基于图像域的势函数正则化SAR车辆图像增强算法

The algorithm of SAR image enhancement based on image regularization of potential functions

SAR图像增强技术是SAR信号处理中很重要的一门技术。一幅SAR原始图像中除了包含模糊的目标区域和阴影区外还包含了大量的噪声。在目标识别过程中,如果先对原始SAR图像进行增强处理,降低噪声的同时增强模糊的目标区域和阴影区,然后进行特征提取,就可以大大降低识别系统的运算量,提高识别效率。正则化方法广泛应用于图像分割,图像增强等领域。通过对两种正则化算法的比较,结合两种方法的优点,提出了一种改进的基于图像域的势函数正则化图像增强算法。利用MSTAR数据库数据进行仿真试验,取得了良好的效果。

The technology of SAR image enhancement is one of the important parts in SAR signal processing. The original SAR image contains clutter noises except targets and shadows. If clutters are reduced and the targets and shadows are enhanced at the same time, the computational complexity can be reduced, improving the recognition probability greatly. The regularization of potential functions is a good method of enhancing SAR images. The simulation results based on the MSTAR public target database show that algorithm proposed is feasible and effective.