Digikam/Performing Fuzzy Searches In digiKam

    From KDE UserBase Wiki
    Revision as of 17:03, 3 June 2019 by Ognarb (talk | contribs) (Marked this version for translation)
    (diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
    Other languages:

    Performing Fuzzy Searches In digiKam

    Transcribed from Mohamed Malik's blog, 17 March 2011. Last update 2019

    One of the useful tools that can be found in digiKam is the ability to perform fuzzy searches, where the user can guess and sketch the colors that the image may contain. In order to perform a fuzzy search you need to rebuild the image fingerprints in your database.

    To do this go to ToolsMaintainanceRebuild Finger-prints. You can additionally choose if you want to rebuild the entire database or only scan your database for missing fingerprint (default behavior).

    After installing digiKam 2 beta 3 from source a few weeks back I created a new database for use with the developmental branch, in my Dual Core 3.2 GHZ ( FSB 800) & 2GB DDR3 RAM, rebuilding all fingerprints took around one and half hours. Therefore please remember that this operation of rebuilding all fingerprints can take sometime depending upon your system.2 beta 3

    However choosing the Faster option is quicker as it only searches for missing fingerprints. If you have already scanned your images for fingerprints then you should skip this step.

    On the left hand side of digiKam, look for the tool with wand and select that tool. Navigate to the tab Sketch. Choose a color that your image might contain, you can change the size of the pen which is on the scale of 1-10, and you can also choose the number of images that you want to be displayed during your fuzzy search, this can be set to a value between 1-50 items. As you sketch on the area you will see that the fuzzy search related images appears on the right hand side. After performing searches you might want to save the search results, in that case give it a name and press the Save button. See screen capture below.