基于Poisson重建的极化SAR图像对比增强

Contrast enhancement of polarimetric SAR images based on the Poisson reconstructions

为了能够同时增强多类目标,提出一种基于Poisson重建的极化合成孔径雷达(SAR)图像对比增强方法。该方法对多类目标和对应的背景杂波分别进行广义相对最优极化增强(GOPCE),并得到相应的最优极化状态和特征系数;以此定义图像中各像素点的最优局部梯度,并在最小二乘准则下,根据局部梯度建立离散Poisson方程;通过快速Fourier变换求解该Poisson方程,得到最终的多目标增强图像。实验结果表明:利用极化SAR数据,使用该方法增强后的图像的直方图保持应有的峰值,且更加均衡,能够达到增强多类目标的效果,从而有利于目标检测等后续处理。

A contrast enhancement method based on the Poisson reconstruction was developed to enhance multiple targets in polarimetric synthetic aperture radar （SAR） images. Each kind of target is enhanced relative to its corresponding clutter by using the generalized optimization of polarimetric contrast enhancement （GOPCE）. The obtained optimal polarization states and characteristic coefficients define the optimal local contrast for each pixel. A discrete Poisson equation is then constructed using least squares minimization, which is solved using fast Fourier transforms. Tests with polarimetric SAR data demonstrated that the method produces an enhanced image with a more uniform histogram while keeping its peaks. Satisfactory results can be obtained when enhancing multiple targets with various types of clutter for subsequent applications such as target detection.