Defining Color band to recognize from sing layer raster in R

Question

I have a 2 part question, the first part was easy & is answered but I have one issue :

1) Load map in R and convert to Raster : Done

2) Tell R to only show a defined color band : NOT done

As you can see in my example map, it is a map of Namibia, and it is a single layer black and white map, with one blue dot. The RGB code of this blue dot is: Red == 14 , Green == 6, and Blue == 209

What I would like to do is tell R to Isolate this dot, and to leave out the rest, so in essence, when I plot the map, I have only the blue dot.

Here is my current code, and here is an example map.

rm(list=ls())

library(raster)

library(maps)

library(magick)

library(rgdal)

# working directory

setwd("Working here")
# load first image
Species_id<-1
Species_i <- image_read(paste("Species (", Species_id, ").png", sep=""))

Species_i_rotated<-image_rotate(Species_i,0)
image_write(Species_i_rotated, path = "Temporary.png", format = "png")

# create raster
Converted <- raster("Temporary.png", band=1)

#### This is where my problem needs solving
#This defines the threshold to classify where the dot on the image is present as 1
# While removing all the unwanted features

Converted<-Converted == 14 & Converted==6 & Converted==209
Converted[Converted==0]<-NA

plot(Converted) # I would like this to show only the blue dot.

Answer 1

Instead image_read, I used readPNG to solve your problem with less code:

# read image
img <- png::readPNG('~/path/to/image.png')

The output is an array with colors in 0 - 1 format.

# convert to raster
r <- brick(img)

raster::plotRGB(r, scale = 1)

# select blue color (R = 0%, G = 0%, B = 100%)    
plot(r[[1]] == 0 & r[[3]] == 1)

# Use it to mask
raster::plotRGB(mask(r, r[[1]] == 0 & r[[3]] == 1, maskvalue = 0), scale = 1)