计算机应用研究2017，Vol.34Issue(7)：1921-1928,8.DOI:10.3969/j.issn.1001-3695.2017.07.001
图像分割方法综述研究
Survey on image segmentation methods
摘要
Abstract
Image segmentation is an important and fundamental problem in computer vision,meanwhile it's a challenging task.In order to find out the state-of-the-art,main problems and future trends of image segmentation,this paper introduced the mainstream image segmentation methods after 2000 on the basis of extensive research on the existing literatures and the latest achievements.These methods were categorized into four classes: graph theory based methods,clustering based methods,classification based methods,and hybrid methods of clustering and classification.The basic ideas,advantage and disadvantage of typical algorithms belong to each category,especially the most recently published papers were introduced and analyzed.Finally,this paper introduced the datasets which were commonly used as benchmark and evaluation metrics,compared all the algorithms,summarized the work and forecasts some potential future research work.关键词
图像分割/图论/聚类/分类Key words
image segmentation/graph theory/clustering/classification分类
信息技术与安全科学引用本文复制引用
周莉莉,姜枫..图像分割方法综述研究[J].计算机应用研究,2017,34(7):1921-1928,8.基金项目
国家自然科学基金资助项目(61373012) （61373012）
江苏省高校自然科学研究项目(15KJB520016) （15KJB520016）
