能把图像右边的明暗分界线提取出来有什么好的方法。

图像分割。

read_image (Image175923xiubtxzvb2xxun66, 'C:/Users/Administrator/Desktop/175923xiubtxzvb2xxun66.jpg')
threshold (Image175923xiubtxzvb2xxun66, Region, 10, 255)
shape_trans (Region, RegionTrans, 'convex')
erosion_rectangle1 (RegionTrans, RegionErosion, 11, 11)
mean_image (Image175923xiubtxzvb2xxun66, ImageMean, 1, 199)
reduce_domain (ImageMean, RegionErosion, ImageReduced)
edges_sub_pix (ImageReduced, Edges, 'canny', 2, 10, 20)

感谢

可以试试halcon的卡尺工具。大概的代码如下：
create_metrology_model (MetrologyHandle)
    set_metrology_model_image_size (MetrologyHandle, Width, Height)
    add_metrology_object_line_measure(MetrologyHandle, rowS, colS, rowE, colE, MeasureLength1, MeasureLength2, 1, MeasureThreshold, [], [], Index1)
    set_metrology_object_param (MetrologyHandle, 'all', 'measure_transition', MeasureDection)
    set_metrology_object_param (MetrologyHandle, 'all', 'measure_select', MeasureSelectPoint)
    set_metrology_object_param (MetrologyHandle, 'all', 'num_measures', MeasureNum)
    set_metrology_object_param (MetrologyHandle, 'all', 'min_score', MinScore)
    set_metrology_object_param (MetrologyHandle, 'all', 'num_instances', NumInstances)
    apply_metrology_model (ImageEmphasize, MetrologyHandle)
        * ******2.2 结果显示
    get_metrology_object_measures (ContoursDown1, MetrologyHandle, 'all', 'all', Row1, Column1)
    gen_cross_contour_xld (Cross, Row1, Column1, 10, 0.78)
   

    * ******2.3 结果获取
    get_metrology_object_result_contour (ContourUp, MetrologyHandle, 'all', 'all', 1.5)
    get_metrology_object_result (MetrologyHandle, Index1, 'all', 'result_type', 'row_begin', rowUpS)
    get_metrology_object_result (MetrologyHandle, Index1, 'all', 'result_type', 'column_begin', colUpS)
    get_metrology_object_result (MetrologyHandle, Index1, 'all', 'result_type', 'row_end', rowUpE)
    get_metrology_object_result (MetrologyHandle, Index1, 'all', 'result_type', 'row_end', colUpE)
    clear_metrology_model (MetrologyHandle)

这种在工程上非常适合用灰度投影算法，计算它的水平方向灰度投影并确定明暗分界。