| 基于多尺度的动态视觉显著性检测方法 |
| |
| 引用本文： | 楼甜甜.基于多尺度的动态视觉显著性检测方法[J].现代计算机,2011(7):7-9. |
| |
| 作者姓名： | 楼甜甜 |
| |
| 作者单位： | 华东师范大学信息科学技术学院 |
| |
| 摘 要： | 视觉显著性是能在视觉环境中感知突兀区域的能力,以显著性的Center-Surround为前提,提出一种基于多尺度的动态视频视觉显著性检测的方法,用不同尺度视频选取匹配的中心-周边图像块(Patch),并结合基于上下文环境的条件概率分布进行显著性检测。
|
| 关 键 词： | 视觉显著性 条件概率 中心-周边 独立成分(ICs) 动态视频 |
A Dynamic Saliency Detection Method Based on Multi-Scale |
| |
| LOU Tian-tian.A Dynamic Saliency Detection Method Based on Multi-Scale[J].Modem Computer,2011(7):7-9. |
| |
| Authors: | LOU Tian-tian |
| |
| Affiliation: | LOU Tian-tian (Department of Information Science and Technology,East China Normal University,Shanghai 200241) |
| |
| Abstract: | Visual saliency has ability that makes some items in visual scenes stand out from their contexts.According to the center-surround hypothesis of saliency,proposes a dynamic saliency detection method based on multi-scale.In this way,different scaled center-surround patches are obtained,which combined with context-mediated PDs to analyze video saliency. |
| |
| Keywords: | Visual Saliency Probability Distributions Center-Surround Independent Component Dynamic Video |
| 本文献已被 CNKI 等数据库收录！ |
