基于自适应量化的超光谱图像压缩  

Compression of hyper-spectral images based on adaptive quantization

作　　者：吴颖谦[1] 方涛[1] 施鹏飞[1] 杜培军[1] 

机构地区：[1]上海交通大学模式识别与图像处理研究所,上海200030

出　　处：《光学技术》2004年第2期218-222,共5页Optical Technique

基　　金：国家863高技术研究资助项目(2002AA134020-05);上海市科学技术委员会资助项目(015115036)

摘　　要：提出了一种基于谱间预测和小波量化编码的超光谱图像压缩方法。在充分考虑子带图像非平稳特性的基础上,首先实现了子带分类步骤,通过分类设计自适应预测器,提高了谱间去相关效率。由于子带中不同子类表现出了不同的统计特性,所以使用空变均匀阈值量化器完成了量化工作。研究了均匀量化器对不同分布训练样本的率 失真表现,并为相关特性建立了模型描述。基于率失真模型和系数序列的统计特性,提出了一个率分配算法,以便能为不同的子类系数序列设计率失真意义上的最优量化器。实验证明,这种方法能高效地压缩超光谱图像,表现出了优异的压缩性能。An approach for compression of hyper-spectral image based on spectral DPCM and wavelet is proposed. Fully considering non-stationarity in sub-bands, the step of classification in sub-bands is implemented based on which adaptive predictors are designed to improve the efficiency of spectral decorrelation. Since different statistics appear in different classes, the algorithm adopts spatial varying uniform threshold quantizer(UTQ). Based on the research on rate-distortion characteristics of UTQ over the training samples with different probability density, a method for rate allocation is proposed by which a optimal in R-D sense UTQ can be designed for each class of coefficients. The experiments show that the approach can efficiently compress hyper-spectral remote sensing images, and the excellent performance of the proposed algorithm is demonstrated.

关 键 词：超光谱图像 子带编码 自适应预测 率分配 图像压缩 小波量化编码 

分 类 号：TN919.8[电子电信—通信与信息系统]