基于稀疏约束与性能最优化的SAR图像目标增强方法

Target Enhancement of SAR Image Based on Sparse Constraint and Performance Optimization

基于合成孔径雷达(Synthetic aperture radar,SAR)图像的稀疏特性,将目标增强模型设计与性能要求相融合,建立了信噪比性能最优化目标函数与lk范数约束项构成的SAR图像变参数增强模型。通过对模型解的惟一性、收敛性及其增强性能分析,确定稀疏参数k的选择范围,继而通过参数估计的均方误差分析,得到正则参数的选择方法,并设计求解迭代过程。图像处理过程可达到较高的自动程度。MSTAR图像数据处理结果验证了该模型的有效性,经处理后图像噪声得到抑制,目标特征得到明显增强,性能优于传统正则化模型。

The traditional regularization model is modified and a new type of model for SAR image enhancement is proposed. For the need of performance optimization, the model, in which regularization parameters vary with pixels, is established by using lk norm sparse constraint and the performance requirement as a target function. The range of k is obtained by the analyses of the uniqueness, convergence of the solution and enhancement performance. The selection method for regularization parameters is obtained based on the minimum MSE of parameter estimation. The processing speed is fast for basic algebraic operations in the iterative solving process. Experimental results show that the new model is superior to the traditional model. The noise on images is reduced and targets are enhanced after being processed. Furthermore, the new model can be automatically implemented.