基于卷积核网格化二维近程微波全息

    Two-Dimensional Short-Range Microwave Holography Based on Convolutional Gridding

    • 摘要: 引入一种二维近程微波全息成像算法,该方法采集的散射数据不仅包括后向散还包括前向散射,对目标的入射场不做任何假设,完全由测量或仿真获得,具有较强抗噪声性能,成像分辨率较高,较适于近程成像。为了探讨算法在实际应用中的影响因素,利用FEKO 软件模拟目标散射场和入射场,验证了该算法能够获得高质量目标图像,比较和分析了不同扫描频率、扫描间隔以及天线位置的成像效果,为实际应用提供参考依据。针对该算法中外部干扰源等引起的混叠效应,提出利用卷积核网格化方法处理采集的散射数据来纠正网格。给出几种典型的卷积核函数,比较分析了它们的抗混叠性能,得出球谐函数是一种较优的卷积核函数,抗混叠较强。

       

      Abstract: A two-dimensional short-range microwave holography imaging algorithm has been introduced. In this method, the scattered data collected includes not only back-scattered but also forward-scattered data, and without any assumptions on incident fields of the target, which obtained entirely by either measurement or simulation. This method has higher robust to noise performance and imaging resolution, more suitable for short-range imaging. To investigate the factors in the practical application of the algorithm, we using FEKO software to simulate the scattered field and the incident field of the target, validate the algorithm can obtain high quality target image. For the aliasing in the algorithm caused by external interference sources, and so on. We proposed using convolution gridding method to process the collected scattered data to correct the grid. Several typical convolution functions are given, validate the spheroidal function which has stronger rejection aliased is an optimal convolution function by compared and analyzed its performance of rejection aliased.

       

    /

    返回文章
    返回