基于ViBe的改进运动目标检测算法

Abstract

ViBe是运动目标检测的常用算法。针对ViBe运动检测算法在实际环境中易出现鬼影、动态背景等干扰的问题，提出结合灰度特征直方图匹配的方法来检测鬼影，并进一步消除鬼影的干扰。针对动态背景错检测为运动前景的问题，为图像中各个像素设置闪烁属性和突变属性来判断各像素是否属于动态背景。实验表明，改进后的算法能有效弥补传统ViBe算法在鬼影消除和动态背景干扰上的不足。 Background subtraction is a common algorithm for moving target detection. Aiming at the prob-lem that the motion detection algorithm of ViBe is prone to ghost and dynamic background in the real environment, it is proposed to detect the ghost by combining the gray feature histogram matching method and further eliminate the ghost interference. Aiming at the problem that motion background is detected as motion foreground, we set the flickering property and mutation property for each pixel in the image to judge whether each pixel belongs to dynamic background. Experiments show that the improved algorithm can effectively compensate for the deficiencies of traditional ViBe algorithm in ghosting and dynamic background jamming.