CAJ | 학술논문

高光谱图像在以超多波段、光谱分辨率高、信息量丰富等优点给人们提供便利的同时，也带来了实时处理难、计算量大等难题。近年来，基于GPU的通用计算以其强劲计算能力、高性价比和低能耗等优势席卷了高性能计算领域，国内外学者纷纷针对GPU展开高光谱数据的并行处理研究。文章对高光谱遥感和GPU研究现状进行了简要介绍，总结了近几年国内外基于GPU的高光谱数据研究现状，指出存在问题和进一步研究的方向，力求为领域科学家进行高光谱图像并行处理研究提供一定参考。
고광보도상재이초다파단、광보분변솔고、신식량봉부등우점급인문제공편리적동시，야대래료실시처리난、계산량대등난제。근년래，기우GPU적통용계산이기강경계산능력、고성개비화저능모등우세석권료고성능계산영역，국내외학자분분침대GPU전개고광보수거적병행처리연구。문장대고광보요감화GPU연구현상진행료간요개소，총결료근궤년국내외기우GPU적고광보수거연구현상，지출존재문제화진일보연구적방향，력구위영역과학가진행고광보도상병행처리연구제공일정삼고。
Hyperspectral images have many advantages such as multi- band, high spectral resolution, abundant information and so on. They provide convenience for people, but also bring problems like dif iculty of real- time processing and huge computing and etc. Thanks to its strong computing ability, cost-ef ectiveness and low energy consumption, in recent years, general computation based on GPU has been sweeping across the High-performance computing field. Domestic and overseas scholars are get ing down to research Paral el processing for hyperspectral images on GPU platform. This paper gave a brief introduction of Hyperspectral remote sensing and GPU application in several fields, summarized the research status of hyperspectral based on GPU recently, pointed out the existing problems, and looked into the future research prospect, to provid a reference to scientists in the field of paral el processing hyperspectral image.