计算机与现代化

• 图像处理 • 上一篇    下一篇

基于鬼影判断抑制和局部运动补偿的改进ViBe算法

  

  1. 东华大学信息科学与技术学院,上海-201600
  • 收稿日期:2017-03-28 出版日期:2018-01-23 发布日期:2018-01-24
  • 作者简介:章伟明(1992-),男,浙江湖州人,东华大学信息科学与技术学院硕士研究生,研究方向:视频图像处理; 周武能(1959-),男,教授,博士生导师,研究方向:神经网络,图像处理,控制理论。
  • 基金资助:
    国家自然科学基金资助项目(61573095)

An Improved ViBe Algorithm Based on Ghost Judgment #br# Suppression and Local Motion Compensation

  1. College of Information Science and Technology, Donghua University, Shanghai 201600, China
  • Received:2017-03-28 Online:2018-01-23 Published:2018-01-24

摘要: 传统的ViBe算法在检测运动目标时会产生“鬼影”现象,并且对只存在局部运动的目标无法检测出完整的结果。已有的改进算法在克服上述缺陷时没有进行缺陷类型的判断使得在一定程度上降低了算法的性能。因此,针对“鬼影”现象,本文提出先利用边缘相似性进行“鬼影判断”,之后结合三帧差分算法来消除鬼影。针对局部运动目标,本文提出一种结合检测阈值判断的方法,先确定是否存在身体局部运动,之后结合扫描线填充算法对目标进行补偿填充得到完整的运动目标。实验结果表明,本文算法能有效克服“鬼影”及局部运动的缺陷,提高目标检测的效率以及准确性。

关键词: ViBe算法, 边缘相似性, 三帧差分, 局部运动, 扫描线填充

Abstract:

The traditional ViBe algorithm has a “ghost” phenomenon in the detection of moving targets, and it cant detect the complete result for the target with only local motion. Most improved algorithms cant determine the types of defects in overcoming the above defects which reduces the performance of the algorithm. Therefore, for the “ghost” phenomenon, this paper uses edge similarity to make “the ghost judgment”, then combines the three frame difference method to eliminate ghosting. Aiming at local moving target, this paper proposes a new method to determine the existence of local motion, then the whole moving object is obtained by combining the scan line filling algorithm. Experiments show that this algorithm can effectively overcome the defects of “ghost” and local motion, and improve the efficiency and accuracy of the target detection.

Key words: ViBe algorithm, edge similarity, three frame difference method, local motion, scan line filling

中图分类号: