摘要
基于BP神经网络技术设计了一套机器视觉系统,该系统可以用于代替人工识别煤矿用吊装吊耳配合。系统利用对数据归一化、图像HU矩、BP神经网络等算法,将数据集进行BP神经网络训练,最终实验数据表明,在二分类情况下,对正面数据集的识别准确率为100%、对左侧数据集的识别准确率为91.67%、对右侧数据集的识别准确率为89.17%。综合实验结果表明采用BP神经网络可实现吊装吊耳位置的识别。
Based on BP neural network,a set of machine vision system can be used to replace artificial recognition of hoisting lug coordination for coal mines.The system uses algorithms such as data normalization,image HU moments,and BP neural network to train the data set with BP neural network.The final experimental data shows that in the case of two classifications,the recognition accuracy of the positive data set is 100%.The recognition accuracy rate for the left data set is 91.67%,and the recognition accuracy rate for the right data set is 89.17%.Comprehensive experimental results show that the position of the lifting lug can be recognized using BP neural network.
作者
邬永江
赵鹏
刘喜
WU Yong-jiang;ZHAO Peng;LIU Xi(Shenhua Zhungeer Energy Co.,Ltd.,Ordos 010300,China)
出处
《煤炭工程》
北大核心
2020年第S02期121-125,共5页
Coal Engineering
关键词
BP神经网络
吊装
位置识别
HU矩
图像分类
BP neural network
lifting
position recognition
HU moment
image classification
作者简介
邬永江(1986—),男,内蒙古鄂尔多斯人,工程师,现从事行政管理工作,E-mail:408656365@qq.com。
