William E. Roper

Hyperspectral imaging applied to medical diagnoses and food safety

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.)

Waste management policy revisions: lessons learned from the Katrina disaster

The main objective of this paper is to identify debris and waste management policies that need to be changed based on the lessons learned from Hurricane Katrina. Policy issues addressed include fragmented jurisdictional problems, issues related to types of debris, burning of house hold debris, wood infestation with Formosan termites and banning of yard wastes from landfills. Current practices and trends in the building material waste management following disasters are examined from a building life cycle standpoint or cradle to cradle concept. Completing the proper planning before the disaster is critical. Having a plan in place can allow for maximum, integrated recycling, resource optimisation, waste reduction and deconstruction. Examination of the waste management hierarchy and life cycle management of material is used to improve the understanding of reuse and recycle opportunities. Based on the lessons learnt from Hurricane Katrina proposed changes in debris management policy following natural disasters.

Strategies for building material reuse and recycle

Reuse and recycling of building material is a growing area of interest and concern in many parts of the USA. Current practices and trends in the building material waste management area are examined from a building life cycle standpoint or cradle to reincarnation concept. Strategies include zero waste, integrated recycling, international approaches, reuse of materials, resource optimisation, waste reduction, and deconstruction. Examination of the waste management hierarchy and life cycle management of material is used to improve the understanding of reuse and recycle opportunities. Other considerations include cost, economic factors, social factors and environmental factors. All of these assessments are needed to develop a comprehensive waste management plan for a specific project. Four case studies are also presented to illustrate the benefits that can be derived from effective building material reuse and recycle.

Spatial multi-database integration for emergency operations support

Geospatial information systems and their application to emergency response can be very beneficial but many technology adoption barriers need to be overcome for best utilisation. Considerations for the management of infrastructure and transportation systems during a disaster situation are presented. A case study of the experience and lessons learned from geospatial technology applications for the New York City response to terrorist attacks on the World Trade Center is also addressed. The opportunities for information systems leveraging geospatial data sources are reviewed. Remote sensing imaging and non-imaging sources of information are explored. Coordination and sharing of information is explored to improve the ability of multiple organisations to more effectively manage and respond to emergency situations. The use of geospatial information was only a part of the total recovery activity but did provide important assistance throughout the recovery operation. Samples of geospatial technologies used in the recovery are presented. These include 3-D visualisation, thermal infrared imagery, LIDAR data systems, IKONOS one-meter panchromatic imagery, SPOT imagery and the use of digital aerial imagery. The general area of disaster response management is also addressed and the findings of various studies in this area are summarised. Recommendations for the use of information systems and geospatial sensors and products in the future for infrastructure management are presented.

Detection and monitoring of oil spills using hyperspectral imagery

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.

Various continuous harmony search algorithms for web-based hydrologic parameter optimisation

This study compares five different harmony search algorithms that consider continuous variables inherently for the hydrologic parameter calibration. A rainfall intensity assessing model, which can provide stochastic rainfall sizes used for various structure designs, is optimally calibrated with the harmony search algorithms. The results showed that three different harmony search algorithms found better parameter values for the rainfall intensity model than mathematical optimisation technique (Powell method) and evolutionary computation technique (genetic algorithm) with respect to root mean square error (RMSE). Also, in order to enhance user-friendliness in utilising this technique, a web-based technique was developed, which optimally performs parameter calibration and presents its result on a digital map without requiring software installation.

Spectral signatures of hydrilla from a tank and field setting

The invasion of hydrilla in many waterways has caused significant problems resulting in high maintenance costs for eradicating this invasive aquatic weed. Present identification methods employed for detecting hydrilla invasions such as aerial photography and videos are difficult, costly, and time consuming. Remote sensing has been used for assessing wetlands and other aquatic vegetation, but very little information is available for detecting hydrilla invasions in coastal estuaries and other water bodies. The objective of this study is to construct a library of spectral signatures for identifying and classifying hydrilla invasions. Spectral signatures of hydrilla were collected from an experimental tank and field locations in a coastal estuary in the upper Chesapeake Bay. These measurements collected from the experimental tank, resulted in spectral signatures with an average peak surface reflectance in the near-infrared (NIR) region of 16% at a wavelength of 818 nm. However, the spectral measurements, collected in the estuary, resulted in a very different spectral signature with two surface reflectance peaks of 6% at wavelengths of 725 nm and 818 nm. The difference in spectral signatures between sites are a result of the components in the water column in the estuary because of increased turbidity (e.g., nutrients, dissolved matter and suspended matter), and canopy being lower (submerged) in the water column. Spectral signatures of hydrilla observed in the tank and the field had similar characteristics with low reflectance in visible region of the spectrum from 400 to 700 nm, but high in the NIR region from 700 to 900 nm.

Renewable energy production issues with the Cape Cod offshore wind energy programme

This paper evaluates the first offshore wind farm proposal in the USA. It is being put forward by Cape Wind Associates, a Boston-based firm that wishes to build 170 turbines five miles offshore in Nantucket Sound. The proposed development would supply up to 95% Cape Cods energy needs and 2% of the energy requirements for the eastern New England region, preventing the need for additional fossil fuel power plants that would pollute the environment and harm the health of New Englanders. The eco-friendly state of Massachusetts has produced much opposition to this project. Criticism of this project ranges from anticipated bird kills by the turbines to concerns about the criteria and considerations that will be used in the permit process. However, the most important issue appears to be that of visual pollution to the residents of Marthas Vineyard and Nantucket. This paper examines this unique form of pollution and how reactionary not-in-my-backyard concerns can shape the debate.

Pavement management using hyperspectral imagery

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 municipalitys 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.

Earth-observing hyperspectral imaging systems: a 2003 survey

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.