基于区域MRF的SAR图像快速分割算法

A Fast Algorithm for SAR Image Segmentation Using Region-Based MRF

针对受相干斑噪声影响较严重的合成孔径雷达(SAR)图像,提出了一种基于边缘保持(EPR)的区域MRF快速分割算法。基于EPR的SAR图像表示方法包括各向异性扩散的相干斑降噪算法和分水岭变换两部分,该方法在存在相干斑噪声的情况下,能够有效地抑制过分割和在区域边界进行目标边缘的准确定位。将基于EPR的表示方法和区域MRF相结合,能够大幅减少优化过程的搜索空间,获得准确的分类结果和统计特性,同时减少了计算量和分割错误。将提出的算法用于一幅添加了各种不同噪声水平的合成图像和SAR海冰影像的分割中,实验结果证明了该算法的有效性。该算法与现有的区域MRF相比,实验结果证明新算法能够节约计算时间50%,同时提高了分割准确性,尤其是在相干斑噪声较强的区域。

A novel edge-preserving region （EPR）-based representation for synthetic aperture radar （SAR） images is proposed, which is incorporated with region-level MRF model to offer an efficient approach to the segmentation of SAR images. The EPR-based representations of SAR images is constructed by applying the speckle reduction anisotropic diffusion （SRAD） algorithm and the watershed transform, which aims at efficiently suppressing over-segmentation within homogeneous objects while accurately locating object edges at region boundaries in the presence of speckle noise. In combination with a region-level MRF, the EPR-based representation facilitates the segmentation process by largely reducing the search space of optimization process and improving parameter estimation of feature model, leading to considerable computational saving and less probability of false segmentation.