四川大学学报（自然科学版）2012，Vol.49Issue(2)：309-314,6.DOI:10.3969/j.issn.0490-6756.2012.02.011
基于光流场核密度估计的动态目标分割
Detection of moving objects based on optical flow kernel density estimation
摘要
Abstract
In classical kernel density estimation, for the estimation is based on pixels' brightness it can offer the target a better connectivity and the inherent concentrated pixel distribution. To make use of the advantages, the calculation of dimensions can be expanded to be enabled to segment the optical flow relative to a pixel, and the method can be named as Detection of Moving Object Based on Optical Flow Kernel Density Estimation. Experiments show that the method can extract more complete moving objects from a moving background.关键词
光流场/核密度估计/运动背景/动态目标分割Key words
optical flow,kernel density estimation, moving background, detection of moving objects分类
信息技术与安全科学引用本文复制引用
王洋,李征,孙嘉..基于光流场核密度估计的动态目标分割[J].四川大学学报（自然科学版）,2012,49(2):309-314,6.基金项目
四川省科技厅应用基础研究基金(2006J13-092) （2006J13-092）
