一种基于序列SAR图像的海浪对消方法

在高分辨率SAR海洋图像中，海浪是图像中最明显的纹理特征。当需要对海面上非海浪信息进行检测或者反演时，需要克服图像中的海浪杂波干扰。本文提出了一种海浪对消方法，它利用序列SAR图像间海浪条纹的差异与海浪的振动特性对海浪纹理进行抑制，不仅利用了图像的几何信息还利用了时变信息，能够选择性地抑制SAR图像中的海浪纹理，同时很好地保留海面上其他信息。本文还利用真实的机载SAR数据对算法进行了有效性验证。实验结果表明，该方法能够有效地抑制图像中的海浪纹理，同时很好地保留图像中的非海浪目标信息，比传统的仅利用空间信息的滤波方法效果更好。

Ocean wave clutter cancellation method based on series of SAR images

Waves are the most obvious surface stripy features in the high resolution synthetic aperture radar (SAR) images of the ocean. In order to effectively detect or retrieve the non wave phenomena from SAR ocean images, the interference from the waves need to be suppressed. The traditional ocean wave clutter suppression method is mainly based on the geometric texture of the images, by means of filtering in the special frequency domain or wavelet domain. However, the regions in the frequency domain or wavelect domain which contain information of the waves may also contain spectra of other non wave objects, thus leading to an unwanted information losing when filtering on these regions. In this paper, we propose a new ocean wave clutter suppression method based on series of SAR images. We use the differences of the two images in combination with the wave dispersion relation to cancel the wave stripe in the images. The main point of our method is to use both the geometric information and the temporal information, thus enable us to suppress the waves selectively while leaving the non wave information unharmed. In addition, we use real SAR data to validate it. The result of the experiments shows that the proposed method can effectively suppressed the wave stripes and maintain the non wave information in the images. Our method outperformed the traditional filter based method as well.