Book Review - High Performance Computing in Remote Sensing

High Performance Computing in Remote Sensing introduces the most recent advances in the incorporation of the high-performance computing (HPC) paradigm in remote sensing missions. Eighteen well-selected and organized chapters cover the entire spectrum of current and future data processing techniques in remote sensing applications, with a focus on the use of HPC techniques. This book begins with an introduction chapter that describes the major innovative contributions covered by this book and its individual chapters. The book is divided into four parts. The fi rst part describes basic concepts of HPC in remote sensing and provides a detailed review of existing and planned HPC systems in this area. Each of the remaining three parts illustrate a specifi c parallel computing paradigm, including multiprocessor (clusterbased) systems, large-scale and heterogeneous networks of computers, and specialized hardware architecture for analysis and interpretation of remotely-sensed data. Part One begins with an extensive review of the state-ofart design of HPC systems for remote sensing missions. It also includes a case study in which the pixel purity index (PPI), a well-known processing algorithm, is implemented on three different HPC platforms: a massively parallel multiprocessor, a heterogeneous network of distributed computers, and a specialized fi eld programmable gate array (FPGA) hardware architecture. Analytical and experimental results are presented in the context of a real application, using hyperspectral data collected by the NASA Jet Propulsion Laboratory’s Airborne Visible Infra-Red Imaging Spectrometer (AVIRIS) over the World Trade Center area of New York City right after the 9/11 terrorist attacks. Another paper in this part covers multimedia and video data processing as another example of an HPC application that demands high computational power. Part Two describes a compendium of algorithms and techniques for HPC-based remote sensing data analysis using multiprocessor systems such as clusters and networks of computers, including massively parallel facilities. A paper authored by Gillis et al. presents a parallel version of the Optical Real-Time Adaptive Spectral Identifi cation System (ORASIS) developed at the Naval Research Laboratory for the analysis of remotely-sensed hyperspectral image data. A paper authored by Tilton describes a parallel implementation of a recursive approximation of the hierarchical image segmentation algorithm (HSEG) developed at NASA. The paper demonstrates the computational effi ciency of the algorithm by testing the implementation on multispectral remotely-sensed data collected by the Landsat Thematic Mapper instrument. A paper authored by Asner et al. summarizes the major processing steps and challenges involved in collection and analysis of hyperspectral image data, and presents examples of how highperformance computing is used to meet these challenges. It also discusses the emerging use of other HPC techniques, such as data processing onboard aircraft and spacecraft platforms, and distributed Internet computing. A paper authored by Plaza et al. presents the use of parallel neural network architectures for solving remote sensing problems. Finally a paper by Valencia et al. presents the use of HPC-based remote sensing techniques to address wildland fi res. Part Three is devoted to large-scale and heterogeneous distributed computing by focusing on parallel techniques for remote sensing data analysis using large-scale distributed platforms, with a special emphasis on grid computing environments and fully heterogeneous networks of workstations. A paper by Lee begins with an introduction on the fundamental concepts of grid computing, Web services, and service-oriented architectures, and is followed by a survey of current grid infrastructure and science projects relevant to remote sensing. A paper authored by Gasster et al. presents the background, architecture, implementation and examples of remote sensing grids. A paper by Fusco et al. describes some European Space Agency (ESA) activities related to the use of grid technology for Earth Observation: ESA Grid Processing on-Demand environment, and several Earth Observation applications that have been plugged into the environment. A paper by Cafaro et al. presents an overview of grid computing environments and discusses their usefulness in the context of remote sensing. This part ends with a paper by Velez-Reyes et al. that describes the concept of a solutionware system for the solution of hyperspectral/multispectral remote sensing image processing problems. The paper presents continued on page 12