Image analysis in R by using colocr-CodePudding

This is the first time for me to ask a question, hopefully I wrote my script in correct way. I wanted to analyze my immunofluorescence images, each images have 2 different colors for same cells, and basically I want to show all the cells with blue stain which is stained by every cells, and pixel differences of green between images. I load the merged image which include blue and green stains. I used colocr package. However R doesn't recognize the blue stain, I confirm by looking images if all cells have blue. So I was wondering if anyone know what it is the solution to show blue stain. I attached my script below

Thank you

library(imager)

library(magick)

library(colocr)

library(imagerExtra)

>image <- image_load(system.file('extdata', 'image.tif', package = 'colocr'))       # merge
>image1 <- imager::channel(image, 1)  # blue
>image2 <- imager::channel(image, 2)  # green

>par(mfrow = c(1,3), mar = rep(1,4))
>plot(image, axes = FALSE, main = 'Merged')
>plot(image1, axes = FALSE, main = 'blue')
>plot(image2, axes = FALSE, main = 'green')

Created on 2022-04-29 by the reprex package (v2.0.1)

CodePudding user response：

Are you sure you have the index of the green and blue channels correct? Often channel 1 is red and channel 2 is green. To remove doubt, you could use imager::G to extract the green channel and imager::B to get the blue channel:

library(imager)
library(magick)
library(colocr)
library(imagerExtra)

image <- image_load(system.file('extdata', 'example1.png', package = 'colocr'))
image1 <- B(image)  # blue
image2 <- G(image)  # green

par(mfrow = c(1,3), mar = rep(1,4))
plot(image, axes = FALSE, main = 'Merged')
plot(image1, axes = FALSE, main = 'blue')
plot(image2, axes = FALSE, main = 'green')

^{Created on 2022-04-29 by the reprex package (v2.0.1)}