Palette-based Photo Recoloring: Related Work

Design Discourse Two

Related Work: Color Transfer (Example based recoloring)

Researchers such as Erik Reinhard: 2001: University of Utah, suggest that a considerable amount of recoloring methods require an example image to work off.

Simply make full use of the of the lab colour space and apply a statistical transformation to map colors from another image.

Lab color Space: A color-opponent space with dimension L for lightness and a and b for the color-opponent dimensions, based on non-linearly compressed coordinates.

Statistical transformation: A transformation based off the data values of, in this case, an image.

Color Opponent Space: The color opponent process is a color theory where the human visual system interprets information about color by processing signals from cones and rods in an opposing manner. 


  • L-lightness (White to black)
  • A-Cyan to Magenta
  • B- Blue to Yellow

Video provided by Rick Rys:

If the reference image is dissimilar, the user will have to manually point out  the regions of close similarity using swatches.

Huiwen Chang:2005: Princeton, suggested organising pixels into a basic color category which have been experimentally derived, to then match input pixels to reference pixels within the same category.

The results are compelling, however, highly depend on the compatibility between the two images. Where the reference image is used as an input which has to be provided by the user or produced by an other algorithm.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s