Parallel Implementation of Endmember Extraction Algorithms From Hyperspectral Data

Valencia, David; Plaza, Javier; Chang, Chein‐I

doi:10.1109/lgrs.2006.871749

Cited by 55 publications

(28 citation statements)

References 10 publications

(8 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A parallel version of the PPI algorithm has been recently proposed [26] and is used here for illustrative purposes. The inputs to the parallel algorithm are an N -dimensional image cube, f, a number of skewers to be generated, s, a number of endmembers to be extracted, p, and a threshold value, t PPI .…”

Section: Parallel Pixel Purity Index (P-ppi)mentioning

confidence: 99%

“…As shown in Figure 7(b), the endmembers obtained at the end of this process will likely define a simplex that encloses most of the pixels in the input hyperspectral data set. A master-slave parallel version of this endmember search process has been recently developed [26]. The inputs to the parallel algorithm (called P-FINDR) are N -dimensional image cube, f, and the number of endmembers to be extracted, p. The output of the algorithm is a set of final endmembers, {e j } p j=1 .…”

Section: Parallel N-findr (P-findr)mentioning

confidence: 99%

“…In all cases, we carefully adjusted algorithm parameters based on our experience with those algorithms in different application domains [6,22,26] Table I reports the individual and overall classification accuracies (with regards to the USGS reference classifications) produced by the parallel algorithms after using different algorithm configurations. As shown in Table I, both the P-PPI and P-FINDR (combined with P-LSU) produced higher classification accuracies than those found by P-ISODATA.…”

Section: Experiments 1: Urban Area Characterizationmentioning

confidence: 99%

See 2 more Smart Citations

Parallel processing of remotely sensed hyperspectral imagery: full‐pixel versus mixed‐pixel classification

Plaza

2008

Concurrency and Computation

View full text Add to dashboard Cite

SUMMARYThe rapid development of space and computer technologies allows for the possibility to store huge amounts of remotely sensed image data, collected using airborne and satellite instruments. In particular, NASA is continuously gathering high-dimensional image data with Earth observing hyperspectral sensors such as the Jet Propulsion Laboratory's airborne visible-infrared imaging spectrometer (AVIRIS), which measures reflected radiation in hundreds of narrow spectral bands at different wavelength channels for the same area on the surface of the Earth. The development of fast techniques for transforming massive amounts of hyperspectral data into scientific understanding is critical for space-based Earth science and planetary exploration. Despite the growing interest in hyperspectral imaging research, only a few efforts have been devoted to the design of parallel implementations in the literature, and detailed comparisons of standardized parallel hyperspectral algorithms are currently unavailable. This paper compares several existing and new parallel processing techniques for pure and mixed-pixel classification in hyperspectral imagery. The distinction of pure versus mixed-pixel analysis is linked to the considered application domain, and results from the very rich spectral information available from hyperspectral instruments. In some cases, such information allows image analysts to overcome the constraints imposed by limited spatial resolution. In most cases, however, the spectral bands collected by hyperspectral instruments have high statistical correlation, and efficient parallel techniques are required to reduce the dimensionality of the data while retaining the spectral information that allows for the separation of the classes. In order to address this issue, this paper also develops a new parallel feature extraction algorithm that integrates the spatial and spectral information. The proposed technique is evaluated (from the viewpoint of both classification accuracy and parallel performance) and compared with other parallel techniques for dimensionality reduction and classification in the context of three representative application case * Correspondence to: Antonio J. Plaza, Department of Technology of Computers and Communications, University of Extremadura, Avda. de la Universidad s/n, E-10071 Càceres, Spain. † E-mail: aplaza@unex.es studies: urban characterization, land-cover classification in agriculture, and mapping of geological features, using AVIRIS data sets with detailed ground-truth. Parallel performance is assessed using Thunderhead, a massively parallel Beowulf cluster at NASA's Goddard Space Flight Center. The detailed cross-validation of parallel algorithms conducted in this work may specifically help image analysts in selection of parallel algorithms for specific applications.

show abstract

Section: Parallel Pixel Purity Index (P-ppi)mentioning

confidence: 99%

Section: Parallel N-findr (P-findr)mentioning

confidence: 99%

Section: Experiments 1: Urban Area Characterizationmentioning

confidence: 99%

See 1 more Smart Citation

Parallel processing of remotely sensed hyperspectral imagery: full‐pixel versus mixed‐pixel classification

