SimilarImages – Define Similarity Threshold To Find Duplicate Images

Duplicate image finder applications are backed by an algorithm which runs over all specified image folder to show information related with duplicate images and allows you to remove them, but what about all those photos which are not similar but get indexed? It often happens that duplicate image removers show images, which are actually not similar, as duplicates due to inefficiency of algorithms they depend upon. SimilarImages is developed to lets you define image similarity threshold to index only those image files which you want to remove off the system.

Apart from following conventional ways for searching similar images from specified folder, it offers Inclusive and Exclusive search patterns as well, which compares only images located in same directory and compare images not in same folder but against each other, respectively. To add image folder(s), click add button in Folder to Process section. Once added, configure image search and comparison, auto-delete, and folder reading options. Now set image comparison threshold from Threshold section and click Search button to start finding similar images from folders according to defined criteria.

similarimages 1 SimilarImages   Define Similarity Threshold To Find Duplicate ImagesView in gallery

Once process is finished, Compare Images dialog shows both original and similar images side-by-side. It shows some extra information, like, filesize, dimensions, CRC32, type and last modified time stamp, and comparison level with each image. Furthermore, you can delete original or duplicate files, remove both files, overwrite other image files with selected image, and open Windows right-click context menu.

Compare Images SimilarImages   Define Similarity Threshold To Find Duplicate ImagesView in gallery

SimilarImages has the ability to store image specific records to its native database to make sub-sequent searches relatively faster.  It supports Windows XP, Windows Vista, and Windows 7.

Download SimilarImages