Assuming the image name doesn't have any spaces in it, you could do it like this:

if ( $url =~ m!/([^/]+\.(?:jpe?g|gif|png|tiff?)\b)!i ) {
    print "Found $1\n";
}
[download]

The way this works is as follows:

m!            # Use an alternate delimiter.
  /           # Match a '/' character to anchor off of.
  (           # Start a capturing parenth.
    [^/]+     # Match any number of non-'/' characters.
    \.        # Match the '.' character.
    (?:       # Group or constrain without capturing.
        jpe?g # jpg or jpeg.
      | gif   # or gif
      | png   # or png
      | tiff? # or tif or tiff
    )         # End the grouping/constraining parens.
    \b        # Make sure that tiff isn't tiffany, for
              # example.
  )           # End the capture.
!ix           # Case insensitive /i (Ignore the x, it's only
              # necessary in this expanded example.
[download]

If the URL is encoded, so that spaces have become '+' characters, etc., you will need to decode the URL first, or else filenames containing spaces will be mangled. That particular issue isn't a regexp issue, just the way URL's get encoded.