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

Abstract

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.

Parallel heterogeneous CBIR system for efficient hyperspectral image retrieval using spectral mixture analysis

“…In those cases, important aspects such as the size of the spatial-domain partitions or the need to overlap adjacent partitions arise. These aspects have been addressed in the previous work [30,31].…”

Hyperspectral Remote Sensing of Urban Areas

“…Increased experimentation, algorithm improvement, and validation over more urban areas of greater size must also be accompanied by increased research in advanced computing to support the data volume and complexity of hyperspectral algorithm advances, most of which are computationally intensive. For researchers interested in this significant challenge, the work of Plaza (2008, 2007), and Plaza et al. (2006) form an excellent starting point.…”

Improving the scalability of parallel algorithms for hyperspectral image analysis using adaptive message compression

“…In the context of hyperspectral imaging applications, it has been demonstrated that the scalability of parallel algorithms is directly related to the size of the messages to be exchanged through the communication network of the system when the parallel algorithm is run [4], i.e. the latency for large message sizes (typical in hyperspectral imaging applications, since each pixel vector is made up of hundreds of spectral values) is extremely sensitive to the size of the message.…”