摘要
针对构建鲁棒的背景模型和背景模型更新问题,结合ViBe算法提出了一种基于生物视觉特性的背景减除算法.首先,借鉴蛙眼视觉外部特性中的近视特性,阐述了"区域模糊化理解"预处理操作的含义及其实现;其次,从人类对颜色的心理认知特性出发,给出了一种LUV颜色空间中基于韦伯定律的颜色差异度量准则;最后,从背景建模、前景检测以及背景模型更新三个主要方面,介绍了算法的具体实现.实验结果表明,该算法能够提高运动目标检测的精度.
For the problem of how to build a robust background model and update the background model,a background subtraction algorithm based on biological vision characteristics combined with the ViBe algorithm was proposed.Firstly,utilizing the extrinsic nearsightedness characteristic of the frog's visual system,the meaning of a pretreatment method called "region fuzzy" and its implementation were described.Then,considering the characteristics of color cognition by humans,a measurement criterion for color difference based on Weber's law in the LUV color space was given.Finally,specific implementations of the algorithm were introduced from three main aspects:Background modeling,foreground detection and background model updating.Experimental results show that this algorithm can improve the accuracy of moving object detection.
基金
国家自然科学基金(61005091)资助
关键词
生物视觉
区域模糊化理解
韦伯定律
背景减除
运动目标检测
biological vision
region fuzzily understanding
Weber's law
background subtraction
moving object detection