Graham Finlayson and Steven Hordley
Department of Computer Science
The University of York
York Y01 5DD
{graham,sdh}@minster.cs.york.ac.uk
The requirement that the sensor responses of a camera to a given surface reflectance be constant under changing illumination conditions, has led to the development of so called colour constancy algorithms. One such algorithm developed by Forsyth [ 7 ] and later extended by Finlayson [ 4 ] exploits the constraint that under a canonical illuminant all surface colours fall within a maximal convex set. This leads to a feasible set of mappings representing the possible unknown illuminant lighting the scene. Here we address the question of how best to select a single estimate for the illuminant from this feasible set.
We develop our approach in the context of Finlayson's colour-in-perspective algorithm. This algorithm performs a perspective transform on the sensor data to discard intensity information, which without unrealistic constraints (uniform illumination and no specularities) being placed on the world, cannot accurately be recovered. Unfortunately, the feasible set of mappings that is calculated is also perspectively distorted. Here, we argue that this distortion must be removed prior to carrying out map selection and show that this is easily achieved by inverting the perspective transform. A mean-selection criterion operating on non-perspective mapping space provides good colour constancy for a variety of synthetic and real images. Importantly, constancy performance surpasses all other existing methods.
Adrian F Clark