Methods

static KMeans(image, numClustersopt, numIterationsopt, neighborhoodSizeopt, gridSizeopt, forceConvergenceopt, uniqueLabelsopt)

models/Algorithms/Image/Segmentation.ts, line 49

对输入图像进行 K-Means 聚类

Parameters:

Name	Type	Attributes	Description
image	Image		待聚类的影像
numClusters	Number	<optional>	聚类的数目
numIterations	Number	<optional>	迭代次数
neighborhoodSize	Number	<optional>	邻域大小
gridSize	Number	<optional>	网格单元大小
forceConvergence	Boolean	<optional>	未达到收敛状态是否强制抛错
uniqueLabels	Boolean	<optional>	是否分配唯一标签

Returns:

- Image

Tutorials:

• Tutorial: gve.Algorithms.Image.Segmentation

static SNIC(image, sizeopt, compactnessopt, connectivityopt, neighborhoodSizeopt, seedsopt)

models/Algorithms/Image/Segmentation.ts, line 33

基于 SNIC（简单非迭代聚类）的超像素聚类

Parameters:

Name	Type	Attributes	Description
image	Image		聚类的影像
size	Number	<optional>	种子的大小
compactness	Number	<optional>	紧凑型因子
connectivity	Number	<optional>	连接性
neighborhoodSize	Number	<optional>	邻域大小
seeds	Image	<optional>	用于连接性的影像

Returns:

- Image

Tutorials:

• Tutorial: gve.Algorithms.Image.Segmentation

static seedGrid(sizeopt, gridTypeopt)

models/Algorithms/Image/Segmentation.ts, line 18

为聚类选择种子像素

Parameters:

Name	Type	Attributes	Description
size	Number	<optional>	种子大小
gridType	String	<optional>	网格类型（square|hex）

Returns:

- Image

Tutorials:

• Tutorial: gve.Algorithms.Image.Segmentation