摘要
针对超深矿井提升机排绳故障在线监测系统现场光照不稳定,多直线目标检测耗时大、稳定性差的难题,根据排绳过程本身所具有的规律,提出一种基于像素灰度分布统计的图像识别方法,并结合图像灰度分布特点,提出一种新的自适应阈值图像分割方法进行图像预处理,介绍了其基本原理并搭建了实验系统进行验证。结果表明,根据该方法对排绳故障进行在线视觉监测能够适应更宽的光照范围,实时性强,能够适应提升机现场复杂的使用环境,实现提升机排绳故障在线监测功能。
In order to overcome the problem that the illumination in the field is unstable and the detection of multiple lines is also unstable and time-consuming,a new image recognition method based on the statistical of pixel gray value was proposed according to the in-house law of rope-arranging process. Also a new threshold adaptive image segmentation method based on the features of gray distribution was proposed to preprocess the image. The basic principles were introduced and the experimental system was establelished to carry out the tests of recognizing the man-made rope-arranging faults. The result shows that the method for monitoring of rope-arranging fault of hoist is strongly real-time and adaptive to wider scope of intensity of illumination and complex environment in the field so that it can achieve the monitor function of rope-arranging fault of hoist on-line.
出处
《仪表技术与传感器》
CSCD
北大核心
2016年第10期73-75,共3页
Instrument Technique and Sensor
基金
国家重点基础研究发展计划(973)项目(2014CB049405)
关键词
超深矿井
提升机
排绳故障
图像识别
阈值分割
在线监测
ultra-deep mine
hoist
rope-arranging fault
image recognition
threshold segmentation
on-line monitoring