Wolfram 语言与系统参考资料中心

范例

打开所有单元关闭所有单元

基本范例 (2)

测量图像中连通分量的一个属性：

Wolfram Language code: ComponentMeasurements[[image], "Elongation"]

只测量较大分量的属性：

Wolfram Language code: ComponentMeasurements[[image], "Elongation", #Count > 1000&]

范围 (24)

数据 (7)

二维整数标记矩阵的测量：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |   |
| - | - | - | - | - |
| 0 | 0 | 1 | 1 | 1 |
| 2 | 0 | 1 | 0 | 1 |
| 0 | 0 | 1 | 1 | 1 |
| 0 | 0 | 0 | 0 | 0 |
| 3 | 0 | 3 | 3 | 3 |⁠), "Count"]

基于非零像素连通度的二维图像分量的测量值：

Wolfram Language code: ComponentMeasurements[[image], "Count"]

二维图像的像素值测量：

Wolfram Language code: ComponentMeasurements[[image], "Min"]

二维彩色图像的像素值测量：

Wolfram Language code: ComponentMeasurements[[image], "Min"]

二维彩色图像的像素值测量及其对应的标签矩阵：

Wolfram Language code:

ComponentMeasurements[{[image], (⁠|   |   |   |   |   |
| - | - | - | - | - |
| 1 | 1 | 1 | 2 | 2 |
| 1 | 1 | 1 | 2 | 2 |
| 3 | 3 | 3 | 2 | 2 |
| 3 | 3 | 3 | 2 | 2 |
| 3 | 3 | 3 | 2 | 2 |⁠)}, "Mean"]

三维整数标记矩阵的测量：

Wolfram Language code: ComponentMeasurements[{{{1, 0}, {1, 0}, {0, 2}}, {{1, 0}, {1, 0}, {0, 3}}}, "Count"]

在三维中度量属性：

Wolfram Language code: ComponentMeasurements[\!\(\*Graphics3DBox[«8»]\), {"Image", "Count", "Mean", "StandardDeviation"}, All, "Dataset"]

属性 (7)

尺寸测量：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |
| - | - | - | - |
| 0 | 1 | 1 | 1 |
| 0 | 1 | 0 | 1 |
| 0 | 1 | 1 | 1 |
| 2 | 2 | 2 | 2 |⁠), "Count"]

填充分量的尺寸：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |
| - | - | - | - |
| 0 | 1 | 1 | 1 |
| 0 | 1 | 0 | 1 |
| 0 | 1 | 1 | 1 |
| 2 | 2 | 2 | 2 |⁠), "FilledCount"]

质心位置：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |
| - | - | - | - |
| 0 | 1 | 1 | 1 |
| 0 | 1 | 0 | 1 |
| 0 | 1 | 1 | 1 |
| 2 | 2 | 2 | 2 |⁠), "Centroid"]

最佳拟合椭圆的半轴长度：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |
| - | - | - | - |
| 0 | 1 | 1 | 1 |
| 0 | 1 | 0 | 1 |
| 0 | 1 | 1 | 1 |
| 2 | 2 | 2 | 2 |⁠), "SemiAxes"]

以轴为方向的边界盒：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |
| - | - | - | - |
| 0 | 1 | 1 | 1 |
| 0 | 1 | 0 | 1 |
| 0 | 1 | 1 | 1 |
| 2 | 2 | 2 | 2 |⁠), "BoundingBox"]

相邻边界的列表：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |
| - | - | - | - |
| 0 | 1 | 1 | 1 |
| 0 | 1 | 3 | 1 |
| 0 | 1 | 1 | 1 |
| 2 | 2 | 2 | 2 |⁠), "AdjacentBorders"]

计算多个属性：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |
| - | - | - | - |
| 0 | 1 | 1 | 1 |
| 0 | 1 | 0 | 1 |
| 0 | 1 | 1 | 1 |
| 2 | 2 | 2 | 2 |⁠), {"Count", "AdjacentBorders"}]

判据 (5)

根据计算的属性选择：

Wolfram Language code: ComponentMeasurements[[image], "Area", # > 1000&]

根据不同的属性选择：

Wolfram Language code: ComponentMeasurements[[image], "Area", #AdjacentBorderCount == 0&]

根据多个计算的属性选择：

Wolfram Language code: ComponentMeasurements[[image], {"Area", "AdjacentBorderCount"}, #1 > 1000 && #2 == 0&]

根据不同于所测量的和返回的一组属性选择：

Wolfram Language code: ComponentMeasurements[[image], "Centroid", #Area > 1000 && #AdjacentBorderCount == 0&]

指定使用哪个计算的属性作为函数参数：

Wolfram Language code: f[x_, y_] := x > 1000 && y == 0

Wolfram Language code: ComponentMeasurements[[image], "Area", f[#Area, #AdjacentBorderCount]&]

输出格式 (5)

将输出格式化为分量和分量属性值之间的关联关系：

Wolfram Language code: ComponentMeasurements[[image], {"Count", "PerimeterCount"}, All, "ComponentAssociation"]

