自动扶梯是始终或阶段性保持运行状态的电梯,针对其在无人乘坐情况下持续运行所造成的能源浪费,提出了一种基于计算机视觉的自动扶梯节能控制系统,该系统由拍摄自动扶梯全景的摄像单元、微处理器和可编程逻辑控制器(PLC)构成;首先通过摄像单元获取整个自动扶梯的全景视频图像,接着利用混合高斯模型对全景视频图像进行背景建模处理,得到在自动扶梯上的前景对象;进而,根据自动扶梯上有无前景对象,生成相应控制信号,通过电梯PLC模块控制自动扶梯的运行和停止;实验结果表明,本文所开发的自动扶梯节能系统,对于自动扶梯上的对象物体,检测率在99%以上,可以较为可靠地控制自动扶梯运转与停止,达到预期的节能效果。
An escalator still keeps working even without any passengers on it,which leads to huge energy waste.In this paper a system that automates the control of escalator in terms of the actual state is described.The system presents an energy-saving device of escalator based on computer vision(CV) which consists of a photograph unit,computer and Programmable Logic Controller(PLC).Firstly,it gets the panoramic video images which cover the escalator from the camera,and then extracts the foreground objects by Mixture Gaussian Background Modeling.Switching on/off the escalator is determined by whether there are valid foreground objects.The experimental results show that the system is robust enough to detect the moving objects on the escalator effectively in real time and control the escalator correctly.