摘要
显著区域检测是计算机视觉中非常活跃的研究方向,其应用领域极为广泛。如何快速准确地找到图像的显著区域尚未形成完整的理论体系,且与具体应用密切相关,对研究人员来说仍是一个富有挑战的课题。对显著区域检测技术进行了综述。首先深入讨论了自底向上和自顶向下的显著区域检测方法,对方法进行了归类,并对典型方法进行了梳理;其次讨论了算法的评价标准和目前流行的显著性评测数据库;最后对目前存在的问题进行了总结,给出了未来的研究方向。
Salient region detection technology is a very active research area and is applied extensively.How to find salient region of the image quickly and accurately has not yet formed a complete theoretical system.In addition,salient region detection technology is closely related to application.So this technology is still a challenging research topic.A survey on salient region detection technology was given in the paper.Firstly,bottom-up and top-down salient region detection approaches were discussed in detail,including technique classification and typical techniques.Secondly,evaluation criteria and open saliency evaluation databases were discussed.At last,the main problems and challenges were highlighted based on analysis of current research.
出处
《计算机科学》
CSCD
北大核心
2016年第3期27-32,共6页
Computer Science
基金
国家自然科学基金项目(61271369
61372148)
北京市信息服务工程重点实验室开放课题(Zk20201402)
北京市自然科学基金项目(4152016)
北京市教育委员会科技计划面上项目(KM201511417008)
北京联合大学人才强校计划人才资助项目(Rk100201510)资助
