期刊文献+

基于空时域特征的视觉显著图生成算法 被引量:1

Visual Saliency Map Algorithm Using Spatiotemporal Features
下载PDF
导出
摘要 提出了一种新的计算图像空时域显著图的方法,该算法首先用lucas-kanade金字塔算法求绝对运动矢量,用8参数透视模型计算背景运动矢量,再用二者的差值求时域显著图;然后利用颜色对比度和纹理信息计算空域显著图;最后,融合空时域并设置阈值得到总的图像显著图。实验结果表明,新算法能比已有算法更有效地提取视频图像的显著性区域。 A new algorithm for computing spatio-temporal saliency maps is proposed in this paper. Firstly, the optical flow vectors of absolute motion is estimated. Then the background motion vectors to obtain the temporal saliency maps is calculated. Secondly, color contrast and texture information is used to calculate the spatial saliency maps. Finally,spatio-temporal saliency maps by fu- sing spatial and temoral maps is got. Experimental results show a better performance when compared to several state-of-the-art temporal saliency models.
出处 《电视技术》 北大核心 2015年第17期1-4,83,共5页 Video Engineering
基金 国家自然科学基金项目(61471201) 江苏省自然科学青年基金项目(BK20130867) 江苏省高校自然科学研究项目(12KJB510019) 江苏省高校自然科学重大项目(13KJA510004) 南京邮电大学校科研基金项目(NY212015) 南京邮电大学"1311人才计划"资助课题项目
关键词 显著图 运动矢量 颜色对比度 纹理 saliency map motion vectors color contrast texture
  • 相关文献

参考文献18

  • 1LI Zhicheng, QIN Shiyin, ITTI L. Visual attention guided bit alloca- tion in vide() compression [ J ]. Image and Vision Computing, 2011, 29 ( 1 ) : 1-14. 被引量:1
  • 2ZHANG W. An adaptive computational model for salient object de- tection[ J]. Muhimedia, 2010,12 (4) :300-316. 被引量:1
  • 3MAHAPATRA D, WINKLER S, YEN S C. Motion saliency out- weighs other low-level features while watching videos [ C ]//Proc. Electronic Imaging 2008. [ S.l. ] : International Society for Optics and Photonics ,2008:680-689. 被引量:1
  • 4LIU D,SHYU M I. Semantic retrieval for videos in non-static back- ground using motion saliency and globM features[ C ]//Proc. 2013 IEEE Seventh International Conference on Semantic Computing( [C- SC ). [ S.l. ] : IEEE Press, 2013:294-301. 被引量:1
  • 5ZHU Y,JACOBSON N,PAN H, et al. Motion-decision based spati- otemporal saliency for video sequences [ C]//Proc. 2011 IEEE In- ternational Confence on Acoustics, Speech and Signal Processing ([CASSP). [ S.l. ] :IEEE Press,2011 : 1333-1336. 被引量:1
  • 6HUA Z, X1ANG T, YAO W C. A distortion-weighing spatiotempo- ral visual attention model for video analysis[ C ]//Proc. 2nd Interna- tional Congress on Image and Signal Processing, C[SP' 09. [ S.l. ] : IEEE Press,2009 : 1-4. 被引量:1
  • 7WU B, XU L, LIU G. A visual attention model fir news video [C]//Pr. 2013 IEEE International Symposium on Circuits and Systems([SCAS). [S.l. I :IEEE Press, 2013: 941-944. 被引量:1
  • 8CHEN Y M,BAJ[C I V. Motion vector outlier rejection cascade for global motion estimation [ J ]. Signal Processing Letters, 2010, 17(2) :197-200. 被引量:1
  • 9HADIZADEH H, BAJ[C I V. Saliency-aware video compression [ J]. Image Processing,2014,23 ( 1 ) : 19-33. 被引量:1
  • 10WANG Z,LI Q. Video quality assessment using a statistical model of human visual speed perception [ J ]. JOSA, 2007,24 ( 12 ) : 61-69. 被引量:1

同被引文献9

引证文献1

二级引证文献3

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部