This is a card in Dave's Virtual Box of Cards.

Use exiftool to strip EXIF data from shared images

Page created: 2025-05-15

Updated: 2025-05-20

As you may know, the image files you put "out there" on the Web may have info you don’t wish to make known, such as the exact model of the device that took the picture or, worse, the GPS coordinates of the location where you took the picture.

ExifTool by Phil Harvey is the command-line tool for manipulating EXIF data in image files:

https://exiftool.org/

It’s almost certainly available in your Linux or BSD distro’s repo if it’s not installed by default.

Like many powerfull CLI tools, it’s got a huge list of options in the man page. But you can get by with a small subset of its features.

To view EXIF data in an image:

$ exiftool concerned_beanz.png
# ...
# all file metadata and exif data shows here
# ...

Or just the "Software" used to make the image:

$ exiftool -Software concerned_beanz.png
Software                        : www.inkscape.org

To write the EXIF tag "Software":

$ exiftool -Software='Ratware 9000' concerned_beanz.png
    1 image files updated

$ exiftool -Software concerned_beanz.png
Software                        : Ratware 9000

To delete the EXIF tag "Software":

$ exiftool -Software= concerned_beanz.png

To delete all optional EXIF tags from the image:

$ exiftool -all= concerned_beanz.png

Note that all of the above examples will also produce a copy of the file before it was altered:

$ ls
concerned_beanz.png
concerned_beanz.png_original

To strip EXIF data overwrite the original file without making a backup, add -overwrite_original_in_place like so:

$ exiftool -all= concerned_beanz.png

Recurse directories, ignore directories

You can certainly write your own script around exiftool, or use it in a fancy find invocation. But I really appreciate that the tool has directory traversal covered right out of the box.

Just add -r to perform an action on all supported image and media files found in a directory and subdirectories (as with many tools, "r" is for "recursive").

I just navigate to the directory I want and use . for the current directory, so stripping all tags looks like this:

$ exiftool -all= -r .

But I also want to skip a particular directory. Thankfully, that’s easy with the -i option ("i" is for "ignore"):

$ exiftool -all= -i repos -r .

Oh, and I also don’t want to have to go through and remove all of the "_original" copies because I already have these backed up. So I add the -overwrite_original_in_place option.

Put it all together, and this stripped all of the unwanted EXIF data from my website:

$ exiftool -all= -overwrite_original_in_place -i repos -r .

Auto-strip new files

I can see which files are new by asking git.

I’ve updated the rat script that I use to manage this site to have a new rat exif sub-command.

Here’s the relevant part:

my $output = `git ls-files -o -m --exclude-standard`;
my @files = split(/\n/, $output);
my $fcount = @files;
print "Site has $fcount changed files...\n";
foreach (@files) {
    if($_ =~ /\.(jpg|png)$/){
        print "Found: $_\n";
        print `exiftool -all= -overwrite_original_in_place "$_"`
    }
}
print "Done.\n";

Here’s something similar in shell (untested):

for f in $(git ls-files -o -m --exclude-standard  | ag '(jpg|png)$')
do
    exiftool -all= -overwrite_original_in_place "$f"
done