基于均值漂移的SAR图像分割方法研究

摘要

In the paper, the application of the mean shift algorithm in the SAR images segmentation is discussed. It is pointed out that the mean shift algorithm is stable and strong adaptive for ima-ges segmentation. For the SAR images contaminated with noise, segmentation effect of the mean shift algorithm is subjected to the great influence. Thus, according to the features of the noise, a method of using the filter to improve SAR images segmentation based on the mean shift is pro-posed. The key of the method lies to design the proper filter. The experiment of SAR images seg-mentation based on the method shows that it can improve the SAR images segmentation results.%本文主要讨论了均值漂移算法在图像分割中的应用,指出均值漂移算法是一个稳定的图像分割算法,该算法具有很强的适应性.对于遭受噪声污染的图像,它的分割效果会受到很大的影响.因此针对噪声特性,提出了用滤波器来改善基于均值漂移的SAR图像分割的方法.该方法的关键是设计合适的滤波器.基于该方法进行了图像分割实验表明,该方法可以改进图像分割效果.