摘要
本文提出了运用神经网络技术,通过拍摄转炉炉表被监测部位的红外热图像,经过计算机图像处理后,提取图像特征,并检测对应的炉内的蚀损状况,分别作为神经网络模型训练样本的输入和输出。训练好的神经网络系统即可用来监测转炉的炉衬蚀损状况。
The paper a method of detecting etching of furnace lining with artificial neural network technology is introduced. Extract image feature through image processing by computer after photographing infrared thermal image of surface part of converter which is being monitored and check the etching of furnace lining, using the result of them as the input and output of the training sets to train the neural network. Then the neural network having been trained well could be used to analyze the status of etching of furnace lining.
出处
《甘肃冶金》
2007年第1期16-18,共3页
Gansu Metallurgy
基金
江西省教育厅科技项目(编号:2003073)
自然科学基金项目(编号:0650041)
关键词
神经网络
红外热像
转炉
neural network
infrared thermal imaging
converter