基于分类和陪集码的高光谱图像无损压缩  

Lossless Compression of Hyperspectral Images Based on Classification and Coset Coding

作　　者：宋娟[1] 吴成柯[1] 张静[1] 刘海英[1] 

机构地区：[1]西安电子科技大学综合业务网国家重点实验室,西安710071

出　　处：《电子与信息学报》2011年第1期231-234,共4页Journal of Electronics & Information Technology

基　　金：国家自然科学基金项目(60802076;60702058);111基地项目(B08038);中央高校基本科研业务费专项资金(JY10000901007)资助课题

摘　　要：在基于陪集码的高光谱图像压缩算法中,由于按照编码块的最大残差确定整块无损压缩所需的码率存在较大冗余,该文提出了基于分类和陪集码的高光谱图像压缩算法。首先利用前一波段对应位置的预测噪声对当前波段编码块的像素进行分类,将具有相似相关性的像素归于一类,然后对每一类像素分别进行陪集码编码。实验表明分类可以有效地降低码率。和基于陪集码的算法相比,该文算法无损压缩的平均码率降低了大约0.4 bpp。Since the bitrate of the whole block is determined by its maximum prediction error and much redundancy exists in the scalar coset coding based compression method for hyperspectral images,a lossless compression method based on classification and coset coding is proposed in this paper to further reduce the bitrate.The current block is classified using the corresponding prediction errors in the previous band to make the pixels with similar inter-band correlations cluster together.Then each class of pixels is then coset coded respectively.The experimental results show that the classification can reduce the bitrate efficiently.Compared to coset coding based method without classification,the lossless compression bitrate of the proposed method is reduced by about 0.4 bpp.

关 键 词：高光谱图像 无损压缩 分布式信源编码 陪集码 分类 

分 类 号：TN919.81[电子电信—通信与信息系统]