复数SAR图像的块自适应多级矢量量化压缩

发布时间：2018-08-03 12:14

【摘要】：研究了在空间域进行复数合成孔径雷达(SAR)图像的压缩,提出了结合块自适应矢量量化(BAVQ)和多级矢量量化(MSVQ)技术各自优点的块自适应多级矢量量化(BAMSVQ)压缩算法。该算法因SAR复数图像的实部和虚部数据在一定采样规模下具有近似高斯分布的特性,可以运用BAVQ编码;而上一级编码后的残差数据由于已经进行了一定程度的去相关性,根据中心极限定律也具有近似高斯分布特性,因而也可进行块自适应的矢量量化编码,由此可形成BAVQ的多级方案。另外由于实虚部归一化的高斯分布块数据具有相同的分布特性,因此可以共享码书,节省存储资源与传输数据量。仿真结果验证了该算法的有效性。在相同的压缩比下,与前两种空间域压缩算法相比,该算法提升了性能,降低了计算复杂度,节省了计算时间和资源占用量。
[Abstract]:The compression of complex synthetic aperture radar (SAR) images in the space domain is studied. A block adaptive multilevel vector quantization (BAMSVQ) compression algorithm is proposed, which combines block adaptive vector quantization (BAVQ) and multistage vector quantization (MSVQ) technology. The algorithm is based on the real and virtual data of the SAR complex image under a certain sampling scale. BAVQ coding can be used to approximate the characteristics of the Gauss distribution, and the residual data after the first level coding has been correlated to a certain degree, and the central limit law also has approximate Gauss distribution characteristics, so the block adaptive vector quantization coding can also be carried out, thus the multilevel scheme of the BAVQ can be formed. In addition, the real deficiency can be achieved. The normalized Gauss distribution block data have the same distribution characteristics, so the codebook can be shared and the storage and transmission data can be saved. The simulation results verify the effectiveness of the algorithm. Compared with the previous two spatial domain compression algorithms under the same compression ratio, the algorithm raises the performance, reduces the computational complexity and saves the calculation. Time and resources are used.
【作者单位】：中国科学院计算技术研究所;中国科学院大学;
【基金】：总装备部预研资助项目
【分类号】：TN957.52

【参考文献】