gray_closing_shapegray_closing_shapeGrayClosingShapeGrayClosingShapegray_closing_shape (算子)

名称

gray_closing_shapegray_closing_shapeGrayClosingShapeGrayClosingShapegray_closing_shape — 使用选定的掩模执行灰度值闭。

签名

描述

gray_closing_shapegray_closing_shapeGrayClosingShapeGrayClosingShapeGrayClosingShapegray_closing_shape applies a gray value closing to the input image ImageImageImageImageimageimage with the structuring element of shape MaskShapeMaskShapeMaskShapeMaskShapemaskShapemask_shape. The mask's offset values are 0 and its horizontal and vertical size is defined by MaskHeightMaskHeightMaskHeightMaskHeightmaskHeightmask_height and MaskWidthMaskWidthMaskWidthMaskWidthmaskWidthmask_width. The resulting image is returned in ImageClosingImageClosingImageClosingImageClosingimageClosingimage_closing。

If the parameters MaskHeightMaskHeightMaskHeightMaskHeightmaskHeightmask_height or MaskWidthMaskWidthMaskWidthMaskWidthmaskWidthmask_width are of the type integer and are even, they are changed to the next larger odd value. In contrast, if at least one of the two parameters is of the type float, the input image ImageImageImageImageimageimage is transformed with both the next larger and the next smaller odd mask size, and the output image ImageClosingImageClosingImageClosingImageClosingimageClosingimage_closing is interpolated from the two intermediate images. Therefore, note that gray_closing_shapegray_closing_shapeGrayClosingShapeGrayClosingShapeGrayClosingShapegray_closing_shape returns different results for mask sizes of, e.g., 4 and 4.0!

In case of the values 'rhombus' and 'octagon' for the MaskShapeMaskShapeMaskShapeMaskShapemaskShapemask_shape control parameter, MaskHeightMaskHeightMaskHeightMaskHeightmaskHeightmask_height and MaskWidthMaskWidthMaskWidthMaskWidthmaskWidthmask_width must be equal. The parameter value 'octagon' for MaskShapeMaskShapeMaskShapeMaskShapemaskShapemask_shape denotes an equilateral octagonal mask which is a suitable approximation for a circular structure. At the border of the image the gray values are mirrored.

The gray value closing of an image i with a structuring element s is defined as i.e., a dilation of the image with s followed by an erosion with s (see gray_dilation_shapegray_dilation_shapeGrayDilationShapeGrayDilationShapeGrayDilationShapegray_dilation_shape and gray_erosion_shapegray_erosion_shapeGrayErosionShapeGrayErosionShapeGrayErosionShapegray_erosion_shape).

注意

Note that gray_closing_shapegray_closing_shapeGrayClosingShapeGrayClosingShapeGrayClosingShapegray_closing_shape requires considerably more time for mask sizes of type float than for mask sizes of type integer. This is especially true for rectangular masks with different width and height!

gray_closing_shapegray_closing_shapeGrayClosingShapeGrayClosingShapeGrayClosingShapegray_closing_shape 可在 OpenCL 设备上执行。 In case of mask sizes of type float the result can vary slightly from the CPU as the interpolation is calculated in single precision on the OpenCL device.

请注意，若使用域缩减后的图像作为输入，滤波器算子可能会返回意外结果。请参阅 滤波器 一章

执行信息

• 支持 OpenCL 计算设备。
• 多线程类型：可重入（与非独占算子并行运行）。
• 多线程作用域：全局（可从任何线程调用）。
• 在元组级别上自动并行化。
• 在通道级别上自动并行化。
• 在域级别上自动并行化。

参数

结果

gray_closing_shapegray_closing_shapeGrayClosingShapeGrayClosingShapeGrayClosingShapegray_closing_shape 在所有参数正确时返回 2 ( H_MSG_TRUE )。

替代

gray_closinggray_closingGrayClosingGrayClosingGrayClosinggray_closing

另见

gray_dilation_shapegray_dilation_shapeGrayDilationShapeGrayDilationShapeGrayDilationShapegray_dilation_shape, gray_erosion_shapegray_erosion_shapeGrayErosionShapeGrayErosionShapeGrayErosionShapegray_erosion_shape, closingclosingClosingClosingClosingclosing

模块

基础