Plaza

2008

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…Two algorithms for endmember extraction have been implemented by parallel approximations [6]: the pixel purity index (PPI) [7] and the N-FINDR [8] algorithm. The inputs to all discussed algorithms are a hyperspectral image F with n dimensions and a number of endmembers to be extracted, e. The output in all cases is a set of endmember pixels, denoted by…”

Section: Parallel Endmember Extraction Algorithmsmentioning

confidence: 99%

Parallel Implementation of Hyperspectral Image Processing Algorithms

Valencia

Plaza

Sánchez-Testal

et al. 2006

2006 IEEE International Symposium on Geoscience and Remote Sensing

View full text Add to dashboard Cite

Abstract-High computing performance of algorithm analysis is essential in many hyperspectral imaging applications, including automatic target recognition for homeland defense and security, risk/hazard prevention and monitoring, wild-land fire tracking and biological threat detection. Despite the growing interest in hyperspectral imaging research, only a few efforts devoted to designing and implementing well-conformed parallel processing solutions currently exist in the open literature. With the recent explosion in the amount and dimensionality of hyperspectral imagery, parallel processing is expected to become a requirement in most remote sensing missions. In this paper, we take a necessary first step towards the quantitative comparison of parallel techniques and strategies for analyzing hyperspectral data sets. Our focus is on three types of algorithms: automatic target recognition, spectral mixture analysis and data compression. Three types of high performance computing platforms are used for demonstration purposes, including commodity cluster-based systems, heterogeneous networks of distributed workstations and hardware-based computer architectures. Combined, these parts deliver a snapshot of the state of the art in those areas, and offer a thoughtful perspective on the potential and emerging challenges of incorporating parallel computing models into hyperspectral remote sensing problems. I. INTRODUCTIONHyperspectral imaging has been transformed in less than 30 years from being a sparse research tool into a commodity product available to a broad user community [1]. The development of computationally efficient techniques able to transform massive volumes of hyperspectral data sets, collected on a daily basis, into scientific understanding is critical for space-based Earth science and planetary exploration. To address the computational needs introduced by hyperspectral imaging applications, a few efforts have been directed towards the incorporation of high-performance computing models and platforms in remote sensing missions. For instance, the Center of Excellence in Space and Data Information Sciences (CESDIS), located at NASA's Goddard Space Flight Center in Maryland, developed the concept of Beowulf cluster with the aim of creating a cost-effective parallel computing system from commodity components to satisfy specific computational requirements for the Earth and space sciences community [2]. Although most dedicated parallel machines for hyperspectral image information extraction and mining have been chiefly homogeneous in nature [3], a current trend in the design of systems for analysis and interpretation of high-dimensional data sets is to utilize highly heterogeneous distributed platforms [4],

show abstract

“…Due to its vast data volume, it is desirable to apply parallel processing to hyperspectral image processing and analysis when parallel computing facilities are available [1][2][3][4].…”

Section: Introductionmentioning

confidence: 99%

Parallel Data Compression for Hyperspectral Imagery

Yang

Zhu

et al. 2008

IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium

View full text Add to dashboard Cite

The high dimensionality of hyperspectral imagery challenges image processing and analysis. It has been shown that hyperspectral compression can be achieved by principal component analysis (PCA) for spectral decorrelation followed by the JPEG2000-based coding. This approach, referred to as PCA+JPEG2000, provides superior ratedistortion performance and can preserve useful data information. However, its main disadvantage is high computational complexity in the PCA process which entails the calculation of the data covariance matrix and its eigenvectors. Parallel processing is an appropriate approach to relieve the computation burden of such a PCA-based compression. In this paper, several parallel PCA implementations are proposed and their processing speed and resulting compression performance are investigated.

show abstract

Parallel Implementation of Endmember Extraction Algorithms From Hyperspectral Data

Cited by 55 publications

References 10 publications

Parallel processing of remotely sensed hyperspectral imagery: full‐pixel versus mixed‐pixel classification

Parallel processing of remotely sensed hyperspectral imagery: full‐pixel versus mixed‐pixel classification

Parallel Implementation of Hyperspectral Image Processing Algorithms

Parallel Data Compression for Hyperspectral Imagery

Contact Info

Product

Resources

About