Color balancing is an essential part of almost any image processing pipeline. People have the ability of perceptual color constancy. This means that the color of objects will appear the same to someone under a variety of different lighting conditions. For example, a piece of blank paper will appear white whether it is under the sun at noon or at sunset. The actual light coming from the paper is quite different these two situations but to the observer, the color of the object remains the same. Color balancing can be thought of as a way to bring this ability to computers, i.e. to make an object in a photo have the same color even under different illuminants.
The process for achieving this consists of two steps:
- Illuminant Estimation
- Chromatic Adaptation