tailored for regional applications as well as high-resolution multispectral (Landsat, MODIS, etc.), hyperspectral, LIDAR and SAR data sets. Broad beam LIDAR technology can provide canopy structure as well as other information for environmental concerns such as the state of wetlands. The data information system is based on a distributed architecture to serve remote sensing and GIS data to a variety of users via the WWW. Several remote sensing and GIS-based environmental and Earth systems science applications projects are discussed here, including flood and fire hazard mitigation, forestry, land use/land cover and epidemiology projects; as well as innovative data fusion, data access and analysis and various tools serving the users and their applications.
Some types of clay, esp. montmorillonite, become slippery when getting wet. Clay movement is very harmful for various constructions and can also cause trouble for both wheeled and tracked vehicles in military operations at some rural areas when raining. We present a summary of a project using hyperspectral imaging in assisting earth roads construction planning and cross-country trafficability analysis. Spectral signature libraries are used to help identify materials and define those areas to be avoided, which have significant montmorillonite content. We perform a case study in this kind of application; some methods of data processing and analyzing are discussed. We also discussed the problems we met in this application. Hyperspectral sensing is a relatively new but mature technology; development of applications and corresponding analyzing procedures will be the major impetus of this technology.
Hyperspectral imaging is a relatively new technology which draws much attention from the scientists now. With information provided on hundreds of narrow and continuous bands, it finds applications in many areas such as mineral identification, environmental monitoring, agricultural survey, medical examination. However, it is not an easy task to utilize the hyperspectral images, due to the large data volume. Certain tools are needed to analyze the hyperspectral image. As an example, the ENvironment for Visualizing Images (ENVI) software is used for the study of a sample hyperspectral image.
Proc. SPIE. 4383, Geo-Spatial Image and Data Exploitation II
KEYWORDS: Hyperspectral imaging, Digital signal processing, Principal component analysis, Image processing, Spectrometers, Data processing, Signal processing, Image classification, Algorithm development, Standards development
The need for fast hyperspectral data processing methods is discussed. Discussion includes the necessity of faster processing techniques in order to realize emerging markets for hyperspectral data. Several standard hyperspectral image processing methods are presented, including maximum likelihood classification, principal components analysis, and canonical analysis. Modifications of those methods are presented that are computationally more efficient than standard techniques. Recent technological developments enabling hardware acceleration of hyperspectral data processing methods are also presented as well as their applicability to various hyperspectral data processing algorithms.
Different research groups have recently studied the concept of wavelet image fusion between panchromatic and multispectral images using different approaches. In this paper, a new approach using the wavelet based method for data fusion between hyperspectral and multispectral images is presented. Using this wavelet concept of hyperspectral and multispectral data fusion, we performed image fusion between two spectral levels of a hyperspectral image and one band of multispectral image. The reconstructed image has a root mean square error of 2.8 per pixel and a signal-to- noise ratio of 36 dB. We achieved our goal of creating a composite image that has the same spectral resolution as the hyperspectral image and the same spatial resolution as the multispectral image with minimum artifacts.