点击次数：

论文类型：会议论文

发表时间：2018-06-09

页面范围：980-985

摘要：Road extraction from complex backgrounds plays a significant role in a wide range of applications. In this paper, a method based on the clustering and morphological operations for road extraction is proposed. First, a novel feature image containing three features of the original image is built for clustering instead of using the original image directly; then, kernel fuzzy c-means (KFCM) clustering is used to extract road area roughly; finally, morphological operations are integrated to eliminate non road areas and improve road smoothness. The experiment results show that, compared with the original image, the feature image can not only reduce the running time of the clustering process, but also improve the anti-interference ability of clustering and robustness in the subsequent morphological processing. © 2018 IEEE.