期刊
IEEE TRANSACTIONS ON BROADCASTING
卷 57, 期 4, 页码 794-801出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TBC.2011.2160106
关键词
Background model; object detection; sudden illumination change; video surveillance
资金
- National Science Council [NSC 99-2221-E-027-108, NSC 99-2221-E-011-145]
Background subtraction involves generating the background model from the video sequence to detect the foreground and object for many computer vision applications, including traffic security, human-machine interaction, object recognition, and so on. In general, many background subtraction approaches cannot update the current status of the background image in scenes with sudden illumination change. This is especially true in regard to motion detection when light is suddenly switched on or off. This paper proposes an illumination-sensitive background modeling approach to analyze the illumination change and detect moving objects. For the sudden illumination change, an illumination evaluation is used to determine two background candidates, including a light background image and a dark background image. Based on the background model and illumination evaluation, the binary mask of moving objects can be generated by the proposed thresholding function. Experimental results demonstrate the effectiveness of the proposed approach in providing a promising detection outcome and low computational cost.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据