rank_nrank_nRankNRankNrank_n (算子)

名称

rank_nrank_nRankNRankNrank_n — 从多个通道返回具有给定秩的灰度值。

签名

rank_n(Image : RankImage : RankIndex : )

描述

算子 rank_nrank_nRankNRankNRankNrank_n returns pixel-by-pixel the result of the rank-function over all channels.

For every pixel in the input image the following is being done: The gray values of all channels at this position are sorted in ascending order. Then the pixel with index RankIndexRankIndexRankIndexRankIndexrankIndexrank_index is selected and placed in the output image at the same position. The output image has one channel.

In the special cases RankIndexRankIndexRankIndexRankIndexrankIndexrank_index = 1 and RankIndexRankIndexRankIndexRankIndexrankIndexrank_index = '(Number of channels)'"(Number of channels)""(Number of channels)""(Number of channels)""(Number of channels)""(Number of channels)" the minimum and maximum are returned. RankIndexRankIndexRankIndexRankIndexrankIndexrank_index = '(Number of channels + 1) / 2'"(Number of channels + 1) / 2""(Number of channels + 1) / 2""(Number of channels + 1) / 2""(Number of channels + 1) / 2""(Number of channels + 1) / 2" returns the median (here, / denotes integer division). Hence, for a five-channel image, 3 returns the median.

算子 rank_nrank_nRankNRankNRankNrank_n should not be confused with the operator rank_imagerank_imageRankImageRankImageRankImagerank_image which computes the rank within a certain mask.

有关平滑滤波器概念的说明，请参阅滤波器 / 平滑一章的引言。

注意

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

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。
在域级别上自动并行化。

参数

ImageImageImageImageimageimage (输入对象) (multichannel-)image(-array) → object (byte / int2 / uint2 / int4 / int8 / real)

Multichannel gray image.

RankImageRankImageRankImageRankImagerankImagerank_image (输出对象) singlechannelimage(-array) → object (byte / int2 / uint2 / int4 / int8 / real)

Result of the rank function.

RankIndexRankIndexRankIndexRankIndexrankIndexrank_index (输入控制) integer → (integer)

Rank of the gray value images to return.

默认值： 2

建议值： 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20

示例（HDevelop）

compose5 (Image1, Image2, Image3, Image4, Image5, MultiChannelImage)
rank_n (MultiChannelImage, ImageMin, 1)
rank_n (MultiChannelImage, ImageMax, 5)
rank_n (MultiChannelImage, ImageMedian, 3)

可能的前趋

compose2compose2Compose2Compose2Compose2compose2, compose3compose3Compose3Compose3Compose3compose3, compose4compose4Compose4Compose4Compose4compose4, compose5compose5Compose5Compose5Compose5compose5, add_channelsadd_channelsAddChannelsAddChannelsAddChannelsadd_channels

替代

mean_nmean_nMeanNMeanNMeanNmean_n

另见

count_channelscount_channelsCountChannelsCountChannelsCountChannelscount_channels, rank_imagerank_imageRankImageRankImageRankImagerank_image

模块

基础

算子