将输出格式化为属性和分量属性值之间的关联关系：

Wolfram Language code: ComponentMeasurements[[image], {"Count", "PerimeterCount"}, All, "PropertyAssociation"]

将输出格式化为分量和分量属性值之间的嵌套关联关系：

Wolfram Language code: ComponentMeasurements[[image], {"Count", "PerimeterCount"}, All, "ComponentPropertyAssociation"]

将输出格式化为属性和分量属性值之间的嵌套关联关系：

Wolfram Language code: ComponentMeasurements[[image], {"Count", "PerimeterCount"}, All, "PropertyComponentAssociation"]

将输出格式化为分量和分量属性值的数据集：

Wolfram Language code: ComponentMeasurements[[image], {"Count", "PerimeterCount"}, All, "Dataset"]

选项 (2)

CornerNeighbors (2)

缺省情况下，如果分量有 corner neighbor，则假定它们相连:

Wolfram Language code: ComponentMeasurements[[image], "Count"]

假定 4-连通性 (four-connectedness)：

Wolfram Language code: ComponentMeasurements[[image], "Count", CornerNeighbors -> False]

默认情况下，一个单元的周围所有8个单元视为相邻单元：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |
| - | - | - | - |
| 0 | 0 | 1 | 1 |
| 0 | 0 | 1 | 1 |
| 2 | 2 | 0 | 0 |
| 2 | 2 | 0 | 0 |⁠), "Neighbors"]

只考虑四个直接的邻单元为相邻单元：

Wolfram Language code:

ComponentMeasurements[(⁠|   |   |   |   |
| - | - | - | - |
| 0 | 0 | 1 | 1 |
| 0 | 0 | 1 | 1 |
| 2 | 2 | 0 | 0 |
| 2 | 2 | 0 | 0 |⁠), "Neighbors", CornerNeighbors -> False]

应用 (5)

找出镰刀细胞：

Wolfram Language code:

m = MorphologicalComponents[img = [image], .3];
ComponentMeasurements[ {m, img} , "MaskedImage", #Elongation > 0.5&]

计算每个单元的平均颜色值：

Wolfram Language code:

i = [image];
ComponentMeasurements[{i, MorphologicalComponents[i]}, "Mean"]

选择完整的和无重叠的单元：

Wolfram Language code: i = [image];

二值化图像，得到前景：

Wolfram Language code: b = Binarize[i, {0, .7}]

计算质心和包含图像的较大单元的等价圆盘的半径：

Wolfram Language code: cells = ComponentMeasurements[b, {"Centroid", "EquivalentDiskRadius"}, #AdjacentBorderCount == 0 && 50 < #Area < 1200&]

显示选中的单元：

Wolfram Language code: HighlightImage[i, Circle@@@cells[[All, 2]]]

用所属分量的属性替换每个像素：

Wolfram Language code: i = [image];

计算图像分割：

Wolfram Language code:

labels = ClusteringComponents[i, 5];
Colorize[labels]

用其平均色彩更换每个分段：

Wolfram Language code: Image[Replace[labels, ComponentMeasurements[{i, labels}, "Mean"], 2]]

使用每个分量的平均强度：

Wolfram Language code: Image[Replace[labels, ComponentMeasurements[{i, labels}, "MeanIntensity"], 2]]

计算地图制图的一个连通图：

Wolfram Language code:

m = MaxFilter[MorphologicalComponents[[image]], 2];
Graph[ComponentMeasurements[m, "Label"][[All, 1]], DeleteDuplicates[Sort /@ Flatten[ComponentMeasurements[m, "Neighbors"]  /. (i_ -> {l___}) :> (i#& /@ {l})]], VertexCoordinates -> ComponentMeasurements[m, "Centroid"], VertexSize -> 1]

属性和关系 (8)

仅计算较大分量的度量：

Wolfram Language code: i = [image];

Wolfram Language code: ComponentMeasurements[i, #Count > 500&]

首先使用 DeleteSmallComponents 删除较小分量：

Wolfram Language code: DeleteSmallComponents[i, 500]

注意到标记被归一化了：

Wolfram Language code: ComponentMeasurements[%, "Count"]

计算不接触图像边界的分量的度量：

Wolfram Language code: i = [image];

Wolfram Language code: ComponentMeasurements[i, "Count", #AdjacentBorderCount == 0&]//Values

首先使用 DeleteBorderComponents 删除边界分量：

Wolfram Language code: DeleteBorderComponents[i]

Wolfram Language code: ComponentMeasurements[%, "Count"]//Values

计算具有一个孔的分量，而不接触图像边界：

Wolfram Language code: i = [image];

Wolfram Language code: ComponentMeasurements[i, "Count", #Holes == 1 && #AdjacentBorderCount == 0&]//Values

使用 SelectComponents 删除不具有孔或接触边界的分量：

Wolfram Language code: SelectComponents[i, #Holes == 1 && #AdjacentBorderCount == 0&]

Wolfram Language code: ComponentMeasurements[%, "Count"]//Values

提取支持的二维图像的完整属性列表：

Wolfram Language code: ComponentMeasurements[[image], "Properties"]