基于图像处理的棚车车门状态自动检测

摘要

This paper provides an automatic detection system for boxcar doors based on image processing. The obtained boxcar image is processed by appropriate treatments. Especially it uses Mathematical Morphology to eliminate the noise, and uses improved Hough transform to detect the horizontal and vertical lines of boxcar doors. It is realized to detect the status of boxcar doors automatically. The measured results show that the detection system is able to achieve higher accuracy detection, and it has certain practicality.%针对目前我国棚车车门状态检测采用人工实现的现状,提供了基于图像处理的棚车车门状态自动检测系统;该系统对所获取的棚车图像进行相应的处理,特别是采用数学形态学进行噪声消除处理,以及利用改进的Hough变换检测车门上的水平直线和门缝的竖直直线,实现了棚车车门状态的自动检测;实测结果表明,该检测系统能够实现较高精度的检测结果,具有一定的实用性.