摘要
智能视觉监控中,背景差分中背景的建立与实时更新和目标感兴趣区域(region of interest,ROI)的快速定位是后续目标跟踪的重要前提,提出了一种新的基于差分的统计背景重构,并在此基础上实现ROI的快速定位算法。实验结果表明,该算法能够很好的克服噪声的干扰,在复杂背景的环境下均能实现目标的准确定位,且算法复杂度低,适合于实时系统的开发应用。
Under intelligent visual surveillance,the construction and real-time update of the background difference and the rapid positioning of region of interest（ROI） are prerequisites to the successive target tracking.Based on the difference,this thesis proposes a new way to realize statistical background refactoring and then achieve rapid positioning.The results of the experiments suggest that,this method enables well-immunity from the noise and rapid positioning even in complicated background,furthermore,its low-level complicacy applies for the development and application of real time system.
出处
《林业科技情报》
2010年第2期133-136,共4页
Forestry Science and Technology Information
基金
南京林业大学科研发展基金项目(2008-10)
