计算机应用研究2017，Vol.34Issue(5)：1548-1552,1589,6.DOI:10.3969/j.issn.1001-3695.2017.05.061
视觉背景提取算法关键技术研究
Research on key technology of visual background extractor algorithm
摘要
Abstract
Aiming at the problems of ghost and poor detection accuracy in complicated scenes which existed in ViBe algorithm,this paper proposed an improved algorithm.The algorithm utilized several pixel values from consecutive frames to train the samples of background during initialization,thus reduced the ghost's effects on precision in detecting objects.In the process of pixels matching,the algorithm calculated the area complexity of pixel to adjust both the classified threshold and the match threshold dynamically,so it could adapt complicated background scenes and diminished the disturbances of incorrect foreground pixels.In the update stage,it used an adaptive time-subsampling factor to reduce the wrong propagation of pixel values and accelerated the recovery of the background samples.Experiments show that compared to traditional algorithm,the presented algorithm performs well in ghost suppression and adaptation to complicated scenes.关键词
目标检测/视觉背景提取/鬼影消除/自适应时间二次抽样因子Key words
object detection/visual background extractor/ghost elimination/adaptive time-subsampling factor分类
信息技术与安全科学引用本文复制引用
郭冲宇,杜慧敏,蒋忭忭,季凯柏..视觉背景提取算法关键技术研究[J].计算机应用研究,2017,34(5):1548-1552,1589,6.基金项目
国家自然科学基金资助项目(61136002) （61136002）
西安市科技发展计划资助项目(CXY1440 (10)) （CXY1440 (10）
