机器人视觉的电梯轿厢门状态识别系统

针对视频监控中人工监督的低效率、传感器在轿厢内安装困难等问题,结合电梯轿厢内场景的特征,设计了一套基于 ARM的电梯轿厢门智能识别监控系统.门状态智能识别系统以三星的 Exynos4412为控制中心,搭载 Linux系统驱动视频采集模块,通过基于计算机视觉的门状态识别算法实现轿厢内门开关状态的识别.电梯轿厢门状态智能识别系统是以数字图像处理为基础,为满足轿厢内监控设备的小型化、安装便利等需求,本文采用前端识别技术.在嵌入式Linux系统上实现图像采集、图像预处理,采用基于 Hough线变换算法来实现开关门状态监测.实验表明,该系统能对轿厢门状态进行准确识别,在检测识别和安防领域具有很好的应用前景.

Recognizing System for State of Elevator Door Based on Robot Vision

There are some problems in the identification system of indoor elevator,such as artificial supervision is inefficient,the installa-tion of sensor is inconvenient.Combining with the characteristics of the scene within the elevator car,an identification system of indoor elevator about the state of elevator door based on ARM is designed.The Exynos4412 is used as the control center,with video capture module.This system detects the state of elevator door by the computer vision.The system of recognizizing for the state of elevator is based on the digital image processing.In order to improve miniaturization of equipment and convenience of Installation,the front-end rec-ognition is used.Acquisiting image processing image are achieved in the embedded Linux,and the detection state of elevator door is a-chieved using the Hough line transformation algorithm.The experiment results verify that the system can obtain state of elevator door accurately,which has a good application prospect in zoom of detection and identification in the future.