Show simple item record

dc.contributor.author曾志宏
dc.contributor.author陈振武
dc.contributor.author肖庆云
dc.contributor.author黄婷
dc.date.accessioned2018-11-26T08:55:27Z
dc.date.available2018-11-26T08:55:27Z
dc.date.issued2017
dc.identifier.citation嘉应学院学报,2017,(2)
dc.identifier.issn1006-642X
dc.identifier.urihttps://dspace.xmu.edu.cn/handle/2288/165750
dc.description.abstract针对视频第一帧中存在待检测的运动物体,利用视觉背景提取算法(ViBe)对该物体后续帧检测,会在第一帧的位置上持续出现鬼影现象,提出了一种改进的ViBe算法.该算法在视频中融合连续N帧图像作为前景点的基础上,采用重采样的方法来初始化背景模型以实现动态背景有效提取.实验结果表明,提出的改进算法能有效地检测出动态背景下移动物体,并能有效地解决图像获取的鬼影现象,从而提高了算法的误检率及鲁棒性,通过改进后的ViBe算法比原算法能够更有效地检测动态背景下的运动目标.
dc.description.abstractAs the moving object to be detected in the first flame of the video, the object using the visual background extraction algorithm (ViBe) will continue the subsequent frame detection of the phenomenon of ghost area at the position of the first frame, an improved ViBe algorithm is proposed . The algorithm uses the resampling method to initialize the background model in order to extract the dynamic background eflficiently by merging the successive N frames in the video as the foreground points. The experimental results show that the improved algorithm can effectively detect moving objects in dynamic background, and can effectively solve the image acquisition of ghost area, thus improve the false detection rate and robustness. Through the improved ViBe algorithm is more effective than the original algorithm in detecting dynamic moving objects.
dc.description.sponsorship福建省自然科学基金项目(2015J01587);福建省科技厅资助高校项目(JK2010056);福建省教育厅中青年项目(JAT160487)
dc.language.isozh_CN
dc.subject视觉背景提取
dc.subject前景点
dc.subject动态背景
dc.subject鬼影
dc.subjectghost area
dc.title一种改进的融合前景点重采样的ViBe算法
dc.title.alternativeAn Improved ViBe Algorithm Based on Fusion of Foreground Points Resampling
dc.typeArticle


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record