| 块聚类的协同显著性检测 |
| |
| 引用本文： | 杨麟,杜吉祥,聂一亮.块聚类的协同显著性检测[J].华侨大学学报(自然科学版),2018,0(3):445-450. |
| |
| 作者姓名： | 杨麟 杜吉祥 聂一亮 |
| |
| 作者单位： | 华侨大学 计算机科学与技术学院, 福建 厦门 361021 |
| |
| 摘 要： | 针对复杂背景的多图像显著性检测问题,提出一种基于块聚类的多图像协同显著性检测方法.该方法通过构建多图像间共同对象的关联性,利用块聚类计算4种显著性测度并融合,获得较好的协同显著性检测效果.实验结果表明:基于块聚类的协同显著性检测方法能够有效提高检测精度,且鲁棒性较高.
|
| 关 键 词： | 协同显著性检测 协同分割 块聚类 显著性测度 测度融合 |
Co-Saliency Detection Using Patch-Cluster |
| |
| YANG Lin,DU Jixiang,NIE Yiliang.Co-Saliency Detection Using Patch-Cluster[J].Journal of Huaqiao University(Natural Science),2018,0(3):445-450. |
| |
| Authors: | YANG Lin DU Jixiang NIE Yiliang |
| |
| Institution: | College of Computer Science and Technology, Huaqiao University, Xiamen 361021, China |
| |
| Abstract: | For the co-saliency detection problem in multiple images with complex background, this paper put forward a co-saliency detection method for multiple images which based on patch-cluster. This method constructing the correlation of the common objects between images, and using block clustering to calculate four kinds of saliency measurement and fused to get a better co-saliency detection effect. Experimental results show that the co-saliency detection method which based on patch-cluster can effectively improve the detection accuracy, and the have higher robustness. |
| |
| Keywords: | co-saliency detection co-segmentation patch clustering saliency cue cue fused |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《华侨大学学报(自然科学版)》浏览原始摘要信息 |
| 点击此处可从《华侨大学学报(自然科学版)》下载免费的PDF全文 |
