中国科学院沈阳自动化研究所机构知识库
Advanced  
SIA OpenIR  > 光电信息技术研究室  > 学位论文
题名: 高光谱图像空谱联合去相关压缩算法研究
其他题名: Joint Spatial-Spectral Decorrelation-based Compression for Hyperspectral Image
作者: 陈永红
导师: 史泽林
分类号: TP391.4
关键词: 高光谱图像压缩 ; 空谱联合去相关 ; 三维预测编码 ; 无损压缩 ; 有损压缩
索取号: TP391.4/C49/2010
学位专业: 模式识别与智能系统
学位类别: 博士
答辩日期: 2010-01-21
授予单位: 中国科学院沈阳自动化研究所
学位授予地点: 中国科学院沈阳自动化研究所
作者部门: 光电信息技术研究室
中文摘要: 高光谱遥感技术也叫成像光谱技术,是20世纪80年代初发展起来的一门新的遥感科学和技术。在精准农业、地质探测、城市调查、军事目标识别等领域得到了广泛应用。 随着成像光谱仪的光谱分辨率和空间分辨率的提高,所获取的高光谱图像数据量急剧增长。海量高光谱图像数据严重制约了星上图像处理、存储和传输技术的发展,因此,有必要开展高光谱图像压缩技术研究,进一步推进星载高光谱遥感的广泛应用与发展。 论文以高光谱遥感应用需求为背景,开展了高光谱图像的空谱联合去相关压缩算法研究,旨在减轻海量高光谱图像数据对星载存储设备容量和有限传输带宽带来的压力。研究内容包括:高光谱图像数据特征分析、Rice熵编码算法在高光谱图像压缩中的应用与优化、空谱联合去相关的高光谱图像三维预测编码、以及基于三维整数小波变换的高光谱图像3DSPIHT有损压缩算法。具体的创新性成果概括如下: 1.    提出了一种空谱联合预测的高光谱无损压缩Rice算法。通过对各种熵编码算法的特征分析与比较,重点研究了经典Rice算法在高光谱图像压缩中的应用。针对Rice算法低维预测器的弱去相关能力,提出了一种空谱联合预测器,有效降低了高光谱图像的空间和谱间冗余,提高了无损压缩比;优化了Rice算法中最优编码参数选择策略,减少了计算量。该方法应用于预测编码和变换编码可进一步提高压缩比。 2.     针对经典三维预测算法3D-CALIC的不足,从高光谱图像的空谱相关性出发,提出了一种改进的空-谱混合预测算法。通过计算局部相关系数进行空谱预测模式选择,并优化了谱间预测系数的计算和空间预测方法,在相同高光谱图像测试集中,取得了比3D-CALIC算法更高的无损压缩比。 3.     将一维最小均方算法推广到三维,提出了适合三维图像处理的自适应三维最小均方算法(3DLMS)。建立了基于3DLMS算法的高光谱图像三维预测模型,并通过去局部因果集均值方法实现了模型优化,可进一步提高无损压缩比。对不同场景的AVIRIS图像测试结果表明,基于3DLMS预测模型的高光谱图像无损压缩算法同时消除了高光谱图像的空间和谱间冗余,取得了更高的无损压缩比。所提出的3DLMS算法可应用到三维信号处理中。 4.    针对3DSPIHT算法直接应用于高光谱图像压缩时存在的复杂度高等问题,提出了一种优化的3DSPIHT高光谱图像有损压缩算法,将波段重组和基于空谱统计特性的三维小波分解结构应用于3DSPIHT算法中,进一步提高了算法的压缩性能。仿真结果表明,在同等码率下,优化后的3DSPIHT算法的解码图像具有更高的信噪比,且更好地保留了原始图像的光谱信息。
英文摘要: Hyperspectral remote sensing, also known as imaging spectroscopy, is a relatively new remote sensing science and technology developed in early 1980s. At present, hyperspectral remote sensing technology has been the major technique applied in many remote sensing projects, such as precise agriculture, mineral exploration, urban investigation, military target recognition, and so on. Imaging spectrometers are advanced enough to provide both high spatial and spectral resolution, therefore the volume of hyperspectral image data is growing rapidly, which poses a great challenge on the development of existing on-board image processing, storage and transmission technology. It is necessary to develop and design effective hyperspectral image compression techniques, further promoting applications and development of hyperspectral remote sensing. This dissertation mainly studies the joint special-spectral decorrelation-based hyperspectral image compression with the background of hyperspectral remote sensing applications. Aiming at alleviating the pressure brought by the massive hyperspectral image on the spaceborne capacity of storage devices and limited transmission bandwidth. Different topics of hypespectral image compression are investigated. They include: detailed analysis of hyperspectral image data characteristics, the optimization and application of Rice algorithm in hyperspectral image compression, spacial-spectral decorrelation-based three-dimensional prediction coding of hyperspectral image, as well as three-dimensional integer wavelet transform-based 3DSPIHT lossy compression of hyperspectral image. Specific innovative achievements can be summarized as follows: 1.       