Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 2 Next »

Demo

Configuration

Duplicate detection has been enhanced with additional detection types.
If duplicate detection is enabled, the chosen duplicate detection types will be used when an asset is uploaded.

In Portal config manager settings:

  1. Click Enable duplicate detection

  2. Choose from the following Duplicate detection types

  • Perceptual hash

  • Sha1 hash

  • Use metadata field

  • Original filename

  1. Depending on the types chosen, customize further options. These will be ignored if the related detection type has not been selected.

    1. Use metadata field > Specify field used for comparison

    2. Original filename > Choose with or without extension

    3. Perceptual hash > Specify threshold for detection

Screenshot 2024-01-12 at 11.17.05.png

Detection types

Perceptual hash

Perceptual hash is a hashing algorithm designed for users to easily identify identical or visually similar/related images. This algorithm is particularly useful for locating duplicate images that have undergone various alterations, such as changes in color, minor cropping, subtle photoshopping, resizing, or skewing. Users can leverage perceptual hashing to compare and categorize images efficiently.

The perceptual hash threshold serves as a crucial parameter for determining when an image is considered a duplicate of the asset being uploaded into the system. This threshold sets the minimum similarity level required for an image to be identified as a duplicate. By default, the threshold is set to 80%, ensuring that only images with a high degree of similarity to the uploaded asset will be returned as duplicates. This default value helps to filter out only the most closely matching duplicates, guaranteeing the identification of the genuine duplicate of the uploaded asset. Users have the flexibility to adjust this threshold as needed to meet their specific requirements for image matching.

Sha1 hash

Use metadata field

Original filename

  • No labels