基于线性预测外推的高分辨率雷达成像

为了提高实验室合成孔径雷达(SAR)成像系统的分辨率,采用自回归(AR)模型,对采集到的数据进行距离向和方位向上的线性预测外推,对外推后的数据进行快速傅里叶变换(FFT)。实验结果表明,相比于传统的快速傅里叶变换成像算法,在同样的硬件条件下,使用该方法可以提高该系统的纵向及横向的分辨率,得到更清晰的雷达图像,或使用较小孔径长度就可以获得相同质量的雷达图像;并且该算法计算速度较快,易于操作,便于在系统上实现实时成像。

High resolution radar imaging based on linear prediction extrapolation

To impr ove the resolut ion of the labor ator y sy nthetic opertune radar ( SAR) system, the auto-regressive( AR) model w as used to ex t rapolate the sampling data bo th in signal bandw idth and aper ture, andthen the ex t rapo lated data w ere processed by Fast Four ier T ransfo rm ( FFT ) . Ex perimental r esult s showthat compar ed w ith the convent ional FFT metho d, the method of auto-regressive ( A R) model combinedw ith FFT can enhance the resolut ion of the radar images both in the range prof ile and the cross rang e forthe same ef fective bandw idth o f t ransmit ted sig nals and of the same aper ture size o r the equal quality imagesf rom smaller aper ture, and that this metho d has fast computat ion speed and is easily operat ional,w hich is convenient for r ealizing real-imaging o n the syst