摘要
方位向降采样滤波器作为机载合成孔径雷达(SyntheticApertureRADAR,SAR)实时成像系统中关键组成部分之一,其性能将直接关系到图像质量的优劣。由于受到运算代价、存储开销的限制,方位向降采样滤波器的阶数不能很高。如何在滤波器阶数受限的情况下,设计出满足实时处理要求的性能优异的滤波器是一个值得研究的问题。针对机载SAR实时成像系统应用,提出了一种新的方位向降采样滤波器设计方法,引入增广拉格朗日函数实现了通带均方误差最小、阻带最大误差最小准则的方位向降采样滤波器。计算机仿真验证说明了该设计方法的优越性,在同样阶数、同样存储开销的情况下,所设计的滤波器性能要优于纯粹的最大误差最小准则设计的滤波器,满足了机载SAR实时成像系统的要求。
As one of key processing of the real-time airborne synthetic aperture RADAR （SAR） imaging system, the performance of Azimuth Pre-filter for data reduction determines the quality of SAR images. A new Azimuth Pre-filter design method was proposed. The optimize criterion of Minimum of the Mean-Square Error in pass-band and Minimum of the Maximum Error in stop：band was proposed, and a novel computing method using extend-Lagrange function for Azimuth Pre-filter in real-time airborne SAR Imaging System was also proposed. The analysis results of computer simulation show that, with this new method, at same computational cost, memory size, and power cost, the better performance of filter can be achieved. Thus, this method is more suitable for real-time airborne applications.
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2007年第3期593-596,600,共5页
Journal of System Simulation
基金
国家自然科学基金(60303017)
863计划项目(2003AA135093)。
