基于改进的K-L变换的多光谱图像压缩算法  

A compression algorithm of multispectral images based on improved Karhunen-Loeve transform

作　　者：徐冬冬[1,2] 付天骄[1] 张宇[1] 张星祥[1] 任建岳[1] 

机构地区：[1]中国科学院长春光学精密机械与物理研究所,吉林长春130033 [2]中国科学院大学,北京100039

出　　处：《光电子．激光》2015年第6期1200-1205,共6页Journal of Optoelectronics·Laser

基　　金：国家"863"高技术研究发展计划(863-2-5-1-13B)资助项目

摘　　要：融合离散小波变换（DWT,discrete wavelet transform）与Karhunen-Loeve变换（KLT）,将图像的能量集中到少数系数上,以达到更好的压缩效果。首先将多光谱图像的每个谱段进行快速9/7 2DDWT,消除多光谱图像的大部分空间冗余;然后对所有谱段产生的小波系数进行改进的KLT,来消除光谱冗余和残存的空间冗余;最后对所得谱段产生的小波系数进行改进的KLT,来消除光谱冗余和残存的空间冗余;最后对所得系数进行熵编码,得到压缩码流。实验结果表明,在码率为0.25-2.0bit/pixel范围内,平均信噪比（SNR）高于41dB,同时缩短了运算时间,从而提升了多光谱图像压缩算法的性能。An effective compression algorithm of multispectral images based on discrete wavelet transform （DWT）and improved Karhunen-Loeve transform （KLT） is proposed in this paper. It could achieve bet ter performance with effective combination of the two methods which convert the energy of images to a small number of coefficients. In order to reduce the most redundancy of space, it deals with each spectrum using fast 9/7 2 D DWT firstly. Then the spectral 1 D KLT is applied on the 2-D DWT coefficients to reduce spectral redundancy and spatial redundancy remnantly. Finally, entropy coding is executed to obtain the compressed code stream. Within 0.25-2.0 bit/pixel range,the average peak signal to noise ratio （SNR） of the proposed compression algorithm could be improved greatly compared with other approaches, which reduces the operating time and improves the performance of multi spectral image compression algorithm,verifies the correctness and feasibility of our method. The work in this paper has a good reference value for similar cases.

关 键 词：多光谱 图像压缩 Karhunen-Loeve变换(KLT) 信噪比(SNR) 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]