【作者】 靳海燕； 熊庆宇； 王楷； 石为人；

【Author】 Jin Haiyan~1 Xiong Qingyu~(1,2) Wang Kai~1 Shi Weiren~1 (1 Institute of Intelligence & Advanced Integration Technologies,College of Automation,Chongqing,400044, 2 School of Software Engineering,Chongqing University,Chongqing 400044,China)

【机构】 重庆大学自动化学院智能科学与先进集成技术研究所； 重庆大学软件学院；

【摘要】 为防止电梯超载,采用摄像头获取视频流,运用图像处理技术对进出电梯的双向人数进行统计,提出了一种通过统计电梯轿厢内的人数判断电梯是否超载的方法。首先建立背景边缘模型,通过减背景边缘法提取人体边缘,根据感兴趣区域(ROI)内的轮廓面积判断人数,并对多个人进行分离。运用运动历史图像(MHI)对目标进行跟踪,根据人体运动方向与水平方向的夹角判断进出电梯的方向,并进行计数。实验结果表明该方法能对进出电梯的双向行人进行准确的计数,电梯轿厢内人数超过最大载客量时进行实时报警,有效地防止了电梯超载。更多还原

【Abstract】 To prevent the elevator overloading,video stream was obtained from the camera,and bi-directional people in/out of the elevator was counted using image processing technoIogy.A method determining whether it is overloaded by counting the number of people in the elevator was proposed.Firstly,background edge model was established and the human body edge was detected by reducing the background edge.The number of people was determined according to the contour area and more individuals was isolated.Motion history image(MHI) was used for target tracking,and the direction in /out of the elevator was determined according to the angle between the movement direction of human body and horizontal direction.Experimental results show that the method can count the bi-directional pedestrian in/out of the elevator accurately.It can real-time alarm when the number of people in the elevator exceeds the maximum capacity of the elevator and it can prevent overloading effectively.更多还原