How to directly read an image file from a url address in R

Question

I'm trying to directly load a .jpeg image from a url address. I'm wondering if there is a basic way to do this by using a url connection.

I first tried:

require(biOps)
con <- url("http://upload.wikimedia.org/wikipedia/commons/9/95/Apollonian_spheres.jpg")
pic <- readJpeg(con)
#Error in readJpeg(con) : Cannot open file.

This other question seems to be along the same lines, but for a .png file. I tried to adapt to a .jpeg, but also got an error.

require(biOps)
require(RCurl)
myurl <- "http://upload.wikimedia.org/wikipedia/commons/9/95/Apollonian_spheres.jpg"
pic <-  readJpeg(getURLContent(myurl))
#Error in readJpeg(getURLContent(myurl)) : Cannot open file.

Any help would be greatly appreciated!

score 22 · Accepted Answer · edited Oct 15 '14 at 22:23

22

Just save the image as a temporary file:

myurl <- "http://upload.wikimedia.org/wikipedia/commons/9/95/Apollonian_spheres.jpg"
z <- tempfile()
download.file(myurl,z,mode="wb")
pic <- readJPEG(z)
file.remove(z) # cleanup

edited Oct 15 '14 at 22:23

Mihai Ionescu

968
1
8
13

answered Jul 29 '13 at 07:48

Thomas

43,637
12
109
140

Thanks Thomas. Can you also provide a way to then delete the file `z` from within R after loading? – Marc in the box Jul 29 '13 at 07:53
@Marcinthebox Just added it. – Thomas Jul 29 '13 at 07:57

score 7 · Answer 2 · answered Jun 10 '20 at 14:43

You do NOT have to save the image into the disk. You could directly read the image through the URL as well as below:

library(magick)
image_url <- "http://upload.wikimedia.org/wikipedia/commons/9/95/Apollonian_spheres.jpg"
pic <- image_read(image_url)
print(pic)

How to directly read an image file from a url address in R

2 Answers2

Linked