DOI: 10.1016/j.jag.2013.08.009
Scopus记录号: 2-s2.0-84897861467
论文题名: Implementation and performance of a general purpose graphics processing unit in hyperspectral image analysis
作者: van der Werff H ; M ; A ; , Bakker W ; H
刊名: International Journal of Applied Earth Observation and Geoinformation
ISSN: 15698432
出版年: 2014
卷: 26, 期: 1 起始页码: 312
结束页码: 321
语种: 英语
英文关键词: Classification
; GPGPU
; Graphicshardware
; Hyperspectral
; IDL
Scopus关键词: algorithm
; data set
; hardware
; image analysis
; image classification
; image processing
; parallel computing
; pixel
英文摘要: A graphics processing unit (GPU) can perform massively parallel computations at relatively low cost. Soft-ware interfaces like NVIDIA CUDA allow for General Purpose computing on a GPU (GPGPU). Wrappersof the CUDA libraries for higher-level programming languages such as MATLAB and IDL allow its use inimage processing. In this paper, we implement GPGPU in IDL with two distance measures frequently usedin image classification, Euclidean distance and spectral angle, and apply these to hyperspectral imagery.First we vary the data volume of a synthetic dataset by changing the number of image pixels, spectralbands and classification endmembers to determine speed-up and to find the smallest data volume thatwould still benefit from using graphics hardware. Then we process real datasets that are too large to fitin the GPU memory, and study the effect of resulting extra data transfers on computing performance. Weshow that our GPU algorithms outperform the same algorithms for a central processor unit (CPU), that asignificant speed-up can already be obtained on relatively small datasets, and that data transfers in largedatasets do not significantly influence performance. Given that no specific knowledge on parallel com-puting is required for this implementation, remote sensing scientists should now be able to implementand use GPGPU for their data analysis. © 2013 Elsevier B.V. Citation statistics:
资源类型: 期刊论文
标识符: http://119.78.100.158/handle/2HF3EXSE/79722
Appears in Collections: 气候变化事实与影响
There are no files associated with this item.
作者单位: Faculty for Geo-Information Science and Earth Observation, University of Twente, Enschede 7500 AE, Netherlands
Recommended Citation:
van der Werff H,M,A,et al. Implementation and performance of a general purpose graphics processing unit in hyperspectral image analysis[J]. International Journal of Applied Earth Observation and Geoinformation,2014-01-01,26(1)