FIVQ algorithm for interference hyper-spectral image compression

CORC > 软件研究所 > 软件所图书馆 > 期刊论文

	FIVQ algorithm for interference hyper-spectral image compression
	Wen, Jia (1) ; Ma, Caiwen (2) ; Zhao, Junsuo (1)
刊名	Optics Communications
	2014
卷号	322 页码:97-104
关键词	Image compression Vector quantization LBG algorithm Arithmetic coding
ISSN号	304018
通讯作者	Wen, J.(448680289@qq.com)
中文摘要	Based on the improved vector quantization (IVQ) algorithm [1] which was proposed in 2012, this paper proposes a further improved vector quantization (FIVQ) algorithm for LASIS (Large Aperture Static Imaging Spectrometer) interference hyper-spectral image compression. To get better image quality, IVQ algorithm takes both the mean values and the VQ indices as the encoding rules. Although IVQ algorithm can improve both the bit rate and the image quality, it still can be further improved in order to get much lower bit rate for the LASIS interference pattern with the special optical characteristics based on the pushing and sweeping in LASIS imaging principle. In the proposed algorithm FIVQ, the neighborhood of the encoding blocks of the interference pattern image, which are using the mean value rules, will be checked whether they have the same mean value as the current processing block. Experiments show the proposed algorithm FIVQ can get lower bit rate compared to that of the IVQ algorithm for the LASIS interference hyper-spectral sequences. © 2014 Elsevier B.V.
英文摘要	Based on the improved vector quantization (IVQ) algorithm [1] which was proposed in 2012, this paper proposes a further improved vector quantization (FIVQ) algorithm for LASIS (Large Aperture Static Imaging Spectrometer) interference hyper-spectral image compression. To get better image quality, IVQ algorithm takes both the mean values and the VQ indices as the encoding rules. Although IVQ algorithm can improve both the bit rate and the image quality, it still can be further improved in order to get much lower bit rate for the LASIS interference pattern with the special optical characteristics based on the pushing and sweeping in LASIS imaging principle. In the proposed algorithm FIVQ, the neighborhood of the encoding blocks of the interference pattern image, which are using the mean value rules, will be checked whether they have the same mean value as the current processing block. Experiments show the proposed algorithm FIVQ can get lower bit rate compared to that of the IVQ algorithm for the LASIS interference hyper-spectral sequences. © 2014 Elsevier B.V.
收录类别	SCI ; EI
语种	英语
WOS记录号	WOS:000335886700017
公开日期	2014-12-16
内容类型	期刊论文
源URL	[http://ir.iscas.ac.cn/handle/311060/16693]
专题	软件研究所_软件所图书馆_期刊论文
推荐引用方式 GB/T 7714	Wen, Jia ,Ma, Caiwen ,Zhao, Junsuo . FIVQ algorithm for interference hyper-spectral image compression[J]. Optics Communications,2014,322:97-104.
APA	Wen, Jia ,Ma, Caiwen ,&Zhao, Junsuo .(2014).FIVQ algorithm for interference hyper-spectral image compression.Optics Communications,322,97-104.
MLA	Wen, Jia ,et al."FIVQ algorithm for interference hyper-spectral image compression".Optics Communications 322(2014):97-104.