I need
1 to check 2 folders full of images for clones, intending here for clones identical images present in the in the same folder but with different name
2 done that i need to confront the 2 folders, to eliminate from the folder 2 or clone of images already in 1
2 to eliminate from folder 2 all images already present in folder 1
(

The problem here is that files names are not relevant (in both folders files names are just randomly assigned progressive numbers with no relation with the content )



so only way to detect the clones is confront the images
are 1300 in the first folder and about 1800 in the other folder, more or less 50% of folder 2 is composed by clones already in 1 to be deleted

well how to do it ?

Clones must have same image same also for image size,
in case have different extension (possible are jpg and png) to be eliminated should be the jpg

PhotoComix:

I think I remember that you are able to use a GNU/Linux OS. My first thought was that as long as the images you want to delete are clones with no difference in processing, then a nice little script comparing md5sum values should be able to find clones and then allow you to delete them. I did a search and found this site with two command-line scripts that you can run in a terminal window:

http://ajayfromiiit.wordpress.com/2009/10/16/one-liner-to-find-and-remove-duplicate-files-in-linux/

What you will have to do is place one folder inside the other folder since the script gets run on the folder of your choice and all folders inside that folder to make the comparison. I tested the scripts on a small set of images where I made a copy of the original. It found the duplicates.

The second script on there allows you to search for the duplicates and will then ask your permission to delete all of the files before doing so.

I only use GNU/Linux so I do not know what may be available for users of Windows.

No problem to use Linux i just updated kubuntu

The md5sum seems a good idea, but i am not sure would be sufficent:

many different images there may have the same size because all variants of the same original
(but as example mirrored along different axes )

I don't want delete the variants on the contrary i need them , i only need to eliminate the perfect clones (same luminosity, same colors , same shapes,same image size...again file names in this cases are not relevant )

Another program that I found is called fdupes. I installed it in Debian and it also can find and delete duplicate image files. I tested it here and it found duplicate image files. Easier to use than the script I found above.

oh this seems perfect going to try now

PhotoComix, if you have the possibility of using Windows or running Wine from within Kubuntu, I suggest you installing VisiPics, a free application that permits to detect not only the real duplicates, but also to find the similiarity between images, as it can occur with pics differing only for level of compression or size.

http://www.visipics.info/index.php?title=Main_Page

The check is performed by comparing the contents of the images, hence regardless of their possibly different names and/or extensions.

For Windows have a look at Anti-Twin. It will do pixel matching.
http://www.joerg-rosenthal.com/en/antitwin/

Antitwin and vispics were both good solution i believe i get rid out of all clones