摘要
该文研究基于矢量量化技术的合成孔径雷达(SAR)原始数据压缩算法,提出了块自适应树型矢量量化(BATSVQ)算法和块自适应球形矢量量化(BASVQ)算法。与块自适应矢量量化(BAVQ)算法相比较,该文提出的算法采用约束型矢量量化技术,能够充分利用SAR原始数据经过自适应块处理后在较小的范围内具有稳定高斯分布的特性。采用以上算法对SAR实测数据进行了验证,并比较图像及其性能参数,结果表明BATSVQ算法和BASVQ算法能够获得算法性能和实现复杂度之间的合理折衷。
This paper deals with the compression algorithms of synthetic aperture radar （SAR） raw data based on the vector quantization （VQ） techniques. The block adaptive tree-structured vector quantization （BATSVQ） algorithm and the block adaptive spherical vector quantization （BASVQ） algorithm are presented. Compared with the block adaptive vector quantization （BAVQ） algorithm, both of the proposed methods using constrained vector quantizer take the full advantage of SAR raw data properties of a Gaussian stationary process after a blockwise normalization. Live SAR data implementations and quantitative analysis of resultant images show that a better trade-off between performance and complexity is achieved by using the BATSVQ and BASVQ algorithms.
出处
《南京理工大学学报》
EI
CAS
CSCD
北大核心
2006年第2期157-161,共5页
Journal of Nanjing University of Science and Technology
