Robust SVA method for every sampling rate condition
Robust SVA method for every sampling rate condition
复制标题
DOI：
10.1109/taes.2007.4285354
复制
发表时间：
2007-04-01
影响因子：
4.4
通讯作者：
Burgos-Garcia, Mateo
中科院分区：
文献类型：
--
作者：
Castillo-Rubio, Carlos;Llorente-Romano, Sergio;Burgos-Garcia, Mateo
线性变迹或数据加权是以牺牲分辨率为代价来改善有限采样信号中的旁瓣电平的传统过程。新的切趾方法，如空间变切趾（SVA），应用非线性滤波的信号，以完全消除旁瓣，而没有任何损失的分辨率。然而，结果受到信号采样率的强烈影响。以前已经发表了一些改进结果的变化，但是由于采样频率没有固定在Nyquist（或倍数），旁瓣消除变得更糟。本文提出了一种新的有效的技术，基于SVA，大大降低了旁瓣电平的每一个采样率条件。该算法本质上是对图像每个像素的变量滤波器的参数优化。一个一维的情况下，一个二维的推广，以及一些应用程序中的合成孔径雷达（SAR）系统的目标检测能力。
Linear apodization, or data weighting, is the traditional procedure to improve sidelobe levels in a finite sampled signal at the expense of resolution. New apodization methods, such as spatially variant apodization (SVA), apply nonlinear filtering to the signal in order to completely remove sidelobes without any loss of resolution. However, the results are strongly influenced by signal sampling rate. Some variations which improve results have been previously published, but sidelobe cancellation gets worse since sampling frequency is not settled at Nyquist (or a multiple).This paper presents a new and efficient technique based on SVA that drastically reduces sidelobe levels for every sampling rate condition. The algorithm is, essentially, a parameter optimization of a variant filter for each pixel of the image. A one-dimensional case and a two-dimensional generalization are presented, as well as some applications to target detection capability in a synthetic aperture radar (SAR) system.