机器视觉下升降机内人数检测算法  

Use of Machine Vision in Counting of the People in Building Lift

作　　者：曹伟东[1] 张中炜[1] 

机构地区：[1]东华大学信息科学与技术学院,上海201620

出　　处：《电子科技》2016年第9期22-25,共4页Electronic Science and Technology

摘　　要：为了限制建筑升降机的人数,防止升降机超载,文中提出了一种基于机器视觉的建筑升降机内人数动态检测方法。该算法是在工人进门的区域内,利用改进的Vibe算法实现视频图像中的运动目标分割,采用领域平滑去除图像噪声,根据人头特征采用椭圆拟合的方式拟合人的头部,利用基于卡尔曼滤波的最近距离匹配法对椭圆的质心进行跟踪并计数。实验表明,文中算法对于升降机内俯视图像中人员计数的实现准确率较高,鲁棒性较好。In order to limit the number of building lifts, and to prevent overloading of the lifts, a method for the detection of the number of workers in lift based on machine vision is proposed in this paper. In the area where the workers enter the entry, the improved Vibe algorithm is used to segment moving objects in video images, and field smoothing to remove image noise. The head is fitted by ellipse fitting, and the eentroid of the ellipse is tracked and counted by the nearest distance matching method based on the Kalman filter. This algorithm has good accuracy and robustness in counting of the workers in the lift.

关 键 词：机器视觉 升降机人数统计 人头分割 目标跟踪 

分 类 号：TP242.62[自动化与计算机技术—检测技术与自动化装置]