计算机应用与软件2017，Vol.34Issue(1)：180-186,257,8.DOI:10.3969/j.issn.1000-386x.2017.01.033
基于超像素分类的显著目标检测
SALIENT OBJECT DETECTION BASED ON SUPER-PIXEL CLASSIFICATION
摘要
Abstract
A new salient detection method combining super-pixel segmentation with boundary-center priors is proposed.Firstly, the image is segmented by SLIC method to get super-pixel region classified as background or foreground.Then, we calculate the saliency of the regions in the respect of color and space.Finally, the fusion of different aforementioned salient value is computed as the total saliency.In the experiments, the importance of foreground, background, color, space in salient calculation are analyzed;on the other side, extensive experimental results show that the performance of this method is higher than the other 8 state-of-the-art saliency detection methods.关键词
显著性检测/超像素分割/边界-中心知识/前景-背景Key words
Salient detection/Super-pixel/Boundary-center priors/Foreground-background分类
信息技术与安全科学引用本文复制引用
李继德,李晓强,沙彩霞..基于超像素分类的显著目标检测[J].计算机应用与软件,2017,34(1):180-186,257,8.基金项目
国家自然科学基金项目(61402279). （61402279）
