Optical comparison of image content (pattern-based)

In contrast to a hash-based search, DoublePics compares the image element itself with other, already existing (e.g. in a database) image elements.
A hash value comparison means, that the source and comparative elements must be identical to be found. The problem is, that a completely new hash value is created by changing just one single pixel. Accordingly, ordinary hash-based scans are powerless against any kind of changes to an element.

However, DoublePics scans your elements and compares the differences/relationship in the image element itself.

It doesn't matter whether the picture elements

• have different dimensions

• have different contrast-/color- or gamma-values or

• are available in different file formats (file extensions).

The desired sensitivity or tolerant adjustable program logic allows to find

• virtually identical image elements, as well as

• only closely related image elements, which have been changed, cropped or modified otherwise (for example, by adding logos).

This search/scan function works with the DoublePics pattern processor.
There are various configuration options available, which may affect the comparison procedure.

• DoublePics Classic

  • color tolerances

  • image interferences

  • image orientations

  • similarities in aspect ratio

  • etc.

• Pixel-Hash

  • Comparison of hashes: comparing the pixel identity