料流轨迹图像中特征半径的分析和提取

Analysis and Extraction of a Characteristic Radius of the Trajectory Charging in a Blast Furnace Charging Process

本文针对目前高炉布料轨迹测量分析的困难,引入模式识别技术,提出激光栅格物理标定方法,对料流轨迹图像的边缘算法进行对比分析,分析不同算法的图像处理结果,采用非固定探测的方法获得边缘算法的阈值。并根据不同激光波长的料流轨迹成像效果,对图像的特征半径的动态分析,对图像的不同特征点采用不同长度的特征半径分析,得到料流轨迹落点分布和料流宽度信息,重建料流轨迹极坐标分布图,为料流轨迹图像的信息获取提供更精确的数据,实现高炉操作的闭环控制自动化和工业信息化的精确要求。

As for the difficulty in analysing the trajectory charging of the current blast furnace,the pattern recognition technique is introduced and the mesurement method of the polar coordinate laser grid is proposed.To contrast the edge analysis of trajectory charging,we calculate the threshold value by linear probing.According to the image quality of trajectory charging in the difference spectrum,the characteristic radius is analysed and defined at the same time to different lengths of the characteristic radius for attaining accurate trajectory charging data information,the trajectory charging in polar coordinate is reconstructed,and the automatic control information of the blast furnace is obtanied.