基于最大变差范数准则的ISAR自聚焦方法

Autofocus Algorithm for ISAR Based on Total Variation Norm

为消除目标平动引起的初相误差,必须进行自聚焦以避免ISAR图像模糊。在分析ISAR回波信号模型的基础上,本文构造了高阶多项式相位信号的初相补偿函数。已有文献多ISAR图像聚焦程度为准则,对该多项式相位信号的参数进行优化。本文利用最大全变差范数作为ISAR方位向成像的聚焦评价准则,该指标值在平动参数空间中的分布具有局部极值点少的优点,利于最优确定初相补偿函数的参数,并采用协同粒子群优化算法加速参数的寻优速度和精度。仿真实验证明了本文方法的可行性和正确性。

To eliminate the phase errors caused by the target translational motion, the autofocus is necessary to avoid inverse synthetic radar （ISAR） imaging blurring. The polynomial model for the translational phase error is derived from the ISAR signal model and is used for constructing phase compensation function. Many techniques described in references are based on the optimization of a contrast function for representing a measure of the focus degree of the image to optimize the parameters of compensation function. A novel method based on the total variation norm（TVN） is presented. Its distribution has less local extremum for finding the optimal target motion parameter. To improve the efficiency and the accuracy of the parameter finding, a cooperative particle swarm optimization（CPSO） is introduced. The simulation result shows the validity of the method.