基于火焰图像信息量度量的钢水碳含量定性分析  

The Quantitative Analysis for Carbon Content of Molten Steel Based on the Flame Image Information Measure

作　　者：刘立[1] 李悛[1] 刘芳菊[1] 汪琳霞[1] 罗扬[1] 

机构地区：[1]南华大学计算机科学与技术学院,湖南衡阳421001

出　　处：《南华大学学报（自然科学版）》2015年第3期69-72,共4页Journal of University of South China：Science and Technology

基　　金：国家自然科学基金资助项目(60475024);湖南省自然科学基金(省市联合基金)项目(13JJ9008)

摘　　要：钢厂炼钢过程中钢水碳含量的多少与炉口火焰的内部结构信息有着密切的关联.本文基于marr的视觉理论,提出了采用harris角点经典算法来测量火焰等非结构运动目标的内部结构信息,以此来定性分析钢水中碳含量.实验结果表明,该方法不但能准确地判断出炼钢所处的三个阶段,而且因算法简单,时间复杂度较低,适合在线炼钢碳含量预测等实时应用.The carbon content of molten steel were closely related with the the internal structure information of flame in steelmaking process. The paper presented the classic algo- rithm such as harris to measure the internal structure information of fire in order to make quantitative analysis for carbon content of molten steel. The experiments showed that the methods could judge the three stages of steelmaking processing. Moreover, for the sake of the simple algorithm the time cost is low and it is suitable for the reahime prediction of on- line steelmaking carbon content.

关 键 词：火焰图像 非结构目标 信息量 含碳量 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]