Search code examples
rcropraster

Warning message couldBeLonLat(x) : CRS is NA using crop function in R

I would like to crop an image of 256x256 pixels into many small images of 64x64 pixels.

enter image description here

I use the function:

crop(raster,extent(xmin,xmax,ymin,ymax))

I got images of 64x64 pixels as I want but there is a white border, means the cropped image is "compressed" into a 64x64 pixels.

enter image description here

It comes from this warning message but I do not know how to avoid it.

Warning message:

In couldBeLonLat(x) : CRS is NA. Assuming it is longitude/latitude

Here my code:

#load picture
setwd("C:/Users/PC/Desktop/R/Image test/New folder/")
img_path <- "Test6.jpg"
raster <- brick(img_path, package="raster", full.names=TRUE)

size = 64
gap <- 8

#number of time to paste image
number_im_h <- dim(raster)[2]/gap-(size/gap)+1
number_im_v <- dim(raster)[1]/gap-(size/gap)+1

r_list <- list()

ind <-0

for(k in 1:number_im_h){
  for(j in 1:number_im_v){
    ind <- ind+1
    r_list[[ind]] <- crop(raster,extent(as.numeric(j-1)*gap,as.numeric(j-1)*gap+size,as.numeric(k-1)*gap,as.numeric(k-1)*gap+size))
    setwd("C:/Users/PC/Desktop/R/Image test/cropped")
    slice <- r_list[[ind]]
    jpeg(filename=paste0("image_",ind,".jpg"), width=size,height=size,units="px")
    plotRGB(slice)
    dev.off()
  }
}

How to solve it ?

Solution

  • Here is a slightly rewritten version, a bit simpler and showing how the warning should go away. For better help you should provide an example dataset (image)

    library(raster)
raster <- brick("Test6.jpg")
# set the crs so that plot knows the data are not lonlat
crs(raster) <- "+proj=utm +zone=1 +datum=WGS84"

size = 64
gap <- 8
n_h <- ncol(raster)/gap - (size/gap)+1
n_v <- nrow(raster)/gap - (size/gap)+1

r_list <- list()
path <- "C:/Users/PC/Desktop/R/Image test/cropped"
ind <- 0
for(k in 1:n_h){
  for(j in 1:n_v){
    ind <- ind+1
    e <- extent((j-1)*gap, (j-1)*gap+size, (k-1)*gap, (k-1)*gap+size)
    r_list[[ind]] <- crop(raster, e)
    f <- file.path(path, paste0("image_",ind,".jpg"))
    jpeg(filename=f, width=size, height=size, units="px")
    plotRGB(r_list[[ind]])
    dev.off()
  }
}