Spacial and spectral joint prediction-based lossless compression Rice algorithm for hyperspectral image is proposed. Based on the analysis and comparison of features of a variety of entropy coding algorithms, we forcus on analyzing the importance of the classical Rice algorithm in hyperspectral image compression applications. To deal wirh its weak decorrelation capability of the low-dimensional predictor, a spacial-spectral joint predictior is presented to remove the reduancy effectively and the selection strategy of optimal encoding parameters is optimized to reduce the computational complexity. This method can also be applied to prediction-based and transform-based coding to further improve the compression ratio. 2.       To cope with the shortage of typical 3D-CALIC prediction algorithm, an improved lossless compression algorithm based on the spatial-spectral hybrid prediction is proposed in the dissertation. We choose the prediction modes between the spatial and the spectral domains by computing the local correlation coefficient. The calculation of the spectral prediction coefficient and the special prediction mode are improved. Simulation results show that the proposed method outperforms 3D-CALIC algorithm with high lossless compression ratio. 3.       Extending one-dimensional least mean square algorithm into three dimensions, an adaptive three-dimensional least mean square (3DLMS) algorithm for three-dimensional image processing is proposed. To improve the lossless compression ratio of hyperspectral image, a novel adaptive prediction model based on 3DLMS algorithm for hyperspectral image lossless compression is presented and optimized by the local casual set mean subtraction method. Experimental results on AVIRIS images show that the proposed algorithm can remove the spatial and spectral redundancy of hyperspectral image concurrently and achieve higher image compression ratios than other state-of-the-art compression algorithms. The proposed 3DLMS algorithm can be applied into other three-dimensional signial processing. 4.       To the high complexity and weak decoding image quality of three-dimensional set partitioned in hierarchical tree (3DSPIHT) algorithm applied direct into hyperspectral image compressions, an improved 3DSPIHT algorithm for hyperspectral image lossy compression is proposed. A new band reordering method and spatial-spectral statistical characteristics-based three-dimensional wavelet decomposition structure are introduced and applied into 3DSPIHT algorithm to further improve the compression performance. Simulation results show that the optimized 3DSPIHT algorithm has higher SNR of decoding image, and good retention of the spectral information of original images at the same compression rate.
语种: 中文
产权排序: 1
内容类型: 学位论文
URI标识: http://ir.sia.cn/handle/173321/9226
Appears in Collections:光电信息技术研究室_学位论文

Files in This Item:
File Name/ File Size Content Type Version Access License
高光谱图像空谱联合去相关压缩算法研究.pdf(2657KB)----限制开放 联系获取全文

Recommended Citation:
陈永红.高光谱图像空谱联合去相关压缩算法研究.[博士学位论文].中国科学院沈阳自动化研究所.2010
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[陈永红]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[陈永红]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

Items in IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

 

Valid XHTML 1.0!
Copyright © 2007-2016  中国科学院沈阳自动化研究所 - Feedback
Powered by CSpace