作者: Huanxin Zou , Xianxiang Qin , Shilin Zhou , Kefeng Ji
DOI: 10.3390/S16071107
关键词:
摘要: The simple linear iterative clustering (SLIC) method is a recently proposed popular superpixel algorithm. However, this may generate bad superpixels for synthetic aperture radar (SAR) images due to effects of speckle and the large dynamic range pixel intensity. In paper, an improved SLIC algorithm SAR proposed. This exploits likelihood information image clusters. Specifically, local scheme combining intensity similarity with spatial proximity Additionally, post-processing, edge-evolving that combines context introduced as alternative connected components To estimate clusters, we incorporated generalized gamma distribution (GГD). Finally, superiority was validated using both simulated real-world images.