How to find pictures that are the same, except for rotation?

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

How to find pictures that are the same, except for rotation?

M. Fioretti
Greetings,

I'm a happy user of digiKam 1.9.0. I have recently added to my
collection a LOT of picture folders I got from relatives and
friends. I have realized that there are a lot of "duplicates"
scattered around those folders, that I would like to remove.

The problem is that they are not really identical files, but rotated
versions of the same original picture, without any other change or
post processing, and sometimes they don't even have the same names...

For example, I may have 3 files called

P1919005.JPG
P1919005bis.JPG
marco.jpg

they're all jpg files generated by some digital camera that included
standard exif tags, and only when you actually **look** at them and at
their EXIF tags you realize that you only need to keep one of them,
because one (but which one???) is the original as produced by the
camera and the others are copies that were manually renamed/rotated
later

So my question is, can digiKam help me to:

- find all pictures that are rotations of the same original
- (ideally) remove all copies and keep only the original?

if not digikam, what would be another (Linux compatible) way to solve
this problem? I have no problem to write shell scripts and/or use
exiftools if there's no other way, but in this specific case I'd
really appreciate some pointer in the right direction.

        TIA,
        Marco
--
Digital Citizenship online course: http://mfioretti.com/node/129

Because your civil rights and the quality of your life depend on how
software is used **around** you
_______________________________________________
Digikam-users mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-users
Reply | Threaded
Open this post in threaded view
|

Re: How to find pictures that are the same, except for rotation?

Ben Staude
Brute force method: write a script that makes three copies of each picture
that are rotated 90/180/270. Now you can use some tool to "search for
duplicates", or, if you're lucky and everybody in the chain used lossless
rotation and didn't change anything else in the file, even md5sum might work.

Could be worth a try.

Ben

Am Freitag 17 Juni 2011 schrieb M. Fioretti:

> Greetings,
>
> I'm a happy user of digiKam 1.9.0. I have recently added to my
> collection a LOT of picture folders I got from relatives and
> friends. I have realized that there are a lot of "duplicates"
> scattered around those folders, that I would like to remove.
>
> The problem is that they are not really identical files, but rotated
> versions of the same original picture, without any other change or
> post processing, and sometimes they don't even have the same names...
>
> For example, I may have 3 files called
>
> P1919005.JPG
> P1919005bis.JPG
> marco.jpg
>
> they're all jpg files generated by some digital camera that included
> standard exif tags, and only when you actually **look** at them and at
> their EXIF tags you realize that you only need to keep one of them,
> because one (but which one???) is the original as produced by the
> camera and the others are copies that were manually renamed/rotated
> later
>
> So my question is, can digiKam help me to:
>
> - find all pictures that are rotations of the same original
> - (ideally) remove all copies and keep only the original?
>
> if not digikam, what would be another (Linux compatible) way to solve
> this problem? I have no problem to write shell scripts and/or use
> exiftools if there's no other way, but in this specific case I'd
> really appreciate some pointer in the right direction.
>
> TIA,
> Marco

_______________________________________________
Digikam-users mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-users
Reply | Threaded
Open this post in threaded view
|

Re: How to find pictures that are the same, except for rotation?

Ignatius Reilly
I solved this same problem easily with a very nice tool: fslint (GUI to
the CLI program fdupes)

HTH


Ben Staude thus spake on 26/06/11 20:23:

> Brute force method: write a script that makes three copies of each picture
> that are rotated 90/180/270. Now you can use some tool to "search for
> duplicates", or, if you're lucky and everybody in the chain used lossless
> rotation and didn't change anything else in the file, even md5sum might work.
>
> Could be worth a try.
>
> Ben
>
> Am Freitag 17 Juni 2011 schrieb M. Fioretti:
>> Greetings,
>>
>> I'm a happy user of digiKam 1.9.0. I have recently added to my
>> collection a LOT of picture folders I got from relatives and
>> friends. I have realized that there are a lot of "duplicates"
>> scattered around those folders, that I would like to remove.
>>
>> The problem is that they are not really identical files, but rotated
>> versions of the same original picture, without any other change or
>> post processing, and sometimes they don't even have the same names...
>>
>> For example, I may have 3 files called
>>
>> P1919005.JPG
>> P1919005bis.JPG
>> marco.jpg
>>
>> they're all jpg files generated by some digital camera that included
>> standard exif tags, and only when you actually **look** at them and at
>> their EXIF tags you realize that you only need to keep one of them,
>> because one (but which one???) is the original as produced by the
>> camera and the others are copies that were manually renamed/rotated
>> later
>>
>> So my question is, can digiKam help me to:
>>
>> - find all pictures that are rotations of the same original
>> - (ideally) remove all copies and keep only the original?
>>
>> if not digikam, what would be another (Linux compatible) way to solve
>> this problem? I have no problem to write shell scripts and/or use
>> exiftools if there's no other way, but in this specific case I'd
>> really appreciate some pointer in the right direction.
>>
>> TIA,
>> Marco
>
> _______________________________________________
> Digikam-users mailing list
> [hidden email]
> https://mail.kde.org/mailman/listinfo/digikam-users
>
_______________________________________________
Digikam-users mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-users
Reply | Threaded
Open this post in threaded view
|

Re: How to find pictures that are the same, except for rotation?

M. Fioretti
In reply to this post by Ben Staude
On Sun, Jun 26, 2011 20:23:01 PM +0200, Ben Staude wrote:

> Brute force method: write a script that makes three copies of each
> picture that are rotated 90/180/270. Now you can use some tool to
> "search for duplicates", or, if you're lucky and everybody in the
> chain used lossless rotation and didn't change anything else in the
> file, even md5sum might work.

OK, I'll try something like that, thanks.

Marco
--
Digital Citizenship basics:
http://mfioretti.com/node/129
_______________________________________________
Digikam-users mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-users