A Supervise Color Constancy Algorithm Based on Finite-dimension Linear Model

Eliminating the influence of illumination is very important in applications such as color image recognition and analysis. In this paper, we proposed a new supervise color constancy algorithm based on finite dimension linear model. It can transform the color values under the abnormal illuminations into a normal illumination. Comparing with the previous methods, the algorithm does not need to know the spectral reflectance functions of color chips and the camera spectral sensitivity, thus can be efficiently applied in practice. We give the main design of the algorithm, its implementation in real imaging systems, and the experimental results.