图  1  高炉料面轮廓检测过程

Fig.  1  Blast furnace surface contour detection process

图  2  高炉料面图像

Fig.  2  Blast furnace material surface image

图  3  高炉料面增强图像

Fig.  3  Blast furnace material surface enhanced image

图  4  分数阶微分算子推导流程图

Fig.  4  Flowchart of deriving ractional differential operator

图  5  3像素$\times$ 3像素邻域(${F(i, j)}$代表该像素点$\ast$的灰度值

Fig.  5  3 pixel $\times$ 3 pixel neighborhood

图  6  基于分数阶的多向微分算子不同分数阶阶次的轮廓检测图

Fig.  6  Contour detection based on fractional order of multiple direction differential operators with different fractional orders

图  7  不同分数阶阶次的评价函数及拟合曲线

Fig.  7  The evaluation function and fitting curve of different fractional orders

图  8  $v =0.81$时检测结果

Fig.  8  Result of $v=0.81$

图  9  本文算法处理结果

Fig.  9  Detection result of ours

图  10  Sobel分数阶微分与本文算法处理结果对比图

Fig.  10  Contradistinction of results of Sobel fractional differential and ours

图  11  经典算子与本文最后处理结果对比图

Fig.  11  Contradistinction of results of classical operators and ours

图  12  Sobel算子检测结果与标准曲线比较图

Fig.  12  Comparison of Sobel operator detection results and standard curve

图  13  Log算子检测结果与标准曲线比较图

Fig.  13  Comparison of Log operator detection results and standard curve

图  14  Canny算子检测结果与标准曲线比较图

Fig.  14  Comparison of Canny operator detection result and standard curve

图  15  本文算法检测结果与标准比较图

Fig.  15  Comparison of ours and standard curves

图  16  斜料轮廓图

Fig.  16  Contour of blast furnace burden sloping

图  17  高炉正常料面增强图

Fig.  17  Normal blast furnace material surface enhanced

图  18  正常料面轮廓图

Fig.  18  Contour of normal blast furnace burden